Report No. 42418-NG Nigeria A Review of the Costs and Financing of Public Education (In Two Volumes) Volume I: Executive Summary March 3, 2008 AFTH3 Human Development Unit Africa Region Document of the World Bank TABLE CONTENTSOF TABLE CONTENTS.................................................................................................................... OF ... 111 ACKNOWLEDGMENTS ...................................................................................................................... V 1 2.. INTRODUCTION ........................................................................................................................ 1 MEETING NIGERIA'S EDUCATIONSYSTEMOBJECTIVES 3 Provision of learning opportunities ......................................................................................... Enrollment ............................................................................................................................... ......................................................... 3 6 Development of cognitive skills .............................................................................................. 7 8 3 10 4.. Setting, enforcement, and financing of educational standards................................................ INSTITUTIONAL FRAMEWORKOFTHE SYSTEM .................................................................... Federal education spending................................................................................................... THE FINANCING FRAMEWORK ............................................................................................... 14 15 15 Private household spending................................................................................................... State and local government education spending.................................................................... Budget formulation and execution ........................................................................................ 17 17 Fragmentationof funding sources ......................................................................................... 19 5. Deviationbetweenapproved budget and actual spending..................................................... 19 EFFICIENCYIN RESOURCEUSE .............................................................................................. 20 6 ESTIMATING THE FUNDING GAP FOR BASIC AND SECONDARYEDUCATION .......................... 23 7. . DATA LIMITATIONS ................................................................................................................ 24 8 STRATEGICOPTIONSFORREFORM........................................................................................... . Enrollment data...................................................................................................................... Expenditure data.................................................................................................................... 24 25 Accountability reforms .......................................................................................................... 27 25 Access reforms ...................................................................................................................... Efficiency reforms ................................................................................................................. 29 31 Policy sequencing.................................................................................................................. Quality reforms...................................................................................................................... 31 ANNEX A MATRIXOF MAINISSUESAND OPTIONSOR RECOMMENDATIONS . 32 ................................ 35 List of Tables Table 2: Nigeria's Gross andNet Enrollment Rates. by Gender and Level of Education. 2006 ....4 Table 1: Basic Indicators for the Education Sector. 2001/02-2003/04 ........................................... Table 3: Mean Class Size inPrimary and Secondary Schools in SEPER States. 2005 Table 4: Characteristics of School Classrooms inEnuguand Kaduna ........................................... ..................49 9 Table 5: Dispositionof Fundsfor'RoomConstruction inEnugu and Kaduna. 2004.................... 23 List of Boxes Box 1: Consequencesof Confused Lines of Authority and Responsibility for Primary Education ....................................................................................................................................................... 12 Box 2: Government Education ReformProgram .......................................................................... 26 iii ListofFigures Figure 1: Enrollment Rates in Select SEPER States. by Age.......................................................... Figure 2: Learning Achievement inSub-SaharanandNorthAfrican Countries. 1990s.................57 Figure 3: Enrollment/Teacher Scatter Plots. a.for Public Primary Schools inJigawa. 2006 and b. for Public Junior Secondary Schools inJigawa. 2005................................................................... 21 iv ACKNOWLEDGMENTS This study was undertakenas part ofthe World Bankmultiyeareducationsector publicexpenditure review, inpartnershipwith Departmentfor InternationalDevelopment(DFID) andthe Governmentof Nigeria.This study incorporatesmaterialdevelopedby federal and state institutionsandby nationaland internationalconsultants.Itwas preparedby a team ledby HalilDundar(senior educationeconomist) and comprisingJacob Bregman(lead education specialist), TanyaZebroff(education specialist), Deborah Mikesell(senior operations officer), and Peter D'Souza (economist, DFID) with contributionsfrom Sue Berryman(consultant) andPaulBennell(consultant). The team gratefully acknowledgesthe contributionsand overallguidance ofthe government'stechnical steering committee, set upby the FederalMinistryof Education. This team was ledby A. Y. Gulma in 2006 and FatimaJiddum Ahmad in2007 andcomprised 0.Olalusi, B. 0.Momah, B.Gana, E. 0. Ihegaranya,Chimdi Ejiogu, 0.Olapeju, FloraAderogba, IfeomaAgunwa, I. Nwagwu, J. B. Olubodun, N. Tony Alabi, Sam Onazi, GeorgeC.Nweze, SamuelW. Aluwong, S. A. Adebayo, N.V. Madubuko, A. D. Isawa, M.M.Rabiu, K.Obasa,A. E. Anyanwu, S. I.Ohaeri, S. B. Harry, B. T. Feese,DanielRogger, L. Y. Aboki, HusseinPai, IsaMohammedAri, MusaAdbullahi, H.D.Kyong, IbrahimJamo, Grace Bello, KabiruM.Sarigi, E.N.Lamorde, S. E. Ameh, Ahmadu B.Nasiru, Zainab Mai-Bornu,and M.A. Alao. This study was preparedover a periodof 18 months(November 2005- June 2007). This reportis based on a reviewof federal and state educationexpenditures and a quantitativeservice delivery survey (QSDS). DFID financed the preparationof backgroundpapersat the federal and state levelthrough its Capacity for UniversalBasic Education(CUBE) Project. Eachstate reportreviewedthe fundingof the education sector, includingthe budgetcycle, expenditure levels and patterns, resource efficiencyand equity, andeducationalaccess and attainment, at the primary, secondary, andtertiary levels.A synthesis reportof state educationexpenditurereviews summarizingthe mainfindings andrecommendationsof state reportswas also preparedas a backgroundreportby a team of nationalconsultants ledby Paul Bennell(consultant). This report draws extensively from these backgroundreports. Thesereportswere preparedintwo phases,with their government counterpartsby a team ofnationaland international consultants. Phaseone concentratedon the federal level, with a team of consultants includingFredric Golladay, TeresaHarnett, PhilipOmegerie, Sarah0.Anyanwu, P.N.Awuqwo, D. H.Bolarin, A. E. Igeims, andKayodeAdesanmi. Phasetwo concentratedon a reviewofeducationexpenditures ineight states andthe FederalCapitalTerritory (FCT), which was ledby Paul Bennelland a team of Nigerian consultantsthat included: Borno: Amino AyubaandMurtalaSagagi Cross River:LouisCheteandNdemAyara FCT:BenAigbokhanand SarahAnyanwu Kano:Kabir isa Dandagoand MurtalaSabo Sagagi Enugu: OnyukwuOnyukwuand Sunday Ojayide Jigawa: KabiruIsaDandogoand MustaphaMuhammed Kaduna: HaliduAbubakar Kwara: LouisChete andAyodeleJimoh Lagos: SheriffdeenTella, AyodeleJimoh, andYusufBashir. A quantitativeservice delivery survey (QSDS) was preparedduringphasetwo. This surveywas ledby Sue Berrymanwith Anna Gueorguieva, NguyanFeese, and a team of enumerators.It was co-financedin part by the MillenniumDevelopment Goals (MDG) Officethroughthe InstitutionalDevelopmentFund V (IDF) Grant andDFIDthrough its CUBEProject. The findingsandkey recommendationswere usedin the preparation ofthe mainreport. Victoria Kwakwa(WorldBank country managerfor Rwanda), RobinHorn(educationsector manager), and MichaelMertaugh(leadeducation economist) were the peer reviewers for the study. The team gratefully acknowledgesthe suggestionsand significantcontributionsreceivedfrom Lev Freinkman(lead economist); SueBerryman(consultant) andPaulBenne11(consultant) throughout the preparationofthe study, includingthe executive summary. The team is also gratefulfor the valuable comments and suggestions received fromTunde Adekola, (senior education specialist); NguyanFeese(education specialist, consultant); EmmaDonnelly(WorldBank partnership coordinator); Peter Hawkins(HDteam leader, DFID), RichardArden (educationadvisor, DFID), RenwickIrvine(governanceadvisor, DFID); IanAttfield (education advisor, DFID); GillRogers(economist, DFID), ElsaDuret(educationspecialist, CUBE); RaphaelleMartinez(education specialist, UNESCO); andPai Obanya(education specialist, consultant). The participationofeight state governments (Cross River, Kano, Kaduna, Enugu, Borno, Lagos, Kwara, Jigawa) andthe FederalCapitalTerritory (FCT) is greatly appreciated. Useful comments were also receivedfrom many individualsduringstakeholdersmeetingsheldperiodicallyduringthe preparation of the report.The team would likethank the many government officialsoutside ofthe steering committee who providedtheir valuable advice, significantcontributionsand cooperationthroughout the study. These includeAmina Ibrahim,officials fromthe state ministriesof education, includingthe honorable commissioners, permanent secretaries, and directors, StateUniversalBasic EducationBoardchairmen and permanentmembers, executive secretaries of localgovernment educationauthorities, and staff from the states of Borno, Cross River, Enugu, Jigawa, Kaduna, Kano, Kwara,Lagos, and FCT. ElizabethForsythassistedwith editingthe report, while Anne Anglio andNgoziMalife provided excellentadministrativeand productionsupportthroughout the process. vi NIGERIA: A REVIEW OF THE COSTSAND FINANCING OFPUBLIC EDUCATION EXECUTIVE SUMMARY 1. INTRODUCTION 1. Sincethe reintroductionof multipartydemocracyin 1999, the federaland state governmentsinNigeria haveembarkedon a series of major educationalreforms.These are intended to achieve universal basic education and improve the quality and relevance o f post- basic education. Faced with large rural-urban, gender, and regional disparities inenrollment and generally poor learning outcomes, the Federal Government o fNigeriaintroduced the Universal Basic Education (UBE) Program in 1999,with the aim o f providing nine years o f free, compulsory basic education comprised o f six years o f primary and three years o fjunior secondary education to all children inthe country by 2015. The UBElaw, which was passed in 2004, sets out the key roles and responsibilities o f public agencies at all levels o f government. The law also established a new parastatal-the Universal Basic Education Commission (UBEC)-and stipulated that 2 percent o f the consolidated federal revenue be allocatedto basic education. Prompted by a desire to improve international competitiveness o f key economic sectors and sustain economic growth, the federal government also adopted a national strategic vision for the reform o f tertiary education, with a strong emphasis on science andtechnology programs. 2. Despitesignificantefforts duringthe past eightyears, muchremainsto be done in the educationsector, includingattainmentof the educationfor all (EFA) and education MillenniumDevelopmentGoals (MDGs). Unlikeother countries inAfrica, the introductionof free universal basic education has not resulted ina surge inenrollment, especially inthe north o f the country. At the same time, the overall quality o f education remains poor at all levels and varies considerably within and across states. The Federal Ministry o f Education (FME), along with all other major stakeholders, recognizes that "the education sector inNigeria is ina state o f crisis" and that "nothing less thanmajor renewal o f all systems and institutions i s required."' To this end, the government launched a major educationreformprogram in2006, which stressesthe importance o f institutional reforms to improve the efficiency and effectiveness o f service delivery at all levels o f education (see FME2007a, 2007b; 2007~). 3, This report reviews the level andpattern o fpublic expenditures on education by federal, state, and local governments and assesses the key efficiency and equity effects o f public spending for education. Since the World Bank's last education expenditure review, which was undertaken in2002, sufficient changes have taken place and more data have become available in the education sector to warrant further analysis. Intendedto assist the government inits reform agenda, the report i s based on a review o f federal and state education expenditures and a quantitative service delivery survey (QSDS) o fprimary education intwo states: 0 Educationexpenditure reviews were carried out at the federal level as well as ineight states (Borno, Cross River, Enugu, Jigawa, Kaduna, Kano, Kwara, and Lagos) and the Federal Capital Territory (FCT). These states and the FCT are referred to here as the state IThe PresidentialForum on the Education Sector, October 28,2006. 1 education public expenditure review (SEPER) states. The reviews collected data and examinedthe management o f public spendingon education across subsectors and tiers o f government, tracing it from the top down through federal systems to the state and local levels. A QSDS was conducted inKaduna and Enugu.Inaddition to this comprehensive survey, observation data were collected from all local government authorities and a representative sample o f primary schools and primary schoolteachers, and a public expenditure tracking survey (PETS) was conducted o f school construction.2 The study assessedthe availability o f resources at the school level and traced the management o f public expenditures from the bottom up, through the school to the local government levels. 4. This report also benefited from several recent studies, including a public expenditure management and financial accountability review, PEMFAR (World Bank 2006a) and a poverty assessment (World Bank 2007). The PEMFAR reviewsthe trends inexpenditures and financial management since 2001 and assesses the impact o fthe ongoing efforts. It covers fiscal policies and performance o f both the federal government and a sample o f states (Bauchi, Cross River, Enugu, and Kaduna). The poverty assessmentuses a variety o fanalyses, including household surveys, to assess poverty across sectors. The chapter on education examines the relationship between education andpoverty, including the impact o f direct costs on education outputs by income quintile, gender, and regional factors. 5. Volume 2 of this report addresses in depth the leveland pattern of public expenditureon educationinNigeria. The first two chapters review objectives, methodology, data sources, and limitations and provide an overview o f the country and the sector. Both chapter 3, on the costs and financing o f education, and chapter 4, on efficiency, analyze available data on how public resources are allocated and used withinthe education sector. Chapter 4 also examines equity inpublic spending and householdexpenditures on education. Chapter 5 provides a conclusion and offers policy recommendations. These chapters are not intended to provide a full external reviewo fNigeria's education expenditure policies. Rather they are designedto highlightthe key priorities basedon available data and information that shouldbe addressedas Nigeria moves forward with the education reform program. 6. This volume summarizes the results detailed involume 2 and highlightsthree broad needs that Nigeria will have to address inthe short to medium term for its educational agenda to succeed: Improve access to basic education, especially for girls and the poor, and improve the quality and relevance o f education at all levels. Originally, the plan was to conduct a PETS, not a QSDS. However, the conditions for a PETS did not exist, except for school construction. A PETS measures coincidence or discrepancies (leakages) between intended and actual allocations along a known finding chain. Inorder to measurethese discrepancies, policies have to specify how muchshould be allocated for some input to a beneficiary, such as a student, school, household, or clinic. InNigeria for many inputs it is not clear how much, ifanything, is supposed to be allocated per school, student, or teacher, and finding chains are not known, except inthe case o f classroom construction. 2 0 Improve governance and accountability arrangementsand strengthen institutional capabilities of the state ministriesof education to improve management, planning, and monitoring; 0 Improve the allocation and efficiency of public spending for education. 7. The following sectionselaborate these issues and identify a set of key options for addressingthem. 2. MEETINGNIGERIA'S EDUCATION SYSTEMOBJECTIVES 8. All education systemsaim to enroll childrenequitably inschool, keeptheminschool, anddevelop their cognitive skills. Iflearning opportunities are not provided or are ofpoor quality, dropout rates are high, and attendancerates are low. Enrollingchildrenfor the first time becomes significantly more difficult, especially when they are from householdsthat perceive attendance as entailing either highrisks (for example, families that worry about the safety of girls) or highcosts (for example, poor families that needchildren's labor). InNigeria, available data suggest that the education system is not achieving its basic mission: equitably enrolling childreninschool and producing learning outcomes that compare acceptablyto regional and international benchmarks. Enrollment 9. As shown intable 1,during the 1990sNigeriamade significant progress inimproving enrollment rates inprimary education and inaddressing gender parity inprimary and secondary education. Still, the primary enrollment rate i s slightly lower thanthe averageof Sub-Saharan African countries, but significantly lower than the average of middle-income countries. As shown infigure 1 andtable 2, however, enrollment rates at all levels of education vary by gender and by state. Basedon the available data, only 63 percent of school-age boys and 60 percent of school-age girls nationwide currently attendprimary school. Net enrollment rates are much lower than gross enrollment rates, indicating a large number of under-age or over-age childrenin the system. The highest estimate of the number of primary school-age children who are not in school i s 8 million, suggestingthat around 40 percent are not attending primary school and 60 percent are not attending secondary school, with markedly lower rates ofparticipation inthe northern states. 3 Table 1: Basic Indicators for the Education Sector, 2001/02-2004/05 Source: Based on most recent available data from various sources. a NBS.2006; FME.2006; basedon partial estimates of public spending for educationat the federal and state level. See chapter4 for details. -Not available. Table 2: Nigeria's Gross and Net Enrollment Rates, by Gender and Level of Education, 2006 Gross enrollment rates Net enrollment rates Education level Female I Male Female I Male ~~ ~~~ Primary 88.5 96.1 59.8 63.0 Junior secondary 67.9 74.6 27.8 28.1 Senior secondary 65.2 69.5 25.4 29.2 Post-secondary(18-20 years old) 22.1 32.8 4.9 6.3 4 Figure 1: Enrollment Rates in Select SEPER States, by Age Source: Benne11et al. 2007. Note: Estimatedfrom the 2006 CWIQ 10. Only around one-third o f the relevant age group attends junior secondary schools. The gross enrollment rate at the tertiary level i s about 10 percent o f the age cohort. This compares well with the Sub-Saharan average, but poorly with that o f other developing countries, such as South Africa and Brazil (17 and 12percent, respectively). Gender, income-related, urban and rural and regional disparities inenrollment are large at all levels o f education. 11. Indicators vary considerably by gender and betweenthe southern and northern states. Despite progress inredressing gender inequalities inenrollment, gender gaps remain large, especially inthe southeast states for boys and inthe northern states for girls, where only about 20 percent o f school-age girls are enrolled inprimary schools. Gross enrollment rates for tertiary education also vary enormously, from a low o f 1.4 percent inJigawa to a higho f 11.8 percent inLagos. The southern states generally have (a) almost universal enrollment o f children betweenthe ages o f 6 and 13; (b) a fairly sharp decline inattendance rates after 14 years o f age, although more than two-thirds o f teenagers are inschool; (c) virtually no gender differences in attendance rates at any age; and (d) enrollment o f 3 0 4 0 percent o f young adults intheir early 20s. By contrast, the northern states have (a) generally low enrollment rates for both boys and girls o f primary school age (6-1 1 years); (b) a sharp decline inenrollment rates from age 13 onward; (c) large gender differences inenrollment rates, with only about 40 percent o f school- age girls enrolled inprimary schools; and (d) enrollment o f less than 5 percent o f females and 10 percent o f males intheir early 20s. 12. The low enrollment and attendancerates at all levels are largely due to a mix of demand- and supply-side constraints. On the demand side, the highprivate out-of-pocket and opportunity costs o f education are a major cause o f low enrollment and o f dropping out o f school. Household income and school enrollment are strongly linked. Children who have never attended school come mainly from the poorest households, especially inthe northernregions. In Kaduna, for instance, among the 7-1 1 age group, 48 percent o f girls from the poorest 20 percent o f households have never enrolled in school compared to 14 percent o f girls from the richest 20 5 percent o f households. The corresponding figures for Cross River, where nearly all children attend primary school, are 5 and 0 percent, respectively. School fees (formal and informal) and expenditures on books and supplies represent the main barriers to attending school, especially in the north, where poverty is highestand enrollment is lowest. The factors affecting gender disparities are still unclear, but social factors, together with the direct costs o f education and poverty, may contribute. 13. Onthe supply side, the availability and quality o f education institutions can play a significant role in determining the rates andpatterns o f enrollment, especially inbasic and secondary education. First, the quality o f primary and secondary education is weak, as indicated by inadequate inputs (lack oftextbooks, materials, andsupplies), processes (poor quality of student-teacher interaction), and outcomes (low student achievement). This i s likely to reduce the demand for schooling. Second, the shortages o f well-maintained education facilities and qualified teachers are a particular concern for basic education. Many school age-children are unable to find accessible education facilities, especially for junior secondary grades. Provisionof learningopportunities 14. Once students are enrolledinschool, preconditionshaveto be metfor them to learn. For primary schools inKaduna and Enugu, the QSDS assessedthe preconditions for learning, finding a picture that was encouraging insome respects anddiscouraginginothers. 0 Schools have to be open. Schools inEnuguand Kaduna were visited unannounced and found to be open, unlike the findings in2002 for Nigeria's health clinics (Khemani 2004). Students are allowed to attend classes even iftheir parents cannotpay the schoolfees. More than a quarter o f Enuguschools and about one in 15 Kaduna schools that charge fees o f any kindbar students from attending classes iftheir parents cannot pay. The practice o f suspending students for inability to pay is more prevalent inEnuguthan in Kaduna. About one in six Enuguschools has a suspension policy, compared with one in 40 inKaduna. 0 Enrolled students have to attend school. On any givenday, 27 percent o f Kaduna students-one out o f every four students-are absent; inEnugu, 14 percent-one out o f every seven students-are absent. There i s little difference inabsenteeism by grade, gender, or urban-rural residency. 0 Teachershave to bepresent in the school. InEnuguand Kaduna, almost all teachers were present in school on the survey day. 0 A teacher has to be in the classroom. Almost all teachers were present inthe classroom on the survey day. InEnugu, only 1.5 percent o f teachers were not present; inKaduna, only 0.8 percent were not present. 0 Theclassroom teacher has to be actively engaged in teaching. Observations o f Enugu and Kaduna classrooms found that, ifa teacher was inthe classroom, most teachers were 6 engaged inteaching. At the same time, even inclassrooms withjust a single teacher, the teacher inone out o f every 20 classrooms was not actively engaged inteaching. The probabilities increased as the number o f teachers inthe classroom increased. In 10 percent o f classrooms with two teachers, one or bothteachers were not engaged in teaching. In30 percent o f classrooms with three teachers, one teacher was not engaged in teaching. Development of cognitive skills 15. The education sector aims to develop students' cognitive skills. New, persuasive evidence indicates that (a) both minimal and high-level cognitive skills are powerhlly relatedto individual earnings, to the distribution o f income, and to economic growth and (b) it i s the cognitive skills produced by the schools, not school enrollment or years of education per se, that result inthese outcomes. 16. The primary grades determine whether students build the cognitive foundation that enables them to progress through the grades and benefit from additional learning. Studentswith weak basic skills get significantly less out o f the same learning opportunities than students with strong basic skills. However, the learning assessment data for Nigeria's primary grades show low student achievement and no significant change in learning achievements from 1996 to 2003 (see Adekola 2007). Inthe 1990sNigeria participated inthe UnescoLJnicef-supported MonitoringLearning Achievement (MLA) Project, which measured student learning competencies inliteracy, numeracy, and life skills ingrades four and six across a number of Sub- Saharan and NorthAfrican countries. Nigerian students performed the worst of all 22 Sub- Saharan and North African countries participating inthe assessment, by a noticeablemargin (see figure 2). Figure 2: Learning Achievement in Sub-Saharan and North African Countries, 1990s 80 70 60 50 40 30 20 10 0 i Suurcc: World Bank, Nigeria Country Status Report, LOO4 7 17. In2003, the FME,again withUnescokJnicefsupport, replicatedthe MLAatthe same grade levels and inthe same subjects. In 1996 fourth-grade students could answer correctly only 25 percent o f literacy test items, 32 percent o f mathematics test items, and 38 percent o f life skills test items. In2003 fourth-grade students had higher, but still low, literacy scores, but only slightly improved mathematics scores. Mean scores for literacy changed from 25 to 35 percent between 1996 and 2003, and mean scores for mathematics changed from 32 to 34 percent. 18. The UBEC carried out national assessmentsin2001 and 2003. These were criterion- referenced tests on four core subjects inthe primary school curriculum. Tests were administered at grades four, five, and six. The UBEC assessment o f fifth-grade students in2001 found that only one infive students was able to answer correctly more than 30 percent o f the test items; fewer than 1 percent o f studentswere able to answer correctly more thanhalf o f the test items.In 2003 UBEC carried out a second assessment ingrades four, five, and six. Mean scores across the four subjects were low at each grade level, with students able to answer correctly about one in four test items inEnglishand social studies and about one inthree test items in science and social studies. Fourth-grade scores ranged from 25 to 50 across the four subjects, fifth-grade scores rangedfrom 25 to 39, and sixth-grade scores rangedfrom 21 to 40. Results at each grade level were lowest for Englishlanguage, a subject inwhich test items are highly dependent on literacy and language use (Adekola 2007: 5-8). Setting, enforcement, and financing of educational standards 19. Standards vary widely across schools and localgovernment areas. There is no agreement on standards as the national minimumstandards that exist in law or inregulations are not enforced, or the resources required to meet established standards are not provided. To improve learningachievements, standards have to be set, followed, and enforced. This i s particularly necessary for class size, educational facilities, teaching and learning materials, health and safety, andteacher provision andqualifications. 0 Classsize. Average class size for primary schools varies significantly among states, in some cases beingmuchhigher and inothers being lower thanthe 35 and 40 students per class inprimary and secondary education, respectively, recommendedby the National Policy on Education. This standard balances adequate conditions for learning and the efficient use o f facilities andteaching staff. Inthe SEPER states, the range i s from a high o f 145 inBorno to a low o f 32 inLagos (see table 3). The modal range i s between 50 and 75 students. The average class size for junior secondary schools is inthe range o f 35 to 55 students ineight o f the SEPER states. Within states, average class size varies considerably from one local area to another for bothpublic primary and secondary schools. 8 Table 3: Mean Class Size in Primary and Secondary Schools in SEPER States, 2005 Jigawa 67 93 52 42 33 44 Kaduna 70 65 54 90 41 70 Kano 93 93 48 41 40 38 Kwara 52 50 36 25 30 25 Lagos 32 5 70 25 52 20 0 Classroom characteristics. Based on observations o f a sample o f classrooms inEnugu and Kaduna, prevailing standards regarding issues such as cleanliness, furniture, and lightingvary considerably between schools (see table 4). Table 4: Characteristics of School Classrooms in Enugu and Kaduna Source: QSDS Access to core textbooks and teaching materials. Although all nine SEPER states aim to provide every primary school student with exclusive access to the prescribedtextbook for each of the four core subjects, the ratio o f students to core textbook was higher than 2.3:1 inprimary schools inall nine states in2005, andthe ratio o f students to core textbook increased as the student moved through the grades. The QSDS found that only 20-50 percent of students inEnuguhad the textbook, depending on the subject, and no more 9 than one in seven had the textbook on any subject inKaduna. About one in 10 Enugu teachers and two in 10 Kaduna teachers did not have the textbook for a given subject. About 45 percent o f Enuguteachers and 75 percent o f Kaduna teachers did not have the teachers' guide for a given subject. Although almost all teachers inboth states had a writing implement,paper, and chalk, only half of Kadunateachers and 15 percent o f Enuguteachers had a blackboard or a blackboard inreasonable condition. Only about one ineight Enuguprimary schools and one in30 Kaduna schools had a library. Health and safety standards.Although staff may have personal cell phones, site visits found that virtually no school ineither Kaduna or Enuguhad a telephone or radio transmitter for communicatingwith the outside world incases o f fire, medical emergency, or other dangers. The national ratio o f pupilsto toilet i s 292:1,compared to the recommendedratio o f40:1.About 70 percent o fEnuguschools and 75 percent of Kaduna schools did not have a toilet. More than a third o f Enuguschools and almost half o f Kaduna schools did not have access to water, either protectedor unprotected. About halfo f Enuguschools and a third o f Kaduna schools had only an unprotected supply o f water. More than half o f Enuguschools and two-thirds o f Kaduna schools had at least one safety hazard. Teacher qualifications. A significant percentage,ofteachers do not have the minimum requirement of at least three years o fpostsecondary teacher education (theNigeria Certificate inEducation ,NCE). The QSDS found that the Enuguteaching force for the primary grades is generally better educatedthanthe Kadunateaching force, but a quarter o f the teachers inEnuguand half o f those inKaduna have no more than two years o f postsecondary education. Too many primary teachers have such limitededucation that they are unable to understandthe subjects they must teach. 20. Insummary, standards exist inseveral areas, butno standards exist inthe provision of key inputs. Implementation ofthe minimumstandards is often constrained by the lack of resources to enforce and the capacity to monitor at the state and local government levels. 3. INSTITUTIONALFRAMEWORK OF THE SYSTEM 21. The Nigerianeducation system i s decentralized under a federal structure, and public education i s a concurrent responsibility o f the federal, state, and local governments, with a fast- growing private sector. The complex set of institutional and intergovernmental relations for the provision of education is particularly opaque in Nigeria, as it does not define the roles and responsibilities among the three tiers of government, leaving no government or agency clearly accountable for results. Over the last eight years, these institutional relationships have become increasingly complicated for several reasons: (a) increased decentralization, (b) the creation o f new agencies and other organizations (for example, UBEC and state universal basic education boards, SUBEBs), (c) new policy initiatives (for example, the UBEIntervention Fund), and (d) the emergence ofprivate education. Unclear roles andresponsibilities, especially for expenditure and management, together with frequent policy changes, especially regarding basic education, have caused confusion, duplication, and sometimes rivalry inthe discharge o f responsibilities. 10 22. The federal government, which provides policy leadership andregulatory oversight through the FMEand the National Council on Education, has a collection o f 21 parastatals to implement its function^.^ The federal government also finances a multitude o f parastatals, such as UBEC and the EducationTrust Fund(ETF), which runvertical programs supporting state and local governments inthe delivery of education. The federal government plays a dominant role inthe provision of tertiary education, state governmentsprovide mainly secondary and regionally focused tertiary education, and localgovernments provide basic education. The FMEandparastals have overlapping functions; for example, the Monitoringand Evaluation Unit o f UBEC, the Federal Inspectorate, the Department o f Planning, Research, and Statistic~,~ UBEC, and the National EducationResearch and Development Council all have mandates to collect data. 23. The state ministries o f education have a similar proliferation o fparastatals, with functions even less well definedthanthose at the federal level, and 774 local government authorities represent state government. Inthe education sector, they are responsible for delivering and managingprimary education through local education authorities, which report to the local government area on administrative and financial matters and to the SUBEB on operational and technical matters. 24. Although inprinciple local government areas are responsible for delivering basic education, they have little ability to control the budget. Officially, they receive federal transfers for basic education. Few local governments have additional resources for the education sector. Dueto the highandrising share o fteacher salaries inexpenditures, some local governments receive no transfers once teacher salaries have beensubtracted from their federal share. This provides a perverse incentive for local governments: ifenrollments go up, local discretionary income goes down, as education expenditures are deducted automatically. 25. The funding and managementof basic educationare highly fragmented. Complex and unregulated institutional arrangements that are prone to political interference govern the relationships between state ministries o f education and SUBEBs. As a ministerial agency, each state education ministry is directly responsible for its SUBEB. However, the underlying rationale for establishing SUBEBs was to create a semi-autonomous agency that i s more efficient and effective thanthe traditional ministry model. SUBEBs are closely linked to UBEC, especially through implementationo f the UBE Intervention Fund, but they rely on the state ministry o f education for salaries and operating resources (see box 1). The federal government has announcedits intention to restructure the educationparastatals.The draft Education ReformBill, which was approvedby the FederalExecutive Council, among other things, consolidatesthe 21 federal parastatalsinto 13. The bill is yet to be approvedby the National Assembly. Proposedreforms are currently under hrther review by the new Government. 4In2006the DepartmentofPlanning, Research, and Statisticswas restructuredand is now called Policy, Planning, Managementand ResearchDepartment. 11 Box 1: Consequences of Confused Lines of Authority and Responsibility for Primary Education The QSDS found that headteachers, the directors o f localgovernment authorities, and directors of localgovernment education authorities have no shared understanding about who has the power to make, and therefore is accountable for, key decisions for primary education. The QSDS asked head teachers and directors o f local government educational authorities to identify the principal and final decision maker for 16 issues. Inboth Kaduna and Enugu, a significant percentage o f head teachers gave answers that differedfrom those given by the educational director for their area, with headteachers inKadunabeing more apt to differ intheir understanding relative to their director than head teachers inEnugu. I Enugu Kaduna I L"." 34.6 1.8 84.2 74.2 04.5 29 33.4 2 30.2 80 60 40 Percent of schools 40 60 80 90 26. Such inconsistencies create inadequate accountability to stakeholders, make it difficult for the systemto maintain good records, and fail to assign clear powers for decision-making. Accountability i s weak for various reasons: (a) insufficient voice and participation on the part o f stakeholders, (b) inadequate information (about outputs, outcomes, and the sources and uses o f education expenditures, especially at the state and local government levels); (c) weak management; (d) confusing roles and responsibilities; and (e) minimal incentives for efficiency at all levels o f government. L o w accountability and limitedtransparency have a negative effect on the efficient and effective delivery o f educational services. 27. Inthe absence o fa well-coordinated, unified approach, some key areas are neglected and others are duplicated. Inbasic education, inefficiencies inresource use and weak accountability occur particularly inthe areas o f management and financing across tiers o f government, teacher salaries, teacher recruitment and deployment, infrastructure development and maintenance, and inspectorate services. Managementandfinancing. Three tiers o f government share responsibility for primary education (see NERDC 2004: 55). The UBElaw assigns the federal government with responsibility for assisting state and local government authorities inthe delivery o f basic 12 education and assigns SUBEBs, under the supervision o f state ministrieso f education, with responsibility for planning, supervision, and service delivery (inspection and monitoring, salary administration, construction, provision o f instructional materials, and staff development). As with management, the responsibility for fundingprimary education across levels o f government has never beenfully resolved. Although theoretically local government authorities fund primary education, the UBElaw altered the share ofresponsibilitiesfor federal, state, and local governments. Initially, state and local governments contributed funds for primary education at a ratio o f 1:9,respectively while the federal governmentprovided funds only on an adhoc basic. The UBElaw expanded the federal government's role by allocating 2 percent o f the consolidated federal revenue through matching grants to states. Teacher salaries. For primary school teachers, local government allocations are sent directly to state governments, and teacher salaries are deducted there and paid via the SUBEB.Having clear responsibilities andmechanisms for paying teacher salaries has been a significant improvement, with virtually all Enuguand Kaduna teachers reporting that they had received their pay for all o fthe previous 12 months, regardless o f school location. Teacher recruitment. The UBE law divides the responsibility for recruitment by salary grade between SUBEBs (seventh grade and above) and local authorities (first through sixth grades). Localeducationauthorities cannot recruit staff at or above the seventh grade without permission from the state government. Since the entry point for a newly qualified NCE teacher i s the seventh grade, local government areas do not have the authority to recruit qualified teachers for junior secondary education. Local education authorities periodically inform SUBEB about the number o f teachers required, especially for pressing vacancies, but these have to be vetted and approved by the Office o f the State Governor. This constraint may explain the existence o f relatively large numberso f support staff insome areas and the limiteddecentralization o f primary education in practice. Inzastructure development and maintenance. Local governments are responsible for constructing and maintaining schools. However, few have sufficient funds for either. For example, about half o f Enuguand Kaduna schools reported not having performed any school maintenance over the previous three-year period. This i s a problem because small repairs that can be done cheaply become major repairs inthe future when ignored. 0 Inspectorate services. Each level o f government has inspection services, with each carrying out its own inspection according to its own procedures. There i s no coordination among inspection units, with much duplication o f function (FME 2007a). InKano, for example, 11 agencies provide some form o f inspection. 28. Responsibility for senior secondary education is relatively clear. Roles and responsibilities are less fragmented, and there are fewer overlapping functions in senior secondary than inbasic education. However, the disarticulation o fjunior from senior secondary education as part o f the unified system of basic education management poses a major challenge for effective service delivery. Junior and senior secondary education often share physical 13 facilities. When the UBE law shifted responsibility for junior secondary education to the SUBEB, state governments interpreted this shift invarying ways. Some states have no intention o f disarticulatingjunior from senior secondary education; others are adding stand-alone junior secondary schools or expanding existingprimary schools to housejunior secondary students. However, the UBE law does not necessarily expect SUBEBsto establish new infrastructure on the same sites as existingschools. There seems to be a contradiction betweenthe spirit and implementation o f the law. Some schools have common premises but function as two schools, with duplicate administrative staff and leadership, butwithout adequate resources. 29. The Nigerian publictertiary education sector is dividedinto federal and state institutionswith little or no coordinationbetweenthe two. At the federal level, most tertiary education institutions are locatedwithin the FME, coordinated through three parastatals (National Universities Commission, National Board for Technical Education, andNational Commission for Colleges o f Education). Inaddition, several other line ministries (e.g., agriculture, health, science and technology) coordinate specialized tertiary education institutions. Mandates for different agencies vary, with duplication, compartmentalization, and overloading o f some agencies and lack o f inclusion o f other agencies. This arrangement i s overly complex and difficult to govern, resulting inplanning without adequate budgeting, lack o f clear roles, murkyaccountability, and overlapping mandates (see World Bank 2006b). Most state tertiary education institutions have a highdegree o f autonomy, and the quality and relevance o f state tertiary educationare questionable. 4. THE FINANCING FRAMEWORK 30. Rapid fiscal decentralizationbetween 1999 and 2002 altered the arrangements for sharing revenuefrom the FederationAccount, doubling the share o f subnational spendinginthe consolidated budget from 235 percent in 1999 to 46 percent in2007. 31. There is no accurateinformationon total publicspendingon educationinNigeria. Nigeria i s a data-poor country; the lack o f information on state and local expenditures for education makes accurate estimates o ftotal spending impossible. The federal governmenthas sought to improve its capacity to monitor state expenditures, but these efforts have been constrained by the lack o f constitutional authority to oblige state to employ proper reporting and disclosure. States and local governments enjoy considerable political and fiscal autonomy, and their fiscal and budgetary systems are independento f the federal funds. Although much o f their funding comes from the FederationAccount (through revenue-sharing arrangements), subnational governments are not requiredto inform or seek approval from the federal government on their budget or fiscal performance. No national fiscal framework encompasses all budgets. No statutory mechanisms are inplace to ensure that state plans and fiscal arrangements are aligned with national goals. This is true inall sectors, notjust education (World Bank 2006a). 32. Between1996 and 2005, total public expenditureon education grew from about 2.3 to 5.4 percentof GDP. Based on incomplete estimates from nine states, between 1998 and 2001 education expenditures increased from 14.2 to 17.5 percent o f government expenditures and from 2.3 to 6.2 percent o f GDP (World Bank 2004a: 71). However, the National Bureauof Statistics does not estimate GDP at the state level. Furthermore, no federal agency gathers information on state and local government expenditures for education. It i s not possible, 14 therefore, to derive accurate estimates o f total public education expenditure as a proportion o f national and state GDP. Based on partial estimates, Nigeria's total public spending for education in2006 was anestimated 5 percent ofGDPand 12.5 percent oftotalpublic spending (FME 2007d). Aggregate public spendingas a percent of GDP is higher thanthe average o f Sub- Saharan countries (3.9 percent), but slightly lower than inSouth Africa and Kenya (5.3 and 7 percent, respectively). Federal educationspending 33. The trends in aggregateexpendituremask considerablevariations in the composition and levelof spendingbetweenfederal and state governments.Inreal terms, the federal education budget increased 47 percent over the past five years, indicating the federal government's commitment to the sector. Direct federal expenditure on education amounted to N 249.2 billion (real 2006 naira) in2007, o f which 32.5 percent was devoted to capital expenditures and 67.5 percent to recurrent expenditures. Federal recurrent expenditures on education increased sharply from 2001 to 2007. A significant share o f this increase occurred in 2005 and 2006 with introduction o f the UBE Intervention Fundand the Virtual Poverty Fund. The budget for both o f these programs represented 20.7 percent o f the total federal budget for education in2007. State and localgovernmenteducationspending 34. The budgetshares of primary, secondary,and tertiary public educationvary considerablyacross states. Among the nine SEPER states, the share o fprimary education in state and local government expenditures ranges from a low o f 36 percent inLagos to a higho f 62 percent inEnugu.The share o f secondary education ranges from a higho f 50 percent inLagos to a low of only 18 percent inCross River. 35. Some o f this variation i s reasonable. Local government areas with higher primary school allocations hire more teachers per student, and those with dense populations usually spend more on urbanprimary education thanrural areas because they tend to attract more expensive, qualified teachers. However, since local government areas determine the number o f primary schools and the number o f teachers to be employed intheir jurisdictions, some o f the variation probably reflects the level o f local commitment to primary education. For example, some areas prefer to hire cheaper, unqualified teachers. The variability inlocal government funding highlightsa potentially serious problemwiththe decentralized provision ofprimary education, especially when minimumnational standardso f service delivery are not enforced. 36. Fundingshortages are acute, especiallyat the state level. Inreal terms, public expenditure on education decreased duringthe 1 9 9 0 and ~ ~ despite improvements since the end o f militaryrule, the educationsector remains seriously underfunded. Excluding direct federal spendingthrough UBEC andthe ETF, total state education expenditures declined appreciably in real terms inall nine SEPER states (withthe exception o f Kwara) between 2001 and 2005. As a result, spending on essential inputs, such as textbooks, instructional materials, and in-service teacher training, and operations and maintenance are inadequate. About half of primary schools require major rehabilitation, with an additional 254,000 classrooms needed countrywide. 15 37. Comprehensive information is not available in any of the nine SEPER states on total overhead expenditures in public primary schools. Neither state ministries o f educationnor ministries o f local government keep adequate records on overhead expenditures. Inthe past, local governments were expected (but not legally obliged) to allocate 10 percent o f the primary school wage bill to the costs of runningprimary schools. However, from what little evidence i s available (includingvisits to schools), most local governments spend little on essential materials. Per student expenditure on instructional materials varies substantially across the states, from N 67 per student inCross River to N675 per student inthe FCT. Infrequent supplies o f chalk, registers, and exercise books represent the limit of most support from local government areas. 38. The share of federal transfers to local government authorities that are spent on primary schools varies considerably between and within states. Local governments inthree out o f the four states inthe north (Borno, Jigawa, and Kano) allocate considerably less than 20 percent o f their budgets to primary education, whereas the other SEPER states spend at least 25 percent. Within each state, fundinglevels are markedly different from one local area to another. InKano, for example, primaryeducation absorbs less than 15 percent o fthe federal allocationin 16 out o f 44 local government areas, but more than 30 percent inthree local government areas. 39. Overhead and capital expenditures are inadequate, especially in primary schools. State and local financial support for essential overhead and capital expenditures i s minimal in most states, which raises questions about the availability o f funds to expand basic education. For example, around 90 percent o f total public expenditure on primary education i s absorbed by salaries, whereas the benchmark i s about 67 percent. Capital expenditure on education accounted for only a relatively small share (less than 10 percent) of total capital expenditure inmost o f the SEPER states over the last five years. Creation o f the UBEC Intervention Fundi s an important attempt to provide additional resources for implementation, but it has not yet had a significant impact, as disbursement has been slow. 40. At the state level, capital expenditures on education are not based on a single investment plan and tend to be fragmented. The main sources o f funding for capital expenditures are state governments, UBEgrants (mostly for rehabilitation), and the ETF. Local governments generally spend little on school construction or other capital investments. Physical infrastructure projects are mostly ad hoc, not based on a physical investment plan, and funded from various sources. 41. In the absence of a comprehensive, clearly defined state policy and strategy for tertiary education, the budget proposals emanate largely from the needs and requirements of each institution. Occasionally, the state government specifies budget ceilings for each institution, but these are determinedlargely by the financial position o fthe state rather than by the development, size, or needs o f the institution. 42. State-funded tertiary education institutions are in serious financial crisis. The fundingoftertiary education is low, with universitiesand other institutions receiving less than 12 percent o f state education spending infive o f the nine SEPER states. All institutions are strugglingto generate sufficient revenue to meet eventheir most basic operating costs, let alone their capital costs. Several states (for example, Lagos) have allowed institutions to diversify revenue sources and sought to attract funding from students and the private sector, while others 16 (for example, Kano) have maintained a policy o f free education, limitingthe ability o f institutions to generate revenue through student fees. Privatehousehold spending 43. Privatehouseholdspendingon schoolingis significantat all levels.Basic education is free and compulsory, and schools are not supposed to charge fees. Inpractice, parents pay parent-teacher association feesY5school fees, and other indirect costs at all levels o f education. In primaryeducation, about 51percent o f households pay school fees-for example, the QSDS found that 63 percent o f Enuguschools and 38 percent o f Kaduna schools charge two types o f fees, with about 82 percent o f households purchasing textbooks. The average education-related expenditure per student inprimary education was N3,607 (or US$28) in2005, which represents about 12 percent of averageyearly per capita expenditure(World Bank 2007a). The burdeno f householdexpenditures is especially significant for the poorest students inrural areas. Similarly, tertiary education institutions are generating a growing share o ftheir revenues internally, including student fees, but these revenues often are not part o f the budgetary process and are usedfor discretionary spending. Public spending on education is regressive, especially at the tertiary level, indicating that the richer households benefitdisproportionately from tertiary education relative to the poor. Budgetformulationand execution 44. Shortcomingsin the qualityof publicspendingfor educationare primarilya result ofweak budgetmanagementand accountability.The current process o f formulating and executing the budget i s unreliable and poorly organized. Financing o f key inputs i s unpredictable and vertically and horizontally unbalanced, exacerbating the lack of clarity regarding who is responsible for financing which activity. Untilthe allocation o f functions i s untangled, it will be difficult to rationalize the financing framework. 45. The budgets ofmost states are based on an outdated system: policy, programming, and budget functions are not fully integrated; there i s little policy debate on spending decisions; budget execution i s far from effective; budget performance i s hamperedby the short time horizon o f the annual budget; and investment budgets are not integrated. Inaddition, states face an acute shortage o fthe technical andhumancapacity neededto design and implement reforms becausethe civil service i s poorly paid and unable to attract qualified, experienced professionals. Several states (Kano, Lagos) are attempting to improve their budget systems. At the same time, states vary intheir reform efforts, with progress primarily dependent on the degree o f local political commitment. 46. Federalfiscal transfers to state and local governmentsvia the FederationAccount are the mainsource of public revenuein most states for bothprimary andsecondary education.The federal government also funds basic education throughthe UBEIntervention Fundandthe Virtual Poverty Fundand funds all levels o feducation through the ETF.These programsare not based on state sector investmentplans anddo not take into account inequities among state and local governments inthe provision o f funds. 5The QSDS found that in Enugu8 percentofthe parent-teacherassociations charge fees, while inKaduna62 percentcharge fees. 17 47. The per capita revenues receivedby the state and local governmentsvary considerably.Allocations o fthe FederationAccount are basedon population, landarea, andthe contribution o f the state to oil revenues. The objective o f the allocation formula i s to ensure that federal resources are distributed reasonably equitably to the states and local governments, while recognizingthat, since most federal resources are from petroleum taxes and royalties, the four states that generate these revenues should receive additional funding. The unevenness inper capita allocations between states probably reflects reliance on an outdated census o f population (1991), which i s now being updated. Among the nine SEPER states, per capita revenue ranges from a low o fN9,800 inKano to a higho fN23,600 inCross River. Moreover, federal revenue per capita varies considerably among local areas inevery state. 48. The fiscal relationshipbetweenstate and local governmentsis nottransparent, despiteclaimsto the contrary by the states surveyed.The Nigerianconstitutionrequires that funds transferred from the federal government to local governments be channeled through state governments. State Houses o f Assembly have enacted laws that enable them to withhold part o f these funds for various purposes, such as payment o f salaries and pensions. Inat least one state inthe sample, anadditional deduction includedfunds for the construction and maintenance of rural roads, and the provision o fprimaryhealth care andprimary education. In2005 the state withheld an average o f 87 percent of federal funds intended for local governments each month. Accountability i s weak because states do not report the deductions intheir budgets. States are constitutionally required to transfer 10 percent o f their total revenues to the local governments undertheirjurisdiction, but those that do so transfer only 10percent o ftheir internally generated revenues, not their total revenues. 49. Traditionally,state budgetshavebeen based on incrementalbudgeting.Budget estimates for the next fiscal year are based on the approved estimates for the current fiscal year. Only new expenditures (as percentage increases inthe current budget) are closely scrutinized, and these new expenditures tend to focus on individual projects, not systemic changes. Incertain states, budgetingis more ad hoc than even incremental budgetingwould imply. However, within the last two years, several states, andin2007 the FME,have started to preparelong-term sector plans, inexpectationthat these will fully estimate costs and set priorities. 50. The existingfinancingmechanismsresultin underfunding,inefficiency,and inequity.Most local government areas have inadequate capacity and financial resources with which to manage primary education. They pay staff salaries, which state governments deduct from their share o f the FederationAccount. As a result, there i s no incentive for local government authorities or school directors to improve the quality o f education or to use resources efficiently, as reflected in low ratios o f students to teacher, highratios o f unqualified to qualified teachers, and low ratios o f teaching to non-teaching staff. Measures are needed to improve the efficiency and equity effects o f public spending. Similar constraints also exist inthe financing o f secondary education, even though secondary schools appear to be better resourced than primary schools inmost states. Some countries address these issues by implementing capitation-based financing, which Nigeria might consider for bothprimary and secondary education. 18 Fragmentationof funding sources 51. School-levelinputsare funded from a wide range of sources. Fundingsources are fragmented, and the financing system i s inchaos, especially for the financing o f basic education. For example, the QSDS found that schools inEnuguand Kaduna have, on average, three funding sources for each input, including no sources for expenses that do not entail purchases (repairs to toilet blocks), four sources for teacher's learning materials and student's textbooks, and as many as five sources for new rooms, toilet blocks, and renovation o f roofs and windows. 52. The fragmentation of fundingsources affectsthe use of resourceswithin basic education.First, it is difficult to ascertain (a) overall public spendingfor education, (b) spending across levels o f education, and (c) functions o f different government agencies. Monitoring, regulation, and policy making are very challenging. Second, different funding sources bring different incentives to the table, makingit difficult to ensure efficiency and equity inpublic spending. Third, administrative costs are highbecause different sources operate as stand-alone, vertical funding sources. Fourth, multiple sources are unpredictable, makingit difficult to ensure sustained financing for medium-term strategies at the state level. Finally, fragmentation without effective coordination compromises the effectiveness o f the budgetas a tool for achieving accountability. Deviation betweenapprovedbudget and actualspending 53. Actual spendingdeviatessignificantly from the approvedbudget, especially for capitalexpenditures.Overall, Nigeria has improved budgetpredictability (World Bank 2006a). Inthe education sector, deviationsbetweenthe approvedandactual budgetfor recurrent expenditures have become rare because the bulk o f recurrent expenditures are emoluments for current staff that have to be met first. However, important deviations exist for capital expenditures. For example, over the past few years, less than a third o f the approved capital budget has been spent. 54. InSEPER states, divergences are also large for most statetertiary educationinstitutions. Once the budget has been approved, it is supposed to be released to each institution at clearly specified intervals. However, this rarely happens. For the last two years, the School of Agriculture inKano has receivednothing, even though the approved budget for overhead was N 121 million. In2004 only 55 percent o f the approved budgetwas released. Only the Kano University o f Science and Technology and Kano Polytechnic have consistently receivedportions o f their overhead budgets, possibly because they have strong trade unions. 55. The low execution o fthe capital budget results from three factors. First, it reflects the limitedabsorptive capacity o fthe implementingunits. Evenwhenresources are available, the ability to prepare and carry out activities ina timely manner i s often limited. Second, although this has improved inrecent years, allocated funds have not been fully releasedon time, preventingthe completion o f planned work. Third, the budget i s poorly formulated. Inmany Nigerianstates, the appropriation bill often includes project ideas at such early stages that the feasibility o f their being launched duringthe concerned budget year i s low. Budget formulation should reflect the capacity o f the spending level, and funds should be disbursed ina timely fashion to facilitate implementation. 19 56. Fundingis unpredictable.Budgets proposed by spendingauthorities are significantly revised before being enacted by the National Assembly for federally funded institutions or by state governments for state-funded institutions. As a result, much o f the logic that underpinsthe formulation o f budgets is lost before the final allocations are made. Nigeria implements its budgeton a cash basis, inwhich cash receipts determine the release o f funds. The recent introduction o f the excess crude account and its use as a stabilization fund are major advancements and should improve the implementationo f budgets at the federal level. 5. EFFICIENCY INRESOURCEUSE 57. The federal government has made impressive progress inexpandingbasic education and providing additional resources through the federal budget and debt-relief gains. However, the quality o f education expenditures has not improved, inpart because state and local governments have weak capacity to plan, manage, and monitor the delivery o f education services. Despite significant efforts, insufficient attention has beenpaid to the allocation and use o f public resources for education, especially at state and local government levels, where about 80 percent o f total public education spendingoccurs. Consequently, social indicators have remained virtually unchanged for about a decade (see, for example, UNDP 2006; World Bank 2006a). The four main oil-producing statesjointly received about 90 percent o f all derivation oil payments, or about N265 billion (US$2 billion), in2005. Despitethese additional revenues, they did not perform much better thanthe rest o fthe country, asjudged by indicators o f service delivery and human development. This suggests that major problems exist inthe expenditure management and accountability systems o f these states. 58. The core problemwithin Nigeria's budgetsystem, common to all levels of governmentand all servicesectors, includingeducation,is one of inefficientspendingmore than inadequateamounts of funding. To understand this situation, this report examinedthe internal efficiency o fthe education sector at both the federal and state levels. Available data show that expenditure per student varies substantially among federal institutions and across and withinstates, which suggests inefficiencies inthe use of resources. For example, at the state level, unit costs for public primary schools vary across states from N5,000 per student inBorno to N24,000 per student inLagos. These differences are mainly due to variations in(a) teacher salaries, which are closely related to teacher qualifications; (b) the proportion o f qualified teachers; (c) the number o f students per teacher; and (d) non-salary expenditures. Salary expenditures as a share o f recurrent expenditures account for more than 90 percent o f recurrent expenditures inprimary and secondary education. 59. Recurrent funding per student for federal universities (expressed as a share o f per capita GDP) i s 2.9 times higher than per capita GDP, which i s similar to the regional average for Africa, but much higher than the average for industrial countries (44 percent o fper capita GDP). Information on institutional productivity intertiary education i s limited, but budgetallocations per student show substantial variations across institutions. This wide range suggests that some budgetsare inadequate, while others are muchhigher thanthe national average. Assuming that the budgetreflects cost structures, there is substantial opportunity to improvethe cost- effectiveness of service delivery. 20 60. By international standards, Nigeria has low repeat rates, which are relatedto the phenomenon o f automatic promotion and low dropout rates, and low transition rates from primarytojunior secondary education. The main indications o finternal inefficiency haveto do with the structure o f expenditures, the deployment o fteachers, the salaries o fteachers, the salaries o f support staff, and the inability to track fundingfor infrastructure investment. Thesector has an inefficient expenditure structure, specifically, inadequate spending on maintenance and operations and inflated spending on payroll at the expense o f recurrent inputs.The current method o fdetermining teacher recruitment encourages oversupply, especially o f unqualified teachers, and inequitable teacher deployment within the state. Schools and local governments have little incentive to use teachers efficiently and strong incentive to claim undersupply and request additional resources. Teacher deployment is a source of ineflciency. All state ministries o f educationhave norms that stipulate staffing requirements relative to the number o f students enrolled. Thus there should be a reasonably close linear relationship betweenthe number of students enrolled ina school and the number o f teachers inthat school. However, as illustrated for Jigawa (see figures 3 and 4), student-teacher ratios vary widely among local government areas and individual schools. Some o f this variation can be explained with a large disparity betweenurbanand rural schools, resulting inshortages of qualified teachers inmany rural schools. Figure3: Enrollment/TeacherScatter Plots,a. for PublicPrimary SchoolsinJigawa,2006 and b. for PublicJunior Secondary SchoolsinJigawa, 2005 a. b. 0 200 1M m em Enrollment 1000 ,.- 200 aa m 5o Enrollment 1000 Wm Source: Bennell, Dandogo, and Mohamed. 2007. Managing expenditures on teacher salaries is essential to achieving sustainable levels of spending and improving the quality of education at all levels. Three main factors affect expenditures on teacher salaries: the number o f teachers hired, the number o f qualified teachers hired, and the level o f other benefits such as pensions and allowances. Teacher remunerationaffects teacher motivation and morale inpublic schools, but the relationship 21 between teacher salaries and student performance i s mixed. InNigeria, average salary levels are low and have been declining for primary school teachers inmost states, although this i s partly due to an increase inthe number o f unqualified primary schoolteachers insome states. As a ratio o f national GDP per capita, the average pay o f a qualified primary schoolteacher is lower than the suggested ratio (3.5 times GDP per capita) inmost states, except Lagos. The variation inthe average teacher salary i s considerable, rangingfrom 1.5 times inKaduna to 5.0 times inLagos. Although several states have introduced allowances and other measures to attract and retain qualified teachers, the scope for further increases inbasic pay i s limited. The benefits and allowances receivedby teachers are inconsistent across states. Some benefits are straightforward: either they are provided by the local government authorities or they are not; ifthey are provided, they are taken up by all or most teachers, such as transport allowance and housing benefits. However, other benefits vary within a given area and do not cover all or even most teachers. Insome cases, pensions cover only about 40-60 percent o f teachers ina given area, and for areas that offer medical benefits, the teacher take-up rate ranges from 5 to 100 percent. These inefficiencies must be remedied, as the costs o f education will continue to rise as the sector expands. Support staffare a source of ineflciency, but variations exist across and within states. If norms regarding the ratio o f teachers to support staff were enforced, there would be little variation among individual schools and thus among local government areas and states. However, these ratios vary considerably among the nine SEPER states. Although secondary schools generally needto employ relatively more support staff because they have more science laboratories, offer specialist technical subjects, and often have boarding facilities, in some states, teacher-support staff ratios are more than 10 times lower in secondary schools than inprimary schools, while inothers they are the same. Thefunding of infrastructure investments is inadequate and difficult to track. An excessive number o f investment projects are inadequately funded year to year, resulting inextended completion periods, cost overruns, and a large number o funfinishedprojects. The QSDS reviewed the construction o f schools, tracing the flow o f funds for school room construction from state government (Enugu) or from local government areas (Kaduna) to the schools. InEnugu, 70 percent o f the state-funded construction projects were completed. Ofthe 30 percent that were not completed, 10 percent were inuse and 20 percent were not usable. InKaduna, less thanhalfo f the room construction projects fundedby local government areas were completed. About a quarter were neither completed nor usable. 61. As table 5 shows, the state government of Enuguandthe local government areas o f Kaduna disbursed more thanN60 millionto construct schools that were not inthe local government registries and could not be found by the QSDS survey team. Of all schools that were supposed to receive funds for classroom construction in2004, 15 percent were not registered locally as an existing school inEnuguand 9.2 percent were not registered inKaduna. Either school registriesare poorly maintainedor funds are beingdisbursedto "ghost" schools. 22 Table 5: Dispositionof Fundsfor Room Construction in Enuguand Kaduna, 2004 I Indicator IEnugu IKadunaI Percent with no construction observed at the school 53.3 40.0 Funds disbursed to schools with no present construction (naira, millions) 29.0 39.0 Percent of schools that are not inthe local government list o f schools 15.2 9.2 Funds disbursedto schools not inthe local government list of schools (naira, millions) 8.8 53.6 Total funds that cannot be traced (naira, millions) 37.8 124.2 Fundsthat cannot be traced as a percent of state room construction budget 64.9 12.0 62. For schools that were supposedto receive construction funds and could be found, more than 50 percent inEnuguand 40 percent inKaduna had no construction. For Enugu, almost two- thirds ofthe state's budget for schoolroomconstruction couldnot betraced. For Kaduna, 12 percent could not be traced. 63. Post-basiceducationdoes not meet the demands of a competitivelabor marketor promoteeconomic growthinkey non-oilsectors. Although available labor marketinformation i s insufficient to explain the apparent mismatch between supply of and demandfor science and technology graduates, postsecondary institutions do not appear to be responding to labor market needs. Even students who complete their education are likely to remain unemployed. About 60 percent of the country's 13 million unemployed are post-basic education graduates. Improving the quality andrelevanceofpost-basicscience andtechnology education is therefore apriority. 6. ESTIMATING THE FUNDING GAP FORBASICAND SECONDARYEDUCATION 64. The gap infundingneededto achievethe EFA and education-relatedMDGsis substantialdespiterecentincreasesin publicspendingon education.World Bank (2005) estimatesthat the total cost to offer universal basic education covering six years of primary and three years ofjunior secondary education would be approximately US$29.4 billion between2005 and 2015, without taking into account the needto improve quality or efficiency. The annual funding gap for the sector would be aroundUS$2.3 billion (50 percent) in2007 and US$7.3 billion (74 percent) in2015, also without taking into account the needto improve efficiency (FME2007d). 65. While data are preliminary, evenwith improved efficiency, the funding gap would be significant and would have to be filled at the state level. Expanding basic education and meeting minimumstandards of quality would require massive increases inbothrecurrent and capital expenditures. Inaddition to annual recurrent expenditures, based on preliminary estimates, about US$178 million (or 6.7 times larger thantotal state and local government spending in2005) will be neededannually inthe nine SEPER states to construct newandmaintain existing infrastructure during the next 10years. Inaddition, basic education facilities inmost states are in needof substantial reinvestment, as the quality of classrooms is unsatisfactory for the large majority of students. 66. The FMEis committed to increasing government spending on education from less than 5 percent of GDP to at least 10 percent by 2020 (FME 2007b: 31). The growth ineducation 23 expenditures may be unsustainable, especially since these estimates are conservative and do not assume that significant additional resources will be devoted to improving the quality and the relevance o f education. Several options would helpto contain education expenditures: (a) improvingefficiency by reducing unit costs inrealterms (without lowering the quality of education), (b) increasinghousehold expenditures, especially at the secondary and tertiary levels, and(c) expandingprivate education. 7. DATA LIMITATIONS 67. Data limitations are the main constraints to analyzing the costs and financing of education inNigeria. As documented ina number o f studies (see, for example, Hinchliffe 2002;World Bank 1996,2004,2006b) and the recent education sector analysis by the FME, "There are serious concerns about the reliability o f the underlying data on which the system i s built andthere is a longway to go before numbers from the system can be fully trusted" (FME 2007a: 12). As a result, critical data on expenditures, inputs, and outcomes are missing or inconsistent. No one really knows how well the system i s performing. Expenditure data 68. The PEMFAR-sponsored budgetmonitoring survey o fprojects financed with the Abacha funds (repatriated from Switzerland) confirmed weaknesses inbudget accounting, reporting, and monitoring at the sectoral level. The ministries, departments, and agencies that participated inthe program have not reportedthe full list o f projects financed with the repatriated funds allocated to them. The share o f non-reported projects was the highest inprograms managed by the federal ministrieso f education, health, and power and steel (rural electrification program). 69. The federal government reports expenditures on education inthe annual federal budget appropriations. However, state and local governments do not routinely gather, analyze, or report information on education expenditures since there is no constitutional requirement for them to report budgetary information to a central unit at the federal level. Comprehensive information could not be obtained inany o f the nine SEPER states on total overhead expenditures inpublic primary schools because centralized financial records are not kept by either the state ministries o f education or local government authorities. Because data on state and local expenditures are missing or cannot be trusted, there are no accurate estimates o f total public spending on education. 70. The QSDS survey found that local government authorities inEnuguand Kaduna had documented records for, on average, 60 percent o f their revenuecategories, but incomplete records on expenditures for schools intheir jurisdiction. For schools that charge fees, only 18 percent o f head teachers inEnuguand 13 percent o f head teachers inKaduna reportedly keep any accounts o f how these revenues are spent. Almost no head teachers ineither Kaduna or Enugucould estimate the costs o finputs consumed by the school, such as salaries for teachers and support staff and for repairs. Two-thirds o f head teachers inEnuguand 90 percent o f head teachers inKaduna do not estimate expenditures for the coming year. 24 Enrollment data 71. Enrollment data vary significantly by source or in inexplicableways. Estimates o f enrollment rates from the NLSS are different from estimates o f the annual school census and the EdData Survey 2004. Inthe case o f primary education, the N L S S reports lower enrollment rates (both gross and net) thanthose reported on the EdData survey. Although the pattern across regions i s similar, with much better indicators for the southern states, the actual estimates differ substantially by source. At the federal level, reportedenrollment intertiary education institutions rises and falls from year to year by more than the typical entering cohort. For example, the reported enrollment for one federal university-the University o f Abuja-increased by 24,000 studentsina single year. On closer examination, the increase resulted from the inclusion o f part-timeand distance learning students inthe second, but not inthe first, reporting period. At the tertiary level, distinctions among full-time,part-time, undergraduate, graduate, and non-degree students have not been preserved. The definition o f "enrollment" also varies across institutions, making simple comparisons impossible. The National Universities Commission i s aware of the problem and has launched an initiative to standardize reporting. This initiative should be accorded highpriority. 72. Data at the state level are available, but largely indicative. Duringthe past eight years, the federal government sought to improvethe collection and analysis o f education-related data, with a focus on primary and secondary education. Most state ministries o f education havemade considerable progress ininstitutionalizing the Annual School Census. However, coverage varies, and reliability i s questionable. As a result, full data are not gathered or reported against key indicators such as enrollment, staff, and textbook ratios. 73. Reliable data about the provision of private education are not available. Registred andunregistered private primary and secondary schools account for a significant share o f enrollment but data are not routinely collected on private education. Available estimates o f enrollment at private (both registered and non-registered) are unreliable and vary by source. For example, the CWIQ survey indicates that only 10percent o f primary and secondary school studentsattend nongovernment schools. Onthe other hand, the Annual School Census showed that private education accounted for as highas 16 percent o f primary and secondary enrollment based on the private schools coveredby the census (FME2006). . 74. The absence of adequate and verifiable data poses a major constraint on policy analysis, evidence-based decision making, and accountability. The main difficulties lie in gathering, standardizing, and aggregating expenditures and education outputs, especially at the state and local levels. Record keeping i s poor at the school or institution, local government, and state government levels, with data often unavailable, incomplete, or not standardized. At the same time, there i s no clearly assigned responsibility for monitoring and evaluating policy at the federal and state levels. These shortcomings impede a thorough assessment o f performance and limit evidence-based decisionmaking. 8, STRATEGICOPTIONSFORREFORM 75. Poor performance and deficient indicators call for major reforms o fNigeria's education sector. To address the key challenges inthis regard,Nigeria's government developed and 25 launched a major education reform in2006. According to the government's Seven-Point Action Plan, education reform "is intendedto overhaul the entire education sector and to promote quality education for life-skill acquisition, job creation, and poverty alleviation. It will put in place a sound framework that will enable implementingauthorities to widen access, increase equity, and enhance the quality o f education" (NPC 2007: 62). The reform program stresses the importance o f institutional reforms to improve the efficiency and effectiveness o f service delivery at all levels o f education (see box 2). It also underlines the need for additional resources to close the funding gap. Box 2: Government Education ReformProgram Reforms within Nigeria's education system have beenunder way since 1999. In2006 the government launched a major education reform program, with a revised vision statement, a comprehensive education sector analysis, a national framework, and the 10-Year Education Sector Plan. The revisedvision for the FME as set out inthe Vision 2020 document is "to become an emerging economy model, delivering sound education policy and management for public good." It highlights the role o f the education sector andthe need for radical policy reforms to make Nigeria one o fthe top 20 economies inthe world and considers education as the foundation for this transformation.. On the basis o f this reform program, a draft Education Reform Bill was prepared. Building on the analysis, the Federal 10-Year Education Plan identified the following priority areas inthe education sector: (a) access to quality basic education by all, with particular focus on the attainment o f the Education For All goals; (b) the use o f information and communications technology in every sphere o f education, including achievement o f universal computer literacy for students in secondary and tertiary school; (c) the introduction o f innovative and entrepreneurial courses relevant to the needs o f the economy; (d) achievement o f an agreed set o f minimumquality standards, which includes restructuring the inspectorate system; (e) finance reformto underpina larger, more equitable education system; and (0development o f human capital throughout the education sector by focusing on both capacity development and effective management o f human resources. The strategic planning process yielded a comprehensive reform agendathat seeks to "leverage a combination of strategic human and infrastructure capital reforms, aided by specific tactical interventions.".. These interventions include (a) a conditions survey for secondary and basic schools, (b) a public-private partnership o f the 102 federal secondary (unity)schools, (c) improved transparency and accountability through more effective community involvement in education, (d) improved vocational and technological training to enhance growth, (e) coordination and delivery o f information and communications technology support services and resources to ensure development o f uniform standards and skills acquisition, and (0consolidation o f tertiary institutions by converting all federal polytechnics and colleges o f education into campuses o f neighbouring federal universities and consolidation or grouping o f education-related parastatals. Some acts will needto be repealed and some legislation amended to align parastatals undertheir respective regulatory authorities underthe proposed reforms. Source: FME2007a;2007b and 2007c 76. Reform i s difficult to implement insuch a fragmented and complex education system as Nigeria's. Giventhe importance o f education for both poverty alleviation and economic development, and the public resource constraints, the sector reforms should focus on (a) improving accountability, (b) increasingefficiency inthe use of resources, (c) reducing the inequities inaccess, especially inbasic education, and (d) improving the quality o f education services. 26 Accountabilityreforms 77. The overarchingrecommendationis to move towardan accountablesystem. Nigeria's education system i s neither transparent nor accountable and i s fragmented across agencies and tiers o f government. Improving the transparency and accountability o fthe education sector entails addressing the following problems: (a) responsibilities and powers for key decisions across all tiers o f government are ambiguous and overlapping, makingit difficult to know who i s responsible for financing what; (b) the financing ffamework is unpredictable, resulting invertically and horizontally imbalanced financing o f key inputs; (c) input standards do not exist, are not enforced, are divided across many parastatals, or are not adequately resourced; (d) data onthe system's performance-expenditures, inputs,andoutcomes-are missing, inconsistent, or not verifiable; and (e) incentives inthe systemdo not encourage government stakeholders to improve the delivery o f responsibilities under their jurisdiction. 78. Only by improving accountability for performance can outcomes and efficiency be improved. Inthe short to medium term, there is scope to improve the use o f existing allocations. Stronger transparency and accountability are critical. The experience o fNigeria and other countries shows that, unless further progress is made inthese areas, no amount o f additional resources will solve the country's social problems. At the same time, even if efficiency were improved, there would be a shortfall to achieve the government's UBE and other education targets. However, there remainsnojustificationfor morespendinguntiluse of the current expenditureenvelope is improved,meaningthat money reachesthe intendedbeneficiaries and is spent efficiently. 79. Accountability reforms are perhaps the most difficult to implement and cannot be done in isolation from improvements inaccess, quality, and efficiency. The main issues include (a) removing ambiguity, overlap, and unassigned roles and responsibilities; (b) setting, resourcing, andenforcing minimumstandards; (c) improving expenditure information; (d) improving instrumentsto assess education quality and learning achievement; and (e) improving data flows and routine monitoring and evaluation. 80. The roles andresponsibilitiesof each level of governmentneedto be clearly defined, especially inthe provision o f basic education and the arrangements for monitoring and evaluation. The key policy options are the following: Clarih the roles and responsibilities of the various tiers of government, especiallyfor the provision of basic education. Policies should be developed and implemented that consolidate the fragmented roles across all tiers o f government to help to rationalize the supply chain. Ensure that the right kinds of incentives are inplacefor allforms of serviceprovision. The absence o f anappropriate accountability framework, based on a system o fboth rewards and sanctions, needs to be addressed through a review and subsequent revision o f enforceable incentives and disincentives for deliveringsustainable, effective services across all tiers o f government. 27 Improve the overall planning and management of public education, by implementing strong incentive structures and an accountability framework. This requires strong leadership and capacity across the system, especially at the school, local, and state levels, where management capacity i s weak in(a) data collection, analysis, and use; (b) school leadership; (c) strategic planning, budgeting, and financial management; (d) staff performance targets and outputs; (e) evaluation; and (f) teacher recruitment and deployment. There is also a need for more key policy studies, management training, and a system or mechanism for publicizing education budgets as well as the results o f inspections and annual educational reviews. Improve the understanding of the institutional relationships between all of the key stakeholders in the education sector. Both the federal and state governments need to articulate a clear division o f labor, and responsibility needs to be assigned and implemented across the subsector. Review the requirements of capitationfinancing to meet those standards. Maintenance requires recurrent outlays, and an effective communications strategy is needed to explain the per capita formula to all stakeholders inthe community. At the state level, review and confirm responsibility for school autonomy inpersonnel managementandprocess decisions. This includes the hiringo f teachers, choice o f textbooks, and allocation o f the budgetwithin the school. 81. Minimumstandardsneedto be set, resourced,and enforced.At the federal level, minimumstandards for key educational inputs are neededto provide guidance to states and inform the planning and budgeting cycle. This i s an important part of the development o f longer- term state-level strategic planning. Inthe short term, states could develop less-demanding quality assurance and accountability mechanisms by providing additional training to head teachers and local officials and giving them effective monitoring tools, such as school report cards and a simple set o f indicators (enrollment rates, student-teacher ratios, unit costs). 82. Improvedinformationon the government's spendingand educationoutcomes should be madeavailableto all key stakeholders.The ultimate goal o f collecting information on actual spending is to ensure that (a) every recipient o f public funds produces a regular report on how much money is received and how these funds are spent and that (b) these reports are publicly available. There also is a need to include key stakeholders at the school, community, and local level inthe receipt, understanding, and acceptance o f education expenditure and outcomes information. This can be achieved through (a) publishingtest results inthe form o f a school report card that informs parents and parent-teacher associations o f the status o f learning within schools, local government areas, and states and (b) disseminating detailed report cards on the execution o fthe previous year's investmentplanandthe current year's planto newspapers, parent-teacher associations, capital funding source(s), local government authorities, and school heads. 83. M o r e effective instrumentsare neededto assess progresstoward improvedquality. Nigeria has made significant progress indeveloping an accreditationmechanism for the tertiary education system. Inprimary and secondary schooling, quality assurance needs to focus on 28 assessing schools' intended outputs and outcomes, not inputs. The key policy options for assessing learning achievement and thereby improving educational quality are the following: Develop capacity to assess student learning at the national level Participate in international sample-based learning assessments (such as the Trends in International Mathematics and Science Study) and state-level learning assessments Create mechanismsfor state governments to use the test results andprovidepublic feedback on the design ofpolicies, strategies, and specific interventions to improve learning outcomes. 84. Federal and state governments need to improve their capacity to collect and interpret data. As with other line ministries, data limitations are systemic and not uniqueto the education sector. Improvements indata should be a long-term priority, linked with institutional reform and an accountability framework. Along with the reforms proposed by the previous government, there i s a needfor stronger capacity o f federal, state, and local staff to collect, manage, analyze, report against, and disseminate data more effectively. Inpracticalterms, this means makinginformation accessible that i s o f sufficient quality to use as a planningtool and makinggovernment more accountable over the mediumto longterm. Improvedbudget reports from government as a whole, and withinthe education sector specifically, would allow the sector to provide better detail o f costs, with greater accuracy, and ina consistent manner across all three tiers o f government. Ifthese documents are available from the federal government, the education sector would be well placed to provide detailed costing o fthe amount o f money allocated, received, and spent withinthe period. This can be accomplished through the following: Improve monitoring and regulation at the federal and state levels inthe short to medium term and at local government areas inthe longer term. Inthe short term, the federal government needs to strengthen its capacity to collect, analyze and report data on key education indicators (including private andnon-fonnal education) and education expenditures, and monitor across-state differences ineducationperformance. Link the allocation of funds to different tiers of government, contingent on each tier providing speci9ed andperiodically audited data that accurately report expenditures and measure the timely performance o f responsibilities under theirjurisdiction. Strengthen consultative arrangementsfor selectingprojects underfederal programs in basic education. Introduce a uniped reportingframework on educational spending. Finance and efficiency reforms 85. Managing expenditures at all levels i s crucial to achieving sustainable levels o f spending. This is difficult to implement, as the attainment o fUBEandother major educational goals continues to be constrainedby financial andpolitical economy realities. Education cannot be expanded or improved without efficiency reforms. However, efficiency reforms are politically 29 difficult because they may entail closing schools, freezing hiringuntil attrition shrinks the size o f the workforce, and reassigning teachers to rationalize the ratio o f studentsto teachers and of teachers to classrooms. Reinvestingthe savings would help, but those resources may not return to the sector. Therefore, although implementing efficiency reforms i s technically straightforward, the benefits must be combined with other reforms to maintain buy-infrom stakeholders and to have a sustainable impact. 86. Key recommendations for efficiency reforms can be broken into reforms that (a) improve budgetformulation and implementationand reforms that (b) improve the quality o fpublic spending. The key policy options to improve budget formulation and implementation are as follows: Develop medium-term education sector strategies (or sectorplans) at the state level, which are already beingprepared in a number of states such as Lagos, Kaduna, Kano, and Kwara,. Consider using these plans for the allocation o f federal fiscal transfers through the UBE Intervention Fund, the ETF, and the Virtual Poverty Fund Improve budgeting and transparency by implementing a medium-term expenditure framework for setting multiyear priorities (based on sector plans), capitation-based financingfor basic and secondary education, and performance-basedfunding for tertiary education (that is, link budgetingto performance or enrollment goals); Refine and clariJLfunding sources by moving to activities-based budgeting, to remove the incentive structures that make it difficult to ensure efficiency and equity inpublic spending for education. The current fragmentation withinpublic education spending increases the administrative costs, particularly as different funding sources operate as vertical funding sources. Improve thepredictability offinancing by streamlining and bringingfunds on budget. This is especially important for supporting medium-term strategies andplans at the state level. Clarifying and streamlining the roles and responsibilities o f each level o f governmentand consolidating the funding mechanisms, use, and management o f funds will improve both accountability and effectiveness o fpublic education spending. 87. Key policy options for increasing efficiency and ensuring financing sustainability inthe education sector are the following: Review the deployment of teachers and support staff and salary policies, which are a major source o f inefficiency and inequity inmost states Review and revise the mechanismfor financing the salaries of primary schoolteachers by local government to create incentives for efficient and effective use o f teachers Consider consolidating and rationalizing underusedprograms and institutions (especially inpolytechnics, monotechnics, and colleges of education) DiversiJL theprovision of tertiary education institutions through the establishment o f Centers o f Excellence with two or three universities designated as the leader inmajor areas o f teaching and research inselect fields (agriculture, petrochemicals) 30 Create an enabling environmentfor theprivateprovision of education by (a) clarifying and simplifying the processand standards for private sector entry; (b) reviewingand refiningthe regulatory framework, including introducing adequate quality assurance mechanisms, especially at the state level; and (c) formalizing the current cost-recovery mechanismwith transparent collection of user fees Increase resourcesfor the expansion of basic education through (a) the more efficient use of resources and (b) the mobilizationof additional resources. Access reforms 88. Efforts to expand access to basic and secondary schooling are perhaps the easiest to implement.Achieving universal basic educationremains amajor challenge inNigeria, especially for girls inthe northand for the poor, andrequires significant additional resources. Key recommendationsfor access reforms are the following: Review and monitor fee policies and user charges at all levels, including (a) reviewing fee and user charges (both formal and informal); (b) introducing policies to ensure that the poor are not affected adverselyby the private costs of education; and (c) reviewing existing regulations with a view to addressing governance and equity issues Strengthenfederal equalizationprograms (for example, UBEC matching grants) to provide additional opportunities for citizens inless developed states. A review should examine implementation ofthe UBEIntervention Fundandthe ETF and refine the allocation formula and guidelines to improve equity among states, by taking into account (a) state performance and (b) local needs (for example, poverty) Introduce measures to address barriers to girls' education, especially in the northern regions, through a mix of supply and demand factors, includingrecruitment and deployment of female teachers; provisionof schools and water and sanitation facilities in schools; provisionof education subsidiesthrough conditional cashtransfers, and advocacy with communities and community leaders. Quality reforms 89. The quality of education inNigeriais weak andvaries across the country. Specific policy actions to improve the quality of education provisioninclude the following: Improve the quality of teachers by providing additional trainingfor unqualified teachers and enforcing standardsfor the recruitment of new teachers. Centralized recruitment at the state level and the enforcement of standards for teacher qualifications and workload would helpto improve teacher quality. Increase theprovision and improvement of physical infrastructure in aplanned manner. This includesclassrooms, libraries, fixnishings, and school environment and could include double shifts and multigrade teaching, where local conditions are conducive. Provide adequate recurrent resourcesfor essential inputs affecting student learning such as textbooks, instructional materials, and teacher training. This requires setting, 31 resourcing, and enforcing minimumstandards for the delivery o f basic and secondary education. 90. Reforms that aim to improve quality are difficult to implement and take much longer than reforms to improve either access or efficiency. First, improving educational quality i s an elusive goal. It requires understanding how inputs are transformed into learning achievements and what skills and incentives predict a more effective transformation. Quality reforms are politically more difficult to institute than access reforms. Quality reforms may not spark the immediate resistance generated by efficiency reforms, which usually impose losses on some stakeholders within government, but neither do they create the broad political support associated with access reforms. Second, this type o f reform places great demands on the institutional and managerial capacities o f the sector for a long period o f time. Third, benefits are less visible, more limited, and distant. Ifthe economy i s stagnating, beneficiaries o f improved educational quality- graduating students-may realize only limited returns. Society benefits ifthe economy i s able to make effective use o f better humancapital, but this payoffoften i s delayed. Higherquality can improve efficiency because systems with weak quality spend significant resources on what is really remedial education, and doing it right inthe first place lets the system obtain the same results with fewer years o f school. Finally, quality reforms are difficult to implement because they must be kept on track over a long period o f time. But the most difficult challenge i s to change behaviors-how teachers teach and how learning i s organized-which means creating new capacities among thousands and, inNigeria, thousands o f teachers and headteacheddirectors scattered across the country. Policy sequencing 91, The top priority for education reform is to improve the accountability of the sector's performance in the context of access and quality reforms. GivenNigeria's federalist system and emerging democracy, getting the institutional fundamentals right has to be based on a carefully built consensus among voluntary participants and protected over a 15-20-year period. The strategies for improving accountability, efficiency, access, and quality are complementary and, ideally, should be implementedsimultaneously as a package. However, giventhe extent and complexity o f the reform agenda and the limited implementation capacity, especially at the state and local levels, the government may have to select some o fthe most pressing policy options in each o f these reforms. Policy sequencing thus requires careful examination. The attached matrix o f main issues and options or recommendations summarizes the strategic direction o f policies and suggests the timing o f specific reforms. 32 REFERENCES Adekola, Olatunde A. 2007. "Language, Literacy, and Learning inPrimary Schools: Implications for Teacher Development ProgramsinNigeria." Working Paper 96. World Bank, Africa Region, HumanDevelopment Department,Washington, D.C. Bennell, P. et al. 2007. A Synthesis Report of the Main Findings and Recommendationsfrom NineState Reports. A background report preparedfor theNigeriaEducation Public Expenditure Review. Bennell, P., Dandogo, K.and Mohamad, M.(2007). "Jigawa State Education Public Expenditure Review". Background Report preparedfor the Education Expenditure Review. FME(Federal Ministryof Education). Basic andSenior SecondaryEducationStatistics in Nigeria 2004 and 2005. Education Data Bank. Federal Ministry of Education. Abuja. ,2007a. "Education Sector Situation Analysis". Draft 5. January, 24, 2007. Federal MinistryofEducation. Abuja. .2007b. "10 Year Strategic Plan". Draft 7. FederalMinistry of Education. Abuja. .2007c. "Vision 2020: The Role oftheNigerianEducation Sector". FederalMinistryof Education. Abuja. .2007d. "The 10 Year FederalEducation Plan(2006-201 5)." National Stakeholders Workshop, Abuja, March 29-30. Hinchliffe, Keith.2002. "Public Expenditures on Education inNigeria: Issues, Estimates, and Some Implications." Working Paper. World Bank, Africa Region, HumanDevelopment Department.Washington, DC. Khemani, S, 2005, "Local Government Accountability for Health Service Delivery inNigeria". Journal of African Economies, pp 1-28. National Bureauof Statistics (NBS). 2006. Core Welware Indicator Questionnaire Survey 2006: Key Findings. National Bureau of Statistics. Abuja National Population Commission (NPC) and ORC Macro. 2004. Nigeria DHS EdData Survey 2004. Calverton, Maryland, USA: National Population Commission and ORC Macro. National Planning Commission (NPC). 2007. "The Nigeria Project Agenda." National Planning Commission. Abuja. Nigerian Educational Researchand Development Council (NERDC). 2004. National Policy on Education.4th ed. Lagos: NERDC Press. UNDP. 2006. Niger Delta HumanDevelopment Report. Abuja. World Bank. 1996.Federal GovernmentExpenditure Review.Report 14447-UNI. Washington. .2004a. "School EducationinNigeria: Preparing for Universal Basic Education. Country Status Report." Working Paper 53. World Bank, Africa Region Human Development Department,Washington, D.C. 33 .2005. "Nigeria's Opportunityof a Generation:Meetingthe MDGs, Reducing Indebtedness," PRMED ReportPreparedfor the Africa Region. 2006a. "Nigeria: A FiscalAgenda for Change; PublicExpenditureManagementand FinancialAccountability Review." 2 vols. Report 36496-NG. World Bank, Washington, D.C. .2006b."Nigeria Science andTechnologyEducationat Post-BasicLevel (STEPB): A Reviewof S&T EducationinFederallyFundedInstitutions." Report 37973-NG.World Bank, Washington, D.C. . (2007). Nigeria: PovertyAssessment. Report40903-NG. World Bank, WashingtonDC. 34 ANNEX A. MATRIX OF MAINISSUES AND OPTIONS OR RECOMMENDATIONS Issue and strategic direction of reform I Policy action Timing A. Governance, management, and accountability Improvemanagement, governance, and Review, clarify, and streamline the assignment of roles of Short term accountabilityarrangements and responsibilities for the federal, state and local Short term Short term Strengthen the capacityfor strategicplanning, policy Medium term implementation, and monitoring and evaluation, especially at the state and local levels Strengthen school management and accountability in Mediumterm primary and secondary education by creating school-based management committees Develop more effective instruments of quality assurance Longterm to improveaccountability,including developingschool report cards, monitoringschool performancebasedon a set of simple indicators, and developingasystem of student assessment and monitoring Mediumterm well as at the schoolor institution level, with a uniform core base and add-ons for each specific type of institution Strengthen the capacity of afederal unit to collect, Medium term analyze, and reportkey nationalstatistics and aunit in each state to do the same for key state statistics Publish information on the status of education at regular Medium term intervals to improveaccountability, inform the formulationofbudgets, and strengthen governance B. Use of public resourcesfor education Improveformulation and Develop medium-term sector strategies (or plans) at the Short term implementationofthe budgetto achieve state level, based on the modeldevelopedby the FME; sector priorities consider usingthese plans to allocate federal transfers through the UBEInterventionFund, the ETF, and the Virtual Poverty Fund II Develop a clear vision and a soundpolicy for tertiary Short term education to guide decision makingand action at the federal and state levels Improve budgeting and resource allocation through Medium term implementationof amedium-termexpenditure framework Introduce capitation-basedflnancing for primary and Medium term secondary education and performance-basedfunding for 35 Issue and strategic direction of reform Policy action Timing tertiary education to achievenationalgoals and policies Improvethe efficiency o f education Revise the current mechanismfor funding primary teacher Short term expenditures salaries to create incentives for the efficient and effective use ofteachers Commission a comprehensive review of institutional Short term productivity to examine the efficiency and effectivenessof teaching and researchproducts o ftertiary education institutions, includingthe wide variations in unit costs Review teacher and support staff deployment and salary Medium term policies Increase non-wage allocations, by providingfiee basic Medium term education materials (including a set of core textbooks) Useprimary andsecondary schools more cost effectively, Medium term by consideringdouble shiftsor multigradeclasses, dependingon the local situation Consider consolidating and eliminating underused Medium term programs and institutions Diversijj the provision of tertiary education institutions Longterm by establishingcenters of excellencein major areas of teaching and researchin select fields C.Inequities in access Improveaccess to basic and secondary Ensure allocation of a minimum of 20 percent of total Short term education, especially for girls and the state and local government expenditures (benchmark) for poor, and address regionaldisparities basic education Review and monitorfee policies and user charges at all Medium term levels to ensure that the poor are not affected adversely by the higher private costs of educationand introduce policiesto address governanceand equity issues Review implementation of the UBEIntervention Fund and Medium term the ETF and refinethe allocation formulaand guidelines to improveequity among states, taking into account state performanceand localneeds Introduce measures to address barriers to girls' Medium term education, especially in the northern regions, through a mix of supply and demand factors D.Resource mobilization Mobilize additionalresources for Review the role of the private sector in education, by Medium term education clarifying and simplifying the process and standards for private sector entry and introducingquality assurance mechanisms, especially at the state level Devote more resources to the expansion of basic Medium term education, by usingresources more efficiently and mobilizing additionalresources 36