Report No. 59755-GH REPUBLIC OF GHANA Education in Ghana Improving Equity, Efficiency and Accountability of Education Service Delivery February 23, 2010 AFTED Africa Region Document of the World Bank This document has a restricted distribution and may be used by recipients only in the performance of their official duties. Its contents may not otherwise be disclosed without World Bank authorization. GHANA - GOVERNMENT FISCAL YEAR January, 1 – December 31 CURRENCY EQUIVALENTS (Exchange Rate Effective as of March 30, 2010) Currency Unit = Ghana Cedi GH¢1 = US$0.70 US$1 = GH¢1.42 Weights and Measures Metric System ABBREVIATION AND ACRONYMS AFUF Academic Facility User Fees BECE Basic Education Certificate Examination CAGD Controller and Accountant General Department CBO Community Based Organization CCT Conditional Cash Transfer CDD Center for Democratic Development CIDA Canadian International Development Agency CL Cluster Leader COTVET Council for TVET CREATE Consortium for Research on Educational Access, Transitions & Equity CSA Civil Service Agency CWIQ Core Welfare Indicators Questionnaire DACF District Assembly Common Fund DEO District Education Office DEOC District Education Oversight Committees DFID Department for International Development DP Development Partner DWAP District-Wide Assistance Project EDI EFA Development Index EFA Education for All EMIS Education Management Information System ERP Economic Reform Program ERRC Education Reform Review Committee ESP Education Strategic Plan ESPR Education Sector Performance Report FCUBE Free Compulsory Universal Basic Education FTI Fast Track Initiative GAR Gross Admission Rate GDHS Ghana Demographic and Health Survey GDP Gross Domestic Product GER Gross Enrolment Ratio GES Ghana Education Service GET-Fund Ghana Education Trust Fund GLSS Ghana Living Standards Survey GNAT Ghana National Association of Teachers GNI Gross National Income GoG Government of Ghana GPI Gender Parity Index GPRS Growth and Poverty Reduction Strategy GRATIS Ghana Regional Appropriate Technology Industrial Service GSFP Ghana School Feeding Program GSS Ghana Statistical Service HDI Human Development Index HE Higher Education HEI Higher Education Institute HIPC Heavily Indebted Poor Country HIV&AIDS Human Immunodeficiency Virus and Acquired Immune Deficiency Syndrome HND Higher National Diploma ICCES Integrated Community Centers for Employable Skills ICT Information and Communication Technology IMF International Monetary Fund IQR Inter-Quartile Range JHS Junior High School JICA Japan International Cooperation Agency KG Kindergarten KNUST Kwame Nkrumah University of Science and Technology LEAP Livelihood Empowerment Against Poverty LESs Less Endowed Schools MDBS Multi-Donor Budget Support MDG Millennium Development Goals MDRI Multilateral Debt Relief Initiative MLGRD Ministry of Local Government and Rural Development MMDAs Metropolitan, Municipal and District Assemblies MoE Ministry of Education MOESS Ministry of Education Science and Sport MoFEP Ministry of Finance and Economic Planning MoLG Local Government MPs fund Member of Parliament fund MTEF Medium Term Expenditure Framework NAB National Accreditation Board NAR Net Admission Rate NCTE National Council for Tertiary Education NDC National Democratic Congress NEA National Education Assessment NER Net Enrolment Ratio NERIC National Education Reform Implementation Committee NERP National Education Reform Program NESAR National Education Sector Annual Report NGO Non-Governmental Organization NPP New Patriotic Party NVTIs National Vocation Training Institutes OECD Organization of Economic Cooperation and Development PBME Planning, Budgeting, Monitoring and Evaluation PCE Per child recurrent expenditure PE Personnel Emoluments PER Public Expenditure Review PPP Purchasing Power Parity PRSCs Poverty Reduction Strategy Credits PRSP Poverty Reduction Strategy Paper PTA Parent Teacher Association PTE Per teacher recurrent expenditure PTR Pupil Teacher Ratio PTTR Pupil Trained Teacher Ratio RFUF Residential Facility User Fees SHS Senior High School SIML Social Impact Mitigation Levy SMC School Management Committee SPAM School Performance Assessment Meeting SSA Sub-Saharan Africa SSSCE Senior Secondary School Certificate Examination SSSS Single Spine Salary Structure TED Teacher Education Department TIMSS Trends in International Mathematics and Science Study TTIs Technical Training Institutes TTL Task Team Leader TVET Technical and Vocational Education and Training UBC Universal Basic Completion UBE Universal Basic Education UCC University of Cape Coast UNDP United Nations Development Programme UNESCO United Nations Education, Scientific and Cultural Organization UNICEF United Nations Children’s Fund UPC Universal Primary Completion USAID United States Agency for International Development VAT Value Added Tax VTIs Vocational Technical Institutes WAEC West African Examination Council WASSCE West African Senior Secondary Certificate Examination WB World Bank WDI World Development Indicators Acknowledgments This report was prepared by Peter Darvas (TTL) and Alexander Krauss based on statistical analysis, as well as information collected from a number of background analyses that have been provided by the Ministry of Education, the Ghana Education Service, Ghanaian and international experts. In particular, Chapter 2 on student enrollment, transition and completion is based on National Education Sector Annual Reports and data from the Education Management Information System as well as the Ghana Living Standard Survey. Prof Albert Akyeampong provided contribution to the chapter on access, Quentin Wodon and George Joseph provided valuable contributions for the assessment of learning outcomes. The sections on teacher policy are based on a report by Paud Murphy, maps were created with the help of Peter DeVries, the section on expenditures and efficiency is based largely on analyses by Michele Savini, the section management and administration is based on a report by Jesper Steffensen, the section on higher education is based on a report by James Anyan. Several sections, especially higher education, benefitted from a rapid sector review by Maurice Boissiere. Discussion on TVET benefitted from a review by co-authored Robert Palmer and the TTL. Burton Bollag edited the text. Valuable suggestions, feedback and reviewer comments were provided by Ishac Diwan, Peter Moock, Katherine Bain, Richard Hopper, Ash Heartwell, Eunice Dapaah, Qaiser Khan, Maureen Lewis, Venkatesh Sundararaman, Aidan Mulkeen, Sukhdeep Brar, Atou Seck, Peter Materu, Susan Hirshberg, Anke Weber, Hiryouki Hattori and Burton Bollag. Ghana Ministry of Education officials also provided helpful comments to this report including Charles Aheto Tsegah, Kwame Agyapong Agyen and Joseph Annan. Table of Content EXECUTIVE SUMMARY ................................................................................................. i 1. INTRODUCTION, CONTEXT AND ORGANIZATION......................................... 1 A. Demographic and Social Context ..................................................................... 3 B. Historical Evolution of Education .................................................................... 7 C. Organization, Structure and Normative Framework ................................... 10 2. EDUCATIONAL OUTCOMES: ACCESS, LEARNING AND RELEVANCE .... 12 A. Coverage and Access at Various Levels ......................................................... 12 B. Learning Achievement ..................................................................................... 36 C. Relevance........................................................................................................... 49 3. INPUTS AND RESOURCES ................................................................................... 58 A. Physical Resources: Schools, Infrastructure and Textbooks ....................... 59 B. Human Resources: Teacher Supply, Allocation, Policy ............................... 64 C. Public Expenditures on Education ................................................................. 75 4. SECTOR PERFORMANCE..................................................................................... 91 A. Equity and Disparity in Education Services .................................................. 91 B. Efficiency ......................................................................................................... 121 C. Management Accountability and Social Accountability ............................ 138 5. HIGHER EDUCATION ......................................................................................... 150 6. EDUCATION STRATEGIC PLANNING, POLICY OPTIONS .......................... 164 List of Tables Table 1-1: Ghana’s Progress towards Selected Millennium Development Goals ....... 4 Table 1-2: Structure of Ghana’s Education System .................................................... 11 Table 2-1: Reasons for children not attending and stopping to attend school, percent ........................................................................................................................................... 14 Table 2-2: Pre-tertiary GER, NER and Completion, by Gender, Region and Quintile ........................................................................................................................................... 14 Table 2-3: Trends in Kindergarten Enrollment ........................................................... 16 Table 2-4: Trends in Primary Education Enrollment, public and private, boys and girls ................................................................................................................................... 17 Table 2-5: Trends in Primary Enrollment, Private/Public ......................................... 17 Table 2-6: Trends in Primary Enrollment, Admission and Gender Parity Index .... 18 Table 2-7: Primary Education, Completion, Repetition and Dropout ...................... 21 Table 2-8: Progress towards education related MDGs ............................................... 22 Table 2-9: Trends in JHS Enrollment, Private/Public and Boys/Girls ...................... 25 Table 2-10: Trends in enrollment to junior high school education ............................ 25 Table 2-11: JHS Completion, Repetition and Dropout ............................................... 26 Table 2-12: SHS enrollments ......................................................................................... 31 Table 2-13: Enrollment in Technical Training Institutes ........................................... 33 Table 2-14: National Functional Literacy Program .................................................... 36 Table 2-15: Overall NEA Mean Scores, Minimum-competency and Proficiency .... 39 Table 2-16: Overall NEA Gender Comparisons .......................................................... 39 Table 2-17: Mean Resource in 2005 and 2007, by Class 6 English Proficiency ........ 42 Table 2-18: Mean Resource, by class 9 BECE English Pass Rate in 2008/09 ........... 45 Table 2-19: Mean Resource, by class 9 BECE Math Pass Rate in 2008/09 ............... 46 Table 2-20: Employment Status and Level of Education in 2008, Female/Male ...... 50 Table 2-21: Level of Education and Occupation in 2008, Male/Female .................... 52 Table 2-22: Level of Education and Teenage Pregnancy/Motherhood, 2008............ 54 Table 2-23: Level of Education and Desire for no more Children, 2008 ................... 55 Table 2-24: Mother’s Education and Early Childhood Mortality Rates, 2008 ......... 55 Table 2-25: Women’s Participation in Decision-Making and Level of Education, 2008................................................................................................................................... 56 Table 3-1: Examples for inputs influencing quality outcomes.................................... 58 Table 3-2: Preschool Facilities ....................................................................................... 60 Table 3-3: Primary School Facilities ............................................................................. 60 Table 3-4: JHS School Facilities .................................................................................... 61 Table 3-5: Trends in English and math textbook distribution ................................... 62 Table 3-6: Indicators on Kindergarten Teachers ........................................................ 65 Table 3-7: Indicators on Primary Teachers ................................................................. 66 Table 3-8: 20 Districts with the Highest and Lowest Primary PTTRs, 2008/09 ....... 68 Table 3-9: Indicators on JHS Teachers ........................................................................ 68 Table 3-10: Number of SHS Teachers, Qualified/Unqualified ................................... 69 Table 3-11: Enrollment in Teacher Training Colleges (TTCs) .................................. 70 Table 3-12: Number of teachers on study leave in relation to output from training colleges ............................................................................................................................. 71 Table 3-13: Admission in Untrained Teachers Training Program ............................ 71 Table 3-14: Percentage of Trained Teachers............................................................... 72 Table 3-15: Teacher questionnaire responses to management-related statements (rounded percentages) .................................................................................................... 73 Table 3-16: International Comparative Analysis of Public Spending on Education 76 Table 3-17: Expenditures in million GHc in 2008 prices, 2003 – 2008 ...................... 76 Table 3-18: Total expenditure in million GHc by source and level, 2008 .................. 79 Table 3-19: Recurrent expenditures in million GHc by source and subsector, 2008 79 Table 3-20: Capital expenditures in million GHc by source and subsector, 2008 .... 80 Table 3-21: GET Fund releases in million GHc by type and subsector, 2008 ........... 81 Table 3-22: Comparative Average Unit Costs as a Share of GDP Per Capita by Level ................................................................................................................................. 81 Table 3-23: Per student recurrent expenditure by level, 2008.................................... 82 Table 3-24: Per student recurrent expenditure by subsector, 2005 - 2008 ................ 82 Table 3-25: Reported Sources ........................................................................................ 83 Table 3-26: Reported years ............................................................................................ 84 Table 3-27: Funding Sources and Budgetary Items .................................................... 84 Table 3-28: Per Child Expenditure in 2008 GHc measured at district level ............. 85 Table 3-29: Sources of information on teachers........................................................... 86 Table 3-30: Payroll numbers GES 2008....................................................................... 86 Table 3-31: Staff in the basic cycle, 2008 ...................................................................... 87 Table 3-32: Median years of service as of 2009 by region ........................................... 89 Table 3-33: Estimated Cost of Trained/Untrained Teachers in Primary 1-3............ 90 Table 4-1: Enrollment in Secondary and Tertiary Public Schools, by Quintile ..... 100 Table 4-2: Gross and Net Enrollment in Primary and Secondary in 2008, by Urban/Rural Location .................................................................................................. 114 Table 4-3: Existing Deprived Districts and Those Calculated Using 2008/09 Data 117 Table 4-4: Targeting performance of school lunches using district level allocations ......................................................................................................................................... 119 Table 4-5: Share of the benefits that go to each quintile using CWIQ and admin. data and education needs in CWIQ ............................................................................ 119 Table 4-6: Share of School Lunch Expenditure by Deprivation Index 1................. 120 Table 4-7: Efficiency in classroom distribution in the Basic Cycle, 2004/05 - 2008/09 ......................................................................................................................................... 123 Table 4-8: Primary Pupil Teacher Ratio at the Regional Level ............................... 125 Table 4-9: Percentage of children in districts with sub-standard PCE in 2008 ...... 130 Table 4-10: Percentage of children in districts with substandard PCE at KG level ......................................................................................................................................... 130 Table 4-11: Percentage of children in districts with substandard PCE at Primary level ................................................................................................................................. 131 Table 4-12: Percentage of children in districts with substandard PCE at JHS level ......................................................................................................................................... 131 Table 4-13: Percentage of children in districts with substandard PCE at SHS ...... 132 Table 4-14: Per unit expenditure by region and deprived, 2009 .............................. 133 Table 4-15: Share of schools affected by teacher absenteeism in 1988 and 2003, (%) ......................................................................................................................................... 135 Table 4-16: Ghana’s Average Percentages of Time Use, by Primary Grade .......... 136 Table 4-17: Time-on-task for classroom activities ..................................................... 136 Table 4-18: Instructional Time Use in 4 Countries.................................................... 137 Table 4-19: Hours spent in primary and junior high school in past week, percent 138 Table 4-20: Frequency of head-teacher and circuit supervisor activities, 2003 (percent) ......................................................................................................................... 145 Table 4-21: Level of MMDA Discretion...................................................................... 147 Table 5-1: List of Accredited Tertiary Education Institutions as of February, 2009 ......................................................................................................................................... 150 Table 5-2: Male/Female Tertiary Enrollment ............................................................ 156 Table 5-3: Graduate Output by Broad Fields of Study, 2001 & 2007 (ISCED 5-6) 158 Table 5-4: Student/Academic Staff Ratios and Norms for Public Universities, 2003/04 and 2007/08 ...................................................................................................... 159 Table 5-5: Countries with the Highest Share of GDP p.c. per Tertiary Student (2005-2007)..................................................................................................................... 161 Table 6-1: Expected changes in PTR by levels ........................................................... 166 List of Figures Figure 1-1: African Comparison of Human Development Index, 2007 ....................... 6 Figure 1-2: Sub-Saharan Comparison of GNI per capita and GDP Growth, 2008 .... 7 Figure 1-3: Trends in Share of Trained Teachers in Primary and JHS, 1987-2008 .. 9 Figure 2-1: Median Years of Education Completed in 2008, by Female/Male ......... 12 Figure 2-2: Share of the Population without Education in 2008, by Female/Male ... 13 Figure 2-3: Share of children who have never attended school, by background characteristics .................................................................................................................. 13 Figure 2-4: Trends in Kindergarten Enrollment, Private/Public ............................... 16 Figure 2-5: Capitation grant’s impact on trends in primary net enrollment ............ 19 Figure 2-6: Trends in Primary GER and NER ............................................................ 19 Figure 2-7: Country Comparison in Primary Net Enrollment and Gender Parity, 2007................................................................................................................................... 20 Figure 2-8: Sub-Saharan Comparison of Primary Completion Rate, 2008 .............. 21 Figure 2-9: Public Primary Schools in 2008/09, by Age-grade ................................... 23 Figure 2-10: Share of new entrants to 1st grade who are at least 2 years over-age, by background characteristics ............................................................................................ 24 Figure 2-11: Enrollment Trend in JHS 1980-2008, by Grade .................................... 26 Figure 2-12: Gross and Net Admission Rates in JHS 1 ............................................... 26 Figure 2-13: Public JHS in 2008/09, by Age-grade ...................................................... 27 Figure 2-14: Student flow in 2008/09 ............................................................................ 29 Figure 2-15: International Comparison in Total Secondary GER and NER, 2006 .. 31 Figure 2-16: International Comparison in Secondary NER and GNI per capita ..... 32 Figure 2-17: Enrollment Trends in Special Education, Male/Female ....................... 35 Figure 2-18: Share of Students with Correct Reading Comprehension, 3rd and 5th Grade ................................................................................................................................ 37 Figure 2-19: Access and Zones of Exclusion in Primary and Secondary .................. 38 Figure 2-20: Mean NEA scores for both 3rd and 6th grade by gender and by locality ........................................................................................................................................... 40 Figure 2-21: NEA test scores and school lunches ......................................................... 44 Figure 2-22: Regional BECE Pass Rates in 2008/09 .................................................... 44 Figure 2-23:JHS Pupil (Trained-) Teacher Ratio and BECE English Pass Rates, by District .............................................................................................................................. 47 Figure 2-24: International Trends in Math (left) and Science Scores (right) of Eighth-graders, 2003/2007 ............................................................................................. 48 Figure 2-25: WASSCE examination results 2005/06 – 2008- 09 ................................. 48 Figure 2-26: Impact of Education on the likelihood of Wage Employment (vs. no Education) and Non-Agriculture Self-Employment (vs. no Education), Urban ....... 51 Figure 2-27: Wage Regressions and Non-Agriculture Self-Employed Regressions with Sample Selection, Ghana 1991-2006, Urban ........................................................ 53 Figure 2-28: Level of Education and Fertility Rates, 2008 ......................................... 54 Figure 3-1: Number of Pupils per School at the Basic Level ...................................... 59 Figure 3-2: Trends in Preschools with Toilets and Potable Water ............................ 60 Figure 3-3: Trends in the Number of Pupils per JHS, Public/Private ....................... 61 Figure 3-4: Textbooks per Pupil at Basic Levels.......................................................... 63 Figure 3-5: Trends in Primary Enrollment and Share of Unqualified Primary Teachers ........................................................................................................................... 64 Figure 3-6: Trends in PTR at the Basic Level, 2001/02-2008/09 ................................ 65 Figure 3-7: International Trends in Trained Primary Teachers and Primary PTRs ........................................................................................................................................... 66 Figure 3-8: Trained JHS Teachers in 2008/09, Public/Private and Male/Female .... 69 Figure 3-9: Trend in Expenditure on Education as a Share of GDP ......................... 75 Figure 3-10: Expenditure by level of education, 2004 and 2008 ................................. 77 Figure 3-11: Trends in MOE Expenditure ................................................................... 78 Figure 3-12: Composition of per unit recurrent expenditure from all sources by subsector, 2008 ................................................................................................................ 83 Figure 3-13: The GES salary scale ................................................................................ 86 Figure 3-14: Per unit annual expenditure by role and subsector, 2008 ..................... 87 Figure 3-15: Distribution over the GES salary scale, 2009 ......................................... 88 Figure 3-16: Distribution of salaries and entry levels, 2009 ....................................... 88 Figure 4-1: Net enrollment at the basic level, by quintile ........................................... 93 Figure 4-2: Mean Age-Grade ‘Delay’ per Pupil, by Household Welfare Quintile ... 95 Figure 4-3: Primary and JHS Completion Rate, by Quintile ..................................... 96 Figure 4-4: Primary PTTR in Districts, by Quintile ................................................... 97 Figure 4-5: NEA learning results in districts for primary 6, by quintile ................... 98 Figure 4-6: BECE English Pass Rate in Districts, by Income..................................... 99 Figure 4-7: Non-attendance and net enrollment in basic education, by quintile .... 100 Figure 4-8: Gender Parity Index in Primary and JHS, by Income .......................... 101 Figure 4-9: Average NEA English scores by region .................................................. 106 Figure 4-10: Primary P(T)TR and Completion Rate, by District in 2008/09 .......... 106 Figure 4-11: Comparison of Mean Age-grade Delay, by Region.............................. 107 Figure 4-12: Portion of Primary Teachers with training, by Region/District, in 2008/09 ........................................................................................................................... 109 Figure 4-13: PTTR in Primary and JHS, by Region in 2008/09............................... 111 Figure 4-14: Junior High P(T)TR and Completion Rate, by Region in 2008/09 .... 112 Figure 4-15: Regional distribution of S.H. Schools by type, 2008 ............................ 115 Figure 4-16: Visualizing the disparities: Deprived districts 2006-2008 ................... 117 Figure 4-17: Effect of increased enrollment in junior high schools ......................... 121 Figure 4-18: Senior High Schools Enrollment distribution ...................................... 122 Figure 4-19: Pupils per Classroom and Teacher by school size in Primary, JHS, 2008................................................................................................................................. 123 Figure 4-20: Efficiency in core textbooks distribution in Primary and JH, 2004/05 - 2008/09 ........................................................................................................................... 124 Figure 4-21: Students - Core textbooks scatter plot in the Secondary Cycle 2005/06 - 2008/09 ........................................................................................................................... 125 Figure 4-22: Efficiency in teacher distribution in Primary and JH by deprivation status, 2004/05-2008/09 ................................................................................................. 126 Figure 4-23: Students - Teachers scatter plot in Primary, 2008/09 .......................... 127 Figure 4-24: Teacher Allocation in Junior High Schools in 2008/09, by District (left) and by Schools (right) ................................................................................................... 127 Figure 4-25: Pupils - Teachers- Classrooms in SH, 2005/06 - 2008/09..................... 128 Figure 4-26: Students - Teachers scatter plot in the secondary cycle 2005/06 - 2008/09 ........................................................................................................................... 128 Figure 4-27: Visualizing the disparities – The districts in 2008 ............................... 129 Figure 4-28: Visualizing the disparities – Between and within regions in 2008 ...... 129 Figure 4-29: Visualizing disparities - Deprived Districts in 2008 ............................. 133 Figure 4-30: Expected and Actual Number of Days Students were Engaged in Learning ......................................................................................................................... 137 Figure 4-31: Education expenditure by source, 2007 ................................................ 139 Figure 4-32: Item 1, GoG Core Budget- PE .............................................................. 140 Figure 4-33: Item 2, GoG Core Budget Administration. .......................................... 141 Figure 4-34: Item 3, Core GoG Budget – Services ..................................................... 141 Figure 4-35: Item 4, GoG Core Budget – Investment. .............................................. 142 Figure 4-36: Capitation Grant ..................................................................................... 142 Figure 5-1: Enrollment in Public Universities and Polytechnics 2001/02-2007/08 . 153 Figure 5-2: Overview of Primary, Secondary and Tertiary GERs, 2006 ................ 154 Figure 5-3: Region of Origin of Undergraduates Students at UG, 2002/3 and 2007/08 ......................................................................................................................................... 155 Figure 5-4: Distribution of Tertiary Gross Enrollment Ratios by household Wealth and Gender, 2006 .......................................................................................................... 157 Figure 5-5: Trends in Enrollments in Public Universities by Broad Fields of Study, 2001/02-2007/08 ............................................................................................................. 158 Figure 5-6:Full-Time Teaching staff in Public Universities by Qualifications and Gender, 2006/07............................................................................................................. 160 List of Boxes Box 2-1: Abolition of school fees compensated by capitation grants ......................... 18 Box 2-2: A Case Study of the Causes of School Dropout in Northern Communities 28 Box 4-1: An Outlier in Learning Achievement - Talensi-Nabdam’s success at a Glance............................................................................................................................... 98 Box 4-2: Reaching the marginalized: the ‘School for Life’ program....................... 102 List of Maps Map 2-1: NEA scores of Primary 6 English in 2007, by district................................. 41 Map 2-2: Public Trained Primary Teachers and NEA Math P6 Scores ................... 41 Map 2-3: BECE English and Math Pass Rates, 2008/09 ............................................. 45 Map 3-1: District Level average PTR in Primary Schools, 2008/09........................... 67 Map 4-1: Kindergarten Net Enrollment in 2008/09................................................... 102 Map 4-2: Primary NER in 2006/07 and 2008/09, by District .................................... 103 Map 4-3: Primary Net Admission Rate in 2008/09 .................................................... 104 Map 4-4: Primary Completion Rate in 2008/09 ......................................................... 105 Map 4-5: Primary PTRs in 2006/07 and 2008/09, by Region .................................... 108 Map 4-6: Classrooms Needing Major Repairs and Schools with Potable Water, 2008/09 ........................................................................................................................... 110 Map 4-7: JHS Net Enrollment in 2006/07 and 2008/09 ............................................. 113 Map 4-8: Share of Trained Public JHS Teachers in 2008/09 ................................... 113 Map 4-9: Government’s Identification of Deprived Districts and District Level Poverty Count................................................................................................................ 116 List of Annexes Annex 1: Political and Administrative System ........................................................... 177 Annex 2: Map ................................................................................................................ 179 EXECUTIVE SUMMARY Purpose of the Report This report is intended to provide a thorough analytical foundation for strategic planning and sector dialogue among education policy makers, development partners and civil society representatives at a key turning point in the development of education in Ghana. During the last decade, access has radically improved. Between 2002/03 and 2009/10 enrollments in primary education grew by 1,284,673 or 50.9 percent. This expansion has outstripped the progress made by virtually all other countries, and is a testimony to Ghana’s deep commitment to a policy aimed at achieving universal education. Pre-primary and Junior High School are the fastest expanding areas. Between 2002/03 and 2009/10, the pre-primary gross enrollment rate went from 21.8 to 97.3 percent (net enrollment, from 19 to 63.6 percent for 2008/09). At junior secondary school, gross enrollment rose from 64 to 79.5 percent (net, from 30 to 47.8 percent) during this period. As enrollment and completion have increased at basic levels, demand for post-basic education has also increased but supply hasn’t followed suit. Only about 35–40 percent of students who finish Junior High School (officially the end of basic education) go on to enroll in a Senior High School, and even this limited transition would be lower were it not for the increase in the capacity of private senior high schools (SHSs). Extensive growth in basic education seems to have reached a point where services cannot be stretched much further. The gross admission rate to primary education is now over a hundred percent, gross enrollments are considerably above 90 percent, and upper secondary enrollments have reached the maximum capacity within the available infrastructure. The education sector employs about 40 percent of the total civil service and uses about a third of the public budget. At the same time, persistent problems (including out-of-school children) and emerging issues (including persistent disparities and limited learning outcomes) require new solutions. Meeting the challenges requires a new definition of performance focusing on more equitable and more efficient services, more informed policies, and strengthened accountability. The present analysis is an attempt to bring together key outcomes, resource allocations, and a deeper look at equity, efficiency and accountability in education service delivery. The great progress in education made by Ghana in the last decade is evidenced by (a) coverage expanded to about 84 percent of the relevant age population in primary, (b) basic services extended to 11 years, (c) expansion of access at secondary and tertiary levels, and (d) growth of public expenditures both in scale and in the share of Government expenditures. At the same time, great challenges remain such as: (a) socio- economic disparities in access and quality; (b) significant and stubborn regional differences; (c) poor quality of learning; (d) inefficient use of resources; and, (e) i insufficient measures to assure that schools and administrators are accountable for delivering the education services in line with expectations. Communities, parents and students, especially in poor areas have legitimate concerns about sub-standard schools, sub-standard supply of human and financial resources, missing teachers, poor exams, and lacking services. Despite investments, many classes are still held under trees, many children are turned away from higher levels of schooling, and many fail comprehensive exams and do not find post-basic schooling places or job opportunities. It is becoming clear that addressing these challenges and achieving further progress cannot be achieved through a simple algorithm between inputs, resources, standards and projected educational outcomes. The source, scale, allocation and use of resources – physical, human and financial alike – have reached a level of complexity in which central planning and decision making can no longer determine how much of these resources are allocated and used, when, where and by whom. New approaches promoting greater effectiveness and accountability are required. The dramatic increases in admission and enrollment rates to basic education during the past decade are due mostly to the elimination of school fees, followed several years later by the introduction of the capitation grant to help schools make up for those fees. Ghana’s 1992 Constitution declared that basic education “shall be free, compulsory and available to all.� The Free Compulsory Universal Basic Education (FCUBE) reforms of 1996 were designed to push this goal forward. FCUBE was extended from 6 to 9 years by adding the 3-year JHS level. In 2002, it was extended to 11 years, when the government added two years of pre-primary, since it is seen as a key element to help prepare children for primary school. However during this period, in the context of flagging public expenditures on education, many schools found themselves forced to impose indirect fees, e.g. for registration, uniforms, textbooks, etc. In response, the government in 2004 introduced capitation grants for 40 deprived districts, extending the grants to the whole country in 2005. One effect of these efforts has been a dramatic increase in the willingness of parents to start their children’s schooling at the appropriate age. As late as 2004, only about a fourth of 6-year-olds were enrolled in first grade. But by 2009, three fourths of 6-year- olds were enrolled. Still, almost thirty percent of children are older than the official age when they start first grade. Being over-age is a prime factor contributing to repetition and drop out. Many over-age children do not complete the cycle. Yet the overall rates mask stark disparities. The prospects of children in the poorest quintile never having gone to school are about 10 times greater than those of children in the richest quintile. If a child lives in a rural area, his or her chances of not having attended school are over three times higher than an urban child. A rural girl from the poorest quintile is 13.9 times more likely to not have attended school than an urban boy from the richest quintile. This trend grows sharper further up the education ladder, culminating in very restricted entry to higher education. A few secondary schools dominate tertiary intake. A recent survey of 1,500 students in the country’s five oldest public universities ii confirms that 50% of students came from only 29 of the 493 public senior high schools in the country. Gender parity has nearly been achieved at primary level, where girls enroll at almost the same rates as boys. But at higher levels, girls’ participation drops off significantly. Poor families often “sacrifice� their girls’ education, and have them work at home or on the family farm, in order to keep their boys in school. When girls do manage to attend secondary school, it is not unusual for them to face a hostile environment of sexual harassment from male students and even some male teachers. Such harassment may be tolerated by school authorities; in any case it contributes to a higher female dropout rate. The gender gap is much bigger at the country’s Technical and Vocational Education and Training (TVET) institutions, which are increasingly seen as having a key role to play in preparing young people for a vast array of in-demand skilled jobs. As for children with various handicaps, there is as yet very little special education to help them develop their potential. Ghana’s stake in gender parity goes beyond giving girls more opportunities. Greater education levels among women are closely associated with a delayed start of child bearing, smaller families and healthier children. For example, women with no education give birth to about three times more children than women with secondary or higher (WHO 2008:92). And a mother’s education is also inversely associated with a child’s risk of dying from disease. Whatever the difficulties, education has become a core priority for Ghana. Government expenditures on the sector have reached unprecedented levels, proportionally double the average for Africa and other developing regions. Ghana’s education spending hovered around a respectable 5 – 6 percent of GDP since 1990, but has been steadily rising over the past decade. Spending reached 11 of GDP percent in 2008 -- about 30 percent of the government budget. A number of initiatives have been established or expanded: the school feeding program; programs for school infrastructure, textbooks, and uniforms; supplements for teachers; and a high profile annual award for the best performing teachers. Education has become one of the most frequently discussed topics of public concern in newspapers, radio and TV programs, next to politics, the economy and of course sports. The country was among the first in Africa joining the Education for All movement, subscribing to the EFA Fast Track Initiative guidelines. Table 1: Trends in key education indicators in Ghana Kindergarten Primary JSS SSS 2002/03 2009/10 2001/02 2009/10 2001/02 2009/10 2003/04 2009/10 GER 21.8 97.3 80 94.9 64 79.5 26.6 36.1 NER 19 63.6** 59 83.6 30 47.8 GAR 88* 101.3 71.6* 86.6 39.4 NAR 24.5* 71.1 10.6* 43.8 GPI 0.98* 0.99** 0.93* 0.96 0.88* 0.92 Note: *refers to data from 2003/04; ** 2008/09. Source: Calculations based on EMIS. iii The completion rate in Ghana is relatively high but not high enough to achieve the Millennium Development Goals (MDG). School census data for the latest year (2008/09) indicate that the primary completion rate reached 87.1 percent, which represents a 9.2 percent increase since 2003/04. This is all the more relevant considering that with primary NER at 83.6 percent, over 16 percent of children never even start school, representing a NER decline of about 5 percent relative to the academic year 2008/09. Although the enrolment and completion rates are relatively high for Sub- Saharan Africa (SSA), they also show that achieving universal primary completion (UPC) is unlikely. As UPC nears, improvements in completion are more and more difficult. Children living in extreme poverty and in geographically isolated settings tend to drop out first as they are expected to contribute to household income generation, and girls in particular are expected to take care of younger siblings and children with special needs. Being over-age is also a key cause of repetition and dropout. The problem is exacerbated by classroom overcrowding, which has in turn been worsened by the rapid enrollment increases of recent years. Drop out is higher in JHS, averaging about 30 percent of students who start the cycle. On average, children from households in the poorest quintile (largely concentrated within the three Northern regions) are about three years over the official age in their respective grade. Children from households in the richest quintile tend to be less than two years older than the official age for their particular grade. In general, the poorer the families are, the higher the risk is for having overage pupils in school. (The poorest quintile tends to be 3 years over age, but the richest are still almost two years over age.) The phenomenon is due to, for instance, the need to care for younger siblings or to work during times of harvest, which may lead to postponed entry into primary 1 or frequent repetition as a result of lost instruction time. The consequences of high age-grade delay are a straining of the pedagogy process, higher public costs of schooling, reduced learning results and increased dropout rates. It is essential to provide incentives to, in particular, poor and rural households, to ensure that children enroll at the official age. Poorer quintiles have much lower enrollment and completion rates. Students from economically deprived households have not only fewer chances than their rich counterparts to be enrolled in school, but those that are have an even smaller probability of completing a given cycle. At the same time, there are serious concerns about the quality of education. In overcrowded classrooms, often instructed by teachers who have not had any pedagogical training, large numbers of students complete primary education without attaining functional literacy. Poor learning outcomes mean large numbers of students who gain little from their time at school and are excluded from the benefits of higher levels of schooling. Poor educational quality also undermines Ghana’s ability to compete internationally. Less than a third of primary school children reach proficiency levels in English or in Mathematics, according to the National Education Assessment tests of 2005, 2007 and 2009. These tests were conducted on a nationally representative random sample of 3.5 iv percent of Ghana’s primary schools (NEA 2007:2). The 2009 results for Primary 3 showed 20 percent of students proficient in English and 25.2 percent in Math (improvements in both subjects relative to 2005 results). For Primary 6 the results were 35.6 percent proficient in English and 13.8 percent in Math (improvements in both subjects compared to 2005 results).i Moreover, there are great geographical disparities in the quality of education , with students in rural schools, especially in the north of the country, at a big disadvantage. While access rates between rural and urban areas are already large, the gap in learning outcomes appears even larger. The highest concentration of 6th graders with low NEA English scores (38 or below) went to school within districts of the Northern region, while the highest concentration of students with high scores (44 or above) attended schools in the districts of Greater Accra. This coincided with other key factors. High performing schools had over 90 percent trained teachers -- 36.4 percent more trained teachers than low performing schools. The proportion of high performing schools in urban settings with over 5,000 residents was 72.2 percent, and among those with electricity was 47.3 percent, while the share of low performing schools in urban areas was 8.5 percent and among those with electricity was 9.1 percent. The data show that the districts with the poorest performance on NEA were in the Northern regions; scores in math were 15 points lower than the national average and less than half compared to the top scoring districts in the country. These districts had primary Pupil Trained Teacher Ratios (PTTR) of 150 to 200. In these remote areas, schools are typically small, with enrollments typically less than 100. This means that often only every second or third school in the region has a trained teacher. Among other hardships, more than half of students don’t have access to a treated water source, a situation linked to higher incidence of disease, including guinea worm infection and typhoid. The same pattern holds true for the Basic Education Certificate Examination (BECE), a comprehensive leaving test at the end of lower secondary school. In 2008/09, the bottom five performing districts on BECE English exams were all found within the Northern region. BECE performance correlates more strongly with the share of trained teachers than to the degree to which students have access to school materials and basic infrastructure. Human Resources: Teacher Supply and Allocation The Pupil Teacher Ratio (PTR) at the primary level has remained unchanged during the past decade, just above 32, despite the sharp increase in enrollments, thanks to efforts to recruit a large number of new teachers, most of them unqualified. Between 2001/02 and 2008/09 the number of primary teachers grew from 80,552 to 114,421. However, the portion of primary teachers who were untrained grew from 35.1 percent to 52 percent, resulting in more students for each qualified teacher: the primary Pupil Trained Teacher Ratio (PTTR) rose from 49.3 to 67.5. This has negative v implications for the quality of education, since the number of trained teachers is widely seen as the single most important factor in learning outcomes. Table 3-8: Indicators on Primary Teachers 2001/02 2002/03 2003/04 2004/05 2005/06 2006/07 2007/08 2008/09 TEACHERS 80,552 80,459 82,833 89,278 88,461 105,257 112,443 114,421 Untrained Teachers 28,238 29,866 32,296 37,608 38,654 49,459 57,266 59,488 % of Private 31.7% 30.0% 39.6% 40.8% 23.1% 20.7% 23.2% 22.2% % of Untrained 35.1% 37.1% 39.0% 42.1% 43.7% 47.0% 50.9% 52.0% % of Untrained Public 22.8% 26.5% 26.2% 27.7% 32.9% 37.9% 40.6% 41.6% % of Untrained Private 83.0% 84.7% 84.5% 85.0% 85.5% 83.6% 87.6% 88.4% % of Female 32.1 32.8 32.0 31.1 33.9 27.7 33.0 33.9 PTR 32.1 31.4 34 34.9 35.7 33.5 34.1 32.4 PTTR 49.4 49.9 53.2 56.7 62.7 60.3 65.5 67.5 PTR in Private 29 27 27 27 26 26 25 26.3 PTTR in Private 170.3 178.6 174.0 178.3 180.2 157.6 203.7 227.6 PTR in Public 33 32 34 35 38 33 34 34.2 Source: EMIS. Worse, the distribution of teachers, and especially of trained teachers, is very unequal. Teachers generally prefer to work in better-endowed urban centers, and avoid being deployed to schools in remote, impoverished areas, with their poorer infrastructure and sanitation. Both in the Primary and JH cycles the highest PTR are more than 30% higher than the lowest. Higher PTRs are found in the North and in the Greater Accra region, in the latter mainly due to migration patterns: the influx of large numbers of families from the rural areas. Disparities in the PTTR are much greater. At the national level, primary level PTTR in 2008/09 was 67.5. But in about one quarter of the districts, the average PTTR was around 100. In ten of these districts, all of them located in the Northern, Upper West, Upper East and Western regions, there were over 155 students for every trained teacher. Most of these districts are at the same time classified as belonging to the poorest quintile. Within individual districts, qualified teachers prefer to stay in urban areas, leaving rural schools with even fewer of their numbers. Also in 2008/09, 75% of the 37,700 kindergarten teachers were unqualified. The proportion of untrained teachers in JHS increased between 2001 and 2008 from one fifth to over a third while their total number grew by 160 percent. Among the approximately 70,000 unqualified teachers employed in Ghanaian basic schools in 2007: (i) about 22,000 were new university graduates doing one year’s national service in Ghana’s schools; (ii) about 12,000 more, young and sometimes older, were people teaching as volunteers under a National Volunteer Scheme and (iii) 9,000 were students in the final year in teachers’ colleges teaching whole-time in schools as part of their teacher training programs. vi Efficiency in schools: Absenteeism and Time-on-task High teacher absenteeism is one of the key inefficiencies in the education sector. It is more common in rural schools, apparently associated with higher occurrences of a poor work environment and poor teacher morale. The average teacher absentee rate, as a survey by the Center for Democratic Development (2008) indicates, was 27 percent. Teachers in Tunisia and Morocco miss 11.6 days and 13.4 days a year, respectively, while teachers in Ghana miss 43 days (Abadzi 2007). Among the main underlying reasons for the high absentee rate are (i) lack of supervision, (ii) sickness/medical care, (iii) collection of salary at a bank located at a distance, (iv) frequent funeral attendance, (v) long distances to school, (vi) religious practices (for instance, Friday prayers among Muslim teachers), (vii) schools lacking facilities, especially sanitation: toilets and potable water, (viii) schools located far from lorry/bus stations and healthcare facilities, (ix) rural teachers supplementing their income by engaging in activities related to farming. Low attendance is compounded by the problem of low instructional time in school, i.e. low time on task, which appears to be widespread. A simple observation of school activities provides a good picture of the causes of low time on task. Frequent breaks, when teachers are separated from students, slow transition to instruction, down time spent disciplining students, collecting homework etc., all create an illusion of children learning while in school. 4.1 Together, teacher absenteeism, poor ToT and short duration of school year can result in the loss of as much as 50-60 percent of teaching time, clearly making this as a key constraint to learning. Figure: Expected and Actual Number of Days Students were Engaged in Learning 250 197194 190185 200195 204203 200 171 175 176 148 145 150 126 109 100 76 50 0 GHANA Tunisia Pernambuco (Brazil) Morocco Days in School Year Days after Closure No. Days Schools Operated Days Students Engaged in Learning Source: Based on Abadzi 2007. Patterns of Resource Allocation Allocation of teachers Efficiency in teacher allocation generally decreased between 2005/06 and 2008/09. In 2008/09 only 29, 46 and 55% of teacher presence could be explained by enrollment in vii Kindergarten, Primary and Junior High School, respectively. In the primary subsector, 55% of the number of teachers observed in school was related to the number of pupils in 2005/06 (falling to 46% in 2008/09). In other words “randomness� in teacher deployment increased from 45% to 54% between 2005/06 and 2008/09. Teachers are even less efficiently distributed in deprived districts, where in 2008/09 “randomness� reached 62%. Efficiency in teacher deployment has been decreasing in the Junior High subsector as well. Whilst 63% of teacher presence was explained by enrollment in 2005/06, 55% of teacher presence could be explained by enrollment in 2008/09. Efficiency in teacher presence in deprived districts fell by 10% in the last year. In effect, this means that for the last four years, efficiency of teacher deployment was always lower at all levels in deprived than in non-deprived schools. Evidently, the Government efforts to assure better education services to deprived districts have not been effective. Allocation of Financial Resources More than 60% of children in the Northern region attend primary school in districts where the per child expenditure (PCE) is within the bottom third of the nation. This means that there is negative discrimination: deprived districts benefit less from public resource allocations than non-deprived districts. The proportion of Junior High School students suffering from under-spending in the Northern region is even higher, reaching 83% of the total number of students in the region. Percentage of children in districts with sub-standard PCE in 2008 KG Primary JHS ASHANTI 28% 29% 23% BRONG AHAFO 0% 35% 45% EASTERN 0% 15% 0% GREATER ACCRA 17% 7% 0% NORTHERN 49% 65% 83% UPPER EAST 59% 66% 63% UPPER WEST 93% 45% 59% VOLTA 15% 0% 12% WESTERN 24% 18% 5% The conclusion is that a key source of deprivation for the deprived districts is the allocation of public expenditures. These expenditures, instead of compensating for the deprivation, exacerbate them by allocating fewer resources per child to the regions where the majority of deprived districts are located. Management Accountability -- Fragmented System of Flow of Funds Education financing is fragmented among a number of sources and among an even larger number of flows of funds. This is the most fundamental reason why planned and executed budgets differ. The authority to allocate the key resources are divided among four agencies: The Ministry of Finance and Economic Planning (MoFEP) is viii responsible to set the overall budget and to determine (through the Civil Service Agency) the remunerations; The Ghana Education Trust (GET) Fund is responsible for investments; Ghana Education Service (GES) is responsible for allocating recurrent expenditure and to set teacher numbers; and the Ministry of Education (MOE) is responsible for coordinating donor funds and proposing the annual education budget to the Government. A large majority of the public spending on education is channeled through the budget of the Ministry of Education. However, this budget only covers teachers’ salaries, new school construction and some inputs such as textbooks. A significant portion of public resources benefitting the education sector is still channeled through the budgets of other ministries, departments and agencies of the Government, among others, the Social Investment Fund and the GET Fund. Education funding is also derived from the budgets of the MoFEP and the Ministry of Local Government. Some of the resources from the “MP Fund,� allocated to Members of Parliament to promote development in their constituencies, also support school projects, scholarships and the like. There are significant delays in several funding flows, particularly for the service component of the budget, due to various systemic and capacity related factors. There is inadequate capacity at the Metropolitan, Municipal and District Assemblies (MMDA) level to account and report in a timely manner for the use of funds. Moreover, these problems are worsened by the multiple funding flows, planning systems, accounting procedures, reporting systems, auditing requirements and bank accounts. With the exception of salaries, fund flow is complex and often leads to fragmented and unpredictable planning and budgeting in the MMDAs. This has a spill-over effect on service delivery institutions/schools, which have to face lack delays of funding for operations and maintenance particularly at the beginning of the financial year. Capitation grants are supposed to address some of these problems and have improved the funding for operations of most of the schools. Division of Roles and Responsibilities Too many vertical levels ‘blur’ responsibility. MOE represents the sector in strategic (Government, and Development Partner -- DP) dialogue whereas GES is responsible for service delivery including deployment of teachers, allocation of textbooks, supervision of schools and teachers. GES has no mandate to manage the payroll. It does not have effective tools to assure that teachers who are assigned to one school or even to one district take the assignments or remain there. Funds and resources to local level flow through a variety of channels, limiting the capacity of agencies to pursue a coherent spending strategy. In the meantime, when teachers leave their posts for one reason or another (such as study leave or relocation) GES has the responsibility of replacing them. As enrollments have increased in the last decade, the gap between the planned and executed payroll has widened from about 10 percent to over 35 percent. ix The District Education Department (DED) system is diffused rather than decentralized, inasmuch as District officials, including the Director are appointed by GES headquarters and report through regional offices to the center. The DEDs were created in 1995 when the Ghana Education Service (GES) reviewed its management structures at headquarters, regional and district levels in order to bring authority and responsibility for service closer to communities. The strong connections between central, regional and local levels limit the capacity of locally elected officials to oversee the sector and confront problems. Under new legislation, District Education Directors are expected to report to the MMDAs. This should strengthen local accountability; however, it is unclear how the standards for curriculum and teaching and other regulations set by the center would be enforced. At school level, headteachers are selected from the ranks of serving teachers. Teachers who have reached the level of Principal Superintendent (with at least 12 years experience) are eligible for consideration. No special training is given. Teachers are interviewed at district level and appointed to the schools. Supervision of teachers by circuit supervisors and head teachers is uneven, but can have a significant positive impact. However, many circuit supervisors don’t observe and assess teachers during instruction time, but rather visit schools to simply verify student and teacher attendance figures. There exist several mechanisms to promote community and stakeholder involvement and oversight in local schools. Parent Teacher Associations (PTAs) are the oldest and most widespread, though they have limited powers. School Management Committees and School Performance Assessment Meetings were established more recently, but have had trouble gaining a real foothold and tend to operate sporadically. The most influential form of community involvement typically comes from local traditional authorities – headed by the local chief. When the traditional authority is interested and active, it can provide powerful oversight on educational issues, including whether teachers are present and the extent to which students spend their school time on task. The Education Management Information System (EMIS) to date is basically an annual school census, usually carried out at the end of the calendar year or early in the following year. This EMIS design is that it maintains a fragmentation between types and sources of data and that, in turn, limits its usefulness for decision makers. Tertiary education Since the early 1990s tertiary education has increased at an exponential rate. Within a decade (1997/98-2007/08) total enrollment in the public HEIs alone (excluding the colleges of education) has more than tripled from 40,000 to 132,604. A 104% increase x occurred in public universities while polytechnics recorded a 69% increase over the period 2001/02-2007/08. The rapid expansion has placed substantial pressures on staffing and on infrastructure. Also, as a consequence of the growing enrollment numbers, there are also increasingly more unemployed graduates, some of which eventually seek job opportunities abroad. The private higher education institutions are playing a significant role by easing the social demand for higher education. The number of accredited private HEIs reached 54 in 2009. But, while private enrollment is in fact rising, public enrollments are expanding at a faster pace and represent a growing share of total tertiary students. The polytechnics in Ghana had operated as second-cycle institutions until they were upgraded to tertiary level status in 1993. They are oriented to technical and vocational education and the training of middle level human resources for national development. Some Policy Options Overall, the learning outcomes have some important messages for policy makers: The educational system is not able to address or overcome the initial differences in reading comprehension and basic math skills; indeed, these initial differences are further intensified overtime. The improved access has brought into the system large groups of children who are now accessing education with various forms of additional challenges including overage children, children coming from poor families, families where parents have little educational background and comprehension of the importance of education and children in poor remote, rural communities and deprived districts where access to infrastructure, work and services are limited. Many of these children have a high risk of falling into various zones of exclusion. Meanwhile, the education system has its standards for transitioning students from one level to the next, there is a strong likelihood that many of those who gained access to education over the last decade have left limited opportunities to truly benefit from it. Even with Ghana’s extraordinary levels of spending on education, it is clear that public funds cannot cover all minimum needs. Under these conditions, allocations should be better prioritized and provisions adjusted to local conditions and need. A key example is in teacher provision. In remote locations the use of various incentives to attract teachers, especially trained ones, may be more important in affecting learning outcomes than other resources, like class supplies or the number of classrooms. The government has been trying to direct more resources to deprived areas. Most of the efforts have been based on externally financed projects, including the pilot programmatic scheme by IDA or lately funding for the deprived districts by USAID. But current administrative structures have often left these efforts unsuccessful. In reality, the complexity of resource allocation has led in many cases not to neutral allocation, but to negative targeting, as poorer areas receive not more provisions but less. Therefore, xi decisions about prioritizing the allocation of resources and the setting up of services will be most effective when they follow some principles: 4.2 they are evidence based, 4.3 they promote policy goals by providing adequate incentives for key participants including students, parents, teachers, headteachers, and district officials, and 4.4 they are targeted so that those who need the most get the largest share of public resources. Equity and Disparity in Education Services: If access is inequitable, it diminishes the performance and compromises the social and economic benefits. That is why the Ghanaian government has been committed to equitable access. A number of policies support this goal, including, among others, the elimination of fees, the provision of capitation grant, textbooks, school feeding and the delineation of deprived districts. But, by the same token, inequitable services, disparities in access and in educational inputs could not only compromise the effect of these policies but also result in greater social inequity and fragmentation. There are three levels of potential and desirable policy actions in the context of disparities: 4.5 Avoiding negative discrimination: At a minimum, the allocation of resources should not discriminate along social, cultural or geographical grounds. If x amount is available for education per student, the Government needs to assure that x amount is allocated so that all students see the benefit of the inputs equal to this amount. 4.6 Establishing specific policy incentives to improve supply of services: At a higher level, the Government resources should be allocated in a way that it addresses specific difficulties in education service delivery. If, for instance, enrollment and attendance is more challenging in some areas, typically in the rural and poor Northern regions, the Government conceives policies that address this specific educational problem. For instance, it is important to top up teacher salaries for those who are willing to teach in remote areas; accommodations or subsidized transportation can be provided to these teachers. 4.7 Creating positive discrimination to stimulate demand: At a highest level, the Government may develop policies that address underlying social and economic problems by stimulating demand and thus access to education. These policies may include school feeding, cash transfers or any type of extra provisions that relieve the burden on families for sending children to school. It is critical to the success of these measures how they are targeted and monitored. xii Quality of Learning As overall access to education improves -- especially at primary and basic levels -- the Government and other stakeholders need to pay increasing attention to quality issues. Even if increasing number of children find ways to gain admission and increasing proportion of them are enrolled, new forms of exclusion emerge as many of them simply do not attend classes, learn enough, drop out, repeat, fail to pass exit exams or fail to transit to post-primary levels. International assessments show that big disparities in learning outcomes in a country pull down overall performance even if there are some significant groups that demonstrate impressive achievements, claim successful academic or professional careers. Ghana’s commitment to education is shown by its willingness to participate in international learning assessments, by the national assessments of learning outcomes and by the numerous innovations that are sponsored and promoted through Government policies or non-governmental initiatives. These will gradually direct the public attention and the accountability of education providers to quality outcomes. This report shows that while the resources for education are impressive, they are not distributed equitably, targeted according to need and used efficiently. Quality standards are defined in terms of input with little understanding as to which inputs and which methods have the most impact on learning outcomes. Among the inputs, two key items appear to show major impact: the access to quality teaching and quality teachers and the instructional time. Based on the findings we identify three levels of potential and desirable policy actions: 4.8 Continue strengthening the Government capacities and increase investment in monitoring learning outcomes. While there is an impressive array of innovations and options for policy to improve quality, it is impossible to assess their effectiveness and scale up the successful ones if learning outcomes are not measured by using solid international and reliable national standards and methodologies. 4.9 Ensure that resources with strong impact on quality reach the classrooms. The available assessments confirm that trained teachers have the biggest impact on learning. Further, there are indications that high level of absenteeism, low time on task significantly compromise learning. To overcome these problems, both the central and local governments as well as the communities need to be given authority, improve capacity to supervise, monitor deployment of teachers, attendance and, indeed, classroom activities. Standards, expectations and regulations need to be stringent to avoid that new generations of students complete education without effective learning. 4.10 Give central support to local innovations. Improving quality in crowded urban schools requires different approach than in small rural schools often with limited infrastructure, facilities shared by different grade levels, poor communities and local language and cultural conditions. The Government xiii needs to provide services to schools with a variety of conditions, empower and motivate local governments, communities and non-governmental agencies to initiate these innovations and monitor their impact. Efficiency and Accountability in Education Management and Administration Resources will have to be better managed if the government wants to achieve further improvements in access to and quality of education services, given the already very high overall level of public expenditures. The Government has been considering some strategic changes at the local, school-level and central levels but has not fully implemented them. 4.11 Local authorities need stronger capacities for decision making, supervision and resource allocation to more effectively address the diverse social and economic conditions of education. These include more consistent promotion of innovative practices, strengthening the supervision by district level authorities and strengthening the role of local communities through improved accountability. 4.12 School management needs to be professionalized through training, strengthened authority and adjustment of the career patterns in education to make sure that those become headteachers who are trained and qualified for the post rather than those who are the most effective or senior teachers (teaching and school management are two different professions). 4.13 District authorities need to be given authority to hire, dismiss and deploy teachers and more resources to better supervise their attendance and performance. 4.14 The central Government needs to build capacities to develop and monitor policies and incentives, assess learning, monitor performance and guide those responsible for service delivery as it devolves management to district and school levels. Improving access to and relevance of higher education Ghana needs policies to deal with social pressures to increase higher education enrollment and economic pressures to improve relevance; without these, most social and economic expectations, and also the Government’s strategic agenda will remain unfulfilled. Differentiation of the sector is also key to increasing access, widening participation, increasing relevance, and reducing government spending. It’s also clear that effective policies require a change in the way higher education is financed and governed, how institutions and services are managed, and how quality is assured. Already almost half of the higher education revenues come from non-state resources; it is imperative that both state and non-state resources are used more strategically. A few policy options may include the following: xiv 4.15 Government financing should be more closely linked to performance, norms, and strategic priorities. The financing of inputs including staff and other costs based simply on enrollment, and the separation of investments from recurrent expenditures limit the Government’s ability to stimulate improved performance or progress along the desired strategic directions. 4.16 By the same token, Government financing should also be more targeted to meet specific social, economic or other strategic objectives. 4.17 Institutions need to have more management authority, responsibility and capability to run higher education effectively and efficiently. There are well tested instruments that can help the improvement of institutional performance including institutional planning, performance contracts, but also partnerships with other HEIs or the private sector. 4.18 Independent quality assurance, accreditation and performance monitoring should be established and strengthened, with the participation of both the academic community and the Government. These processes should involve members of the international academic community, the business community, social partners and civil society. xv 1. INTRODUCTION, CONTEXT AND ORGANIZATION 1. The purpose of this report is to provide sound analytical foundations for strategic planning and sector dialogue between education policy makers, development partners and civil society representatives at a key turning point in the development of education in Ghana. During the last five years, access has radically improved. As the GOG refines its economic development and poverty reduction agendas, it has become increasingly clear that better quality learning and increased educational attainment across the country are essential part of that agenda. After six years of accelerated growth in the sector, the Government and the stakeholders have come to the conclusion that an earlier strategic plan needs to be revised for a number of reasons. Extensive growth seems to have reached a point after which the scale of education services cannot be stretched too far. Gross admission rate to primary education is over a hundred percent, gross enrollment rate at primary level is in the upper 90 percentile, upper secondary provisions reached the maximum capacity within the available infrastructure, the sector employs about 40 percent of the total civil service and uses about a third of the public budget. In the meantime, persistent problems (including out-of-school children) and newly emerging issues (including widening disparities and limited learning outcomes) require new solutions. Further build-up of the system is unlikely. Attending to the challenges requires a new definition of performance by focusing on more equitable and more efficient services, more informed policies and strengthened accountability. 2. The present analysis is an attempt to bring together the key outcomes, the resource allocations and a strengthened look at equity, efficiency and accountability in education service delivery. There have been numerous strategic and analytical exercises about Ghana completed by policy makers, independent experts and development partners. Most notably, the 2003 Education Strategic Plan was preceded by a wide-ranging sector review and analysis. Since the ESP, National Education Sector Annual Reports looked at the sector performance and measured it against benchmarks set in the ESP. These reports mostly looked at the key outcomes measured through annual school censuses in terms of enrollment rates, gender parity and inputs including teachers, school infrastructure and textbooks. However, as noted earlier, further improvements in enrollment, admission rates, further investments into scaling up education are not possible unless some of the key challenges are addressed, including equitable and efficient resource allocation and service delivery, targeted policies and strengthened accountability. 3. Few countries in Sub-Saharan Africa show as much progress in education as Ghana does. Enrollment has increased sharply; in fact, gross enrollment for primary education has come to the upper 90 percentile and net enrollment is also near 90 percent. Enrollment rates at secondary and tertiary levels, although behind the averages of other continents, are among the top Sub-Saharan averages. Gender parity again at primary level is almost at the maximum, girls’ education appears to be catching up with boys’. The public and private investment is unprecedented. A third of the budget, over 10 percent of the GDP is used for the sector and other progressive policies have been pursued since the 1980s: universal basic education has been gradually extended to cover 6, then 9 and now 11 years including pre-primary years. Fees were eliminated for basic level, a school grant (“capitation grant�) was introduced to compensate for the fees, school feeding, public programs for school infrastructure, textbooks, uniforms and supplements for teachers and 1 a high profile annual award for the best performing teachers were introduced. The country was among the first in Africa joining the Education for All movement, subscribing to the EFA Fast Track Initiative guidelines and developing a long-term strategy early on that aims to reach the MDG goals by 2015. Education is one of the most frequently discussed public concern in newspapers, radio and TV programs, next to politics, the economy and of course sports. 4. Yet, the education sector is under constant pressure from expectations that surpass the achievements. Budgets are pressured and appear to be out of effective Government control. Communities, parents and students especially in poor areas have legitimate concerns with sub-standard schools, missing teachers, botched exams, lacking services. Despite the investments, many classrooms are still held under trees, many children are turned away from higher levels of schooling, and many fail completing exams or finding post-basic or job opportunities. 5. The pressure is well exemplified by the sector’s role and agenda within Ghana’s poverty reduction strategy, in the Growth and Poverty Reduction Strategy (GPRS) and a connected multi-donor budget support program (MDBS), which is the major channel for financial support from external development partners’. As it is universal, and indeed normal, the education sector has a key role in the GPRS. Also, as expected, the sector’s MDBS targets are closely linked with the EFA FTI guidelines as well as with the country’s first EFA-inspired strategy, the 2003 Education Strategic Plan (ESP). According to the ESP, the education community each year carries out an annual sector review, looking at the key outcomes, including enrollment and completion rates, gender parity, the resources used to achieve the outcomes and makes assessment of the sector’s performance. With connected MDBS, the country has high stakes in this review. If the country fails to achieve its MDBS triggers, it can lose some critical external funds. Consequently, each year anxious politicians are watching whether the milestones in education outcomes, in the resource allocations and in other measures of sector performance service delivery have been met. 6. The Books, Building and Learning Outcomes, a WB report which was completed in partnership with the Government in 2004, provided evidence that gains in educational outputs are directly linked to better school quality, manifested in improved infrastructure and greater availability of school supplies. It highlighted the importance of textbooks and school infrastructure in improving test scores and enrollments. Over the last seven years, since the report was completed, the Government further strengthened its commitment to Education for All and the Millennium Development Goals and its policy regime through measures such as the elimination of fees, the capitation grant, further improvements in infrastructure, textbook supply and in the conditions of teaching, gave rise to a rapid expansion in enrollment. 7. However, amidst these improvements in inputs and policies and the gradually increasing demand for education – partly as a consequence of these policies – it is becoming clear that further progress cannot be achieved by a simple algorithm between inputs, resources, standards and projected educational outcomes. There are four key reasons for this development: 2 1. Quantitative expansion has brought about persistent and sometimes worsening quality problems; 2. Aggregate improvements in inputs hide persistent disparities in resource allocation; 3. Various types of resources have different level of impact in addressing the disparities in outcomes; 4. The delivery of education services will only improve in the long- term if services become more equitable, efficient and accountable. 8. Educational outcomes are now also determined by deep-seated social and other external factors that cannot be easily changed by some short-term well-intended actions (e.g. gender parity index will not change quickly by carrying out sensitization workshops or even by distributing bicycles to girls to facilitate commute to schools). 9. Also, the source, scale, allocation and use of resources – physical, human and financial alike – have reached a level of complexity in which central planning and decision making can no longer determine how much of these resources allocated and used, when, where and by whom. Here, too, political, economic and social factors play roles in where families, students, teachers migrate and where resources move, accumulate or evaporate. 10. Consequently, sector performance and quality of education service delivery is driven by complex rules. These rules will require new roles from central and local decision makers including efficient instruments, new institutional capabilities to make sure that deep-rooted inequities are addressed, policies and resources are targeted, service delivery is efficient and those managing services and allocating resources are held accountable. 11. In the subsequent chapters we address the major challenges facing Ghana in the service delivery of not only but especially basic education. The report will follow the following sequence (i) progress in the outcomes during the last decade, with a particular focus on trends over the period 2001/02-2009/20101, (ii) resources allocated to the sector, (iii) performance achieved in improving equity, efficiency and accountability, and (iv) policy recommendations. A. Demographic and Social Context 12. Projections suggest that the population in Ghana will rise from 23.4 million people in 2008 to about 29.6 million by 2020, placing large demands on infrastructure and social services. Demographic trends indicate that over the last 50 years, between 1960 and 2008, Ghana’s population grew on average by 2.6 percent annually. Analysis 1 Upon the completion of the first draft of this report, the Ministry of Education (2010) released aggregate EMIS data for 2009/10. This data, which is predominately used here in the sections that focus on aggregate historical trends, shows notable variations on some indicators from previous Government performance reports and is to be read with caution. However, more comprehensive analysis using disaggregated data applies 2008/09 data throughout this study. 3 conducted by the United Nations Population Division (2009) estimate that Ghana’s population will exceed 45 million by 2050. In 2008, population growth was approximated at 2.1 percent. Key determinants of population growth in Ghana are fertility, mortality, migration and HIV trends – although the prevalence of HIV in Ghana is low compared to many sub-Saharan countries, affecting 1.9 percent of its population aged between 15 and 49. Ghana’s fertility rate was 4.3 in 2007. About 7.6 percent of all people living in Ghana are immigrants, some of who emigrated from conflict-ridden countries in the West African region. In 2007, the number of refugees residing in Ghana reached 35 thousand. 2008 was the first year in which the majority of the world’s population lived in urban areas. In Ghana, the share of the urban population is estimated at 51.5 percent for 2010, up from 23 percent in 1960 (UNDP 2009). 13. As in most low-income countries, children and youth represent a very large proportion of the total population in Ghana. About 40 percent of the total population was less than 15 years old in 2005. In that year, the median age among the general population was 19.8 years old. Given the current population structure and growth patterns, this figure is estimated to reach 21.5 years by 2015. However, Ghana’s development process is being accompanied by demographic transition, i.e. fertility rates and death rates are simultaneously falling, as trends since the second half of the last century indicate. 14. Ghana is a politically stable and growing low-income country. Committed to education, the government has been currently allocating about one-third of public expenditure to education, and it spends about one tenth of the GDP on the sector. Key political and economic issues are summarized in the Annex. 15. Prospects of achieving some of the key UN Millennium Development Goals (MDGs) appear positive. The 1st MDG, which aims primarily to halve 1990 levels of poverty by 2015, has arguably received the most international attention, technically, financially and politically. Given favorable political conditions and high economic growth rates, the number of Ghanaians living under the national poverty line sank from about half of the population to nearly a quarter, i.e. from 51.7 percent (1992) to 28.5 percent (2006). Thus, current trends in poverty reduction are encouraging and suggest that Ghana is on track to meet this target by 2015, if not before. The prevalence of malnutrition among children under 5 years in Ghana is about one in every seven children (13.9 percent) in 2008, lower than the sub-Saharan average of 26.5 percent and the global average of 23.1 percent. Table 1-1: Ghana’s Progress towards Selected Millennium Development Goals 1990 2008 Goal 1: Eradicate extreme poverty and hunger Poverty headcount ratio, national poverty line (% of population) 51.7 28.5 Poverty headcount ratio at $1.25 a day (PPP) (% of population) 49 30 Poverty gap at $1.25 a day (PPP) (%) 17 10 Income share held by lowest 20% 7.0 5.2 GDP per person employed (annual % growth) -1 4 Employment to population ratio, 15+, total (%) 69 65 Employment to population ratio, ages 15-24, total (%) 40 40 Prevalence of undernourishment (% of population) 34 9* Malnutrition prevalence, weight for age (% of children under 5) 24.1 13.9 Goal 2: Achieve universal primary education 4 Literacy rate, youth female (% of females ages 15-24) .. 76 Literacy rate, youth male (% of males ages 15-24) .. 80 Persistence to last grade of primary, total (% of cohort) .. 83 Primary completion rate, total (% of relevant age group) 61 78 Total enrollment, primary (% net) .. 73 Goal 3: Promote gender equality and empower women Proportion of seats held by women in national parliaments (%) .. 11 Ratio of female to male enrollments in tertiary education 30 54 Ratio of female to male primary enrollment 84 99 Ratio of female to male secondary enrollment 65 89 Ratio of female to male in primary and secondary education 78.5 95.2 Source: Based on World Bank, World Development Indicators database. Note: Figures in italics refer to data closest to the year specified; *refers to the year 2005. 16. Although the provision of basic infrastructure and the delivery of social services are expanding, large gaps in access continue to exist. For instance, one in every five Ghanaians lacked access to an improved water source in 2006. As a result, twenty percent of the population remains prone to higher incidences of illness and disease (e.g. typhoid, Buruli Ulcer and guinea worm infection). Higher disease prevalence is exacerbated by the fact that the share of the population having access to improved sanitation facilities was only 10 percent in 2006. Roughly 35 percent of Ghana’s population over 15 is illiterate and about one in every five children failed to complete the primary cycle in 2008, as data from the World Development Indicators database indicate. Thus, about 20 percent of children fail to acquire the basic competencies which help to lay the foundation for cognitive development, better economic prospects in life and increase human capital. Large segments of the population, particularly in rural areas, have restricted access to markets, a product of the fact that half of all Ghanaians did not have access to a mobile telephone in 2008 and only about 15 percent of all roads in Ghana were paved in 2005. Since 1991/92, when less than 30 percent of the population had access to electricity, accessibility to a power grid line has increased to about 50 percent by 2005/06 (GLSS). As a consequence, about half of the population lives without electricity, making it difficult for many children to do homework at night and, among other things, limiting the efficiency of some agricultural activities (e.g. water irrigation system). 17. The Human Development Index (HDI) highlights that Ghana has been progressing and in 2007 ranked 152 out of 182 countries analyzed, placing it thereby as a country with medium human development (UNDP 2009). Error! Reference source not found. illustrates that Ghana, with a human development index of 0.53, lies slightly below the average for the African continent of 0.55. It is important to note that Ghana’s human development may, in the long-term, be more sustainable than other higher performing, oil-producing countries such as Equatorial Guinea, Sudan and Angola. In particular, progress in the education sector offers positive signs for Ghana’s HDI performance in the short- and long-term. 5 Figure 1-1: African Comparison of Human Development Index, 2007 0.90 HIGH MEDIUM LOW 0.80 0.70 0.60 0.50 0.40 0.30 0.20 0.10 0.00 me NA e & oro ca tsw pt Su ya Ma go yc ibya Sie an ali Se itrea Ca HA a Le . Ca G uin ia ga a Cô B wi Pr cco Za oire s er Ma An nd Ta dan a-B que ng Ch i Dji nia tor Al on Co ongo Bu o (D ad rde Gu zam opia d' I n Ga ia Mo Eth ea So ami a Er ia Le outi Nig e Nig da Lib bia Ma lles pe ea Gu ria Ug tho Ke r o Rw egal Ga nda aF ) eria C pe az s uri n ut h bia rra Rep Bu au d a rkin RC tiu G ni da gol t e eni N an Bo Eg y on as Sw oro Ma roo r mb run nis fric M n M fri la sc To ila b ial ge in an e a i ta iss m Ve so inc uri v he b ine bi L A a nz i n Tu m Se .A Co om ua ntr oT Eq Ce Sa Source: Based on data from UNDP, 2009. 18. With the average person living 56.8 years, Ghana has made substantial progress in raising life expectancy from about 46 years in 1960 and surpassed the sub-Saharan average of 52 years in 2008. Improvements in the access and quality of education service, along with advancements in the health, water and sanitation as well as agriculture sectors, delivery have played a fundamental role in this progress. 19. The reduction of poverty levels remains a crucial challenge confronting the Government of Ghana. 51.7 percent of the population lived below the national poverty line in 1992, but by 2006 the share was nearly reduced in half to 28.5 percent. In contrast, using international norms, more than half of the population still lives in poverty and almost a third lives in extreme poverty. The most recent data indicate that 53.6 percent of Ghanaians live on less than US$2 a day and 30 percent live on less than US$1.25 a day (UNDP 2009). The highest levels of poverty are heavily concentrated in the three Northern regions of the country. In the Upper West region, the poverty headcount is such that about 4 out of 5 people (81 percent) live below the national poverty line, and in the Upper East region it is about 3 in every four (73 percent) and in the Northern region slightly more than half of the population (56 percent). In contrast, about 10 percent of the population in the Greater Accra region lives in conditions below the nationally established poverty line. 20. Higher incidences of poverty are not only strongly concentrated in the North, but also in most rural areas of the country, with 39.2 percent of the rural population and 10.8 percent of the urban population living below the national poverty line in 2006. Geographic as well as socioeconomic disparities pose central challenges in Ghana. In 2006, the poorest 20 percent of the population (lowest quintile) held 5.2 percent of the country’s income share, whereas the share of the richest quintile reached 48.6 percent. 21. Analysis results throughout this report suggest that the correlates of poverty in Ghana are closely related to low human capital indicators. Increasing the supply and demand of education service delivery among poor households could help to promote social cohesion and economic productivity. However, in the absence of some critical enabling factors in rural and Northern Ghana, making a greater impact in education service delivery in these areas will be more difficult. 6 22. Living standards in Ghana have risen both based on key human development indicators as well as on levels of income. Gross National Income (GNI) per capita has increased from US$360 ($470 using PPP) in 1980 to US$670 ($1,430 using PPP) by 2008. While Ghana’s GNI per capita is lower than the sub-Saharan average of US$1,082, Figure 1-2 illustrates that the level of welfare of the average Ghanaian still surpasses a number of its sub-Saharan neighbors. Between 1961 and 2008, GDP growth averaged 3.0 percent annually. However, averages are known to hide large economic fluctuations. Some years the economy retracted such as in 1975, when GDP shrank by 12.4 percent, while in other years the total market value of all final goods and services produced in Ghana grew substantially such as in 1970 by 9.7 percent. Since 2000 the gross domestic product has grown on average by 5.3 percent, reaching 6.2 percent in 2008. Figure 1-2: Sub-Saharan Comparison of GNI per capita and GDP Growth, 2008 GNI per capita, PPP (current international $) GDP growth (annual %) 2000 12 10 GNI per capita 1600 GDP growth 8 1200 6 800 4 2 400 0 0 -2 C en i pe nz e ca L e as l i M pi a G bi a Ke i re C o ui E g er an Su a bi NA Iv l ep e N c U had a s ria Li d i a G GH in Za nia Bu au Bu Com ea w r l A Si e mb o P n B a N ho oz To i Be a e ga Et da ag a aw Ta , Th R ca fri rra i qu li in ro R n s i nd g d da a - re ny er n n ad M ub S c o Fa n eo be n t iss o m ot e am A a an rk o i ru so ri n ne rit ig ui al ga hi a d' Le a M M e G m To tra o Sa en C Source: Based on data from the World Development Indicators database. B. Historical Evolution of Education 23. Ghana has been considered to have had a solid and respected tradition in its education system in Africa and beyond. Since the 19th Century, schools had been set up and run by local communities and the churches, rather than by colonial powers (Foster 1965). A key driver in the expansion of education was the Christian missionary work, considering the widening of educational access as a prerequisite for missionary activities. By 1881, approximately 5000 children attended one of the 139 schools. The first plan to universalize primary schooling goes back to 1945, the year when the colonial administration put forth a proposal of a 10-year strategy to attain universalization within a 25 year period on the basis of modest financial projections (Akyeampong 2007). But the history of free and compulsory basic education only began in 1951. As a direct consequence, primary enrollment in the sector jumped by more than sevenfold, growing from 153,360 (1951) to 1,137,495 (1966) primary students over a 15 year time span (Hayford 1988:35). During this period of rapid expansion, the lesson was that widening access and improving quality are to be carried out in unity in order to reap the benefits of education (Akyeampong 2007), still a very valuable imperative in Ghana today. 24. In 1952, five years before full political independence of the country, the Ghanaian government, by then exercising limited self rule under the British, announced the introduction of fee-free primary education under an Accelerated Development Plan for Education. At the time Ghana obtained independence in 1957, it was repeatedly viewed as a country having a viable educational sector, especially in a regional comparison. In 7 that era, international institutions such as the World Bank (2004a) and UNESCO commonly held Ghana’s education system to be the best in Africa. The 1961 Education Act was introduced to reinforce the commitment to free primary education. However, in the late 1970’s, the Ghanaian economy and consequently the education system retracted due to institutional mismanagement and poor economic policy (Akyeampong 2006). Between 1976 and 1983, the proportion of GDP allocated to the education sector sank from 6.4 percent to about 1.5 percent (World Bank 2004a). As a result of scarce financial resources, textbooks and other school materials became in short supply, school infrastructure deteriorated and the necessary data for strategy planning and policy purposes were not compiled anymore (Akyeampong 2006). In the years following 1975, when over 2.3 million children attended primary schooling, sector-wide deterioration led to a decline in over one million enrolled children by the early eighties (World Bank 2004a). Thousands of Ghanaian teachers were lured to Nigeria by more attractive wages.(Akyeampong 2007). 25. With support from the World Bank (WB) and International Monetary Fund (IMF), the Government of Ghana (GoG) in 1983 launched the Economic Reform Program (ERP). Soon thereafter, in 1987, the Government launched a major reform of the education system, again with heavy involvement of the WB. The 1987 education reforms were intended to improve access, raise quality, increase vocational content for secondary curriculum, lower costs and increase cost recovery for secondary and tertiary education. 26. In the 1990s, in response to the Jomtien Declaration, GoG launched a new program focusing primarily on access, the Free, Compulsory and Universal Basic Education (FCUBE) program. Over the 15-year period from 1987 to 2002, the WB and other donors provided close to US$ 600 million in soft loan and grant financing to support the series of reform programs in education. In the mid-1990s, IMF and WB designated Ghana as a Heavily Indebted Poor Country (HIPC), one of 41 countries worldwide in need of special assistance to foster economic growth and reduce poverty by means of channeling debt repayment to key development sectors including education and health. To qualify for assistance, each HIPC country needed to prepare a Poverty Reduction Strategy Paper (PRSP). Ghana submitted its “Growth and Poverty Reduction Strategy� (GPRS), updated four years later as GPRS-II, and began receiving HIPC funding in 2003, through a series of one-year projects (Poverty Reduction Strategy Credits, or PRSCs) from WB. Starting in 2007, HIPC funding is being phased out, and a new “enhanced� modality for debt relief, the Multilateral Debt Relief Initiative (MDRI), is gradually replacing HIPC funding. 27. As will be seen later, a key factor in education policy as well as in the evaluations of the sector’s performance is the supply of qualified teachers. In the little over twenty years since 1987, this supply has showed two different storylines. While between 1987 and 1997 the pupil to trained teacher ratio, the number of students per teacher with formal qualifications, gradually decreased by about 20 percentage points for primary as well as for junior secondary (called junior high school since 2007) levels, since then it has lost more than what it gained, in fact, it lost over thirty percentage points at primary levels. The underlying logic is that decent economic growth and a reduction of the poverty rate lured a large number of children into the school system, increasing admission rates 8 sharply. The supply of qualified teachers could not keep up with the growth in student numbers and the Government substituted unqualified teachers for qualified ones. As it will be seen later, while this substitution maintained a relatively solid trend of improved access, it negatively affected other performance indicators and increased disparities in inputs (geographical, social and gender) and in educational outcomes. Figure 1-3: Trends in Share of Trained Teachers in Primary and JHS, 1987-2008 100 90 87 84 80 83 79 80 79 79 76 72 75 74 77 80 73 74 75 73 70 72 76 77 71 71 68 66 69 69 69 68 65 65 64 71 61 65 63 61 58 60 57 60 65 56 50 53 49 48 40 30 Primary 20 JSS 10 0 87' 88' 89' 90' 91' 92' 93' 94' 95' 96' 97' 98' 99' 00' 01' 02' 03' 04' 05' 06' 07' 08' Source: Based on RECOUP; EMIS. Note: The indicated year reflects the beginning of the academic year. 28. At the same time, the eleven major donors that support education have been trying to move away from the traditional modality of project support and to assist development in the country through the provision of budget support within the framework of a rolling Medium Term Expenditure Framework (MTEF). Ghana’s MTEF is updated annually to cover the next three-year period. Through an instrument known as the Multi-Donor Budget Support (MDBS), participating donors release funds when satisfied that certain key development indicators (“triggers�) have been met. Budget support of this kind is intended to lower transaction costs, by eliminating the need for multiple reporting requirements by the government and duplicative supervision activities by the donors.2 29. The other new development of particular relevance for Ghana’s education sector has been the Catalytic Fund of the EFA Fast Track Initiative (FTI), discussed above. To qualify for FTI funds, a country has to prepare a credible strategy for achieving EFA goals by 2015 and demonstrate the existence of a funding gap. Ghana prepared its Education Strategic Plan (ESP) in 2003. The local donors endorsed the plan and made a recommendation to the FTI partners for funding. Ghana was given an initial grant of US$ 33.2 million, disbursement of which began in 2005 and by the middle of 2008 had reached a cumulative total of US$ 19 million. FTI funding is supposed to be flexible – 2 HIPC, MDRI and MDBS funding have impacted [possible to indicate in a phrase HOW this funding has impacted the sector? E.g. by providing an incentive for GoG policy makers to do more to meet donor priorities in the sector?] on Ghana’s education sector, given that about 20 percent of the external funds flowing to the Government via these mechanisms are notionally earmarked for education. Of course, in a strict sense, “budget support� comprises money paid into the general revenues of MoFE P and cannot, by definition, be earmarked, but the real world is generally more complicated. In the case of DFID’s commitments to MDBS, these flows are, in fact, linked to expenditures on education, and while it is true that money in the overall budget is fungible, DFID’s MDBS disbursements could be counted as part of education’s total allocation from the budget. When one of the 2007 MDBS triggers (primary education’s share of the education budget) had not been met, DFID temporarily suspended its disbursements to GoG through MDBS. 9 able to respond to demonstrated need. If Ghana can demonstrate effective use of the funds, additional funding might be made available. 30. As part of the ESP, Ghana established the National Education Sector Annual Review (NESAR), a regularly scheduled meeting to review performance in the sector. The Department of Planning, Budgeting, Monitoring and Evaluation (PBME) in MoE compiles an annual Education Sector Performance Report (ESPR) and presents this at the review, which is held in about June of each year, to other staff in the Ministry and the Ghana Education Service (GES), as well as to education donors and other stakeholders, for discussion and deciding on priorities for the coming year. 31. In a parallel track, largely independent of the ESP (which is arguably a donor- driven document), in 2001 the new Government (after the elections in 2000) set up an ad hoc Education Reform Review Committee (ERRC) to re-examine the “goals and philosophy� of Ghana’s education system. In late 2002, the Committee released its report, “Meeting the Challenges of Education in the Twenty-First Century.� This led to the issuance of the “White Paper of the Report of the Education Reform Review Committee� in 2004, to the establishment of the National Education Reform Implementation Committee (NERIC) in early 2007, and to the launch of the National Education Reform Program (NERP) later in 2007. The parallel development of the ESP 2003 and the NERP was finally brought into a coherent track through the 2009 ESP that defined a single financing plan for 2010-2020. C. Organization, Structure and Normative Framework 32. The 1992 Constitution of Ghana ensures comprehensive provision to education. The right of every Ghanaian to education is laid out in Article 25 of the Constitution: “Basic education shall be free, compulsory and available to all�. However, legislation needs to be complemented by sustainable and well targeted policies, such as the capitation grant paid to schools, school feeding program, provision of textbooks and uniforms, incentives for improved teaching, management and supervision, salary supplements for teachers and cash transfers to the needy, all of which need to be targeted, monitored and adjusted as necessary. There is no enforcement of Article 25; when deciding whether or not to send children to school, families take into account indirect and opportunity costs such as registration fees, transport expenses and school uniforms, which are a greater burden for the poor and for the extremely poor in rural, remote locations. In addition, legislation mandates that all forms of secondary education are to be made generally available and accessible to all Ghanaians, especially through the progressive introduction of free secondary education. It is stipulated that higher education is to be made equally accessible to all Ghanaians based on capacity, and its provision without fees should be progressively promoted. As regards to functional literacy, the Constitution specifies that it should be ‘encouraged or intensified’ to the extent this is possible. 10 Table 1-2: Structure of Ghana’s Education System Years Appropriate Level of Academic Focus Age Education 1 4-5 Kindergarten 2 5-6 (free, compulsory and universal) 3 6-7 4 7-8 Basic Education 5 8-9 Basic Literacy, Numeracy, Science Primary 6 9-10 and Social Studies 7 10-11 8 11-12 9 12-13 English, Math, Social Studies, 10 13-14 Junior High Science, 11 14-15 School Technical, Vocational, Agricultural Courses 12 15-16 Specialization in Grammar, Secondary 13 16-17 Senior High Technical, Upper 14 17-18 School Agricultural, Vocational, Apprenticeship Program 15 18-19 University, Polytechnic, Technical 16 19-20 University, Professional Institutes, Tertiary 17 20-21 College of 18 21-22 Education, Teacher Education Source: Based on the content of this report. 33. Structure of the education sector. The structure of the pre-tertiary education system of 6,4,7 had been replaced by the 1986 education reform program with a 6,3,3 system, in turn reducing pre-tertiary education from 17 to 12 years (World Bank 2004a). By reducing the duration of pre-tertiary education, these policy reforms led to improved transition rates and school efficiency. By 2002, the Government adopted a new system that included two years of pre-primary schooling. As of 2009, the formal school system in Ghana follows a 2-6-3-3-4 structure: 2 years of kindergarten, 6 years of primary, 3 years of junior high school, 3 years of senior high school and typically 4 years of university level education. The decision to transfer to a four year senior high system was taken in September 2008. As a result, the new fourth year of SHS was officially planned to begin in September 2011. However, as of August 2009 the Ministry of Education made public that they had plans to readopt the three year SHS system. This planned extension to a four year system would have caused a roughly one-third increase in enrollment in senior high school, needing the rise in the recurrent costs of about 30 percent and further pressures on the already constrained SHS infrastructure. 11 2. EDUCATIONAL OUTCOMES: ACCESS, LEARNING AND RELEVANCE 3 A. Coverage and Access at Various Levels 34. According to the Ghana Demographic and Health Survey (2009), the average Ghanaian female has completed 7.2 years of education, while males completed 8.5 years, in rural areas females completed 5.1 years and men 7.4 years, in urban areas 8.5 years for women and 9.3 years for men. Education attainment for urban men is almost fifty percent more than for rural women. In the Northern region, the education attainment of men is 0.9 years and for women is as low as 0.1 years. Figure 2-1: Median Years of Education Completed in 2008, by Female/Male 12 Female 10 Male 8 6 4 2 0 Vo a Ea lt a l H rth Fo le R n U ort o A i U er rn M nd 15 4 20 9 25 4 30 9 35 4 40 9 45 4 9 t G Ce rn er st Br As rn .A l al Se est ta st g nt es at tra r N haf a -2 -1 -2 -2 -3 -3 -4 -4 cc d pp Ea ur he pp he te To e on ha rb u co id W 15 re n w st es U ig Lo W Age Residence Region Wealth Quintile Total Source: Based on data from Ghana Demographic and Health Survey, 2009. Note: Figures are based on weights applied by DHS. 35. In the meantime, according to the same survey, 21.2 percent of the female population and 13.3 percent of the male population have never received education, 30.8 percent of women and 19.9 percent in rural areas and 10.9 percent among women and 5.6 percent among men in urban areas. The share was the highest in the Northern region, with 65.7 percent of women and 48.4 percent of men with no education; as it is expected, the share of those without education is the highest at 60 percent among women in the bottom quintile. 3 If any data is used throughout this report without citation it is administrative data from either EMIS or ESPR, unless otherwise indicated. National sources of data are derived from: • Education Management Information System (EMIS); • Education Sector Performance Reports (ESPR), adjusted data; • District education expenditure data; • Payroll data; • GLSS; • Demographic and Health Survey. Any adjustments to administrative data such as in the Education Sector Performance Reports are made by the MOE, which for instance give non-responding schools district averages. Other sources of data are referenced respectively. 12 Figure 2-2: Share of the Population without Education in 2008, by Female/Male 70 Female 60 Male 50 40 30 20 10 0 Vo a Ea lt a l H rth Fo le R n A i U rt o U er rn t M nd 15 4 20 9 25 4 30 9 35 4 40 9 45 4 9 e r st G Ce rn Br As rn .A l al Se est ta st g nt es at tra r N haf a -2 -1 -2 -2 -3 -3 -4 -4 cc d pp Ea ur he pp he te To e on ha rb u co id W 15 re n w st es U ig Lo o W Age Residence Region Wealth Quintile Total Source: Data based on Ghana Demographic and Health Survey, 2009. 36. Even though gross enrollment rates in primary are relatively high in Ghana (94.9 percent in 2008/09), the prospects of children in the poorest quintile never having gone to school are about 10 times greater than those of children in the richest quintile. If a child lives in a rural area, its chances of not having attended school are over three times higher than an urban child. A rural girl from the poorest quintile is 13.9 times more likely to not have attended school than an urban boy from the richest quintile. As enrollment rates increased rapidly over the last decade, educational disparities have decreased during this period, strongest in terms of gender inequalities, followed by North-South differences and then those related to welfare levels, and trends indicate that disparities are likely to continue decreasing. Figure 2-3: Share of children who have never attended school, by background characteristics 40 37.8 35 30 25 19.7 20 14.9 13.6 15 10 6.3 3.9 5 0 Female Male Rural Urban Poorest Richest quintile quintile Source: Based on data from Ghana Demographic and Health Survey, 2003. 37. Household and school survey data, shown in Table 2-1, support the view that a large portion of parents (45.9 percent) and head-teachers (78.2 percent) believe that the (in-)direct and opportunity costs of schooling are the central cause of non-attendance in school. While nearly half of parents indicate that economic factors influence the decision to not send their children to school, about a quarter (22.4 percent) state socio-cultural factors, i.e. they do not value the importance of education especially for girls. About 7 percent of parents point to geographic factors associated with school distance and to low 13 school quality. Yet nearly all parents who point to costs as the main hurdle to attendance live in remote and difficult-to-reach areas, underscoring geography as a principal determinant. Data on parents, who were asked why their children (between ages 6 and 21) stopped attending school, illustrated that: 37.1 percent stated because of not doing well at school, 35 percent cannot afford school and/or their children need to work at home, 4.1 percent do not value education, for 3.1 percent of parents school was too far or the quality too low, and the remaining 20.6 percent indicated ‘other’ reason (World Bank 2004a). Although policies such as the capitation grant4 have been adopted since this data was collected, it is likely that these causes still largely reflect the barriers facing the remaining out-of-school population. Table 2-1: Reasons for children not attending and stopping to attend school, percent Reason for children’s Reason for children non-attendance in school stopping to attend school Primary Households Households school head- (parents)b (parents)b teachers Parents cannot afford/children need to work at home 78.2 45.9 35.0 Parents do not value education 11.3 22.4 4.1 School too far/low quality 2.1 7.1 3.1 Other 8.4 24.5 20.6 Not doing well at school a a 37.1 Total 100 100 100 No. of surveyed 417 289 97 Source: Based on data from World Bank, 2004a. Note: a. The question ‘not doing well at school’ was not included in the reasons for children’s non-attendance in school, therefore percentages still total 100 percent. b. Only children aged between 6 and 21. 38. The prototypical, most disadvantaged child would be a girl coming from a very poor household living in a rural area of the Northern region. In terms of enrollment and completion, throughout Ghana’s pre-tertiary system geographical inequalities tend to be the most important barriers, followed by but closely associated with socioeconomic disparities. Poor students, especially at higher levels of basic schooling, have much fewer education opportunities. In districts falling into the richest quintile, net enrollment in JHS (i.e. the ratio of correct-age enrollment to the school age population) averaged 60.5 percent, whereas at the other end of the socioeconomic spectrum, districts in the poorest quintile had an average junior high NER of 30.9 percent. Table 2-2: Pre-tertiary GER, NER and Completion, by Gender, Region and Quintile GER NER Completion Rate Primary Primary Primary SHS SHS SHS JHS JHS JHS Percent Indicator National 94.9 80.6 33.9 88.5 47.8 17.7 86.3 75.0 By gender Boys 97.0 83.9 36.7 89.6 47.9 18.0 89.3 79.7 Girls 92.8 77.0 30.8 87.4 47.7 17.4 83.2 70.1 Parity Index (girls/boys) .96 .92 .84 .98 1.00 .97 .93 .88 4 The capitation grant is analyzed in more details later in the report. 14 By region Ashanti 96.0 84.0 41.8 90.0 43.0 22.6 86.0 84.0 Brong Ahafo 99.0 78.0 33.1 92.0 41.0 15.2 85.0 73.0 Central 1.07 96.0 50.1 96.0 70.0 29.5 99.0 93.0 Eastern 95.0 78.0 47.7 86.0 44.0 26.3 83.0 75.0 Greater Accra 1.05 90.0 21.1 86.0 64.0 12.9 97.0 81.0 Northern 92.0 63.0 25.5 81.0 31.0 10.5 79.0 56.0 Upper East 95.0 69.0 20.3 81.0 27.0 8.3 83.0 62.0 Upper West 99.0 75.0 25.7 79.0 35.0 7.8 84.0 60.0 Volta 93.0 75.0 35.1 86.0 42.0 15.7 83.0 70.0 Western 99.0 84.0 27.4 93.0 47.0 15.2 92.0 75.0 Parity index (North./Central) .86 .66 .51 .84 .44 .36 .80 .60 By level of welfare quintile Q1 (poorest) 95.0 67.4 80.9 30.9 81.3 58.5 Q2 94.0 70.9 85.5 36.7 79.7 66.4 Q3 99.1 80.1 92.7 43.7 88.1 76.8 Q4 94.7 84.2 88.5 46.4 86.3 81.2 Q5 (richest) 103.5 91.1 90.9 60.5 95.7 84.4 Parity index (poorest/richest) .92 .74 .89 .51 .85 .69 Source: Calculations based on EMIS 2008/09 and GLSS 2005/06. 39. Ghana ranked 102 out of 128 countries analyzed in 2007, which was classified as low in UNESCO’s (2010) EFA Development Index (EDI). The EDI is calculated using the arithmetic mean of its four components: primary adjusted net enrollment rate (73.3 percent), adult literacy rate (65 percent), gender-specific EFA Index (89.6 percent) and survival rate to grade five (88.6 percent). These results confirm that although much progress has been achieved in the education sector over the past years, Ghana is still lagging behind internationally. 40. In the following section, we will analyze changes in educational access, transition and completion, by education levels. Access to Kindergarten 41. The Government introduced two years of kindergarten, also known as preschool, on a voluntary basis in 2002. Since 2007, formal basic education encompasses an 11-year cycle, two years of kindergarten, followed by six years of primary, then three years of junior high school (GOG 2004). Nationally, gross enrollment in kindergarten has expanded enormously. Ghana provides now higher pre-primary coverage than the vast majority of sub-Saharan countries and even several high-income Arab states (UNESCO 2008). Since 2002/03, when only 21.8 percent of children had access to kindergarten schooling, the gross enrollment rate increased to 97.3 percent by the school year 2009/10, translating into an annual 10.8 percent growth rate over this period. Yet the net enrollment rate remains troubling low; only 63.6 percent of children of the official age group (i.e. 4 and 5 year olds) had access to kindergarten schooling in 2008/09. 15 Table 2-3: Trends in Kindergarten Enrollment5 2001/02* 2002/03 2003/04 2004/05 2005/06 2006/07 2007/08 2008/09 2009/10 Total 702,304 769,199 687,643 778,109 1,065,963 1,142,784 1,262,264 1,338,454 1,440,732 GER 21.8% 54.6% 60.1% 85.3% 89.0% 89.9% 92.9% 97.3% NER 19.0% 34.4% 38.5% 50.0% 55.8% 62.6% 63.6% Boys 355,385 389,478 320,939 369,019 498,524 552,995 632,421 671,822 Girls 346,919 379,721 316,176 362,312 500,295 551,784 626,062 666,632 Source: EMIS. Note: *GER and NER data are not available for 2001/02. 42. The share of private sector enrollments has declined over the past years, mostly in parallel with the implementation of the policy which made KG part of compulsory basic education. However, this trend indicates that the twofold expansion of the public enrollments was achieved through some private kindergartens becoming public, which made them eligible for the capitation grants, which were introduced in 2005. Many private programs were simply attached to public schools. However, these newly public KG programs are seriously short of publicly paid teachers leading to large disparities in pupil teacher ratios. Figure 2-4: Trends in Kindergarten Enrollment, Private/Public 1,200,000 1078973 Public Private 1016606 1,000,000 896522 807369 800,000 600,000 516734 457597 453561 482251 400,000 244,707 252,084 249,080 241,877 259,481 183,554 191,450 208,257 200,000 0 2001/02 2002/03 2003/04 2004/05 2005/06 2006/07 2007/08 2008/09 Source: EMIS. Access to Primary Education 43. Between 2001/02 and 2009/10 enrollments in primary education experienced sharp growth by about 1,284,673 or 50.9 percent. Yet in spite of this surge in enrollments, a considerable portion of children don’t attend school, repeat a grade, drop out before completing or simply don’t make it to junior high schools. 5 Depending on the data source and methodology used, respective data values may vary. In this chapter on access, EMIS census data is used over the years 2001/02-2002-3, for which no adjustments are made. EMIS data for 2003/04-2009/10 are derived from administrative performance reports, which are adjusted such that non-reporting schools receive the district average, which could lead to possible overestimates. It is also important to note that there is the likelihood that some schools report higher GERs than actual values as the amount of capitation grant funding received, which is per capita based, is higher for schools with higher reported gross enrollment rates. 16 Table 2-4: Trends in Primary Education Enrollment, public and private, boys and girls 2001/02 2002/03 2003/04 2004/05 2005/06 2006/07 2007/08 2008/09 2009/10 Total 2,586,434 2,524,585 2,686,133 2,929,536 3,122,903 3,365,762 3,616,023 3,710,647 3,809,258 Public 2,113,749 2,121,740 2,196,774 2,328,324 2,647,616 2,824,407 2,990,773 3,041,895 3,099,234 Private 472,685 402,497 489,359 601,212 475,287 541,355 625,250 668,752 710,024 % of Private 22.4% 19.0% 22.3% 25.8% 18.0% 19.2% 20.9% 22.0% Boys 1,359,150 1,323,518 1,403,913 1,525,548 1,606,178 1,732,162 1,860,289 1,908,232 Girls 1,227,284 1,201,067 1,282,220 1,403,988 1,516,725 1,633,600 1,755,734 1,802,415 % of Girls 47.45% 47.57% 47.73% 47.93% 48.57% 48.54% 48.55% 48.57% 48.7% % of Girls in 48.70% 48.70% 48.80% 48.70% 49.40% 49.30% 49.50% 49.50% Private Source: EMIS. 44. Ghana has a consistently high share of private enrollment in primary education. It is estimated that in sub-Saharan Africa eight percent of all primary students are enlisted in private schools, compared to about 22 percent in Ghana. The share of the private sector actually grew to 25 percent by 2004/5, after which private provision dropped to 18 percent from which it gradually increased to the present level of 22 percent. A study in the Ga district found that school fees on a monthly basis ranged on average from about US$4 in unrecognized private schools to about US$7 in recognized schools for the early primary grades (Tooley 2005). The private sector’s expanding coverage area in rural areas has, as Akyeampong (2008) notes, taken place largely in areas where the majority are not employed as subsistence farmers who tend to be poorer, but rather employed in the fishing and trading industries. Table 2-5: Trends in Primary Enrollment, Private/Public 4,000,000 3,500,000 Public Private 625,250 668,752 3,000,000 541,355 475,287 601,212 2,500,000 472,685 402,497 489,359 2,000,000 2,990,773 1,500,000 2,647,616 3,041,895 2,113,749 2,196,774 2,824,407 2,121,740 2,328,324 1,000,000 500,000 0 2001/02 2002/03 2003/04 2004/05 2005/06 2006/07 2007/08 2008/09 Source: Based on EMIS data. Grade specific enrollment trends 45. The following figure also shows changes in enrollment in primary education. The trend shows sharp increase for first grade in 2005 at the introduction of the capitation grant, a contraction in 2007 at the same grade at a time when inflation started to depreciate the grant value. The introduction of the capitation grant, in the meantime, had no impact for later grades and even negative ones for the 6th grade. Transition throughout primary levels and ultimately completion of the primary cycle has not improved over time and has even gotten worse since the introduction of the capitation grant, suggesting the grants have not been large enough to cover the costs of educating ever-growing numbers of pupils, not to mention to improve quality and implement measures (like smaller classes, tutoring, etc) to improve transition and completion. 17 Figure 2-5: Evolution of Primary Enrollment 1980-2008, by Grade 800000 P1 700000 P2 600000 enrolment 500000 P3 400000 P4 300000 P5 200000 P6 100000 0 1994 1980 1982 1984 1986 1988 1990 1992 1996 1998 2000 2002 2004 2006 2008 Year Source: Akyeampong 2009 based on EMIS. 46. Enrollment and Admission Ratios: Between 2001/02 and 2009/10, gross enrollment grew from 80 percent to about 95 percent. During the same period net enrollment progressed at an average rate of 3.1 percent per year, increasing from 59 percent to 83.6 percent. Table 2-6: Trends in Primary Enrollment, Admission and Gender Parity Index 2001/02 2002/03 2003/04 2004/05 2005/06 2006/07 2007/08 2008/09 2009/10 GER 80% 75.7% 86.50% 87.50% 92.10% 93.70% 95.20% 94.9% 94.9% NER 59% 55.9% 55.6% 59.10% 69.20% 81.10% 83.40% 88.5% 83.6% GAR 89.4% 92.3% 103.2% 107.3% 102.9% 101.3% NAR 26.2% 61.7% 69.1% 74.3% 72.1% 71.1% GPI 0.93 0.93 0.93 0.96 0.96 0.96 0.96 Source: EMIS. Box 2-1: Abolition of school fees compensated by capitation grants The impact of the ‘Free Compulsory Universal Basic Education’ (FCUBE) reforms adopted in 1996 was mitigated by the reduction of public funding as schools began levying indirect school fees, i.e. registration fees, uniforms, textbooks etc. The government adopted capitation grants in the academic year in 2004 in 40 deprived districts and scaled up to all regions in 2005. The capitation grant had a significant effect on first year admissions (see table above on GAR for 2006). However, this increase was almost fully counterbalanced by increasing dropouts and limits in learning outcomes (to be discussed later). Altogether, the effect of capitation grant on net enrollment was an increase of slightly more than 2.2 percent, but the effect on deprived districts was not significant given the high level dropout and prevalence of overage enrollment. 18 Figure 2-5: Capitation grant’s impact on trends in primary net enrollment Source: UNESCO 2008 based on Fredriksen 2007; Gershberg and Maikish 2008. 47. A Gross admission rate greater than 100 percent (that is, admission into primary 1), together with a relatively low net admission rate underscores a widespread problem of late entry: about 30% of P1 enrollment is made of children older than the appropriate age category. As late as in 2004/05, overage children represented more than 70 percent of the first year admissions. Figure 2-6: Trends in Primary GER and NER 100% 92% 94% 95% 95% 87% 88% 89% 80% 80% 81% 83% 76% 69% 60% 59% 59% 56% 56% 40% GER 20% NER 0% 2001/02 2002/03 2003/04 2004/05 2005/06 2006/07 2007/08 2008/09 Source: Based on EMIS data. 48. Official gender parity index has been relatively high and improved to .96 by the end of the decade. The percentage of girls enrolled has always been high and is even higher in private schools. Only a select few sub-Saharan countries have reached or nearly reached gender parity in primary education, among them are Ghana, Namibia, Tanzania and Botswana (UNESCO 2007). 19 Figure 2-7: Country Comparison in Primary Net Enrollment and Gender Parity, 2007 Note: Gender parity in primary is measured using gross enrollment ratios. Source: UNESCO 2010:65. 49. However, these indicators do not show some key factors leading to lower girls participation in junior high school education (to be discussed below), including the high initial admission but higher dropout and lower survival rate up to P6. Higher incidences of girls’ non-attendance in schooling tend to be concentrated within rural and economically deprived areas of the country, a phenomenon strongly affecting the North. Indicators of internal efficiency: completion, dropout, repetition and the key causes 50. Completion rate in Ghana is relatively high but not high enough for achieving the MDG goals. School census data on the completion rate for the latest year (2009/10) indicate that it reached 87.1 percent, which represents a 9.2 percent increase from 77.9 percent in 2003/04.6 Different methodologies also show differing results. According to 6 This report uses the methodology set by the Ministry of Education for calculating the completion rate based on enrollment into the 6th grade divided by the 11 year old population. Thus, students who repeat are not calculated and the share of completers would drop by several percentage points. [UNCLEAR] 20 the WDI, 77.7 percent of students completed primary education. Although this is high in SSA, it still doesn’t make achieving the MDG goal likely. Figure 2-8: Sub-Saharan Comparison of Primary Completion Rate, 2008 120.0 111.7 100.0 77.7 75.6 80.0 69.0 54.7 60.0 37.1 40.0 30.5 20.0 0.0 pe e so A ic ria ia Th AN l an ci ub Fa be ri n a, nz H ep Li a P bi G Ta in R am d rk an an Bu G c e fri m lA To tra o Sa en C Source: Based on World Bank, World Development Indicators database. Note: Completion rate is based on the share of the relevant age group. Table 2-7: Primary Education, Completion, Repetition and Dropout 2001/02 2002/03 2003/04 2004/05 2005/06 2006/07 2007/08 2008/09 2009/10 Total 2,586,434 2,524,585 2,686,133 2,929,536 3,122,903 3,365,762 3,616,023 3,710,647 3,809,258 Completers 998,849 2,305,545 2,360,915 2,695,975 3,091,699 Comp. rate 77.9 78.7 75.6 85.4 88.0 86.3 87.1 Comp. rate girls 74.0 75.1 72.4 79.6 82.4 85.5 84.3 Repeaters 177,330 148,518 152,726 170,777 135,288 197,520 243,625 139,820 Rep. Rate 6.9 5.9 5.7 5.8 4.3 5.9 6.7 3.8 Leavers 130,645 453,214 626,700 472,267 280,699 Dropout Rate 4.9 15.5 20.1 14.0 7.8 Source: EMIS. 51. There has also been significant disparity between boys and girls completion rate. We note that the radical improvement reflected in girls’ completion rate in 2008/09 is statistically improbable; this level of improvement would have required more girls to complete P6 in 2008/09 than the number of girls enrolled in P5 in the previous year. By 2009/10, the share was 84.3 percent. 52. Completion rate, according to the EMIS data, would have to grow by 2.3 percent annually to reach UPC by 2015. Historical evidence has shown that linear growth projections may be unrealistic. As UPC nears, improvements in completion are more and more difficult. Children living in extreme poverty and in geographically isolated settings tend to drop out first as they expected to contribute to household income generation, girls taking care of younger siblings, and children with special needs. 21 Table 2-8: Progress towards education related MDGs Predicted Year for 2015 2003/04 2008/09 Low High Goal 2: Universal Primary Education Net enrollment ratio in primary 59.0a 88.5 100 100 Primary completion rate, total 77.9 86.3 91.3 96.4 Primary completion rate, girls 74.0 85.5 92.4 99.3 Adult literacy rate 53.4 62.0c 69.5 77.1 Goal 3: Promote gender equality and empower women Ratio of girls to boys in primary 93.0 96.0 97.8 99.6 Ratio of girls to boys in lower secondary 88.0 92.0 94.4 96.8 Ratio of girls to boys in upper secondary 76.0b 80.0 84.0 88.0 Ratio of girls to boys in tertiary 47.0 49.0c 50.8 52.5 Note: Predictions for 2015 are based on simple extrapolation, assuming growth rates over the period are sustained for high scenario and that it is halved for low scenario. a. Indicate data for 2001/02; b. for 2005/06; c. for 2007/08. Source: Based on EMIS data. 53. In 2005/06, dropout in primary education rose to 20.1 percent as the new capitation grants allowed schools to accept more students, many of them older and less prepared to study. 54. Some of the underlying factors were overcrowded classrooms and more pupils per trained teacher, mostly an impact of the capitation grant being adopted in 2005 and the subsequent explosive jump in enrollment numbers. By 2007/08, the dropout rate in primary had reduced to about 7.8 percent. Among the 6-11 year old children who dropped out of primary, more than 30 percent of these dropouts did so after they completed the 1st grade, as data from the Core Welfare Indicators Questionnaire indicates (GSS 2003). While enrollments have been consistently on the rise, a real challenge seems to be in reducing early drop out, but also overage enrollments and repeaters. To ensure that fewer children abandon the school system, it is pivotal to get a student into primary school (but also higher levels) at the appropriate age, since being overage places incremental pressure on the student to begin working and generating income. Data collected from household surveys show that in Ghana more than 60 percent of children in primary are older than the expected age for their given grade (UNESCO, 2008; GDHS, 2003). Girls appear to be dropping out more than boys in deprived districts relative to non deprived. 55. What are the determinants of dropout? At the primary level, those who drop out tend to come disproportionately from rural settings and poorer households, and be girls. As survey data suggest, how well students perform in class, is the most influential ingredient in completing school. As we will see in more detail throughout this report, various interlocking risk factors influence the decision to drop out, including late enrollment, poor teaching and learning outcomes, past short-term withdrawals, repetition, low attendance, poor school facilities and infrastructure, overcrowded classrooms, early entry into the labor force, illness and malnutrition, gender, geographic remoteness and economic deprivation. It is therefore necessary to address both school-specific aspects 22 and household and community related issues, in order to reduce drop out but also high absenteeism and repetition and to tackle other inefficiencies, topics we turn to later. 56. Patterns among primary repeaters. Over the past seven years, the primary repetition rate has fluctuated between 6.9 percent (2001/02) and 3.8 percent (2008/09), reaching its lowest point in the last year for which data is available. However, data reliability is questionable and clear trends are difficult to decipher. For P1 alone, EMIS data indicate as high as 20 percent for both repetition and dropout rates over the past few years. 57. Being overage (older than the age corresponding to the grade level) is a key cause of repetition and dropout. Figure 2-9 shows that the age-grade profile in primary is considerably wide – especially between 3rd and 6th grade. In 2008/09, a large majority of students in primary 6, about 84 percent, were older than the appropriate age: about 60 percent were 12 to 14 years old (i.e. JHS official age) and about a quarter (24 percent) were 15-17 years old (i.e. SHS official age). Overage students, who, either begin late, repeat often or undergo a combination of the two, are the most likely to later drop out, most commonly due to labor demands in the community. Figure 2-9: Public Primary Schools in 2008/09, by Age-grade .25 .2 .15 Density .1 .05 0 5 10 15 20 age P1 P2 P3 P4 P5 P6 Source: Akyeampong 2009 based on EMIS. 58. In a comparative analysis of over twenty sub-Saharan countries, Ghana had the smallest share of new entrants into primary education who were of the normal age in 2005 as the new capitation grants allowed schools to accept more students, many of them older and less prepared to study. Only about one third of new entrants into primary are six years olds, less than one third are 1 year over-age and over one third are 2 or more years older than the normal age Figure 2-10 illustrates that not enrolling at the normal age is a systemic and pervasive problem facing Ghana’s school system and society, at large. Survey data indicate that poor rural boys are more likely to be two years over-age or more when entering primary 1. In contrast, countries such as Mauritius and Tanzania have been very successful in ensuring enrollment at the normal age (UNESCO 2007:43). 23 Figure 2-10: Share of new entrants to 1st grade who are at least 2 years over-age, by background characteristics 60 53.7 48.3 47.5 47.9 50.5 45.8 50 42.9 41.6 40 30.4 30 20 10 0 Source: Based on the Ghana Demographic and Health Survey, 2003. 59. In addition to completion, dropout and transition rates, school attendance is also a critical factor in access.. No official administrative data on school attendance exists in Ghana, for both students and teachers. But case studies suggest that student attendance rates may be very low. Using household survey data, a study carried out by the World Bank (2004a) found that primary students attended class on average 82.5 percent of the instruction time and junior high school students only 78.7 percent. Another investigation using one year attendance records for more than 100 primary schools reported that students attended four out of five days (80 percent), (Fobih et al. 1999). Access to Junior High School Education 60. About 90 percent of primary 6 pupils transition to JHS (grades 7 to 9). Since basic education is compulsory and seen as an essential part of basic public services, its fee-free structure has helped in tearing down some but not all of the economic barriers to accessing education at the level of JHS. Since a selection examination at the end of primary schooling is no longer mandatory, all students, in principle, completing primary 6 can progress to JHS. 61. Enrollment numbers in JHS are on the rise. JHS has become one of the fastest growing types of services. Enrollments rose from about 866,000 students (2001/02) to 1,302,000 (2009/10), resulting in a 50.4 percent increase over this period. As a consequence of rapidly opening access to primary education, JHS enrollment figures will potentially experience similarly rapid growth in the coming years. 62. The share of private enrollments has slowly increased but their proportion and the growth rate is slower than at primary level. There hasn’t been a significant increase in the proportion of girls who, like at primary level, represent a higher proportion of enrollment in private JHS-s. 24 Table 2-9: Trends in JHS Enrollment, Private/Public and Boys/Girls 2001/02 2002/03 2003/04 2004/05 2005/06 2006/07 2007/08 2008/09 2009/10 Public 741,895 740,753 773,982 822,205 883,060 952,151 1,015,489 1,064,088 1,075,036 Private 123,741 124,455 145,352 188,041 157,942 180,167 208,521 221,489 226,906 private % 14.29% 14.38% 15.81% 18.61% 15.17% 15.91% 17.04% 17.23% 17.43% Total 865,636 865,233 919,334 1,010,246 1,041,002 1,132,318 1,224,010 1,285,577 1,301,940 Boys 468,514 468,934 498,786 548,156 557,261 605,086 652,146 685,099 Girls 397,122 396,299 420,548 462,090 483,741 527,232 571,864 600,478 girls % 45.88% 45.80% 45.74% 45.74% 46.47% 46.56% 46.72% 46.71% 47.0% % of Girls in Private 48.6 48.6 48.5 48.1 48.6 49.2 49.3 49.3 Source: Based on EMIS data. 63. As fast as the JHS sector has grown, it also reflects some of the most typical and persistent problems in education in Ghana. These include disparities in access by gender, geographical location, wastage and inefficient use of resources (to be discussed later). 64. Despite increasing, enrollments in JHS remain relatively low; about one in every five children in Ghana is excluded from junior high schooling. Gross enrollment in JHS has increased from 64 percent (2001/02) to 79.5 percent (2009/10), indicating progress but also that much more focused efforts need to be taken to come closer to the goal of universal access to basic education. Quite significantly, the net enrollment rate has contracted since 2004/05. This indicates that the elimination of the fees and introduction of the capitation grant brought a large number of overage youth back in the system. Educating these groups brings about its own challenges (as discussed below). 65. Gender parity, although improved, remains persistently low. When very poor and rural households have to decide between girls’ schooling and their services at home or their contribution to generate household income, girls’ education opportunities tend to be “sacrificed.� When girls do remain in school, sexual violence can drive them to drop out. An analysis of gender violence conducted in Ghana found that many female students reported that older male students and teachers make sexually aggressive gestures towards them. It also found that teachers commonly do not question acts of boys’ sexual harassment towards girls (Leach 2006). Table 2-10: Trends in enrollment to junior high school education 2001/02 2002/03 2003/04 2004/05 2005/06 2006/07 2007/08 2008/09 2009/10 Total 865,636 865,233 919,334 1,010,246 1,041,002 1,132,318 1,224,010 1,285,577 1,301,940 GER 64% 63.4% 70.2% 72.8% 74.7% 77.4% 78.8% 80.6% 79.5% NER 30% 36.9% 29.5% 70.3% 74.5% 52.4% 53.0% 47.8% 47.8% GPI 0.88 0.88 0.93 0.91 0.92 0.92 0.92 Source: EMIS. Enrollment trends by grade show significant increase in dropout between grades 2 and grades 3 and a widening of the gap since 2003. 25 Figure 2-11: Enrollment Trend in JHS 1980-2008, by Grade 500000 Total Enrolment in 400000 M1/JS1 (Public & Private) 300000 Total Enrolment in 200000 M2/JS2 (Public & Private) 100000 Total Enrolment in 0 M3/JS3 (Public & 1980 1983 1986 1989 1992 1995 1998 2001 2004 2007 Private) Source: Akyeampong 2009 based on EMIS. 66. The net admission rate to JHS 1 remains very low, even though the gross and net admission rates have to some extent begun to converge since 2004/05. From that year net admission has grown from 12.2 percent up to 43.6 percent by 2008/09. Figure 2-12: Gross and Net Admission Rates in JHS 1 90.00% 84.60% 84.60% 80.00% 77.30% 80.20% 75.60% 76.60% 70.00% 60.00% 50.00% 44.60% 40.00% 43.60% GAR NAR 30.00% 28.80% 25.30% 20.00% 10.00% 12.20% 0.00% 2003/04 2004/05 2005/06 2006/07 2007/08 2008/09 Source: Based on EMIS. Internal efficiency measures, completion, dropout, repetition. 67. A third of all children fail to complete junior high school. Even though this represents a rapid improvement over the last six years, it stills keeps universal basic education, a key Government priority, in the far distance. Table 2-11: JHS Completion, Repetition and Dropout 2001/02 2002/03 2003/04 2004/05 2005/06 2006/07 2007/08 2008/09 2009/10 Total 865,636 865,233 919,334 1,010,246 1,041,002 1,132,318 1,224,010 1,285,577 1,301,940 Repeaters 35,242 28,156 37,550 47,963 38,051 53,183 54,544 40,233 Rep. Rate 4.1 3.3 4.1 4.7 3.7 4.7 4.5 3.1 Leavers 348,570 356,135 192,010 344,261 340,812 Dropout Rate 37.9 35.3 18.4 30.4 27.8 Completers 533,214 606,148 810,941 734,874 828,654 Comp. Rate 58.0 60.0 77.9 64.9 67.7 75.0 66.0 Source: EMIS. 26 68. As is the case in primary, the trend within junior high classrooms is a wide range of children of different ages in any given grade. Figure 2-13 reflects the wide age-grade profile in JHS. Principally due to late primary starters and issues of school inefficiency and associated high repetition rates, the vast majority of the junior high school age population (i.e. 12-14 year olds) is attending primary school. In 2008/09, approximately 60 percent of students in primary 6 were between 12 to 14 years old. Also, while trends over time indicate that gross admission in junior high school is slightly on the rise, net admission has stagnated at low levels of about 44 percent over the past few years. It is also important to note that about 85 percent of students in JHS 3 were 15 to 17 years old (i.e. senior high school official age), of which many are unlikely to start SHS. Figure 2-13: Public JHS in 2008/09, by Age-grade .4 .3 Density .2 .1 0 10 12 14 16 18 Age JSS1 JSS2 JSS3 Source: Akyeampong 2009. 69. A key factor limiting JHS completion is the high dropout rate. It went from 37.9 percent (2003/04) to 27.8 percent (2007/08), averaging 30 percent over the past years. It has nonetheless remained constant at around 340 thousand students annually through the period. Dropout from JH School has a strong gender component. In both deprived and non-deprived districts girls drop out consistently more than boys. Dropout appears to be a particularly strong phenomenon in the Northern region, where most deprived districts are. 70. The key to achieving UBC is not only arresting the contraction of net enrollments, but also controlling repetition and dropout. Some of the factors contributing to attrition include low learning attainment and high absence rates. Student absenteeism is associated poverty and the indirect costs of schooling, as well as with geographic conditions such as distance and difficulty accessing the school. Distance however seems to be slowly 27 becoming less of a problem due to the construction of new school buildings and infrastructure development. Box 2-2: A Case Study of the Causes of School Dropout in Northern Communities A study conducted by Ampiah et. al. (2009) surveyed 89 children aged 7-16 who had dropped out of school within one of six communities in the Savelugu-Nanton district (Northern region). Nearly half of the examined children (about 47 percent) left school in primary 1-3 and about a third (35 percent) in primary 4-6, while the remaining children (18 percent) dropped out in JHS. Hence, about four in every five children examined (82 percent) abandoned the school system while in primary. Study findings show that multiple and interlinking causes led to the decision to drop out: “Mainly they are factors associated with child labor, poverty, death of parent, fosterage, poor performance in school, parental lack of interest in formal education, pregnancy, teacher attitudes to children, distance from school, persistent learning difficulties, corporal punishment, poor attendance, and repetition� (Ampiah et. al. 2009). A series of events with several of such interlocking factors increases the likelihood of a child to drop out of school, more so than one individual incident or factor. These variables don’t predetermine that a child leaves school, but they increase the probability that a child is at risk of dropping out. The underlying cause most commonly identified for dropping out, accounting for about one third of all children examined, was a high prevalence of poverty and child labor. All of these poor and working children also indicated several or all of the above cited risk factors. Thus, reducing the indirect costs of schooling and increasing the pro-poor targeting of subsidies and programs, e.g. the capitation grant, could help to continue deterring some children from entry into child labor. In addition, local education authorities could adopt a more flexible school operation schedule, as Ampiah et. al. (2009) note, in order to adapt to household needs during agricultural periods, i.e. allow children to provide family assistance on farms during times of harvest. Every child surveyed was overage for the respective grade when they dropped out, strongly emphasizing the need for community sensitization to ensure that parents enrol their children into primary 1 at the proper age. Nearly all examined students (about 94 percent) cited having difficulties in learning. Children frequently referred to physical abuse (about 65 percent of those analyzed) as one of the most horrific and reoccurring incidences in school, in spite of legislation abolishing corporal punishment. What also unmotivated children is when teachers were absent from class instruction. To that end, more effective teacher monitoring is needed through PTAs, frequent visits from circuit supervisors and ensuring head teachers register teacher attendance. Consequently, the key to reducing the number of children at risk of dropping out seems to be both in adopting policies and practices by school administration and local governing bodies aimed at minimizing risk factors, as well as in reducing the opportunity costs of schooling and raising school demand by increasing the impact of pro-poor targeting schemes (Ampiah et. al. 2009). 71. For 15-18 year olds who have only completed primary education, the prospects of transitioning to junior high school become significantly diminished at each higher age, 28 given external factors such as increased pressures to generate income. In junior high schools, only 24 percent of 8th grade students were the appropriate age (i.e. 13) or below at the end of the 2000s. Empirical evidence shows that primary students who are at least 3 years overage perform poorer on English examinations in comparison to other students within the proper age range (Akyeampong 2009a). A wide age-range in a given grade presents obstacles for teachers trying to mediate core subject matter to children and youth with large variations in learning capacity, interests and, amongst others, levels of maturity (see also Akyeampong 2007). 72. Passing four core subjects of the Basic Education Completion Examination (BECE) is the key requirement of a successful transition from JHS. As shows, out of those admitted to JH1 (the first year of JHS), about 71 percent reaches the last year, 10 percent doesn’t take the BECE exam and 26% fails at least one subject, leaving about one third of those admitted to JHS who pass the BECE with eligibility to continue at senior high school level. Figure 2-14: Student flow in 2008/09 Kindergarten GER 92.9% NER 63.6% Primary GAR 102.9% NAR 72.1% Repetition 3.8 Dropout Completion 10% of P6 pupils 7.8* rate 86.3% do not transition to JHS (no exit exam) 90% of P6 pupils transition to JHS JHS Repetition GAR 84.6% NAR 43.6% 3.1 Reach JH3 - Don't - 29% 71% Fail at least 1 - 26% Take BECE Don't - 10% (exit exam) 61% Pass the 4 core subjects – 35% SHS GAR 35% WASSCE (exit exam) WASSCE pass rate in the 4 core subjects 74% 26% don't pass Tertiary GER 6% Source: Illustration based on EMIS data and WAEC examination records. Note: *reflects data from 2007/08; enrolment data is used where no data is available for admission rates. 29 Access to Post-Basic Education: Senior High Schools and Technical Training Institutes 73. 2.40 Throughout the decade the Government maintained a strategic priority to expand post-basic education with the dual purpose of improving the livelihood of people and boosting the human capital to reach middle-income status for the country. However, this strategic priority is being pursued through two different but somewhat overlapping agenda: 74. the expansion of formal post-basic school based programs that link basic education and tertiary education into a coherent school system with appropriate regulation of curriculum, transition from one level to the next and planning the necessary resources to meet the goal and the reform and expansion of formal and informal TVET programs to partly address the social agenda of improving livelihoods while advancing the economic agenda of providing skills to improve employment and productivity. Access to Senior High School Education 75. Currently, the transition rate from JHS to SHS is around 35-40 percent. Annually, approximately 100,000 junior high school students fail the leaving examinations taken by all students in the third and last year, known as the Basic Education Certificate Examination (BECE). This has driven many students to repeat or even drop out. Among those students who pass the BECE -- used also as an entrance examination to continue onto SHS -- only about half actually go onto to senior high school. While gross enrollment in JSH reached 79.5 percent, about a third of youth were enrolled in senior high schools in 2009/10. With an expansion in SHS enrollments from about 328,400 students (2003/04) to about 537,300 (2009/10), upper secondary has been taking in on average nearly 34,800 new students per year, largely thanks to the increase in private sector provisions. In terms of cumulative expansion, enrollment in SHS grew 64 percent since 2003/04, a rate well above that of population growth over this time span. Between the academic years 2006/07 and 2007/08, the senior high system experienced a decline in enrollment figures, although it appears to have recovered. 30 Table 2-12: SHS enrollments 2003/04 2004/05 2005/06 2006/07 2007/08 2008/09 2009/10 Total 328,426 333,002 384,455 485,742 454,681 490,334 537,332 GER 26.6% 25.6% 29.1% 35.8% 32.2% 33.9% 36.1% GAR 1 33.3% 31.5% 35.0% 39.4% Source: EMIS. 76. During 2003/04 and 2009/10 gross enrollments grew annually at a rate of 1.6 percent on average, having increased from 26.6 to 36.1 percent. 77. UNESCO comparative data looks at secondary education and Ghana’s total gross and net enrollment rates for secondary education are estimated at 49 and 45 percent and significantly higher relative to sub-Saharan Africa averages (32 and 25 percent respectively) Figure 2-15: International Comparison in Total Secondary GER and NER, 2006 Niger 9 11 % of Total Secondary NER Chad 15 % of Total Secondary GER Pakistan 30 30 Benin 32 SSA 25 32 Nigeria 26 32 Comoros 35 Lao DPR 35 43 GHANA 45 49 Kenya 42 50 India 54 El Salvador 54 65 Botsw ana 56 76 0 20 40 60 80 Note: Secondary schooling refers to lower and upper secondary education. Source: Based on UNESCO, 2008. 78. Figure 2-16 shows that among countries with a similar Gross National Income (GNI) per capita, which have net enrollment data available, Ghana has the highest total secondary NER at 45 percent (UNESCO 2008), in spite of it being the second income- poorest country with a GNI per capita of US$670 (World Bank 2009). 31 Figure 2-16: International Comparison in Secondary NER and GNI per capita 50 GHANA 45 Secondary Net Enrolment Rate Kenya 40 35 Lao DPR Pakistan 30 Nigeria 25 SSA 20 15 10 Niger 5 0 0 200 400 600 800 1000 1200 1400 GNI per capita (US$) Note: GNI data is for 2008, while NER data is for 2006. Secondary schooling refers to lower and upper secondary education. Source: Based on data from UNESCO, 2008; World Bank, 2009. 79. The increasing demand for SHS education is met with limited supply. The increasing access to basic education has increased demand for post-basic education, especially for the senior high school (SHS) programs. In turn, the increasing education opportunities in SHS tends to boost enrollments at basic level, as the prospects of accessing SHS serve as incentives. Altogether, there is a relatively high and persistently increasing demand to access SHS programs. Almost all students successfully passing the four core subjects of the BECE exam attempt to continue at SHS level, though only half find a place at SHS. 80. To assure a level playing field, since 2007 admission into SH Schools has been regulated by a computerized central system by which students are asked to rank their 6 preferred schools and admission is granted or denied according to the student’s BECE results and the schools’ minimum standards. Applicants are normally expected to enroll at 15 years of age in the SHS, which is made up of 3 years, and upon successful completion, to sit the West African Senior Secondary Certificate Examination (WASSCE. Results of the examination are very important for admission into university and students with insufficient grades are given the possibility to re-sit the WASSCE in December. 81. There are only 493 public senior high schools in Ghana, unevenly distributed across the territory, which can only accommodate 78% of BECE passers, although entrance also depends on performance on students. The three best served regions: Ashanti, Eastern and Volta account for close to half of the schools in the country. The public sector had about half of all senior high schools in urban areas and the other half of public SHS-s are in rural communities. Meanwhile, 73.4 percent of the private senior high schools are located in rural areas (26.6 percent in urban areas), trying to fill the gap in demand and supply there. This is an interesting finding, because it is often the contrary in other countries. 32 Access to the Technical and Vocational Training Institutes 82. Ghanaians pursuing technical and vocational education or training can either attend public programs offered through Technical Training Institutes (TTIs) that extend from junior secondary through to post-secondary, administered by the Ghana Education Services, or participate in a broad variety of private and public programs. They can choose from formal and non-formal, school-based, community-, or enterprise-based training programs. A new inter-governmental institution, the Council for Technical and Vocational Education and Training (COTVET) is responsible for creating a coherent qualification and certification framework that would span the various training forms. COTVET is also responsible for establishing a coherent database for all training providers, participants and activities. 83. There are approximately 200 public TVET institutes, including the 26 TTIs under the Ministry of Education (MoE), 137 other institutions under the Ministry of Employment and Social Welfare and the remainder under different other ministries. The 26 TTIs under the MoE accounted for about half of total enrollments at public TVET institutes in 2007/08. With regard to enrollment per institute, the MoE technical training institutes are far larger than other public training institutes; the average number of trainees in a TTI was 642, compared to 330 in the Opportunities Industrialization Centers, 181 in the National Vocational Training Institutes (NVTI), 147 in the Community Development Vocational/ Technical Institutes and about 70 in both the Youth Leadership and Skills Training Centers and the Integrated Community Centers for Employable Skills (ICCES). 84. The majority of the enrollment in public TVET is male trainees; for example only 16.7 percent of the 18,432 TTI trainees are female, and about 26 percent of the 6,710 NVTI trainees are female. However, the Community Development Vocational/ Technical Institutes and the Opportunities Industrialization Centers have more female than male trainees: 73% of 3,520 trainees and 55% of 1,010 trainees are females respectively. 85. At all pre-tertiary education levels, the gender gap has been closing in Ghana, but the gap is still much larger and the pace slower at the TTIs. Female enrollment in TTIs saw an increase of 32 percent between 2001/02 and 2006/07, while male participation in TTIs contracted by 13.7 percent over this period. In spite of these evident strides in closing the gender gap due to diminishing male attendance, female students constituted less than one in every five students in 2006/07. In addition, the relationship between levels of welfare and non-attendance in technical and vocational education is such that a Ghanaian from the poorest quintile is seven times less likely to have vocational education than a Ghanaian from the richest quintile. Table 2-13: Enrollment in Technical Training Institutes % increase 2001/02 2002/03 2003/04 2004/05 2005/06 2006/07 2007/08 2008/09 2009/10 over period Total 17,934 19,777 18,672 21,424 20,303 18,432 17,210 17,280 20,694 13.3% Male 15,603 17,060 15,889 18,440 16,933 13,467 -13.7% Female 2,331 2,717 2,783 2,984 3,370 3,081 32% % Female 13% 13.7% 14.9% 13.9% 16.6% 16.7% 22% Parity Index .15 .16 .18 .16 .20 .23 35% 33 Source: EMIS. 86. Enrollment in public TVET institutes has been declining over the last few years. For example, total enrollment in the TTIs was lower in 2006/07 than in 2004/05. In the NVTI total enrollment 2005-2007 fell by just less than 10% from 7,297 to 6,710. 87. Information on internal efficiency, measured by drop-out and completion rates, is not routinely collected by the majority of training providers. What evidence there is suggests that somewhere between 10-30% of trainees drop out of training annually, and this implies completion rates for a four-year course of between 24% and 65%. 88. The 2008 Education Management Information System (EMIS) report included 629 public and private TVET institutions in its database; if there were 200 public TVET providers, and then the total number of private providers was approximately 430. Total enrollment in a partial selection of private TVET institutes in 2007/08 was, according to EMIS 23,452 trainees; however this only represented the enrollment in 143 private TVET institutions, covered by the EMIS sample. 89. Most formal public and private TVET providers target lower secondary school (JHS) graduates and sometimes set a minimum aggregate score at the Basic Education Certificate Examination (BECE) level as an entry requirement. Among all the public TVET providers, the ICCES appears to be the most accessible; it does not have official entry requirements and most trainees are JHS graduates with either poor BECE aggregates of failure at the BECE level. Of the public providers, GRATIS is slightly different as it offers a variety of training programs targeted at a variety of individuals 90. While entry requirements varying from none to passing of the BECE and even the Senior Secondary School Certificate Examination (SSSCE), most formal TVET options – TTIs, NVTIs, and most other public and private vocational training institutes – require that entrants have completed basic education, and often specify a minimum aggregate grade prospective trainees are meant to achieve at the BECE level. This automatically excludes those who were unable to attain this level; most of whom will be the poor or poorest. Since educational attainment in Northern communities, in particular, can be very low, access to formal training by Northern youth is further limited. There are some exceptions of course in both the public and private sectors; e.g. the public Integrated Community Centers for Employable Skills (ICCES) as well as some private (non-profit) vocational training institutes often do not have minimum formal schooling or examination entry requirements. Even though most trainees in formal TVET programs have completed basic education, many are still functionally illiterate. Informal apprenticeship training in some trades attracts more educated, and hence usually less- poor, youth, while other trades attract poorer, less educated ones. 91. The limited numbers of formal training institutions and the distance of these institutions to communities impede access to people in these communities. Training institutes are usually urban based, making it difficult for rural youth – and especially the poor – to attend. The public ICCES centers and some NGO training centers are the exception to this, being located in more rural areas. 92. Formal VOCATIONAL AND TECHNICAL training has a strong focus on long duration (2-4 years) pre-employment courses; this period is far too long for many poor families to pay the costs associated with training for their child and to pay the opportunity 34 cost of their child being out of the labor market. The provision of short or modular courses by formal INSTITUTIONSs remains very limited. Access to Special Education 93. According to the Education Sector Performance Report (2009) it is estimated that 5 percent of Ghana’s population has some form of disability. Sight problems are recognized as the most prevalent at about 59 percent, followed by hearing and speech impediments. There are significant disparities between rural and urban areas in the willingness and frequency of recognition of children with special needs and the facilitation of these children to attend a school. 94. Enrollments in institutions of special education have grown by 87.7 percent from 3,361 students (2001/02) to 6,308 students (2008/09). In 2001/02 there were about twice as many boys attending school as girls but the share of the female population improved slightly to 40.4 percent by 2008/09. Figure 2-17: Enrollment Trends in Special Education, Male/Female 7,000 Total Male Female 6,308 6,000 5,654 5,000 5,092 4,722 4,203 4,435 4,000 4,114 3,760 3,361 3,315 3,000 2,821 3,004 2,617 2,509* 2,666 2,548 2,112 2,339 2,000 1,901 2,088 1,694 1,769 1,497 1,249 1,000 0 2001/02 2002/03 2003/04 2004/05 2005/06 2006/07 2007/08 2008/09 Source: Based on EMIS. Note: *Due to a data discrepancy an adjustment has been made. 95. Various studies generally estimate that at least half of the children with special needs are out of school and for many of such children who are in school their special needs remain unrecognized. A simple visit to regular classrooms, especially in remote areas, reveals that in Ghana there are very few children wearing glasses and none using hearing aids. In Ghana, there are societal prejudices towards children with disability. The importance of providing them with an education is typically undervalued by communities and, they commonly receive less attention from teachers when they attend school. There is a need for greater awareness among families about these children’s potential in life. (Ministry of Education 2008a). Non-Formal Education 96. In Ghana, the adult literacy rate was 62 percent in 2007/08. Non-formal education’s principal goal in Ghana is to improve adult literacy and gender parity among graduates of the National Functional Literacy Program. Data on the program in 2003/05, the last year for which information on graduates and dropouts are available, shows that of 35 the 182,037 learners recruited, a total of 141,769 graduated from the program (78 percent), while the remaining 22 percent dropped out. Recruited learners were taught by about 7,281 facilitators in that same year, most of whom conducted the classes for at least 300-400 hours and the attendance rate among the learners was at least 70 percent of this time (ESPR 2009). About 202,500 people were recruited for the program in 2006/07, while 13 percent dropped out. Trends over time (Table 2-14) show that between 140,000 and 250,000 Ghanaians graduate from the Literacy Program on an annual basis, equipped with necessary literacy tools for many jobs in today’s labor market. Since 1995, the overwhelming majority of the enrollees, i.e. 78-88 percent, graduate yearly from the program. Table 2-14: National Functional Literacy Program 1995/97 1996/98 1997/99 1998/00 2000/02 Feb 03/05 Nov 03/05 Jun 05/Mar Jan 06/Oct 07 07 Learners Recruited 209,226 219,299 207,325 187,550 196,170 312,907 182,037 272,357 202,500 Graduated 175,431 175,383 180,554 163,781 172,453 249,118 141,769 232,761 176,683 Dropout Rate .16 .20 .13 .13 .12 .20 .22 .15 .13 Source: EMIS. 97. Since 1992, about two million youth and adults have graduated and benefited from the Literacy Program. It should be stressed that the vast majority of these graduates have been women, to this effect, aiding to close the gender gap in literacy. With an estimated increased intake of about one million additional learners to the program between 2009-2015, the demands on training 40,000 new facilitators will have to be complementarily addressed (ESPR 2009). B. Learning Achievement 98. Access, enrollment, completion and transition are important but only intermediary steps towards an educated society and knowledge-based economy. The real target is learning, acquiring competencies and skills that help the individuals, the communities and the society at large. Yet, it is not easy to gain an accurate and reliable picture about how effective is the education system in achieving desired learning outcomes and achieve improvements in these from year to year. Ghana’s experience, described below, in line with a large majority of countries, shows that improvements in access put pressure on effectively achieving learning outcomes across various social groups and communities. It is also a more or less universal rule that schools and education are not able to overcome some initial levels of differences in abilities and learning skills, much of which is determined in the early childhood environment by such factors as the education and income levels of parents. Further, schools often inadvertently exacerbate initial differences, helping good performers strive and letting others fall further behind. 99. International tests repeatedly show that overall performance of education systems can be measured by the strength of the average achievers, i.e. the proportion of those who achieve just acceptable level of proficiency as it is defined by national or international standards. This middle group constitutes a key part of the productive workforce, stable consumers with regular income, who participate in and often decide elections, and who maintain social cohesion. While it is almost inevitable and indeed desirable that a professional and intellectual elite emerges from education, the real challenge for 36 education systems is to keep to a minimum the range of underperformers at risk of falling to the margins of society. 100. The goal is, therefore, to expand the share of those students who achieve average ‘acceptable’ results. 7 If that share is big enough, few students fall through the cracks. On the other hand, those systems where a good share of the students perform on the top and another large share of the students perform poorly, typically hide large inefficiencies and masses of students who are in school but benefit very little from their education. Education strategies should aim (i) to keep most of the education performers above (even if just) the minimum level of proficiency and (ii) gradually improve this minimum level of proficiency without letting more students fall behind. In the Ghanaian context, policy makers can rely on a number of national and international tests to assess the system’s performance. Primary School 101. In a relatively small sample over 1700 3rd and 5th grade students in 50 primary schools were tested on basic reading comprehension and mathematics skills. The test was tailored to what the students were supposed to know at 3rd grade. While for instance in reading comprehension about 40 percent of the students provided correct answers to about 80 percent of the questions, a similarly large proportion of the 5th graders provided less than 18 percent of the answers correctly and only about one third of the students provided answers between 40 and 80 percent. What this data shows is that while there is a good number of students on the top and a relatively big number of students on the bottom, the middle in between is missing. Therefore, the question for Ghanaian policy makers is what are the characteristics of the groups that have gained access to education as a result of the last decade of economic growth, decline in poverty and commitment to education but failed to achieve the learning outcomes that are in line with the country’s goals to build learned society or at least one where graduates achieve proficiency in basic skills. Figure 2-18: Share of Students with Correct Reading Comprehension, 3rd and 5th Grade 7 In international assessments, countries like Singapore, Korea or Finland perform on the top and show the least internal variations between high and bottom performers. 37 Source: World Bank Ghana Survey, 2009. Note: See Annex for key results on the World Bank’s 2009 Ghana Survey. 102. Keith Lewin, of the ‘Consortium for Research on Educational Access, Transitions & Equity (CREATE), characterizes the phenomenon of students falling through the cracks by the concept of “zones of exclusion�. These zones include those who don’t get into the system or drop out early but also those who do not learn and gain access to further levels of education. According to CREATE, only about 35 percent of all pupils in primary 1 are not at risk and have secure enrollment, attendance and achievement. By primary 6 this proportion gradually reduces to about 20 percent, meaning that the remaining 80 percent of all children are ‘silently excluded’, at risk or have no access at all. The proportion of Ghana’s youth reaching the 10th grade who have secure enrollment, attendance and achievement reduces near to 10 percent. Figure 2-19: Access and Zones of Exclusion in Primary and Secondary CREATE Zones of Exclusion Source: CREATE 2007. 103. Other learning achievement tests have produced similar outcomes. The National Education Assessment tests were conducted in 2005, 2007 and 2009 on a nationally representative random sample of 3.5 percent of Ghana’s primary schools (see NEA 2007:2). To reach the proficiency level, students need to answer correctly to at least 55 percent of the multiple choice questions (from four possible answers). In spite of progress, learning results remain weak. The proficiency level at primary 3 in 2005 was 16.4 percent in English and 18.6 percent in math. There was an overall increase of 3.6 percent in English proficiency and 6.6 in math proficiency among 3rd graders over the period 2005-2009. In 2005, the proficiency level in primary 6 was 23.6 percent in English and 9.8 percent in math, therefore, registering an improvement of 12 percent in English and 4 percent in math by 2009. Given the very large overall improvements between 2007 and 2009, concerns may arise about the consistency in methodology of the administered exams and/or the accuracy of data collection methods. 38 Table 2-15: Overall NEA Mean Scores, Minimum-competency and Proficiency 2005 2007 2009 Mean Minimum Proficiency Mean Minimum Proficiency Mean Minimum Proficiency Competency Competency Competency English P3 38.1 50.5 16.4 37.6 50.1 15.0 40.2 57.6 20.0 English P6 43.1 63.9 23.6 44.2 69.7 26.1 48.2 76.9 35.6 Math P3 36.6 47.2 18.6 35.0 42.6 14.6 41.8 61.2 25.2 Math P6 34.4 47.2 9.8 35.7 46.2 10.8 39.6 61.9 13.8 Source: NEA 2007; and Ministry of Education 2010. 104. Generally, a country’s education system should strive for 75 percent of children reaching a minimum competency level (ESPR 2008). However, in the case of Ghana that benchmark is far from being met. Nearest to the goal is the Greater Accra area; there in 2007, 64.1 percent of primary 6 students achieved minimum competency in mathematics and 20.2 percent attained proficiency. These two figures at the national level were 46.2 and 10.8 percent, respectively, illustrating the largest regional divide between the Greater Accra region and all other regions. In Greater Accra, the percentage attaining minimum competency in English primary 6 was 86.7 percent, while 50.3 percent reached proficiency in 2007. These two figures at the national level were 69.7 and 26.1 percent. 105. Nationally, girls perform better in English, while boys do so in Math, which is consistent with most assessment results in other countries. Table 2-16: Overall NEA Gender Comparisons Primary 3 – English Primary 3 – Mathematics Mean % % Reaching Mean % % Reaching Proficiency Proficiency Boys 37.1 14.0 35.4 15.2 Girls 38.1 16.0 34.5 13.9 Primary 6 – English Primary 6 – Mathematics Mean % % Reaching Mean % % Reaching Proficiency Proficiency Boys 43.6 24.5 36.1 11.6 Girls 44.8 27.9 35.2 9.8 Source: NEA 2007:17. 106. Disparities in students’ performance based on locality are more pronounced. While access rates between rural and urban areas are already large, the gap in learning outcomes appears to be even larger. 39 Figure 2-20: Mean NEA scores for both 3rd and 6th grade by gender and by locality 42 NEA Test Score by Gender NEA Test Scores by Locality 50 48 40 46 44 38 42 40 36 38 36 34 34 32 30 32 English Urban English Rural Mathematics Mathematics English Male English Mathematics Mathematics Urban Rural Female Male Female NEA English Test Scores by Locality 50 NEA Mathematics Test Scores by 50 and Gender 48 46 Locality and Gender 48 44 46 44 42 42 40 40 38 38 36 36 34 34 32 32 30 30 English Urban English Urban English Rural English Rural Maths Urban Maths Urban Maths Rural Maths Rural Male Female Male Female Male Female Male Female Source: Wodon and Joseph 2010 based on NEA. 107. Children living in more remote, isolated and deprived communities tend to score lower than their counterparts in urban centers on NEA tests. In addition, 6th graders in schools within the Southern areas scored higher on NEA English exams than their counterparts in more Northern areas. The highest concentration of 6th graders with low NEA English scores (38 or below) went to school within districts of the Northern region, while the highest concentration of students with high scores (44 or above) attended schools in the districts of Greater Accra. The three districts with the highest NEA English scores in primary 6 were the Tema district (Greater Accra) with 60 points, the Kumasi district (Ashanti) with 59 points and the Ga West district (Greater Accra) with 55 points, while the Saboba district (Northern) with 25 points scored lowest followed by the Juabeso district (Western) with 26 points and the Zabzugu district (Northern) scoring 27 points. 40 Map 2-1: NEA scores of Primary 6 English in 2007, by district Source: Based on NEA data. 108. Key resources influencing learning results. A USAID (2009) commissioned study shows that the strongest relationship between achievement and independent factors are resource inputs (particularly teachers), location and other school indicators. Table 2-17 shows the proportion of a given resource/input for low, medium and high performing schools on NEA English assessments of primary 6 in 2007. 109. Hence, high performing schools had over 90 percent trained teachers and 36.4 percentage points more trained teachers than low performing schools. The proportion of high performing schools in urban settings, with over 5,000 residents, was 72.2 percent and accessing electricity was 47.3 percent, while the share of low performing schools in urban areas was 8.5 percent and those with electricity was 9.1 percent. Map 2-2: Public Trained Primary Teachers and NEA Math P6 Scores Source: Based on EMIS data from 2008/09 and NEA data from 2007. 41 110. Geographic location also appears to be a significant factor influencing achievement outcomes, since 39.1 percent of high performance schools were within the four most densely populated districts (Accra, Tema, Takoradi or Kumasi) and only 1.6 percent of low performance schools. In general, primary 6 English students in these four districts scored 6.3 points higher on the NEA test than their counterparts in the other districts and the math scores were as much as 15 points higher than the national average (34). It was also less common among high performing schools for students to dropout (2.9 percent) than in low performing schools (5.8 percent). In addition, enrollment rates tended to be greater in high performing schools at 95 percent than in low performing schools at 66 percent in 2007. 111. As data illustrates, the districts with the poorest performance on NEA, which to date has only been conducted in 2005 and 2007, were in the Northern regions; scores in math were 15 points lower than the average and less than half than the top scoring districts in the country. These districts had a primary PTTR of 150 to 200. In these remote areas, schools are typically small, with enrollments typically less than 100. This means that often only every second or third school in the region has a trained teacher. More than half of students don’t have access to a treated water source, which has been known to lead to a higher incidence of guinea worm infection and typhoid. The Northern Savannah, being one of the most deprived areas, affected by droughts, floods and ground erosion, high transport costs, long distance from markets, low population density. Gender marginalization remains part of everyday reality. The remoteness of the Northern region contributes to low teacher motivation to be deployed, compounded with the reality of major infrastructure deficiencies (as further discussed in the chapter on equity). Table 2-17: Mean Resource in 2005 and 2007, by Class 6 English Proficiency Performance level Resource/Input 2005 2007 LOW MEDIUM HIGH LOW MEDIUM HIGH Human Resources Pupil/teacher ratio 41.59 41.76 39.95 36.26 36.99 42.22 % untrained teachers 0.35 0.18 0.16 0.45 0.26 0.09 Physical Resources Math texts per pupil 0.56 0.58 0.62 0.88 0.91 0.81 English texts per pupil 0.27 0.28 0.40 0.88 0.93 0.81 Seats per pupil 0.83 0.83 0.86 0.88 0.87 0.95 Writing places per pupil 0.84 0.86 0.86 0.89 0.83 0.92 % schools needing major repairs 0.26 0.26 0.15 0.25 0.23 0.18 % schools with meals programs 0.09 0.12 0.13 0.12 0.14 0.08 Library books per pupil 0.59 0.64 0.40 0.76 0.55 0.67 Teacher guides per teacher 3.99 3.65 2.38 2.33 2.57 1.50 % schools with functional toilets 0.52 0.49 0.49 0.54 0.59 0.60 % schools with functional urinals 0.39 0.33 0.35 0.55 0.66 0.58 % schools with accessible drinking water 0.49 0.59 0.52 0.60 0.50 0.64 % schools with electricity 0.09 0.25 0.37 0.09 0.23 0.47 Geographic location % urban 0.13 0.39 0.69 0.09 0.32 0.72 42 % in one of the four largest urban districts 0.01 0.14 0.22 0.02 0.08 0.39 Education Indicator P6/P1 enrollment 0.85 0.97 1.01 0.66 0.84 0.96 % dropout 0.05 0.08 0.02 0.06 0.04 0.03 Source: Based on data from USAID, 2009. 112. PTR and physical resources had a weak relation to learning, though they do affect access. In 2007, the pupil teacher ratio was lower within schools performing more poorly than in higher performing ones (although it is unlikely that the negative relationship was statistically significant), going against the conventional wisdom that the lower PTR the better. Indeed, some observers point to negative consequences from hiring teachers without qualifications. (There is…)“[…] considerable concern among both policy makers and head teachers surround ing the employment of “untrained� teachers, often individuals hired by District Assemblies, such as through the National Youth Unemployment Program, or by communities.� (USAID 2009:12). 113. Also, in 2007 low performing schools appeared to have marginally (not significantly) better access to English and math textbooks per pupil (0.88) than high performing schools (0.81), although the relationship was the opposite in 2005. Over this two year time span (2005-2007), the number of mathematics textbooks per student increased on average from 0.55 to 0.89 and English textbooks per student rose as well from 0.24 to 0.87. In contrast, according to EMIS, the amount of teacher guides per teacher declined by nearly 0.5 between these two years. 114. Lower performing schools had almost twice as many teacher guides per teacher as higher performing schools both years, most likely because of the lack of teachers or the lack of qualified teachers in these schools. Also, low performing schools disposed of more library books per pupil (0.76) compared to those with high performance (0.67) in 2007, but also in 2005. Thus, given their relative equal distribution among low- and high- performing schools, learning and teaching resources overall don’t appear to be directly related to achievement outcomes. 115. No relation can be observed between schools with functional toilets, potable water, availability of school furniture and the need of major infrastructure repairs in the schools on the one hand and improved learning outcomes, on the other. However, some of these resource indicators have a strong impact on low admission rates and dropout. For example, schools that lack a water source and sanitation facility are more likely to adversely affect school participation due to higher prevalence of illness. Therefore, they also skew the sample of those tested through NEA. In other words, those who would have performed poorly in the NEA tests because of poor school infrastructure may not have been in school by the time the tests were carried out as a result of not signing up for school or being dropped out of school beforehand. 116. The school feeding program’s impact appears to be limited in improving access and does not seem to lead to increased test scores. One possible reason is that the school lunches are measured as of 2005 and very few schools had a systematic school meals program in 2005 43 Figure 2-21: NEA test scores and school lunches NEA Test Scores and School Lunches NEA Test Scores and School Lunches in P3 and P6 Grades 41.6 41.4 41.0 44.2 36.6 38.6 35.8 34.6 35.2 38.3 43.1 36.2 35.5 37.1 35.3 34.8 English Schools with meals 2005 Mathematics Schools without Schools Schools Schools Schools meals 2005 with without with without meals meals meals meals 2005 2005 2007 2007 Source: Wodon and Joseph 2010 based on NEA. Junior and Senior High School 117. The Basic Education Certificate Examination (BECE) is a key achievement test in junior high school, taken by all students in the third and last year and which serves as an entrance exam to proceed onto SHS. BECE test results are normalized, meaning that the passing grade is adjusted each year so that roughly 60 percent of examined students pass the BECE successfully. This also means, that diachronic analysis of BECE results (i.e. comparing results over more than one year) is not possible. However, variations within a year and even changes in relative distribution of pass rates by various external factors (such as geographical location, socio-economic status, gender) are possible. 118. In 2008/09, BECE pass rates were highest in the Greater Accra region (66.9 percent in math and 84.1 percent in English) and the Ashanti region (80.2 percent in math and 58.8 percent in English). Geographical disparities are significant. The Northern region had the lowest English pass rate of 39.7 percent, while the Volta region had the lowest math pass rate of 40.1 percent. Figure 2-22: Regional BECE Pass Rates in 2008/09 100.0% 80.0% 60.0% 40.0% 20.0% 0.0% GREATER BRONG UPPER UPPER ASHANTI WESTERN CENTRAL EASTERN NORTHERN VOLTA ACCRA AHAFO WEST EAST Math 66.9% 80.2% 65.6% 58.5% 48.4% 47.4% 45.2% 42.8% 41.5% 40.1% English 84.1% 58.8% 58.0% 60.2% 49.9% 57.5% 50.6% 50.9% 39.7% 54.4% Science 74.7% 68.9% 59.6% 55.4% 49.3% 52.7% 55.0% 56.0% 50.0% 52.1% Social Stud. 70.7% 70.6% 61.1% 53.3% 49.6% 52.3% 50.8% 49.1% 40.7% 49.4% Source: Based on EMIS. 119. In 2008/09, the bottom five performing districts on BECE English exams were all found within the Northern region. Some external characteristics that schools within these five poorest performing districts have in common are: remoteness and poor infrastructure, especially access roads, and scarce postal services, making the region an undesirable posting choice for teachers. 44 Map 2-3: BECE English and Math Pass Rates, 2008/09 Source: Based on EMIS data. 120. BECE performance correlates more strongly with the share of trained teachers than to the degree to which students have access to school materials and basic infrastructure. BECE results are also not related strongly to the degree of availability of basic school materials and infrastructure. Table 2-18: Mean Resource, by class 9 BECE English Pass Rate in 2008/09 Performance level* JHS Resource/Indicator (BECE pass rate) (National Average)** LOW MEDIUM HIGH Difference (Average (Average (Average (betw. high 34.8%) 50.7%) 67.0%) and low) Human Resources Pupil/teacher ratio (17.96) 19.34 17.60 16.94 -2.40 Pupil/trained teacher ratio (29.67) 32.96 29.27 26.79 -6.17 % trained teachers (0.63) 0.62 0.62 0.66 0.04 Physical Resources Core textbooks per pupil (2.08) 2.06 2.16 2.03 -0.03 Seats per pupil (0.79) 0.74 0.83 0.79 0.05 Writing places per pupil (0.77) 0.71 0.80 0.80 0.09 % classrooms needing major repairs (0.24) 0.24 0.23 0.24 0.00 % classrooms needing minor repairs (0.33) 0.36 0.35 0.30 -0.06 % schools with toilets (0.57) 0.49 0.57 0.64 0.15 % schools with accessible drinking water (0.62) 0.58 0.63 0.65 0.07 Education Indicators GER (0.79) 0.69 0.79 0.88 0.19 NER (0.44) 0.36 0.44 0.52 0.16 GAR (0.83) 0.73 0.82 0.93 0.20 NAR (0.37) 0.29 0.37 0.46 0.17 Completion rate (0.73) 0.65 0.74 0.82 0.17 Repetition rate (0.03) 0.02 0.04 0.03 0.01 45 Note: *Of the 138 districts, 46 districts are classified as low performing as to English BECE pass rates (scoring between 6.6%-44.3%), 46 districts as medium performing (scoring between 44.3%-57.2%) and 46 as high performing (scoring between 57.5%-89.6%). **The indicated national averages are calculated using district averages and, thus, in some cases may slightly differ from other aggregate data found throughout this report. Source: Calculations based on EMIS data. Table 2-19: Mean Resource, by class 9 BECE Math Pass Rate in 2008/09 Performance level* JHS Resource/Indicator (BECE pass rate) (National Average)** LOW MEDIUM HIGH Difference (Average (Average (Average (betw. high 33.44%) 53.63%) 78.27%) and low) Human Resources Pupil/teacher ratio (17.96) 18.70 18.01 17.16 -1.54 Pupil/trained teacher ratio (29.67) 30.15 29.97 28.90 -1.25 % trained teachers (0.63) 0.65 0.64 0.61 -0.04 Physical Resources Core textbooks per pupil (2.08) 2.03 2.12 2.10 0.07 Seats per pupil (0.79) 0.80 0.79 0.78 -0.02 Writing places per pupil (0.77) 0.77 0.76 0.78 0.01 % classrooms needing major repairs (0.24) 0.24 0.23 0.24 0.00 % classrooms needing minor repairs (0.33) 0.36 0.34 0.31 -0.05 % schools with toilets (0.57) 0.54 0.60 0.56 0.02 % schools with accessible drinking water (0.62) 0.56 0.63 0.68 0.12 Education Indicators GER (0.79) 0.76 0.77 0.82 0.06 NER (0.44) 0.43 0.44 0.44 0.01 GAR (0.83) 0.80 0.81 0.86 0.06 NAR (0.37) 0.37 0.37 0.38 0.01 Completion rate (0.73) 0.70 0.72 0.78 0.08 Repetition rate (0.03) 0.03 0.04 0.02 -0.01 Note: *Of the 138 districts, 46 districts are classified as low performing as to math BECE pass rates (scoring between 6.8%-42.3%), 46 districts as medium performing (scoring between 42.9%-63.5%) and 46 as high performing (scoring between 64.5%-99.4%). **The indicated national averages are calculated using district averages and, thus, in some cases may slightly differ from other aggregate data found throughout this report. Source: Calculations based on EMIS data. 121. Students’ pass rate on English BECE examinations tends to be higher in those districts where the PTR and the PTTR are lower, but the variance is very high. There is also a positive relationship between the numbers of teachers and students’ pass rates on BECE math exams. 46 Figure 2-23:JHS Pupil (Trained-) Teacher Ratio and BECE English Pass Rates, by District 1.00 1.00 BECE English Pass Rate BECE English Pass Rate 0.80 0.80 0.60 0.60 0.40 0.40 0.20 0.20 0.00 0.00 0 10 20 30 40 50 0 20 40 60 80 100 JHS Pupil Teacher Ratio JHS Pupil Trained Teacher Ratio Source: Based on data from EMIS 2008/09. 122. Ghana is one of the few African countries participating in the Trends in International Math and Science Study (TIMSS) survey. Its results in the 2003 and 2007 surveys are, by and large comparable to other developing countries, but this also means scores that are much below international averages. In TIMSS 2003, Ghana performed second poorest as to average mathematics scores of eighth-grade students with a score of 276 after South Africa with 264, whereas the global average was 466. By 2007, the improvements were significant, as Ghana’s average improved to 309, although was still the second lowest to Qatar. Importantly, the international benchmarks appear to be in line with the NEA standards of mathematics proficiency. Like in the NEA, where the proportion of those achieving math proficiency was much below 20 percent, in TIMSS about 17 percent of students aged 16 scored above the low international benchmark. As only about half of the children aged 16 attended school the day the examination was administered, this further distorted the sample size and the true national score may have been lower. Overall, Ghana scored behind countries with a similar GNI in other regions, despite its relatively high share of public spending on the sector at about 30 percent at the end of the 2000s. 47 Figure 2-24: International Trends in Math (left) and Science Scores (right) of Eighth- graders, 2003/2007 GHANA 309 303 276 GHANA 255 340 2007 2007 El Salvador Botsw ana 355 2003 365 2003 Botsw ana 364 387 366 El Salvador 380 408 Colombia Algeria 387 417 Algeria Colombia Egypt 391 408 406 Egypt 421 Tunisia 420 445 410 Tunisia 404 Korea, Rep. of 597 553 589 Korea, Rep. of 558 593 567 Singapore 605 Singapore 578 500 500 Average* 466 Average* 473 0 100 200 300 400 500 600 700 0 100 200 300 400 500 600 700 * Ave ra ge fo r 2003 is inte rna tio na l a ve ra ge ; a ve ra ge fo r 2007 is s c a le a ve ra ge . * Ave ra ge fo r 2003 is inte rna tio na l a ve ra ge ; a ve ra ge fo r 2007 is s c a le a ve ra ge . No 2003 da ta a va ila ble fo r C o lo m bia , Alge ria a nd El S a lva do r. No 2003 da ta a va ila ble fo r C o lo m bia , Alge ria a nd El S a lva do r. Source: Based on TIMSS 2009. 123. At the Senior High School level, the exit exam is the West Africa Senior Secondary Completion Examination. According to Government data Pass rates increased substantially in 2008/09 for both sexes, jumping from 36% to 74%. This fast jump calls for some credible explanation to avoid questions of the reliability of the Government data. There remains however a substantial gender gap in achievements. Figure 2-25: WASSCE examination results 2005/06 – 2008- 09 WASSCE Examination 150000 100% 80% 100000 60% 40% 50000 20% 0 0% 2005/06 2006/07 2007/08 2008/09 Examinations taken Pass Pass rate Source: Waec Examination results, EMIS Report 2005/06 – 2008/09. 124. Overall, the learning outcomes have some important messages for policy makers. First of all, only a minority of students become proficient in basic literacy and mathematical skills. The educational system is not able to address or overcome the initial differences in reading comprehension and basic math skills, indeed, these initial differences are further intensified over time. A lot of the initial challenges and the general pressure on quality of learning come from the radically improved access measured by increased initial admission and overall enrollment rates. The improved access has brought into the system large groups of children who are now accessing education with various forms of additional challenges including: overage children, children coming from poor families, families where parents have little educational background and comprehension of the importance of education and children in poor, remote, rural communities and deprived districts where access to infrastructure, employment and services are limited. Many of these children have a high risk of falling into various zones of exclusion. Meanwhile, in an education system with its standards for transitioning students from one level to the next, there is a strong likelihood that many of those who have gained access to an education over the last decade will end up with very limited opportunities to truly benefit from it. C. Relevance 125. While measuring outcomes in access and learning is relatively straightforward, measuring and monitoring relevance is difficult. Nevertheless, in all societies, policy makers use a number of proxies to assess the relevance of education and, perhaps more importantly, to make critical assumptions about the areas where education should have its desired impact. For the individuals, education can bring improved living standards and earnings, for enterprises it can result in improved performance and productivity, for the economy, education can stimulate growth and competitiveness and bring more jobs. For communities and society at large, education can result in better cohesion and social inclusion, as well as a number of other benefits. These potential gains are equally critical for Ghana, which intends to reach middle income status, accelerate and sustain growth, alleviate poverty and address persistent inequalities between various social and ethnic groups. In this sense, the present Government has continued its predecessor’s focus on improving education outcomes and has even strengthened its resolve to assure that these outcomes are distributed among a wider range of citizens and communities. General Context: Relevance of Ghana’s Education Sector 126. CREATE expresses the fundamental objective of any education system this way: “sustained access to meaningful learning that has value is critical to long term improvements in productivity, the reduction of inter-generational cycles of poverty, demographic transition, preventive health care, the empowerment of women, and reduction in inequality� (Akyeampong et al. 2007). A senior official of the Ghana Education Service stated that most children, who after completing JHS are not able to continue their education, have difficulties integrating into society. The Government characterizes unemployed youth who complete 6 years of primary and 3 years of JHS as “NEET� people, i.e. no education, employment, training. The “NEET� characterization raises questions about the relevance of basic education and also the condition of the labor market that provides limited options for this youth. 127. Language of instruction. English is the principal language of school instruction. However, in the first few years of basic education the local language is used and gradually becomes phased out as pupils progress through primary. When teachers are 49 able to speak the local language, this has shown to improve learning outcomes in the initial years of schooling (World Bank 2004a). As the instructional language is different in different areas of the country, some teachers deployed from other regions have been known to have difficulties communicating with their students. From a strategic perspective, a long-standing policy of the Ghana Education Service (GES) has been a bilingual transitional language approach, through which local language would constitute 80% of instruction in primary 1 and English would be the only language of instruction by primary 4. This policy has been confronted with severe implementation constraints, partially due to a either a lack of appropriate learning material or a lack of teachers who speak the local language where they are posted, or a combination of the two. Economic Externalities 128. A relevant education is pivotal in fostering a smooth transition of young adults from the education system to the labor force, and in training them to enter a versatile and competitive labor market, both nationally and internationally. However, much can be done in Ghana’s education system to improve its skills base. A White Paper, based on the Ministry’s Education Strategic Plan 2003-2015, made the case for a much stronger and targeted focus on technical, vocational, and agricultural education and apprenticeships (MOESS 2004). In the education sector performance review 2006, the Ministry indicated that evidence exists of a large discrepancy between the outcomes of education institutions and the needs of the labor market (MOESS 2006). Thus, by increasing education opportunities to develop needed skills, Ghana will be better equipped with the human capital needed to advance to middle-income status (World Bank 2008). 129. In Ghana, 74.8 percent of women and 79.7 percent of men age 15-49 are currently employed, according to the Ghana Demographic and Health Survey, 2009. As data from the survey shows, being currently employed appears, ironically, to be inversely related to the level of one’s education, falling from 88 percent among women (96.3 percent of men) who don’t have any education to about 63 percent among women (73.1 percent of men) who have secondary education or higher (Table 2-20). In part, this can be explained by the fact that the majority of the labor force is employed in agriculture, which generally requires no or only minimal formal education. Table 2-20: Employment Status and Level of Education in 2008, Female/Male Employed in the 12 months Not employed preceding the survey in the 12 months Number Currently Not currently preceding the of employed Employed survey Total women/men Women No education 88 2.6 9.3 100 1,042 Primary 79.2 2.2 18.7 100 988 Middle/JSS 70.8 2.6 26.6 100 2,039 Secondary+ 62.7 5.4 31.9 100 844 Total 74.8 3.0 22.2 100 4,916 Men No education 96.3 1.2 2.4 100 540 Primary 76.5 1.5 21.9 100 619 50 Middle/JSS 76.1 2.3 21.6 100 1,721 Secondary+ 73.1 4.7 22.2 100 1,167 Total 79.7 2.7 17.6 100 4,568 Source: Based on Ghana Demographic and Health Survey, 2009. Note: Currently employed is defined as having done work in the past seven days. Includes persons who did not work in the past seven days but who are regularly employed and were absent from work for leave, illness, vacation, or any other such reason. 130. Having only a primary or junior secondary education (or less) is generally the norm among self-employed Ghanaians, while wage workers tend to have made greater investments into education – based on the Ghana Living Standards Surveys (GLSS) conducted in 1991/92, 1998/99, and 2005/06. Thus, when controlling for other background traits, the probability of being a wage worker in an urban area is higher as the level of education increases, and this appears to be consistent over time. The higher the level of one’s education, the greater is the likelihood of moving out of the agriculture sector and into wage or self-employment in services or industry (GLSS 2005/06). Figure 2-26: Impact of Education on the likelihood of Wage Employment (vs. no Education) and Non-Agriculture Self-Employment (vs. no Education), Urban Impact of Education on the Likelihood of Wage Impact of Education on the Likelihood of Non- Employment (vs. no Education), Urban Agriculture Self-Employment (vs. no Education), Urban 16 0.6 14 12 0.4 1991/92 10 1998/99 8 0.2 6 2005/06 4 1991/92 0 2 0 1998/99 -0.2 -2 2005/06 -0.4 -0.6 -0.8 -1 Source: World Bank 2008 based on GLSS 1991/92, 1998/99, and 2005/06. 131. Based on the findings of the GDHS 2009, the vast majority of men (about 84 percent) and women (about 59 percent) who have no education are active in the agriculture sector, which represents about 40 percent of GDP. For those with at least a secondary education, about half of women (48.7 percent) were engaged in the sales and services sector and about a quarter (27.4 percent) had a professional, technical or managerial occupation. The corresponding proportions for men were 21.3 and 35.1 percent, respectively. 132. Occupation is not only closely linked to education attainment, but also to levels of income, with Ghanaians in the poorest quintile being principally employed in the agriculture sector and those in the richest quintile being mainly engaged in sales and services or professional, technical or managerial activities. 51 Table 2-21: Level of Education and Occupation in 2008, Male/Female Professional/ Sales Un- Number technical/ and Skilled skilled Agri- of women managerial Clerical services Manual manual culture Missing Total /men Women No education 0 0.1 33.6 6.7 0 58.5 1 100 945 Primary 0.5 0 54.2 11.1 0.2 33 0.9 100 803 Middle/JSS 1 0.3 62.1 14.2 0.1 20.7 1.7 100 1,496 Secondary+ 27.4 11 48.7 7.8 0.5 2.7 1.8 100 575 Total 4.6 1.8 51.4 10.7 0.2 29.9 1.4 100 3,822 Men No education 0.5 2.1 3.7 8.4 0 84.1 1.2 100 527 Primary 1.7 6.7 8.5 20.4 0.3 58.5 4 100 483 Middle/JSS 3.1 13.1 10.8 30.1 1.2 35.9 5.9 100 1,349 Secondary+ 35.1 7.6 21.3 17.4 1.8 12.7 4.1 100 908 Total 11.4 8.8 12.2 21.6 1 40.6 4.3 100 3,276 Source: Based on Ghana Demographic and Health Survey, 2009. 133. Level of education impacts earnings. In the chapter on disparities, we analyzed how income levels strongly influence access to education opportunities. This relationship goes in both directions. Based on the Ghana Living Standards Surveys (GLSS) of 1991/92, 1998/99, and 2005/06, Ghanaians with higher levels of education and training in both rural and urban settings tend to have higher earnings in all three years analyzed. Investments in education capital, which increase workers’ productivity and open doors to more lucrative employment opportunities, have a significant impact on wages and returns. Higher levels of education bring higher income levels by increasing people’s cognitive capacities and by making them more attractive to employers. Those who have primary education and, in many cases even lower secondary education, tend to receive only slightly higher returns when compared to Ghanaians without education. Basic education however opens options to transition and continue through the education system. GLSS data indicate that completing upper secondary or TVET education translates into considerably higher economic returns, both in rural and urban settings and for nearly all types of employment. Ghanaians who have tertiary education yield even higher gains, especially in wage employment (World Bank 2008). 134. Another analysis also using GLSS data supports the view that education yields positive returns at all levels and is incremental at higher levels. Over the decade between 1987/88 and 1998, the increase in levels of education contributed to about a third of the increase in per-capita income (Teal 2001). These findings from GLSS data offer strong economic grounds for further investment in education at all levels. 52 Figure 2-27: Wage Regressions and Non-Agriculture Self-Employed Regressions with Sample Selection, Ghana 1991-2006, Urban Wage Regressions with Sample Selection, Non-Agriculture Self-Employed Regressions with Ghana 1991-2006, Urban Sample Selection, Ghana 1991-2006, Urban 2.5 2 2 1.5 1.5 1 1 1991 1991 0.5 0.5 1998 1998 0 0 2005 2005 -0.5 -0.5 -1 Source: World Bank 2008 based on GLSS 1991/92, 1998/99, and 2005/06. 135. Completing an apprenticeship has gained importance over time, particularly for women, and is by far the most common way to acquire and develop skills for employment. Thus, ensuring the quality of apprenticeships remains a critical issue. They play a key role in widening non-agriculture employment opportunities for youth, and in increasing prospects of being self-employed, thus they also raise the probability of having a job. Many businesses -- and self-employed artisans -- train youth by means of apprenticeships and on-site training, accounting for about 80-90 percent of all basic skills training in Ghana (Atchoarena and Delluc 2001). Ghanaians who have an apprenticeship are found holding nearly all forms of employment, although they are most commonly self-employed in non-agriculture, or wage employed in the informal sector or in the private formal sector. 136. While 21.9 percent of males and 13.9 percent of females aged 15-30 had completed an apprenticeship by 1991/92, the practice increased such that 27.3 percent of males and 26.3 percent of females within this age group had done so by 2005/06 (GLSS). An analysis carried out in 2006 by the Centre for the Study of African Economies found that in Ghana’s urban labor market a third of the population aged 16-65 had acquired some type of training (Monk et al. 2007). In terms of earnings, an apprenticeship is commonly linked to lower economic returns, when compared to Ghanaians who have no apprenticeship (i.e. who have had other forms of education). Despite their current significance in developing skills and opening employment opportunities, formal apprenticeships are most likely to become less important, given that Ghana is a country aiming towards middle-income status (World Bank 2008), which will require a workforce with higher levels of formal education. 53 Social Impact, Social Relevance 137. Fertility rates are one of the many health indicators strongly related to levels of education. Women with no education give birth to about three times more children than women with secondary or higher education. Data from the Ghana Living Standards Measurement Surveys confirms that the amount of schooling a mother receives is a strong predictor of how many children she has later in life. Figure 2-28: Level of Education and Fertility Rates, 2008 4 Total Total fertility rate 2.1 Secondary+ 3.5 4.9 Middle/JSS 6 Primary 7.3 Percentage of women age 15- 4.9 No education 49 currently pregnant 7.4 7.4 9 5.2 Mean number of children ever 3 born to women age 40-49 4.5 5.6 6.2 0 2 4 6 8 10 Source: Based on data from Ghana Demographic and Health Survey, 2009. Note: Total fertility rates are for the period 1-36 months preceding the interview. 138. Mother’s education related to share of births attended by a specialist. Among mothers without any formal education less than a third (only 29.7 percent) gave birth to a child under the attendance of skilled health personnel, based on demographic and health survey data. In contrast, about two thirds of mothers (67.9 percent) who reached secondary or tertiary education had given birth while attended by skilled health personnel (WHO 2008:92). 139. The median age at which women begin bearing children is also linked to education attainment, with women who have no education having their first birth on average at least 5 years earlier than women with secondary or higher education. About 2 percent of households surveyed stated that the reason their child was not enrolled in junior high school was due to pregnancy or already having a child, based on data from the Core Welfare Indicators Questionnaire Survey (GSS 2003). As Table 2-22 shows, the percentage of adolescents who have begun childbearing is many times higher among those with no education – 30.9 percent – than among those with at least secondary education – 1.3 percent. Table 2-22: Level of Education and Teenage Pregnancy/Motherhood, 2008 Percentage who: Percentage who have Have had a Are pregnant begun Number live birth with first child childbearing of women No education 23.4 7.4 30.9 72 Primary 18.8 7 25.9 222 Middle/JSS 7 2.4 9.5 571 Secondary+ 1.3 0 1.3 159 54 Total 9.9 3.4 13.3 1,025 Source: Based on Ghana Demographic and Health Survey, 2009. Note: Figures refer to women age 15-19. 140. Contraceptive use among married women varies with their level of education: about 13.6 percent who have no education currently use a contraceptive method and 0.3 percent specifically use a male condom, while 30.4 percent who have secondary or higher education use a contraceptive method of which 6.2 percent use a male condom. 141. Among women aged 15-49 with no formal education, the mean ideal number of children was 5.8, while it was 4.3 if they had primary education, and fell to 3.4 children if they had secondary education or higher. The desire to limit the number of children tends to be highest among women in urban areas who have secondary or higher education and come from the richest quintile. Table 2-23: Level of Education and Desire for no more Children, 2008 Number of living children 0 1 2 3 4 5 6+ Female No education .. 1 11.3 23.6 34.9 62.8 67.3 Primary .. 0 17.2 32 52.7 63 86.5 JHS .. 2.2 16.5 44 76.8 84.6 88.8 Secondary+ .. 2.9 32.7 46.1 84.7 90.3 100 Total .. 1.7 17.7 36.1 57.5 70.5 76.7 Male No education 0 0.5 10 9.4 31.3 23.2 39.7 Primary 0 0 21.6 29.6 43 37.7 51.8 JHS 0 4.4 24.1 41.4 56.1 52.1 71.8 Secondary+ 0 4.7 26.4 40.8 64.6 88.2 57.2 Total 0 3.6 22.4 33.7 50.5 47.6 56.6 Source: Based on Ghana Demographic and Health Survey, 2009. Note: Percentage of currently married women/men age 15-49 who want no more children. 142. A mother’s education is inversely associated with a child’s risk of dying. Among mothers with no education, there were 61 deaths per 1,000 live births; mother’s with JHS or with secondary education or higher had 46 and 49 deaths per 1,000 live births respectively. Under-five mortality sunk from about 10 to 6 percent for mothers with secondary or higher education compared to those without any exposure to the education system (Table 2-24). Table 2-24: Mother’s Education and Early Childhood Mortality Rates, 2008 Post- Under- Neonatal neonatal Infant Child five mortality mortality mortality mortality mortality No education 38 23 61 44 102 Primary 35 20 55 35 88 Middle/JSS 23 23 46 23 68 Secondary+ (38) (11) (49) (15) (64) Source: Based on Ghana Demographic and Health Survey, 2009. 55 Note: Deaths are per 1,000 live births. Numbers in parentheses are based on 250-499 unweighted exposed persons; an asterisk indicates that a rate is based on fewer than 250 unweighted exposed persons and has been suppressed. 143. Levels of education are associated with the prevalence of malnutrition among children under age five in Ghana. Well educated mothers are six times less likely to have severely underweight children: 3.7 percent of children born to women with no education are severely underweight compared to only 0.6 percent of children were whose mothers have secondary education or higher. But of course more educated mothers also generally come from richer families and hence are more capable to buy more food for their children (GHDS 2009). 144. GDHS data show that better educated women have a higher tendency to influence personal and household decision-making, partially due to being able to better inform themselves and communicate in a more sovereign way. 41.7 percent of women without education participated in making four specific decisions (on health, daily and major purchases and visits outside the home), while 56.6 percent of women did, if they had received at least a secondary education. Table 2-25: Women’s Participation in Decision-Making and Level of Education, 2008 Specific decisions Making Percentage Percentage Making purchases Visits to who who Own major for daily her family participate in participate in Number health household household or all four none of the of care purchases needs relatives decisions decisions women No education 65.4 56 76 82 41.7 8.1 853 Primary 68.1 60.7 78.5 80.2 45.7 8.3 638 Middle/JSS 69.9 65.3 81.5 84.1 49.5 5.3 1,058 Secondary+ 75 72.1 85.1 90.8 56.6 3.9 325 Total 68.8 62.3 79.6 83.4 47.2 6.6 2,876 Source: Based on Ghana Demographic and Health Survey, 2009. Conclusions on Quality of Learning 145. As overall access to education improves -- especially at primary and basic levels - - the Government and other stakeholders need to pay increasing attention to quality issues. Even if increasing number of children find ways to gain admission and increasing proportion of them are enrolled, new forms of exclusion emerge as many of them simply do not attend classes, learn enough, drop out, repeat, fail to pass exit exams or fail to transit to post-primary levels. International assessments show that big disparities in learning outcomes in a country pull down overall performance even if there are some significant groups that demonstrate impressive achievements, claim successful academic or professional careers. 146. Ghana’s commitment to education is show by its willingness to participate in international learning assessments, by the national assessments of learning outcomes and by the numerous innovations that are sponsored and promoted through Government policies or non-governmental initiatives. These will gradually direct the public attention and the accountability of education providers to quality outcomes. 56 147. This report shows that while the resources for education are impressive, they are not distributed equitably, targeted according to need and used efficiently. Quality standards are defined in terms of input with little understanding as to which inputs and which methods have the most impact on learning outcomes. Among the inputs, two key items appear to show major impact: the access to quality teaching and quality teachers and the instructional time. 148. Based on the findings we identify three levels of potential and desirable policy actions: a. Continue strengthening the Government capacities and increase investment in monitoring learning outcomes. While there is an impressive array of innovations and options for policy to improve quality, it is impossible to assess their effectiveness and scale up the successful ones if learning outcomes are not measured by using solid international and reliable national standards and methodologies. b. Assure that resources with strong impact on quality reach the classrooms. The available assessments confirm that trained teachers have the biggest impact on learning. Further, there are indications that high level of absenteeism, low time on task significantly compromise learning. To overcome these problems, both the central and local governments as well as the communities need to be given authority, improve capacity to supervise, monitor deployment of teachers, attendance and, indeed, classroom activities. Standards, expectations and regulations need to be stringent to avoid that new generations of students complete education without effective learning. c. Give central support to local innovations. Improving quality in crowded urban schools requires different approach than in small rural schools often with limited infrastructure, facilities shared by different grade levels, poor communities and local language and cultural conditions. The Government needs to provide services to schools with a variety of conditions, empower and motivate local governments, communities and non-governmental agencies to initiate these innovations and monitor their impact. 57 3. INPUTS AND RESOURCES 149. In general, the Government and service providers need to ensure that all necessary inputs, supply of infrastructure, teaching and learning materials and methodologies, as well as human resources, are made equally available to all students and schools. There are general standards and guidelines for most inputs, including standards of school construction, textbook ratios, teaching methodologies and standards for management. However, even in the context of Ghana’s extraordinary support for education, planners struggle to supply even a minimum of inputs to all schools. The challenge is complicated by the lack of evidence on the effectiveness of various inputs on learning outcomes, making it hard to to decide where to concentrate limited resources and funding. In addition, needs are not uniform. For instance, small remote schools in rural areas with high poverty incidence need a different scale and type of resources than large urban schools with diverse student populations. Table 3-1: Examples for inputs influencing quality outcomes Input Description Physical infrastructure Schools Sufficient number of schools so that pupils do not have to walk large distances to reach school. Classrooms Sufficient number and size of classrooms and pavilions that can be used when raining and that have internal walls to reduce noise in shared classes. Water and toilets Schools with their own supply of running water and separate toilets for both girls and boys. Electricity Access to electricity allows to use light and, increasingly, technologies like computer and other electrical and electronic aids. Other facilities These may include teacher quarters, teacher rooms, library, etc. Materials Textbooks and learning materials Adequate ratio of core textbooks (e.g. math and English) per pupil and policies ensuring their usage. Desks and seats Ratio of seats and writing places per pupil. Chalkboard Accessibility and quality of chalkboards. Chalk Access to sufficient chalk. Teachers and teaching methods Teacher training Minimum number of teachers and equitable share of trained teachers. Methods Ensuring teachers are trained to implement improved, modern and student-centered teaching methods, and that they demand frequent homework. Attendance Measuring attendance and absenteeism among teachers and students. Time-on-task Ensuring time spent in school is used for learning. Curriculum Ensuring locally-relevant curriculum. Teacher morale Working and living conditions influence level of teacher morale. Financial Resources Per Capita Expenditures Expenditures show no disparities, distributed equally to a level where effective decisions about effective use of funds Flow of Funds make financial resources available in a timely manner, allowing pooling resources and efficiency gains Incentives, programmatic, financing Policies are costed, incentives established and financed to compensation schemes stimulate desired behavior, innovation, accountability 58 150. As subsequent sections on resource allocations and resource use will show, public resources cannot provide each school and each classroom with all necessary inputs at an acceptable standard. It is inevitable that some schools will enjoy less resources than others. At the same time, it will be necessary for some resources to be prioritized and adjusted to local conditions, circumstances and need. For example, various incentives may be needed to get teachers to agree to being deployed in remote areas. Likewise, efforts to minimize teacher absence may be more important in improving learning outcomes, than providing an adequate number of classrooms. Therefore, decisions about allocating resources, setting up services to education, and establishing priorities need to be made along the following principles: a. they have to be evidence based, b. they have to create adequate incentives for the key participants including students, parents, teachers, head teachers, and district officials, and c. they have to be targeted so that those with the greatest need get the largest share of public resources. A. Physical Resources: Schools, Infrastructure and Textbooks 151. Basic education is largely provided by small and medium sized schools. Enrollment in the median KG is 80 (mean 81.4), in Primary school 193 (mean 207.5), and JHS 112 (mean 125.9). At all three levels, more than half of the schools have below average enrollment. Junior high schools experienced a significant increase in enrollment in the past years, a phenomenon that affected disproportionately very small schools, the proportion of which decreased significantly between 2005/06 and 2008/09. Figure 3-1: Number of Pupils per School at the Basic Level 250.0 Preschool Primary JSS 204.0 205.1 208.8 207.5 200.0 169.2 167.4 174.8 160.7 150.0 125.6 125.9 116.1 119.0 125.1 114.2 109.9 112.5 100.0 81.5 83.7 80.3 81.4 72.9 68.8 62.9 62.1 50.0 0.0 2001/02 2002/03 2003/04 2004/05 2005/06 2006/07 2007/08 2008/09 Source: Based on EMIS. 152. Prior to systemic efforts by the Government, communities were responsible for constructing their own schools, the majority of which were made of adobe and had a short lifespan. A World Bank (2004a) study reported that in 1988 fewer than half of primary schools (47.6 percent) were able to utilize all their classrooms when it rained, but improvements in school structures meant that by 2003 more than two-thirds (68.1 percent) could do so. School facilities have a direct impact on enrollment figures and, although to a lesser extent, their quality is known to affect school performance. Between 1988 and 2003, 4 percent of enrollment growth (about one third of all growth) was 59 attributed to new and better school facilities, for instance, by shortening distances to school (World Bank 2004a). 153. The Government has adopted a policy to ensure a preschool facility be attached to primary schools across the country. Over the last four years, the number of public KGs grew by 40 percent but, to date, 28.1 percent of preschools are private. This high rate shows that the Government needs to strengthen its commitment to nationalize kindergartens in order to improve equity. Table 3-2: Preschool Facilities 2001/0 2002/0 2003/0 2004/0 2005/06 2006/07 2007/08 2008/09 2 3 4 5 * * * * Schools 9,634 11,177 10,133 11,781 11,931 13,750 15,449 16,439 w/ toilets 4,903 5,002 5,130 5,978 6,585 7,608 8,839 9,684 w/ potable water 4,760 4,751 4,887 6,305 6,766 8,675 10,599 11,137 average # of pupils per 72.9 68.8 62.9 62.1 83.7 80.3 81.5 81.4 preschool % which are private 34.4% 32.1% 36.6% 43.4% 27.4% 27.2% 27.9% 28.1% % w/ Toilets 50.9% 44.8% 50.6% 50.7% 55.2% 55.3% 57.2% 58.9% % w/ Potable Water 49.4% 42.5% 48.2% 53.5% 56.7% 63.1% 68.6% 67.7% Note: *KG data used due to lack of preschool data. While preschool population is 0-5yrs, KG population is 4-5yrs. Source: EMIS. Figure 3-2: Trends in Preschools with Toilets and Potable Water 80% Preschools 18000 % of Schools w/ Toilets 70% % of Schools w/ Potable Water 16000 60% 14000 12000 50% 10000 40% 8000 30% 6000 20% 4000 10% 2000 0% 0 2001/02 2002/03 2003/04 2004/05 2005/06 2006/07 2007/08 2008/09 Source: Based on EMIS. 154. The average number of pupils per primary school is 207.5, while the number of pupils per private primary schools is 153. About 19.3 percent of all primary schools were private in 2001/02, but this share rose to 24.4 percent by 2008/09. 155. Between 2001/02 and 2008/09 the number of primary schools increased from 15,285 to 17,881. This increase represents a large expansion in the supply of education services, with an average growth rate of 2.4 percent of new primary school constructions per year. Given the annual population growth is officially estimated at 2.7 percent, the government’s efforts to construct new schools have not been able to keep pace with natural growth. As a result, some schools have become overcrowded. Table 3-3: Primary School Facilities 2001/02 2002/03 2003/04 2004/05 2005/06 2006/07 2007/08 2008/09 60 Total Schools 15,285 15,708 16,046 16,760 15,307 16,410 17,315 17,881 w/ toilets 8,703 7,955 7,819 8,150 7,990 8,745 9,679 10,370 w/ potable water 6,553 6,241 6,472 7,929 8,183 10,004 11,678 11,890 # of pupils per primary school 169.2 160.7 167.4 174.8 204.0 205.1 208.8 207.5 % which are private 19.3% 20.4% 22.4% 25.0% 20.1% 21.5% 23.5% 24.4% # of pupils per private primary 160.2 125.9 136.1 143.5 154.3 153.4 153.7 153.0 % w/ toilets 56,9% 50,6% 48,7% 48,6% 52,2% 53,3% 55,9% 58.0% % of private w/ toilets 17.2% 14.7% 16.3% 21.7% 79.8%8 79.4% 81.5% 81.2% % w/ potable water 42,9% 39,7% 40,3% 47,3% 53,5% 61,0% 67,4% 66.5% % of private w/ potable water 65.1% 53.2% 56.7% 65.1% 72.4% 77.3% 81.9% 79.9% Source: EMIS. 156. While the average number of students per private JH school is 86.6, the average number of students for all JH schools is 125.9. About 15 percent of all junior secondary schools were private in 2001/02, but this share increased to 25 percent by 2008/09. Between 2001/02 and 2008/09 the number of JH schools increased from 7,582 to 10,213, which reflects an annual rate of 5.0 percent. Among all JH schools 62.2 percent have toilets and 67 percent have access to potable water. Table 3-4: JHS School Facilities 2001/02 2002/03 2003/04 2004/05 2005/06 2006/07 2007/08 2008/09 Total Schools 7,582 7,875 8,170 8,699 8,749 9,054 9,742 10,213 w/ toilets 4,012 3,973 3,916 4,090 4,875 5,195 5,844 6,348 w/ potable water 3,483 3,388 3,604 4,399 4,736 5,612 6,677 6,845 # of pupils per JHS 114.2 109.9 112.5 116.1 119.0 125.1 125.6 125.9 % which are private 15.4% 16.5% 19.8% 22.6% 18.5% 21.3% 23.8% 25% # of pupils per private JHS 105.9 95.6 89.9 95.5 97.6 93.3 89.9 86.6 % w/ toilets 52,9% 50,5% 47,9% 47,0% 55,7% 57,4% 60,0% 62.2% % of private w/ toilet 12.3% 12.2% 14.8% 18.1% 84.9% 85.1% 85.7% 85.9% % w/ potable water 45,9% 43,0% 44,1% 50,6% 54,1% 62,0% 68,5% 67.0% % of private w/ potable water 72.3% 63.3% 63.4% 71.5% 74.4% 78.8% 83.2% 81.3% Source: EMIS. Figure 3-3: Trends in the Number of Pupils per JHS, Public/Private 160 140 136.8 139 133.7 120 122.2 123.9 115.7 112.7 118.1 105.9 100 95.6 95.5 97.6 89.9 93.3 89.9 86.6 80 60 40 Private 20 Public 0 2001/02 2002/03 2003/04 2004/05 2005/06 2006/07 2007/08 2008/09 Source: Based on data from EMIS. 157. Since the school year 2001/02 the percentage of public junior high schools with toilets has decreased by 8.3 percent and those with drinkable water have increased by 8 This large jump suggest data reliability problem. 61 22.9 percent. As to private junior high schools, those with toilets have sharply risen by 73.4 percent and those with potable water have increased by 10.9 percent. Textbooks 158. Core textbooks especially for English and mathematics at primary level are essential components of the Government’s strategic agenda. A comprehensive evaluation of basic education in Ghana found that widening access to textbooks contributed to improvements in English and math scores between 1988 and 2003 (World Bank 2004a). The study reported that in 1988 only 21 percent of primary schools had at least one English textbook per pupil; by 2003 the share had risen to 72.4 percent (Table 3-5). At the junior secondary level, the share of schools that had at least one math textbook per pupil grew from 13 to 71 percent over this period. The World Bank was the main international provider for textbooks during this 15 year period, responsible for the provision of nearly 35 million textbooks. Table 3-5: Trends in English and math textbook distribution English Mathematics Primary JHS Primary JHS 1988 2003 1988 2003 1988 2003 1988 2003 Less than one book between two 58.4 11.3 42.1 22.1 35.7 11.0 34.3 5.5 At least one book between two 20.6 16.3 37.3 42.6 32.9 35.7 52.4 23.2 At least one book per student 21.0 72.4 20.6 35.3 31.5 53.2 13.3 71.3 Total 100 100 100 100 100 100 100 100 No. of observations 286 417 233 289 286 417 233 289 Source: Based on World Bank 2004a. 159. During the last decade the textbook per student ratio revolved around two core textbooks per pupil in primary schools and mildly improved from just below 5 books to over 6 books for JH schools. The pattern of textbook supply shows slow decline until sudden improvement in 2005/06 for primary and 2006/07 for JHS after which between 2006/07 to 2008/09, the ratio declined again. During that time, the average number of textbooks per pupil in primary schools declined more in deprived district than in non- deprived. There are districts, such as Afram Plains Kwahu North District (Eastern) or Tolon-Kumbungu District (Northern) that had on average less than 0.80 textbook per pupil in 2008/09. In 2009, a large number of core textbooks was procured and delivered after the EMIS census had taken place. This suggests that textbook provisions have a campaign-like nature with large procurement of textbooks from time to time, likely once in every four years, but either inefficient distribution or inefficient maintenance of textbooks. 62 Figure 3-4: Textbooks per Pupil at Basic Levels 8.0 7.4 7.0 Primary JSS 6.8 6.6 6.3 6.0 5.0 4.8 4.5 4.0 3.9 3.0 3.0 3.0 2.0 2.2 2.0 2.1 1.7 1.8 1.9 1.3 1.0 0.0 2001/02 2002/03 2003/04 2004/05 2005/06 2006/07 2007/08 2008/09 Source: Based on EMIS. Ghana School Feeding Program 160. The Ghana School Feeding Program (GSFP) started in 2005 with a pilot in 10 schools, one per region. In 2008, the program reached 560,000 students in all 170 districts, which accounts for about 14.3 percent of total primary school enrollment. Plans call for expanding the program to reach 1,040,000 students by the end of 2010. The 2008 budget of the program was 43.6 million GH ¢ but actual expenditures reached only 33.0 million GH ¢. The budget for 2009 at 51.1 million GH ¢ was more than 20 percent higher than the budget for 2008. 161. School level data from EMIS for 2005/06 to 2008/09 show the feeding program had only a small impact on enrollment – from a 5.86 increase in primary schools after three years, to a 2.9 percent decrease in JH schools after three years. However, the lack of more positive effects from the program may be due to initial problems with its implementation. Studies in other developing and developed countries have shown feeding programs to be an effective means of boosting both enrollment and student performance. 162. By using school level data from EMIS for 2005/06 to 2008/09 one can examine whether the Ghana School Feeding Program (GSFP) has had an impact on enrollment. GSFP had a positive impact on enrollment in the second and third year of its introduction. In the first year as well as in the fourth year, the effect was positive but not statistically significant. In schools with meals program for three years, overall enrollment increased by 5.86 percent. The increase in primary one enrollment was about a percentage point less. 163. School meals do not have a statistically significant effect on overall JHS enrollment. The school meals program had a positive impact on enrollment only in the first year of its introduction in JHS1, of a magnitude of 2.8 percent. In the subsequent years, the effect was negative in JHS1. In schools with meal programs for three years, overall enrollment decreases by 2.9 percent while in those with the program for two years, overall enrollment decreases by around 1.6 percent. It should be noted that in those schools with both primary and JHS levels, it is possible that school meals are provided only to the primary level and not JHS. This could possibly explain the results. In addition, 63 the school meal program has no statistically significant effect on non-promotion in primary (Wodon and Joseph 2010; Annex.. for regression results). B. Human Resources: Teacher Supply, Allocation, Policy 164. The Government sets the targets for pupil teacher ratios. According to the 2004 ESP, at kindergarten level, the 2001/02 baseline and the long-term target is 25. However, the sharp increase in enrollment numbers has placed greater pressure on pupil teacher ratios in pre-primary, which rose to 35.5 (2008/09). In primary schools, the strategic goal was 35, whereas the average PTR was 32.4 in 2008/09. According to UNESCO, sub- Saharan Africa had an average pupil teacher ratio in primary of 45 for the school year ending in 2006; Ghana’s primary classrooms have on average about 10 students less per teacher. At the global level, the average share of primary students per teacher was 25 in 2006, and among developed countries it was 14 (UNESCO 2008:118). 165. The primary PTR hasn’t changed during the decade despite sharp increase in enrollments, thanks to efforts to recruit a large number of new teachers, most of them unqualified (to be analyzed later). As a result, both the number and proportion of unqualified teachers grew mostly in line with the large increases in student intake. Figure 3-5: Trends in Primary Enrollment and Share of Unqualified Primary Teachers Total Primary Enrolment % of Unqualified Primary Teachers 4,000,000 60% 3,500,000 52.0% 51% 50% 3,000,000 47% 42% 44% 37% 39% 40% 2,500,000 35% 2,000,000 30% 1,500,000 3,122,903 3,616,023 2,586,434 2,686,133 20% 1,000,000 3,710,647 2,524,585 2,929,536 3,365,762 10% 500,000 0 0% 2001/02 2002/03 2003/04 2004/05 2005/06 2006/07 2007/08 2008/09 Source: Based on EMIS. 166. The national PTR in junior high schools averaged 17 in 2008/09, considerably lower than the Government’s target rate of 25 and lower than the 2001/02 baseline of 18.2. As the Figure shows, the PTR for JHS have historically been very low: while this could be positive in terms of quality, it is extremely expensive. For all three levels, PTR peaked in 2005/06 and have slowly declined since. For primary and JHS levels, the present PTRs are below the Government targets, while for KG, it is above target. 64 Figure 3-6: Trends in PTR at the Basic Level, 2001/02-2008/09 40 -- Government target for 2015 35 -------------------------- Primary, 35 30 25 -------------------------- --------------------------- Kindergarten, 25 Junior High School, 25 20 15 10 5 0 2001/02 2002/03 2003/04 2004/05 2005/06 2006/07 2007/08 2008/09 2001/02 2002/03 2003/04 2004/05 2005/06 2006/07 2007/08 2008/09 2001/02 2002/03 2003/04 2004/05 2005/06 2006/07 2007/08 2008/09 Source: Based on EMIS. Note: KG data for 2001/02 and 2002/03 are preschool data. 167. Teachers in Kindergarten. In 2008/09, 75% of the 37,700 kindergarten teachers were unqualified. In the public sector, about 31 percent of all kindergarten teachers are qualified, while in the private sector only 7 percent are qualified. The large gap in trained KG teachers is attributable to its relatively recent introduction. PTR for KG appear to have stabilized from 2007/08 onwards since KG was declared part of the basic education cycle and every primary school was deemed to have an attached Kindergarten. 168. As we have seen in the previous chapter, the private kindergartens are more prevalent in the rural areas where public services are not available, illustrating that the policy is yet to be implemented successfully. Therefore, these schools can be safely considered as weak substitution for the higher quality public services. While the pupil trained teacher ratio (PTTR) in kindergarten averaged 137 at the national level in 2008/09, there were about 457 private preschool attendees for every trained teacher. Table 3-6: Indicators on Kindergarten Teachers 2001/02* 2002/03* 2003/04 2004/05 2005/06 2006/07 2007/08 2008/09 TEACHERS 27,882 28,382 23,2489 32,712 26,580 34,140 35,736 37,713 % of Private 31.7% 30.0% 39.6% 40.8% 23.1% 20.7% 23.2% 22.0% % of Untrained 77.8% 78.7% 74.3% 80.3% 72.6% 69.2% 74.7% 74.1% % of Untrained Public 71.0% 72.5% 62.1% 71.4% 66.9% 64.3% 69.1% 68.7% % of Untrained Private 92.6% 92.9% 92.8% 93.1% 91.6% 87.8% 92.9% 93.2% PTR 25.2 27.1 27.4 22.4 37.6 32.8 35.2 35.5 PTTR 113.5 127.0 106.5 113.4 137.2 105.1 139.0 136.9 PTR in Private 28 30 20 19 31 29 29 31 PTTR in Private 373.6 418.0 276.9 272.2 372.5 241.3 410.0 456.8 PTR in Public 24 26 32 25 39 33 37 37 Source: EMIS. Note: *KG data indicated for 2001/02 and 2002/03 illustrates preschool data. 169. Teachers in Primary Schools. As mentioned earlier, between 2001/02 and 2008/09 the proportion of untrained primary teacher grew rapidly, given that the total number of teachers increased by nearly 42 percent (from about 80,500 to 114,400), while 9 The reasons for the apparent big drops in the teacher numbers in 2003 and in 2005 are unclear, possibly data error. 65 the total number of unqualified teachers more than doubled between the same period (from about 28,200 to 59,500). The proportion of untrained primary teachers was 35.1 percent in 2001/02 and worsened to 52 percent in 2008/09; 42 percent in the public schools and 88 percent in the private schools. Table 3-7: Indicators on Primary Teachers 2001/02 2002/03 2003/04 2004/05 2005/06 2006/07 2007/08 2008/09 TEACHERS 80,552 80,459 82,833 89,278 88,461 105,257 112,443 114,421 Untrained Teachers 28,238 29,866 32,296 37,608 38,654 49,459 57,266 59,488 % of Private 31.7% 30.0% 39.6% 40.8% 23.1% 20.7% 23.2% 22.2% % of Untrained 35.1% 37.1% 39.0% 42.1% 43.7% 47.0% 50.9% 52.0% % of Untrained Public 22.8% 26.5% 26.2% 27.7% 32.9% 37.9% 40.6% 41.6% % of Untrained Private 83.0% 84.7% 84.5% 85.0% 85.5% 83.6% 87.6% 88.4% % of Female 32.1 32.8 32.0 31.1 33.9 27.7 33.0 33.9 PTR 32.1 31.4 34 34.9 35.7 33.5 34.1 32.4 PTTR 49.4 49.9 53.2 56.7 62.7 60.3 65.5 67.5 PTR in Private 29 27 27 27 26 26 25 26.3 PTTR in Private 170.3 178.6 174.0 178.3 180.2 157.6 203.7 227.6 PTR in Public 33 32 34 35 38 33 34 34.2 Source: EMIS. 170. Ghana has not been able to hire enough qualified teachers to keep up with the rapid expansion of enrollments and has had to rely on many new teachers without training. Ghana now lags behind the Sub-Saharan average of 85 percent trained teachers. While the PTR is in Ghana is significantly lower than the SSA average, the PTTR is higher. Figure 3-7: International Trends in Trained Primary Teachers and Primary PTRs % of Trained Primary Teachers 06’ Primary PTR 06’ % of Trained Primary Teachers 100 70 100 99 96 94 92 90 87 86 85 85 63 60 80 72 Primary PTR 06' 53 70 59 50 45 60 40 40 40 44 50 40 39 37 50 31 35 30 40 24 27 30 21 20 20 10 10 0 0 ts r ria ad or ia n La na Vi y a PR A Sa m n Pa A Bo g e ni ta an AN ad SS El n a ge Ch n wa Be Ni D kis Ke nz lv et H Ni o G Ta Source: Based on data from UNESCO, 2008. 171. Disparities in pupil per trained teacher ratios (PTTR) within Ghana are even more significant. While the national PTTR for primary level in 2008/09 was 67.5, in about one quarter of the disctricts, the average PTTR was 100 and within ten of these districts there were over 155 students for every trained teacher. All of the these were located in the 66 Northern, Upper West, Upper East and Western regions. On the opposite end, 12 districts with a public primary PTTR of 35 or below were found, all within the Eastern, Volta and Ashanti regions. A glance at Map 3-1 verifies that the strongest deployment and retention of trained teachers is found in these three regions and the Greater Accra area. 172. There is significant unwillingness among trained teachers to except postings in more rural areas, for instance, in the Northern region, where 5 of the 10 districts with public primary PTTRs of 155 or greater are located. In 2008/09, the district-level pupil trained teacher ratio in public primary schools ranged widely from 23 to 390. Trained teachers are not inclined to accept postings in many of the districts afflicted with diverse infrastructure deficiencies as well as socio-cultural and geographic hurdles. Trained teacher retention is particularly low in communities affected by poor infrastructure, whereby many teachers lack basic access roads to commute to remote schools. They are less motivated to instruct in environments with a low prevalence of accessible drinking water. They have fewer incentives to reside in communities with limited and in some cases no access to a functioning grid line or with long distances to a hospital or police services. Especially in areas of the North, the fact that many teachers have scarce access to communication facilities helps to deter their retention and, to a larger extent, hinders their initial deployment. The Northern region’s Gushiegu district, where only 21.4 percent of primary teachers had formal qualifications in 2008/09, is known as a conflict zone that consequently experiences many difficulties in attracting qualified – and even unqualified – teachers. Map 3-1: District Level average PTR in Primary Schools, 2008/09 Source: Based on EMIS. Note: Data is for public schools. 67 The following table shows the districts with the highest and lowest primary PTTRs. Table 3-8: 20 Districts with the Highest and Lowest Primary PTTRs, 2008/09 Districts with Highest Primary PTTR Districts with Lowest Primary PTTR 1 Bia, WESTERN 450.9 1 Ho, VOLTA 26.6 2 Karaga, NORTHERN 219.9 2 Akwapim North, EASTERN 28.9 3 Amenfi West, WESTERN 203.4 3 Asuogyaman, EASTERN 34.6 4 Saboba-Chereponi, NORTHERN 200.7 4 Hohoe, VOLTA 35.0 5 Wa West, UPPER WEST 182.0 5 Yilo Krobo, EASTERN 35.3 6 Yendi, NORTHERN 181.8 6 South Dayi, VOLTA 36.1 7 Gushiegu, NORTHERN 175.7 7 North Dayi, VOLTA 36.6 8 East Gonja, NORTHERN 175.4 8 East Akim, EASTERN 38.3 9 Garu-Tampane, UPPER EAST 171.7 9 Akwapim South, EASTERN 39.1 10 Juabeso, WESTERN 169.9 10 Adaklu-Anyibe, VOLTA 39.6 Source: Based on EMIS. Note: Data includes both public and private schools. Teachers for Junior High School Education 173. The proportion of untrained teachers in JHS increased between 2001 and 2008 from one fifths to over a third while their total number grew by 160 percent indicating that as enrollment grows, so does the share of untrained teachers. Their proportion is three times larger in private JH schools than in public JH schools. Table 3-9: Indicators on JHS Teachers 2001/02 2002/03 2003/04 2004/05 2005/06 2006/07 2007/08 2008/09 TEACHERS 47,445 49,270 51,419 55,958 56,485 67,005 73,656 75,409 Unqualified Teachers 10,138 12,585 13,807 16,380 16,565 21,692 25,653 26,427 % of Private 15.7% 15.7% 19.0% 22.4% 18.9% 19.1% 20.8% 21.5% % of Untrained 21.4% 25.5% 26.9% 29.3% 29.3% 32.4% 34.8% 35.0% % of Untrained Public 12.5% 16.5% 15.8% 16.4% 18.9% 22.8% 23.6% 23.3% % of Untrained Private 69.1% 73.9% 74.0% 73.7% 74.0% 73.1% 77.6% 78.0% PTR 18.2 17.6 18.6 19.0 19.4 17.6 17.4 17.0 PTTR 23.2 23.6 24.4 25.5 26.1 25.0 25.5 26.2 PTR in Private 17 16 15 15 15 14 14 13.7 PTTR in Private 53.9 61.6 57.3 56.8 57.0 52.3 60.7 62.2 Source: Based on EMIS. 174. There are also significant regional variations in the share of trained JHS teachers. The Greater Accra region had by far the highest concentration of qualified teachers in public junior high schools; 93.7 percent of male teachers and 98.1 percent of female teachers were trained. In contrast, the Upper East had the lowest proportion of public instructors with formal qualifications, where 64.6 percent of male teachers and 77.5 percent of female teachers had acquired teacher training. It is also noticeable that among female teachers, the share of trained teachers is higher and in some deprived regions, such as the Upper East and to some extent in the North, the share of trained teachers in private schools relative to the share in public schools is much higher than in the rest of the country. 68 Figure 3-8: Trained JHS Teachers in 2008/09, Public/Private and Male/Female 100.0% 80.0% 60.0% 40.0% 20.0% 0.0% GREATER BRONG UPPER UPPER ASHANTI CENTRAL EASTERN WESTERN VOLTA NORTHERN ACCRA AHAFO WEST EAST Public M 93.7% 76.9% 66.6% 75.7% 71.5% 57.3% 77.0% 63.0% 61.7% 64.6% Public F 98.1% 92.8% 90.9% 89.1% 89.1% 87.4% 85.3% 80.0% 79.0% 77.5% Private M 28.4% 20.1% 23.7% 18.0% 14.0% 12.9% 14.2% 32.0% 16.0% 47.2% Private F 39.9% 21.3% 26.7% 16.8% 19.4% 19.9% 15.5% 30.0% 0.0% 57.1% Source: Based on EMIS. 175. At the senior high school level, 18 percent of teachers are untrained (14 percent at public and 42 percent at private schools). Both the share of untrained teachers and the pupil trained teacher ratio have declined in the last four years. Meanwhile, while efficiency of teacher distribution has declined, it is much higher than at various levels of basic education. Table 3-10: Number of SHS Teachers, Qualified/Unqualified 2005/06 2006/07 2007/08 2008/09 Total Teachers 15,658 20,055 21,449 22,521 Qualified Teachers 12,150 16,413 17,674 18,765 Unqualified Teachers 3,508 3,642 3,775 3,756 % of Unqualified 22% 18% 18% 16.7% % of Unqualified Public 20% 15% 14% 13.2% % of Unqualified Private 44% 42% 42% 42.2% PTR 24.6 24.2 21.2 21.8 PTTR 31.6 29.6 25.7 26.1 Source: EMIS. The Challenge of Getting Teachers Trained, Deployed and Motivated 176. A large number of unqualified teachers, about 71,000 in 2006-2007 and 81,000 the following year, are teaching in Ghanaian basic schools. These include (i) about 22,000 new graduates doing one year’s national service in Ghana’s schools; (ii) about 12,000 more, young and sometimes older, people teaching as volunteers under a National Volunteer Scheme and (iii) 9,000 students in the final year in teachers’ colleges teaching full-time in schools as part of their teachers’ course. In addition, qualified teachers are not deployed where they are needed and urban areas and richer districts benefit more than poorer, rural areas. 177. The Teacher Education Department (TED) of the GES manages and funds 38 teacher training colleges in different regions; they offer 2-year single-subject and 3-year dual-subject courses. These courses lead to the award of the Teachers’ Certificate A. In 2005, each of these colleges began to offer a full-time 3-year Diploma in Basic Education. Again, one of these years (the final year) is spent as a teacher-in-training in a school. The Diploma is intended to qualify graduates to teach in basic education 69 institutions (K1 to JHS3) and is accredited by the Institute of Education, University of Cape Coast (UCC). Applicants should have the equivalent of an aggregate score of 24 in six subjects (English, mathematics, science and three others) in the Senior Secondary School Certificate Examination. 178. The 38 public colleges of education graduate about 9,000 teachers each year for primary and junior high schools and from 2009 will provide graduates for kindergarten also. Given the current rates of teacher attrition, estimated at 5 percent, and the population growth, this should be sufficient to maintain the current ratios. The two education universities are graduating more than 3,000 young people annually in education and these are qualified to teach in senior high schools. Table 3-11: Enrollment in Teacher Training Colleges (TTCs) 2002/2003 2003/2004 2004/2005 2005/2006 2006/2007 2007/08 Total 8,167 8,980 8,989 8,538 9,401 Intake % Female 45 43.90 42 42.60 46.60 Total 19,764 23,999 25,116 25,534 26,025 26,100 Enrollment % Female 36 40.50 45 42.10 46.30 47.08 Teachers Produced 7,115 7,532 8,300 8,321 Source: Education Sector Performance Report 2008. 179. There are 6 public and 8 private universities in Ghana. Two of the public Universities, the University of Education, Winneba, and the University of the Cape Coast (UCC) specialize in education and UCC certifies the qualifications of the 38 teacher colleges at diploma level. Each of these 2 universities offers degrees in education. In addition, graduates from other universities can take a postgraduate certificate in education, qualifying them to teach. In 2008, UCC graduated 1,728 students from its whole-time education programs. 180. There are four channels teachers can use to improve their professional competence. First, “Certificate A� teachers can apply for paid study leave and upgrade to diploma or degree status. Second, the current “Certificate A� teachers can enroll in a sandwich program to upgrade to Diploma level. Third, unqualified teachers can enroll in a four-year distance teacher training program to obtain a diploma or degree qualification. A fourth channel available to all teachers is one-off in-service training to support the introduction of a new curriculum or a new topic within an existing curriculum, such as ICT. 181. Study leave has been available to Ghanaian teachers since 1997. At any one time, up to 15,000 teachers are enrolled in universities full-time to study for degrees. This is highly valued by teachers and their unions. The expectation is that teachers will return to teaching following the award of the degree and the Ghana National Association of Teachers (GNAT) insists that numbers returning are high. However, there is some dispute about the numbers that actually do return. The scheme is expensive as teachers continue to be paid their salaries for the duration and Government pays all fees and accommodation allowances. Table 3-12 shows the numbers on study leave to 2004. Since 2001 the numbers have been falling; in 2008, 3,000 teachers were awarded study leave and in all 12,000 teachers were out of their classrooms on paid study leave. 70 Table 3-12: Number of teachers on study leave in relation to output from training colleges Year No. graduating No. approved for % Equivalent of Pool of Available from TTCs in July study leave from TTC output TTC graduates Sept. approved for study leave 1997/98 6114 3149 52.0 +2965 1998/99 6858 4411 64.3 +2447 1999/2000 6417 6343 98.8 +74 2000/2001 6596 10103 153.0 -3507 2001/2002 6362 6247 98.2 +115 2002/2003 6955 6076 87.4 +879 2003/2004 4676 5000 107 -324 Average Addition to Teaching Service Each Year 378 Source: Akyeampong. 182. Sandwich program. The sandwich program for the Diploma in Basic Education is open to all currently practicing “Certificate A� teachers. It takes two years and consists of two 6-week semesters each year, with one semester for 6 weeks over the long vacation and the second in two 3-week periods over the Christmas and Easter vacations. It is managed and overseen by the Teacher Education Department of MOESS and delivered by and in a number of the colleges of education. The Institute of Education, UCC accredits the course. Since the inception of the program in August 2007, 10,950 teachers have been enrolled and were expected to complete the upgrading program by May 2009. It is estimated that there are approximately 60,000 “Certificate A� teachers who are expected to upgrade their qualification to Diploma by 2013. 183. Untrained teachers in-service training The upgrading program for non- professional teachers takes four years, and terminates with the award of a Diploma in Basic Education. The Diploma is the same as that offered full-time in the colleges of education. However, the program uses mainly distance education methods with teachers continuing to teach in schools and provided with printed modules to study at the weekends and in the evenings. Participants attend a one-week induction course and three 3-week residential courses each year. Participants are also supported as they stud4y by district support cadres who work with teams of 20 trainees holding tutorials at regular intervals. 184. The program began in 2005 with an enrollment of 5,000 untrained teachers from the Northern Regions and has now reached nationwide coverage of 27,183 teachers. Table 3-13 shows the number enrolled in recent years. Table 3-13: Admission in Untrained Teachers Training Program Phase Start Date Total 1 - 3 Northern Regions and Afram Plains District 2005 5,119 2 - Ashanti and Brong Ahafo Regions 2006 7,627 3 - Central, Western and Eastern Regions 2007 8,994 4 – Volta and Greater Accra Regions 2008 5,439 Total 27,183 71 185. Cluster-based in-service training. Since 2006, The Japan International Cooperation Agency (JICA) has been assisting TED to pilot a new national mode of school-based and cluster-based in-service training. This has operated successfully in ten districts (one in each region) and it is intended to extend it to a further 71 districts in 2009 and to all districts in 2010. The new modality is demand-centered and involves one Cluster Leader (CL) in each school (chosen by the head-teacher and receiving training at district level). The CL brings together all teachers in each school twice a term. Schools are linked in clusters of three to five schools and the teachers in a cluster meet once a term. At district level there is a district in-service training committee and a district teacher support team 186. At basic education level, teacher numbers are determined nationally based on enrollment in each district. Teachers are assigned to districts on the basis of one teacher and one assistant for 45 pupils at kindergarten and one teacher for 35 pupils at primary level. However, class size is not expected to exceed 45 at either kindergarten or primary levels and when the number exceeds 45 an additional teacher may be appointed. All teachers are deployed by the district and then assigned to schools. Data are aggregated at regional and national levels and this information guides the assignment of newly qualified teachers and of national service and volunteer teachers. At junior high school level teachers are appointed on the basis of five teachers (including the head-teacher) for three classes and class size is expected not to exceed 40. 187. At senior high school level teacher numbers are determined nationally and allocated to schools. The basis of the allocation is one teacher for every 25 students. Teachers are recruited directly by the school. Again, demand is determined at district, regional and national levels and vacancies are advertised. New graduates seek posts in schools with vacancies and the schools interview and appoint. As was described the two main teaching universities, University of Education, Winneba, and University of the Cape Coast, offer a wide range of education degrees and graduate over 3,000 each year. These are sufficient to take up the necessary vacancies. 188. At the primary level, the percentage of trained teachers in public schools has decreased over recent years both nationally and for the 53 deprived districts. The disparities between the deprived districts and the national averages have been increasing over the same period. The same has occurred at the JHS level. These trends can at least partially be explained by the increased number of national youth employment placements in schools. The increase in the number of youth employment teachers is higher than the rate at which trained teachers are graduating and existing teachers are upgrading. Table 3-14: Percentage of Trained Teachers 2003/04 2004/05 2005/06 2006/07 2007/08 Kindergarten 35.3 42.9 Primary National 73.9 72.4 70.8 62.1 59.4 Primary Deprived 55.3 53.2 55.9 42.8 37.2 Junior High National 84.2 83.5 85.5 77.2 76.4 Junior High Deprived 75.9 73.9 77.7 64.2 62.9 Source: Education Sector Performance Report 2008. 72 189. The ten districts with the lowest percentages of trained teachers are all deprived districts. Paradoxically, MOE efforts to post national service teachers to deprived districts drive the trained teacher share down. No national service teacher is trained as they have just completed a first degree. Clearly, it is important that deprived districts have sufficient numbers of teachers and assigning national service and volunteers helps. Equally clearly, the ideal is to deploy qualified teachers to these districts. 190. Student teachers must receive an endorsement from a district before they are admitted to teachers’ colleges. The deprived districts have a larger quota than other districts. When students complete their course work, it is anticipated that they will be assigned to the endorsing district. However, this has only been applied in recent years and it is difficult to determine impact. In addition, national service and volunteer teachers are assigned to deprived districts in accordance with vacancies and the success of this effort is reflected in the overall number of teachers assigned to these districts and within the district in the share of unqualified teachers. 191. The current national PTR averages at each level of education do not provide a great deal of guidance for policy makers. First, as seen before, the national average is different from regional averages and these in turn are very different from district and school averages. Second, the ratios do not show the share of untrained teachers. Third, there is a strong argument that the current average ratios are lower than is necessary for the system and lower than is affordable for Ghana now. 192. Headteachers, also known as principals in other countries, are selected from the ranks of serving teachers. Teachers who have reached the level of Principal Superintendent (with at least 12 years experience, see section on pay later) are eligible for consideration. No special training is given. Teachers are interviewed at district level and appointed to the schools. They receive a modest allowance for their work as head- teacher. Headteachers in one-stream schools teach half-time. Headteachers in two- stream schools have a deputy head and do not teach. Headteachers in three-stream schools do not teach and have two Deputies. In addition to managing the school and its resources, headteachers are expected to support teachers in their work. In addition, inspectors visit schools and work with teachers; circuit supervisors at primary level and subject inspectors at high school level. Table 3-15: Teacher questionnaire responses to management-related statements (rounded percentages) STATEMENT RURAL URBAN Not Not Agree Disagree Agree Disagree sure sure The head teacher of this school leads by 82 9 9 79 7 14 example Teachers salaries are usually paid on time 50 36 14 57 14 29 School inspectors regularly visit this school 77 9 14 85 4 11 The impact of HIV/AIDS on teachers at this 50 9 41 51 21 18 school has not been serious Our head teacher regularly observes classes 87 9 4 85 4 11 Our head teacher is often away from school on 0 95 5 11 75 14 73 private business Our head teacher is often away from school on 41 59 0 29 64 7 official duties Teacher transfers are managed well and fairly 50 27 23 50 25 25 Teachers sometimes come to school hungry 36 41 23 50 43 7 Teacher absenteeism is not a problem at this 86 9 5 68 18 14 school Teachers at this school come to work on time 77 18 5 86 7 7 Source: Akyeampong and Asante 193. Whilst rural school heads seem generally happy with the behavior of their teachers, urban heads complained that teachers were increasingly becoming rude and difficult to manage. They attributed this to general apathy, especially among older teachers. It may be that older teachers having experienced the frustrations in teaching for longer are becoming less concerned about maintaining their sense of professionalism. The school survey also showed that dismissals rarely occur, which could mean one of two things: either the majority of teachers are professionally well-behaved or, more likely, the general laxities in the Ghanaian occupational culture makes enforcement of teacher regulations difficult (Hedges 2002). Most primary headteachers, as one union official pointed out, “lack the authority that goes with being a head�, and therefore are unable to initiate disciplinary measures against teachers. 194. Teachers’ pay compared to international norms. In real terms, teachers’ salaries have increased consistently since 1990. Average median annual salary (at mid-point of level 13) is US$3,600, which is roughly 5 times the per capita GDP. The average cumulative in EFA FTI countries is 3.5 times the per capita GDP. However, the purchase power parity in Ghana is roughly one third lower than in the rest of the EFA community, which could serve as a justification for higher teacher salaries. In short, salaries are higher because prices are higher in Ghana. (See the next chapter on education expenditures for the analysis of the payroll) 195. The Best Teacher Award scheme is a nation-wide competition attracting publicity at district, region and national levels. The national level awards are extremely valuable and Ghanaian firms are given the opportunity to show their appreciation for the teacher (and gain welcome publicity). In addition, Ghana has two teacher Unions and each of these is active in pedagogy and advocacy. Discussions with teachers and Government indicate that the unions are consulted on education reforms as well as on the conditions of their members. While it is often difficult for Governments to meet expectations on teacher pay, it is critical that discussions take place and that the power of the Unions to contribute to education is realized. It appears that this is happening in Ghana. 196. Ghana’s national service scheme prioritizes education. Every Ghanaian student in Ghanaian universities has his/her fees paid and is provided with an allowance for accommodation, food and books. In return, all graduates must provide one year of national service. This scheme prioritizes education and two-thirds of all graduates teach for one year in schools. In addition, the scheme is targeted to deprived districts and helps ensure that these districts and the deprived schools within them have some minimum number of teachers in place. Separately, Ghana also runs a national volunteer scheme, again targeted to education. Young and older people interested in making a contribution to their countries development can volunteer to teach in schools. 74 C. Public Expenditures on Education 197. This review of public expenditures on education covers five areas: the sources of education expenditures, the flow of funds, the aggregate trends in expenditures, the geographical disparities, and the trends in teacher salaries. Overall Expenditure trends10 198. Between 1976 and 1984, Government spending on education plummeted from 6.4 percent of GDP to 1.5 percent considerably constraining physical and human resources and creating challenges in maintaining quality. From 1990 onwards, this share has fluctuated between 5 and 6 percent of GDP. Throughout the last decade, spending gradually increased in parallel with the enrollment increase. In 2008, government spending on education reached about 11 percent of the GDP and about 30 percent of the government budget. 199. Since 2003 total expenditure on education nearly doubled. Internal funding has increased by 40% , and external support has almost quadrupled. Table 3-17 on expenditures in million GHc in 2008 prices presents the series of spending in 2008 prices. Education expenditure in terms of GDP is high. It was 6% in 2003 and reached 11% in 2008, well above the average in Sub Saharan Africa, demonstrating significant commitment to education Figure 3-9: Trend in Expenditure on Education as a Share of GDP Education Expenditure as a % of GDP UNESCO target of 6% of GDP 12.0% 11% 10.0% 9.1% 8.2% 7.3% 7.5% 8.0% 6.2% 6.0% 4.0% 2.0% 0.0% 2003 2004 2005 2006 2007 2008 Source: Based on EMIS and UNESCO data. 200. Internationally, Ghana allocates a significantly larger share of its GDP to the education sector than do other countries. For example, low income countries in sub- Saharan Africa invest on average less than half the share of GDP on education that Ghana does. 10 All financial information reported in this chapter is expressed in 2008 constant prices using the IMF GDP deflator series for Ghana. 75 Table 3-16: International Comparative Analysis of Public Spending on Education Public spending on Total public Public spending on education as % of spending education as % of total GDP as % of GDP public spending Ghana 2007 8.9 31 29 Low-income countries (i) Sub-Saharan Africa 4.3 25 17 Other regions 4.3 31 14 Middle-income countries (ii) Sub-Saharan Africa 5.7 35 16 Other regions 4.2 25 17 High-income OECD 5.6 45 12 Source: Preliminary Education Sector Performance Report, MOESS (2009) based on data from UNESCO Institute of Statistics. Total public spending (middle column) is computed using the other two columns. Note: (i) Countries eligible for lending from IDA. (ii) Countries eligible for lending from the IBRD. Table does not include: i) countries in the World Bank's Europe & Central Asia region, and ii) countries, mostly island states, with population less than approximately 300,000. Expenditure trends by Funding Source 201. In the six years, 2003-2008, the total domestic funding for education doubled, including a near doubling of Government support, a near tripling of funding from the Ghana Education Trust fund, and a 60 percent growth in internally generated support. The GoG budget also includes about 10 percent externally financed budget support. Beyond the budget support, direct external financing of the sector has grown 4.5 times. Given the sharp increase in economic growth, a doubling of the domestically funded spending on education has increased the share of education within the GDP by 50 percent. Table 3-17: Expenditures in million GHc in 2008 prices, 2003 – 2008 2003 2004 2005 2006 2007 2008 Domestic GoG 667 677 729 863 1014 1219 GET Fund 79 95 107 153 192 213 IGF 0 91 104 136 129 164 SIF 0 0 7 5 0 0 Total 746 862 947 1157 1335 1596 External Donor Agency 36 73 93 30 90 96 EFA Catalytic 0 0 5 4 12 4 Debt Relief 22 47 47 59 52 47 HIPC 22 47 47 59 47 23 MDRI 0 0 0 0 5 24 Total 81 191 253 226 347 385 Total (Domestic + External 827 1053 1200 1383 1682 1981 GDP (IMF data) 13064 13794 14603 15541 16423 17618 Domestically funded spending on 6% 6% 6% 7% 8% 9% education as % of GDP Total expenditure on education as % of 6% 8% 8% 9% 10% 11% GDP Source: Education Sector Performance report 2009. Note: All figures are reported in million GHc in 2008 prices. 76 202. Expenditure trends by level. The significant increases in the budgets and in the sources caused surprisingly little changes in the structure of expenditures by levels and types of programs. The share of expenditure allocated to KG education decreased between 2004 and 2008 from 5.4 to 4.5 percent, which shows a significant incongruence between the official policy of making KG part of basic education and public financing. There has been significant pressure on the Government to increase the share of primary education in line with the EFA FTI guidelines. However, the increasing volume of education expenditures meant such an increase has been unneeded. Consequently, the portion of total expenditure assigned to primary education has fluctuated over the period; it sunk from 39.2 to 35.3 percent between 2004 and 2006, then increased by 8.1 percentage points by 2008 to reach a share of 43.4 percent of the total public spending on education. Primary education, hence, almost constitutes half of the entire education budget, which is consistent with the EFA FTI norms. However, this share makes per student spending on primary education relative to the per capita budget almost double of the SSA average. The problems with using the EFA FTI norm in Ghana are several: First, it doesn’t take into account the economic growth and subsequent growth in public budget. Second, the Government budget in Ghana includes internally generated funds, thus distorting the true picture of how the Government priorities are reflected in the allocations. Basically, an increase in IGF in the form of more fees at tertiary level decreases the share of the budget allocated to primary education. Third, using the international norm doesn’t help the focus on the key problem in primary spending, i.e. the regional disparities (to be discussed later). 203. The proportion of expenditure allocated to junior high school remained near the 20 percent mark, while the portion disbursed to senior high dropped from 13 to 8.8 percentage points since 2004. Overall trends indicate that as primary’s and tertiary’s share of total government spending on education increased by 4.2 and 2.0 percentage points, respectively, these additional funds to a large extent were reallocated from senior high school, which shrunk by roughly 4.2 percentage points as a share of total education expenditure. Figure 3-10: Expenditure by level of education, 2004 and 2008 Pre-School Pre-School 2008 0.148 2004 Primary Primary 0.011 0.045 0.013 0.004 0.054 JHS 0.000 JHS 0.041 0.051 SSS SSS 0.006 NFED 0.006 0.168 NFED 0.006 SPED 0.005 SPED 0.392 0.434 Teacher Education Teacher Education 0.088 0.130 0.196 TVET 0.202 TVET Tertiary Tertiary Subvented Agencies Subvented Agencies & Management & Management Source: Based on EMIS. 204. Expenditure trends by Categories: Since 2004 around 93-96 percent of expenditure through the Ministry of Education has been allocated to personnel 77 emoluments. The remaining 4-7 percent is distributed among administration, service and investment. Between 2004 and 2008, the share of expenditure allotted to administration rose from 2.2 to 3.6 percentage points, but has been declining after peaking in 2005, and the portion distributed to service sunk by 2 percent from 2.4 to 0.4 percentage points over this period. The proportion allocated to investment reduced from 0.9 to 0.1 percentage points during this time span, however, most of the investments into education are covered through various statutory funds, especially through the GET Fund and are not counted in this data. As the proportion of administration, service and investment expenditure has been sinking since 2005, public funding is being reallocated into personnel emolument costs. Figure 3-11: Trends in MOE Expenditure PE Administration Service Investment 100.0% 94.4% 93.6% 94.3% 96.0% 6.0% 93.2% 90.0% Personnel Emoluments (PE) 5.0% 5.0% Administ./Service/Invest. 80.0% 4.7% 4.6% 70.0% 4.0% 60.0% 3.6% 50.0% 3.0% 2.2% 40.0% 2.4% 2.0% 30.0% 1.3% 1.3% 20.0% 0.9% 0.7% 0.6% 0.4% 0.4% 1.0% 0.4% 10.0% 0.1% 0.0% 0.0% 2004 2005 2006 2007 2008 Source: Based on data from EMIS. 205. Total Expenditure trends by sources and by level. Total expenditure on the sector in 2008, as seen in Table 3-18, was GHc1,744 million, 6 percent of which from donor funds. Adding HIPC and MDRI funds, classified under “Other� rather than under “Donor� in the MoE ESPR, donor funds reach 10% of the total expenditure. Table.. s how that recurrent expenditures in 2008 was GHc 243 million, 3% of which was donor funded, and Capital expenditure (table..) was GHc183 million, 20% of which was donor funded. These figures do not include expenditures from the MPs fund, DACF and all off budget donor spending on education. It therefore does not represent the full education sector spending. Spending on Primary Education constitutes 43% of total government spending on education and 63% of total government spending on the basic cycle. 78 Table 3-18: Total expenditure in million GHc by source and level, 2008 GoG % Donor Other11 Total % Basic education Kindergarten 54.28 4% 6.02 5.60 66 4% Primary 529.45 43% 50.59 33.62 614 35% Junior High 245.98 20% 25.59 20.85 292 17% Second cycle Senior High 106.99 9% 6.70 57.37 171 10% TVET 13.55 1% 1.24 3.52 18 1% Special education 6.88 1% - 3.79 11 1% Non-formal Ed. 6.33 1% - - 6 0% Teacher Education 50.34 4% 0.32 4.61 55 3% Tertiary education 204.97 17% 7.39 166.25 379 22% Mgt & Subv. 0.26 0% 1.48 128.27 130 7% HIV AIDS - 0% 1.33 - 1 0% Total 1,219.03 100% 100.65 423.89 1,744 100% Percentage 70% 6% 24% 100% Source: Education Sector performance report 2009. Note: Others include: HIPC, GET Fund, IGF, MDRI. 206. Recurrent expenditures by source and by level: Table 3-19 shows the amounts and shares of recurrent expenditures by source and level. The Government recurrent expenditures are mostly used to pay salaries. Primary, JHS and Tertiary levels take as much as 80 percent of the total Government recurrent expenditures. Even though KG is part of the basic education cycle, the Government spends very little on KG salaries at 4 percent -- just over 54 million GHc -- leading to high PTR, large disparities and inequitable distribution of KG, discriminating against poor and remote areas of the country (as will be discussed in the next chapter on performance). The relatively low percentage of SHS relative to the JHS recurrent reflects the low transition rate between basic and post-basic education. Donor funds represent a relatively low level of contribution, however, most (2/3rd) of donor funding to education comes indirectly through budget support. Other funds include large allocations to tertiary and to sector management. Table 3-19: Recurrent expenditures in million GHc by source and subsector, 2008 GoG % Donor Other Total % Basic education Kindergarten 54.28 5% 3.23 2.89 60 4% Primary 529.45 43% 28.20 17.32 575 38% Junior High 245.98 20% 14.51 12.70 273 18% Second cycle Senior High 106.67 9% 0.57 16.82 124 8% TVET 13.49 1% 1.24 - 15 1% Special education 6.83 .5% - - 7 0% Non-formal Ed. 6.33 .5% - - 6 0% 11 Other sources include statutory funds and others that are not directly planned under the sector budget. 79 Teacher Education 50.27 4% 0.32 - 51 3% Tertiary education 204.92 17% 0.53 78.70 284 19% Mgt & Subv. - 0% 0.31 114.43 115 8% HIV AIDS - 0% 0.70 - 1 0% Total 1,218.22 100% 49.62 242.86 1511 100% Percentage 81% 3% 16% 100% Source: Education Sector performance report 2009. Note: Others include: HIPC, GET Fund, IGF, and MDRI 207. Investments to education: The MoE from its own sources spends very little on investment and most of the MOE investment expenditures are externally funded (and made part of the MOE budget) and concentrated in the Senior High cycle. A substantial amount of investment – GHc150 million in 2008 – takes place through the Ghana Education Trust Fund (the GET Fund), a statutory fund constituted of 2% of VAT receipts. Although there are complex formulas regulating GET Fund spending, they are rarely used, and the GET Fund tends to operate on an ad hoc basis. The GET Fund disbursements in 2008 by type and subsector are presented in Table 3-21. Approximately 70% of GET Fund releases are in capital investment, the largest receiver of GET Fund investment is the Tertiary subsector, followed by Senior High School and Primary. Table 3-20: Capital expenditures in million GHc by source and subsector, 2008 GoG % Donor Other Total % Basic education Kindergarten - 0% 2.79 2.72 6 2% Primary - 0% 22.38 16.30 39 17% Junior High - 0% 11.08 8.15 19 8% Second cycle Senior High 0.32 40% 6.12 40.55 47 20% TVET 0.05 6% - 3.52 4 2% Special education 0.05 6% - 3.79 4 2% Non-formal Ed. - 0% - - 0 0% Teacher Education 0.07 9% - 4.61 5 2% Tertiary education 0.05 6% 6.86 87.55 94 41% Mgt & Subv. 0.26 32% 1.17 13.84 15 7% HIV AIDS - 0% 0.63 - 1 0% Total 0.81 100% 51.03 181.03 233 100% Percentage 0% 22% 78% Source: Education Sector performance report 2009. Note: Others include: HIPC, GET Fund, IGF, and MDRI 208. GET Fund investment expenditures More than half of GET fund expenditures continue to go to tertiary education including pre-service teacher education and less than 20 percent goes to regular basic education. This creates distortion for both: The significant investment to tertiary education distorts that subsectors budget and sector performance for two key reasons. The full spending on tertiary does not calculate the recurrent implications of investments often resulting in low maintenance and low use of new infrastructure. Secondly, not enough is being spent on improving the human resources and creating incentives for young professionals to join the teaching force. The 80 low level of investment in basic education perpetuates the difficulties with service delivery especially in areas such as KG services in remote rural schools. It also limits the Government’s ability to react to natural disasters, such as, for instance the 2007 floods in the North, which destroyed over 150 schools and damaged many more. Despite significant budget sources, the Government relied heavily on external funding to re- establish the damaged infrastructure but to this date many classes are being held under the trees in the North. Table 3-21: GET Fund releases in million GHc by type and subsector, 2008 Basic education Recurrent Capital Total % Kindergarten 1.39 2.72 4.10 2% Primary 8.32 16.30 24.62 11% Junior High 4.16 8.15 12.31 5% Second cycle Senior High 2.36 26.80 29.17 18% TVET - 3.52 3.52 2% Special education - 3.79 3.79 3% Non-formal Ed. Teacher Education - 4.61 4.61 3% Tertiary education 19.54 72.77 92.30 48% Mgt & Subv. 25.87 12.26 38.12 8% HIV AIDS Total 61.63 150.91 212.54 % 29% 71% 100% Unit costs and Per Student expenditures 209. Table 3-22 compares unit costs across countries by level of education. It shows that among sub-Saharan countries Ghana has the highest share of unit costs in primary. Also JHS unit costs are higher than the Sub-Saharan African average, whereas SHS and HE unit costs are marginally below the SSA averages. From an EFA perspective, these allocations are considered to be progressive. However, without taking into account the disparities among sub-groups within Ghana (to be discussed below), these averages do not provide an accurate picture. Table 3-22: Comparative Average Unit Costs as a Share of GDP Per Capita by Level LS/JH US/SH Prim Unit HE Unit Unit Unit Cost Cost Country Cost Cost GHANA 19% 35% 55% 297% Sub-Saharan Africa Average 11% 29% 57% 314% Burkina Faso 17% 19% 63% 244% Gambia, The 16% 26% 47% 253% Burundi 15% 64% 136% 363% Nigeria 14% 20% 25% 111% Benin 13% 11% 32% 134% Mali 11% 26% 117% 187% Guinea-Bissau 11% 17% 31% 124% 81 Togo 10% 19% 32% 148% Liberia 9% 21% 26% 98% Sierra Leone 9% 29% 30% 340% Rwanda 8% 51% 63% 790% Cameroon 7% 32% 37% 80% Guinea 6% 11% 9% 153% Chad 5% 19% 25% 420% Cote d'Ivoire 5% 19% 19% 220% Source: Based on World Bank Data, 2008 and Ministry of Education. 210. Per child recurrent expenditure (PCE) on the primary cycle was GHc188 in 2008. This includes the administrative overheads, but not expenditure on management and subverted agencies. PCE for tertiary students is more than 10 times the PCE for primary students. PCE has been on the rise in real terms since 2005 in all subsectors except senior high school, where it decreased from GHc502 to GHc281. Table 3-23: Per student recurrent expenditure by level, 2008 2500 2000 1500 1000 500 0 Pre School Primary JHS SHS TVET* Tertiary Source: Data from Education Sector Performance Report 2009. Note: All figures in 2008 GHc, * calculated from 2007 data. 211. Table 3-24 below shows the evolution of the per student recurrent expenditures. Primary education is likely the most reliable source given that over 95 percent is salary (PE) and per teacher recurrent expenditure (PTE) hasn’t changed significantly. Accordingly the almost 100 percent increase in recurrent expenditures over three years reflects increases in average salaries. KG shows significant underfunding, lack of publicly paid teachers and very low PTE. JHS is relatively higher than Primary. The recurrent expenditure is pushed upwards by the low PTE. SHS is the only level where per student recurrent expenditures declined showing that teacher supply couldn’t keep up with the enrollment increase. Table 3-24: Per student recurrent expenditure by subsector, 2005 - 2008 2005 % 2006 % 2007 % 2008 % Pre School 43 46 Primary 97 100 113 116% 155 160% 188 194% JHS 186 100 207 111% 215 116% 255 137% SHS 502 100 333 66% 238 47% 281 56% 82 TVET 395 100 616 156% 566 143% Tertiary 1588 100 1536 97% 1763 111% 2143 135% Source: Data from the Education Sector Performance report 2009 All figures in 2008 GHc 212. Salaries constitute more the 80% of per unit recurrent expenditure in the entire pre tertiary cycle, as shown in Figure 3-12. Figure 3-12: Composition of per unit recurrent expenditure from all sources by subsector, 2008 100% 80% 60% 40% 20% 0% KG Primary JHS SHS Tertiary Services 4.92 15.56 26.40 40.65 463.00 Administration 1.03 2.96 4.59 10.95 209.34 P.E. 40.48 170.51 225.74 229.51 1,470.55 Disparities in PCE 213. This section will present an estimate of the district level per capita expenditures. The district offices are the smallest spending bodies in the administrative chain for which records are available. Although a more accurate estimate of the disparities between per student expenditures would be obtained by calculating the school level per unit expenditure, there is no systematic record keeping of financial information at the school. 214. Actual expenditures classified by subsector, budgetary item and source were collected from all districts for the financial years 2005 to 2009. These expenditure figures by no means capture the entire spending on the education sector but capture the lowest level for which records are available, removing the overheads from the upper layers of administration. Thus, these unit costs provide (i) a closer picture of what the actual per unit expenditures are, and (ii) a previously unexplored dimension of the geographic variation in education expenditure. Table 3-25: Reported Sources GOG IGF DFID PPS OTHER ASHANTI X X X X BRONG AHAFO X X X CENTRAL EASTERN X GREATER ACCRA X NORTHERN X X X X X 83 UPPER EAST X X X X UPPER WEST X X X VOLTA X X X WESTERN X X X X X Note: OTHER includes HIPC, DONATIONS, GET Fund, and Unspecified Table 3-26: Reported years 2005 2006 2007 2008 2009* ASHANTI X X X X BRONG AHAFO X X X X X CENTRAL EASTERN X X X X X GREATER ACCRA X X X X X NORTHERN X X X X X UPPER EAST X X X X X UPPER WEST X X VOLTA X X X X X WESTERN X X X X X Note: For the year 2009, only the first 3 financial quarters were reported 215. Different funds are allocated across different budgetary items depending on the nature of the funds. Personal Emoluments are exclusively covered by Government funds and constitute the quasi totality of District Education Office (DEO) expenditures. Administration, Services and Investment, on the other hand, tend to be covered by a variety of other sources, mainly internally generated revenues and Donor Support. Donor funds tend to cluster on Items 3 and 4. Table 3-27: Funding Sources and Budgetary Items Source Item 1: Item 2: Item 3: Item 4: PE ADMIN SERVICE INVESTMENT Internal GOG X X X X IGF X X X External DFID X X PPS X X OTHER X X X Note: OTHER includes HIPC, DONATIONS, GET Fund, and Unspecified 216. Per child expenditures12 (PCE) increase over the education cycle, with PCE in the SHS cycle being approximately four times higher the in the Basic cycle. The average 12 The PCE is calculated as GOG expenditures – excluding investment – in a given subsector divided by enrollment in the given subsector12. 84 expenditure on a primary school child, measured at the district level, is GHc125. With respect to the primary cycle: the lower secondary cycle receives 50% more and Kindergarten only about 60% less. 217. The median PCE are consistently lower than the average PCE, indicating that a more than half of the districts spend less than the average PCE. The high standard deviation reported in district PCE indicates a substantial amount of cross district variation. Table 3-28 highlights the value that demarcates the bottom third of the distribution of PCEs in 2008, which is artificially chosen here as the largest acceptable level of disparity between average and low PCEs. This already indicates a significant disparity for KG compared to other levels, as the bottom third in KG is 50 percent of the average PCE whereas it revolves around 70 percent for the other levels. This choice is in line with the definition of deprived district – chosen to be the bottom third performing district under a series of education indicators. PCE below the indicated value are referred to as substandard in this section. 218. It is important to note that per child expenditures calculated from recurrent expenditures aggregated at the national level presented in Table 3-24 are lower than average per child expenditures calculated from district level recurrent expenditures presented in Table 3-28. There may be several reasons behind such a discrepancy. Firstly, expenditure aggregated at the national level may contain administrative overheads that are not present in district level data. Secondly, districts may have poor book-keeping on expenditures which may lead to under reporting of expenditures. Finally, there might be leakages along the funding flow. Only a full audit of government expenditures can shed light on the nature of such a discrepancy. Table 3-28: Per Child Expenditure in 2008 GHc measured at district level KG Primary JHS SHS Average 47.04 125.24 196.23 511.96 Median 36.74 117.53 172.23 462.29 bottom third 23.48 90.09 134.02 363.07 bottom third/average 50% 72% 68% 71% Standard Deviation 38.88 67.4 120.25 332.03 Min 5.16 11.52 20.35 16.09 Max 224.35 396.34 589.64 1,910.11 Teacher Payroll 219. This section will present a description of the structure of the Ghana Education Service (GES) payroll and analyze the variations in teacher remuneration across subsectors and geographical areas. Information on teachers for the purpose of this study is available from three only partially overlapping sources as summarized in Table 3-29. Full information is available but had to be forgone due to time constraints. This section will draw from Payroll information from 2008 and 2009. 85 Table 3-29: Sources of information on teachers EMIS Payroll 2008 Payroll 2009 Subsector X X Salary X X Geographical area X X The GES payroll 220. The GES payroll is made up of a 22-grade scale with a number of steps in each for a total of 211 different salary points, known as the Ghana Universal Salary Scheme (GUSS). This was based on the principle of equal pay for work of equal value. There are four entry points to the payroll. Unqualified teachers with GCE O levels begin at Level 5 (GHc 1279 p.a.). Unqualified teachers with A levels enter at Level 7 (GHc 1719 p.a.). Qualified teachers with a Certificate A begin at Level 10 (GHc 2678 p.a.). Over their career, such teachers can move up to Level 16 (GHc 8237 p.a.). Graduate teachers commence further up the scale at level 14 (GHc 4837 p.a., [US$3,720]). The salary scale, with the correspondent salaries is plotted in Figure 3-13. Figure 3-13: The GES salary scale 221. The GES payroll in 2008 counted approximately 250.000 individuals, classified as teaching and non teaching. Teaching staff constitute 85% of the total and absorb 91% of the total expenditure. 146,000 people, mainly teachers and headteachers, are paid for by GES in kindergarten, primary and junior high schools and these take about 63 percent of all GES salary costs. In addition, at least 30,000 are paid by the National Service fund and a smaller number are paid by the Youth Employment Agency. At senior high school level, GES pays for 33,500 staff, of whom about 19,250 are teachers. It is also of interest to note that 12,650 teachers were on study leave in 2008 and these cost the Government more than the total cost of all 138 district education offices. Table 3-30: Payroll numbers GES 2008 Cost centre Number of staff Total wage bill ‘000 Ghcedis Share of wage bill Headquarters 676 3301 0.4 District Education offices 10309 51262 5.6 Regional Education Offices 780 3967 0.4 Educational Units 1428 510 0.1 Study leave 12650 52376 5.6 86 Trainees 22689 24492 2.7 Training college staff 3371 14721 1.6 Senior High schools 33512 148038 16.1 Technical schools 2161 9328 1.0 Special schools 1162 4696 0.5 Newly trained 11806 24499 2.7 Basic schools 146710 578912 63.0 Total 247254 921103 100 Source: Financial Controller, GES 222. Two thirds of the total teaching and non-teaching staff is employed in the basic cycle, receiving approximately two thirds of the total expenditure. Within the Basic cycle, 95% of staff is classified as teaching. Primary education employs 60% of the total staff, Kindergarten 10% and Junior High 30%. Per unit expenditure on teaching staff are highest in JHS, followed by primary and KG (Table 3-31). Teachers in the upper secondary cycle are, as expected, better paid than those in the basic cycle. Figure 3-14 plots per unit expenditures by role and subsector, including per unit expenditure on teachers on study leave. In 2008, there were 13,993 teachers on study leave; total salary expenditure on them was GHc51 million. Table 3-31: Staff in the basic cycle, 2008 Number Expenditure Per unit expenditure kg teaching 13711 39.37 2,871.09 non teacching 756 1.64 2,172.68 primary teaching 86923 299.57 3,446.42 non teacching 4191 9.72 2,319.46 jhs teaching 41907 164.88 3,934.35 non teacching 2991 7.32 2,446.94 basic teaching 142541 503.82 3,534.53 non teacching 7938 18.68 2,353.51 Source: 2008 payroll. Note: Expenditure in million GHc, All figures in 2008 GHc. Figure 3-14: Per unit annual expenditure by role and subsector, 2008 7,000 6,000 5,000 4,000 teaching 3,000 non teaching 2,000 1,000 - Basic SHS TVET TTC Admin Study leave Source: Payroll 2008 223. Figure 3-15 overlays the active GES payroll in 2009 (red dots) to the relative frequency with which each salary point is found in the 2009 payroll (blue line). The 87 peaks in the blue curve indicate point on the salary scale with a higher concentration of individuals. There are three major salary classes: unqualified, qualified and graduate. Figure 3-15: Distribution over the GES salary scale, 2009 224. Figure 3-16 reproduces the relative frequency of salary points against the entry points. The first salary class, covering approximately 35% of the total number of staff, is made of staff entering the payroll with low qualifications. The second salary class, covering approximately 42% of the total staff, is made of staff entering the payroll with Certificate A, and the last salary class, approximately 21%, is made of graduate teachers. Figure 3-16: Distribution of salaries and entry levels, 2009 Source: 2009 Payroll 88 225. The difference in per unit expenditure is driven, as shown in Table 3-32, by seniority. Teachers in Greater Accra have on average 4 years of service more than in the rest of the country. Table 3-32: Median years of service as of 2009 by region years of service in 2009 ASHANTI 8 BRONG AHAFO 9 CENTRAL 8 EASTERN 8 GREATER ACCRA 12 NORTHERN 7 UPPER EAST 6 UPPER WEST 7 VOLTA 8 WESTERN 10 Total 8 non deprived 9 deprived 7 Source: 2009 Payroll. 226. The GES payroll is very large and has a fairly complicated structure. With the introduction of the Single Spine Salary Structure (SSSS) the GES payroll will have to be suitably adapted. Remunerations on average are within internationals standards, although there are large numbers of untrained teachers on relatively low salaries. An important factor shown in table 3-33 is the substantially higher salary and number of years of service of teachers in the Greater Accra region. It would appear that teachers tend, over the course of their career to converge towards Greater Accra. Teachers in deprived areas, conversely, are less paid and less experienced. An analysis of non teaching staff is beyond the scope of this section, but the large number of non teaching/ auxiliary staff cannot be overlooked. 89 Table 3-33: Estimated Cost of Trained/Untrained Teachers in Primary 1-3 Table 1. Estimated Number and Cost of Trained and Untrained Teachers in Primary Grades 1-3 55% All figures in Cedis, unless otherwise specified Cumulative Annual Avg Cost Number of PTR salary per Estimated % of % of per Teachers estimates Teacher Salaries Salaries Teachers Student GES (Primary) 48,947 2,060 183,337,663 81% 61% 112 Trained Teachers 28,597 59 2,892 150,371,554 66% 36% 92 Untrained Teachers 20,349 34 1,620 32,966,109 15% 26% 21 NSS 17,600 3,960 33,660,000 15% 22% 21 Recent University Graduates 12,100 28 1,800 21,780,000 10% 15% 14 Volunteers (NVP) 5,500 25 2,160 11,880,000 5% 7% 7 0% NYEP 13,200 640 15,364,800 7% 17% 10 Masters degree holders 1,320 25 3,000 3,960,000 2% 2% 2 University Graduates 2,640 24 1,800 4,752,000 2% 3% 3 High school Graduates 9,240 21 720 6,652,800 3% 12% 4 0% TOTAL 79,747 2,914 232,362,463 100% 100% 143 Source: J-PAL and IPA 2009. Conclusion: Inputs and Resources 227. Even with Ghana’s extraordinary levels of spending on education, it is clear that public funds cannot cover all minimum needs. Under these conditions, allocations should be better prioritized and provisions adjusted to local conditions and need. A key example is in teacher provision. In remote locations the use of various incentives to attract teachers, especially trained ones, may be more important in affecting learning outcomes than other resources, like class supplies or the number of classrooms. The government has been trying to direct more resources to deprived areas. Most of the efforts have been based on externally financed projects, including the pilot programmatic scheme by IDA or lately funding for the deprived districts by USAID. But current administrative structures have often left these efforts unsuccessful. In reality, the complexity of resource allocation has led in many cases not to neutral allocation, but to negative targeting, as poorer areas receive not more provisions but less. Therefore, decisions about prioritizing the allocation of resources and the setting up of services will be most effective when they follow some principles: a) they are evidence based, b) they promote policy goals by providing adequate incentives for key participants including students, parents, teachers, headteachers, and district officials, and c) they are targeted so that those who need the most get the largest share of public resources. 90 4. SECTOR PERFORMANCE 228. A variety of reasons can contribute to threats to -- and even decline in -- the quality of learning. In urban classrooms this may include overcrowding; in remote areas lacking access to qualified teachers, inadequate teaching methods that are not adjusted to local conditions. The most critical issue for the Government is to identify those groups and localities which are for one reason or another at risk of low attainment, high dropout and low learning results and assign the available resources to address these potential disparities. 229. As the previous chapters have shown, improvements in access have been the result of a long period of sustained commitment to education both on the supply and on the demand side. In addition, important external factors have played a role, including economic and political stability, a reduction of poverty, and the recognition that improvements in educational attainment bring about improvements in living standards and earnings. But, we have also seen, improved access measured in admission and enrollment has brought about new challenges in learning and education. Importantly, subsequent Governments have maintained a strong commitment to education that had been expressed in improved resources, investments into the physical infrastructure, expansion of the teaching force, efforts to provide qualifications to teachers and mobilizing a variety of financial resources. 230. Education has become the biggest public enterprise in Ghana, taking about 11 percent of the GDP, a third of the public budget, 40 percent of the civil service and enrolling about a quarter of the population in schools and other education services. Education has become a key part of the basic services, everybody has and most people claim the right to receive quality education and expect that this will result in improved living standards and job opportunities. This is good news: it lays the foundation of a knowledge-based economic and social development. In the meantime, it makes education a high risk business for politicians, policy makers, administrators, operators and managers of education service delivery. How can exclusion, dropout, and repetition be minimized, while learning is maximized? How can resources be used more efficiently? How could the sector’s performance be improved? It is clear that paying teachers, providing the capitation grant, printing textbooks, investing in infrastructure and school feeding, maintaining services, information systems, administration, and records are all important. But it is becoming clear that these are not enough. There are persistent and also newly emerging challenges and new demands. Existing resources need to be used more efficiently, policies need to be realigned and services need to be provided more equitably. Also, the increasing number of stakeholders expects better accountability from the providers. This chapter suggests using some of these expectations to measure the performance of the sector and of the delivery of education services. A. Equity and Disparity in Education Services 231. Education is considered to be among the great social equalizers, creating opportunities for young people to improve their living standards, become upwardly mobile, and break out of poverty that perhaps their parents’ generation was mired with. 91 However, education can also be a key factor in reproducing or even strengthening existing social inequities. If services are not accessible at the same levels to unprivileged or disenfranchised groups, they likely fall further behind and risk not only sustained deprivation but also social fragmentation. Access to quality learning has to be provided to all children on equal terms. 232. Equitable education services usually indicate better sector performance, high attainment, strong learning results on a massive scale, strong skills and strong relevance. If access is inequitable, it diminishes the performance and compromises the social and economic benefits. That is why the Ghanaian government has been committed to equitable access. A number of policies support this goal, including, among others, the elimination of fees, the provision of capitation grants and textbooks, school feeding and the designation of deprived districts. But, by the same token, inequitable services, disparities in access and in educational inputs could not only compromise the effect of these policies but also result in greater social inequity and fragmentation. There are three levels of potential policy actions in the context of disparities: 233. avoiding negative discrimination: At a minimum, the allocation of resources should not discriminate along social, cultural or geographical grounds. If x amount is available for education per student, the Government needs to assure that x amount is allocated so that all students see the benefit of the inputs equal to this amount. 234. establishing specific policy incentives to improve supply of services: At a higher level, Government resources can be allocated in a way that addresses specific difficulties in education service delivery. If, for instance, enrollment and attendance is more challenging in some areas, typically in the rural and poor Northern regions, the Government conceives policies that address this specific educational problem. For instance, teacher salaries can be topped up for those who are willing to teach in remote areas, and accommodations or subsidized transportation can be provided to these teachers. 235. Creating positive discrimination to stimulate demand: At the highest level, the Government may develop policies that address underlying social and economic problems by stimulating demand and thus access to education. These policies may include school feeding, cash transfers or any type of extra provisions that relieve the burden on families of sending children to school. Critical to the success of these measures is how they are targeted and monitored. 236. In this section, we’ll analyze the equities and disparities in accessing education services by income, gender and geographical location. Household Income 237. A household’s level of welfare is a good predictor of whether a child is enrolled in school. Although research results point to geography as the strongest determinant of non-participation in basic education, it is closely interrelated with low income levels. The three Northern regions are the most geographically isolated, have lower population density, higher infrastructure and transport costs and a number of other geographic factors that hinder development. Accumulatively this contributes to lower levels of 92 welfare. The real challenge of meeting the Millennium Development Goals (MDG) is in improving the education service delivery and incentive packages for the most poor and rural children. 238. In Ghana, the poorest 20 percent of the districts based on household income, earns roughly three times less than the richest. All districts falling into the poorest group are located within the borders of the Upper East, Upper West and Northern regions, representing a very strong link in Northern Ghana between physical remoteness and income marginalization. 239. Although basic education in Ghana is free, compulsory and universal, socioeconomic inequities help to explain performance of key education indicators, but so do geographic disparities and, to a lesser degree, differences in gender. In terms of out- of-school children, the poorest group of districts accounts for the vast majority. in the negative influence of socioeconomic status increases incrementally with every higher grade. Notwithstanding, the parity index (the ratio of the poorest to richest districts) in primary net enrollment was 0.89. 240. The parity index in household income at the district level was considerably lower at the JHS level, in relation to gross enrollment (.74) and net enrollment (.51). Thus, youth of the appropriate junior high school age, reflected by net enrollment data, were twice as likely to be enrolled in a junior high school if they resided in one of the districts falling into the top group of districts compared to those in the bottom group of districts. Not only do household income levels strongly influence access to education opportunities, but the level of education one receives also tends to determine earnings later in life. What the data in Figure 4-1 illustrates, is that the private costs of going to school persist as a financial barrier for the most disenfranchised, so that targeted measures to decrease school costs can make a difference. Figure 4-1: Net enrollment at the basic level, by quintile 100 90 93 88 91 85 NER Prim. 80 81 70 60 61 NER JHS 50 44 46 40 37 30 31 20 Q1 Q2 Q3 Q4 Q5 (poorest) Average Level of Welfare (richest) Source: Estimates based on GLSS 2005/06 and EMIS 2008/09. 241. Poverty as the most cited explanation for non-attendance. The most common reason given for why children were not in primary school was that the costs were too high, 41.5 percent of all reasons given, based on data from the Core Welfare Indicators Questionnaire Survey (GSS 2003). The second most frequent cause was that school was 93 considered unnecessary or uninteresting, with 34.6 percent of those examined believing that the curriculum content and overall school relevance were limited in real life and employment. Another study found that the indirect costs of schooling are such that they were most commonly given as the primary cause for not being able to attend school (Oduro 2000). Among the most prevalent and, for very poor households, burdensome school expenses are the registration fees, school levies and costs associated with lunches and uniforms (Oduro 2000). Registration fees, according to a study by Transparency International (2009a), were reported being paid by 9 percent of surveyed parents. Within classrooms, acquiring school supplies such as notebooks and textbooks can pose challenges for poor children. 242. Outside classrooms, transport costs can contribute to the economic barriers. In addition, according to a study by Transparency International (2009a), 48 percent of households in Ghana indicated that they pay additional fees for private tutoring. 243. Wealth influences enrollment incrementally at each higher grade. Socioeconomics incrementally gains significance at each higher level of education, resulting in senior high students and, most profoundly, tertiary students coming generally from the top quintile. Therefore, higher education has been characterizes as an institution that tends to exclude the most economically marginalized segments of society, but other factors add to this reality. Residing in urban environments strongly contributes to the background characteristics determining tertiary enrollment, although males are also about twice as likely to be enrolled in tertiary institutions as females. The gender parity index in tertiary has been stagnate and has even declined in recent years, dipping to 0.49 in academic year 2007/08. 244. A child from the poorest quintile is on average 3 years over grade-relevant age. On average, children from households in the poorest quintile, which are strongly concentrated within the three Northern regions, are about three years over the normal age in their respective grade. Within the richest districts, children tend to be less than two years older than the official age for their particular grade. The causes of being overage in school are closely associated with income-poverty due to, for instance, the necessity to help family members care for younger siblings or during times of harvest (i.e. child labor), which may lead to postponed entry into primary one or frequent repetition as a result of lost instruction time. The consequences of high age-grade delay are likely to raise the public costs of schooling, strain the pedagogy process, reduce learning results, and increase dropout rates and early labor force integration. It is essential to provide incentives to, in particular, poor and rural households to ensure that children enroll at the official age. 94 Figure 4-2: Mean Age-Grade ‘Delay’ per Pupil, by Household Welfare Quintile Mean Age-Grade 'Delay' Per Pupil by Household Welfare Quintile 4 3 mean 'delay' in years 2 1 0 1 2 3 4 5 d) welfare quintile Source: Based on GLSS5. 245. Poorer quintiles have much lower completion rates. Students from economically deprived households have not only fewer chances than their rich counterparts to be enrolled in school, but those that are have an even smaller probability of completing a given cycle. 246. While some children drop out and don’t complete as a result of monetary obstacles, other children come to school late and leave early due to local economic activities such as farming and fishing above all in rural and poor-urban communities. A survey assessing the extent of child labor on cocoa plantations in Ghana and Côte d’Ivoire reported cases of children having to use toxic pesticides, work in conditions of very high temperatures and use dangerous implements (Payson Center for International Development 2008). 247. Data in Figure 4-3 illustrate the higher probability of students completing basic education, if they come from districts with higher levels of welfare. On average, about four in every five students complete primary school, if they attend schools in one of the poorest 20 percent of the districts. In contrast, 96 percent of children complete the primary cycle, who go to schools located in districts of the top quintile. Thus, data disaggregated by levels of income paint a clearer picture of the education landscape. If the wealthiest 20 percent of Ghanaians constituted their own country, it would have nearly achieved universal completion in primary. In junior high schools, the likelihood of a student completing averaged between 84 percent for those residing in one of the districts falling into the richest quintile and 58 percent for students in one of the districts of the bottom quintile, all of which are located within the three Northern regions of Ghana. 95 Figure 4-3: Primary and JHS Completion Rate, by Quintile 100 96 90 88 86 84 80 81 80 81 77 70 66 60 Primary 58 JHS 50 Q1 Q2 Q3 Q4 Q5 (poorest) Average Level of Welfare (richest) Source: Estimates based on GLSS 2005/06 and EMIS 2008/09. 248. Trained primary teachers are almost twice as scarce in the poorest districts. The degree of inequity in the distribution of trained teachers throughout the education system is significant, as illustrated earlier in terms of inter-regional variations and displayed in Figure 4-4 as to disparities across quintile levels. Within districts classified into the poorest quintile, there were about 127 students per trained teacher in primary schools in 2008/09, while the primary PTTR reduced approximately to half in the fourth and fifth richest quintiles. Of the 138 districts in Ghana, one finds that 10 of the 20 districts with the highest PTTRs also fall into the group of districts within the bottom quintile, all of which are located within the Northern (7), Upper West (2) and Upper East (1) regions. 249. It is critically important to mobilize trained teachers into such districts afflicted with the poorest conditions in the country, since they tend to be accompanied by severe infrastructure deficiencies and have by and large more difficult living conditions. This is where hardship allowances, offering free housing accommodations, mobilizing newly graduated teachers and other innovative enticements can play a crucial role in incentivizing trained teachers to accept difficult-to-fill postings. 96 Figure 4-4: Primary PTTR in Districts, by Quintile 250.0 Note: Bia District (Western), as an outlier Karaga district with a PTTR of 450.9, is (Northern) 219.9 not illustrated. 200.0 150.0 100.0 50.0 Ho district (Volta) 26.8 0.0 Q5 Q4 Q3 Q2 Q1 (richest) (poorest) PTTR Avg. 67 58.6 95.1 106.9 126.6 Source: Estimates based on GLSS 2005/06 and EMIS 2008/09. 250. Poorest performers come from poorest quintile. As a result of their inferior learning conditions, children from poorer households acquire fewer competencies in the core subjects of English and math, arguably the epicenter of primary schooling. Children residing in the economically most affluent 20 percent of the districts, measured by household income levels, answered an average of 46 percent of the questions on NEA English exams in 2007 correctly, compared to only 37 percent among the children coming from the 20 percent of the income-poorest districts. As to NEA math results, the learning gap between the richest and poorest quintile was nearly as wide, so that the top quintile answered a total of 37 percent of the questions correctly, whereas the bottom quintile averaged 31 percent. 7 of the 20 districts with the poorest results on NEA English and 11 of the 20 lowest performing districts on NEA math assessments are located within the three Northern regions. This link is also consistent in the top quintile, in which many of the income-wealthiest and highest performing districts on both NEA English and math are disproportionately concentrated within and near the borders of the Greater Accra and Eastern regions. Large disparities not only exist across quintiles but also between districts themselves. 97 Figure 4-5: NEA learning results in districts for primary 6, by quintile English Avg. Math Avg. 46 42 41 40 37 37 34 33 32 31 70 Tema District (Greater Accra) 60 60 Talensi-Nabdam District (Upper East) 49 50 40 30 Saboba-Chereponi District (Northern) 25 20 Mamprusi East District (Northern) 21 10 Q5 Q4 Q3 Q2 Q1 Q5 Q4 Q3 Q2 Q1 (richest) (poorest) (richest) (poorest) Source: Estimates based on GLSS 2005/06 and NEA 2007. Box 4-1: An Outlier in Learning Achievement - Talensi-Nabdam’s success at a Glance General trends in learning outcomes appear to be consistent with background factors such as welfare levels, although some outliers do exist. Among them, the most profound is the Talensi-Nabdam District (Upper East), with average learning results on NEA math exams in 2007 making it the country’s leader. Talensi-Nabdam, ranks as the 18th poorest district in Ghana, and it’s learning success is difficult to explain. In 2008/09, its gross enrollment in primary was significantly below the national average of 95 percent, reaching merely 79 percent. And only 66 percent completed. The pupil teacher ratio at the primary level was 43.2 (nationally, 32.4) and the PTTR was 128.1 (nationally, 67.5). Only 34 percent of the teacher work force was trained (nationally, 48%). In some respects, marginalization of girls and women persists in the Northern areas of Ghana, although gender parity in primary is on the rise in Talensi-Nabdam, reaching 0.89 in 2008/09, but lagging behind the national mean of 0.96. Indicators show that the district tends to perform overall poorer than the national average in key areas, although variations from Ghana’s mean are smaller in terms of school materials and facilities. The district did, however, have slightly more core textbooks per pupil and a greater percentage of schools with access to potable water. On NEA English assessments the district also performed better than the national mean in 2007 (41) with a score of 45. What then helps to explain its high learning outcomes? In part, its status as a deprived district has helped in securing additional resources to improve its schools infrastructure and performance. It appears also that district administrators and head teachers place a stronger emphasis on acquiring core mathematic competencies, as is the case in many Ashanti districts. Talensi-Nabdam shows that in spite of geographic, economic and demographic obstacles as well as other development hindrances, income-poor districts can still break the ‘low performance trap’. 251. As data results indicate, location and affluence are very important predictors of a student’s performance, with one’s prospects for educational achievement being 98 influenced in good measure by one’s place of residence and one’s household background. On NEA English exams in 2007, 6th graders who scored on average highest resided in the Kumasi district (Ashanti) and the Tema district (Greater Accra), with students answering 59 and 60 percent of the questions correctly, respectively. In contrast, 6th grade students in the poorest performing districts, the Saboba-Chereponi district (Northern) and the Juabeso district (Western), were only able to answer on average 25 and 26 percent of the questions correctly, respectively. 252. Northern region is poorest as to income and JHS learning results. Differences in learning achievement are significant across regions as well as across quintiles and, although to much lesser extent, also in terms of gender, mainly in math results. On average, about 63 percent of students going to schools located in districts of the top quintile passed the BECE exams in 2008/09, whereas only 40 percent of students who resided in the districts of the poorest quintile passed, the outcome of which is a parity index of 0.63. Figure 4-6: BECE English Pass Rate in Districts, by Income 1.00 BECE English Pass Rate 0.90 0.80 0.70 0.60 0.50 0.40 0.30 0.20 Districts in the Northern region: Gushiegu, Karaga, 0.10 Bunkpurugu-Yunyooa, Nanumba South and Central Gonja. 0.00 - 1,000,000 2,000,000 3,000,000 4,000,000 Level of Household Income (old cedis, 2006) Source: Based on data from GLSS, 2005/06; EMIS, 2008/09. 253. For social cohesion it will be important to tackle large inequities. Viewing the parity indexes, if the Government of Ghana has the intention of strengthening demand- side targeting to the children of the poorer or poorest quintiles to reduce socioeconomic disparities, it would seem more sensible from a policy perspective to ensure that the pupil per trained primary teachers ratio (.53 parity index) is reduced by deploying more trained teachers in districts with the lowest income levels. 254. Figure 4-7 shows the distribution across wealth quintiles of children not attending school, as well as those attending primary and secondary. While 33 percent of Ghanaians in the poorest quintile have not attended a primary school, the corresponding share was 8 percent in the richest quintile. Based on 2008 data from GDHS, net enrollment shares among students in the bottom quintile were 59.1 percent in primary and 22.2 percent in secondary, whereas these figures in the top quintile were 85.9 percent in primary and 60.7 percent in secondary. 99 Figure 4-7: Non-attendance and net enrollment in basic education, by quintile 60.7 Secondary 51.3 41.7 school 34.9 22.2 Highest 85.9 Fourth 81.8 Primary school 76.5 Middle 71.9 59.1 Second 8 Lowest Not attending 13 19 school 27 33 0 20 40 60 80 100 Source: 2003 data on not attending school: Harttgen et al. 2008; 2008 data on primary and secondary school: Ghana Demographic and Health Survey, 2009. Note: Secondary school includes both junior and senior high schools. 255. Income-levels determine access to tertiary. As mentioned, a household’s level of welfare strongly helps to understand why some children have access to higher levels of education, while others don’t. In essence, levels of income determine an individual’s accessibility to tertiary institutions, with 9.6 percent of all pubic tertiary students coming from the poorest two quintiles and 66.9 percent coming from the richest quintile in 2005/06. Table 4-1: Enrollment in Secondary and Tertiary Public Schools, by Quintile Quintile Secondary (2005/06) Tertiary (2005/06) Poorest quintile 12.1 1.8 2nd quintile 17.2 7.8 3rd quintile 23.7 9.4 4th quintile 23.3 14.1 5th quintile 23.8 66.9 Richest to poorest 2.0 37.2 Source: Based on GLSS 2005/06. 256. Identifying new deprived districts based on education performance and income. In terms of policy formulation, there are options for the Ministry of Education to be able to link the promotion of the national development objective of poverty reduction with the improvement of key education indicators. This could be attained by targeting measures to the income-poorest districts that at the same time have the poorest results on key education indicators in the country. Thus, making gains on pro-poor growth and income- poverty alleviation can be combined with achieving education targets, a political aspiration of many developing countries. Meeting economic and education goals simultaneously and with greater impact can be achieved by reducing the number of deprived districts, both in terms of income and education performance, from 53 in 2010 to a significantly smaller number . 100 Gender Parity, Gender Equity 257. Studies have shown that the persistent factors affecting girls’ enrollment and retention patterns include domestic obligations, sexual harassment in school, pregnancy and early marriage, and traditional attitudes, especially in some communities in Northern Ghana, regarding the importance of girls’ education and the role of girls within communities and families. 258. The Northern region presents the lowest ratio of girls to boys in primary school. This is also the case in junior high school. About twice as many boys are enrolled as girls in several districts within the Northern region, where the gender parity index in junior high school only reached 0.50 in Tolon-Kumbungu, followed by Nanumba South with 0.51 and Savulugu Nanton with 0.56. The figure below shows that while income rises, so does gender equality, other social and cultural aspects also have an effect. Figure 4-8: Gender Parity Index in Primary and JHS, by Income 1.20 Ratio of Girls to Boys in Primary 1.00 0.80 Districts in the Northern region: 0.60 Gushiegu, Karaga, Savulugu- Nanton and Tolon-Kumbungu. 0.40 0.20 0.00 - 1,000,000 2,000,000 3,000,000 4,000,000 Level of Household Incom e (old cedis, 2006) Source: Calculations based on GLSS 2005/06 and EMIS 2008/09. 259. While policies addressing the issue, such as sensitization, community outreach, and hiring more female teachers, are very important, the Government also has some effective economic and educational instruments that can help break major gender barriers. The first issue is to recognize the age and grade where interventions are most needed. When girls in Ghana drop out, they typically do so following the fourth grade or else following the end of primary education. Gender parity significantly drops at JHS level. Key concerns for families at this age is hygiene and security. Separate toilets, for instance, have proven to be an effective incentive. Girls are considered from age 12 or 13 to be needed on family farms. Take-home food portions, in addition to school lunches, have proven to be effective incentives to keep girls in school. Similarly the more effective measure of cash transfers to the poorest families for keeping girls in school may be an option for policy makers. Geographical Inequities 260. In parallel with rapid enrollment growth in KG, inequities in access have also grown between the Northern regions and the rest of the country. In 18 of the 25 districts in Northern Ghana the net enrollment rate is 50 percent or below. KG, which has become 101 part of basic education is in fact far from being part of it in the deprived territories. Infrastructure is non-existent, teachers are under qualified if they are present at all, and pupil teacher ratios are typically 80 or above. KG has become part of the basic cycle with the purpose of giving children a head-start during the early childhood years. This head- start is especially important to the poor given that they don’t have the necessary home environment (educated parents, books, electricity) that helps provide a leg up. Yet, precisely these at risk children, families and communities are the ones that fall further behind in absence of adequate KG provisions. Map 4-1: Kindergarten Net Enrollment in 2008/09 Source: Based on EMIS. 261. A UNESCO (2010:161) report focusing on the marginalized, identified the Northern region as the most educationally deprived, with 61.6 percent of its population living in education poverty, i.e. having less than four years of education. The share of the Northern region’s population living in extreme education poverty, measured as having fewer than two years of education, was 16.6 percent: 12.1 percent of men and 20.4 percent of women. Among the population aged 7 to 16, about half (52.5 percent) of all rural girls from the poorest quintile in the Northern region, who tend to be systematically the most disadvantaged, have not had any education. It is important to reiterate that, as Ghana aspires to reach universal access to primary, progress will be contingent on how successful the government and non-government organizations are in focusing efforts on these marginalized groups. Box 4-2: Reaching the marginalized: the ‘School for Life’ program As a means to widen access to marginalized children, School for Life is a program led by non-governmental organizations that target the needs of hard-to-reach children in the North. This nine-month literacy program aims to facilitate the re-entry of 8 to 14 year old children into primary schooling. It provides an intensive and relevant course that takes seasonal labor demands into account when preparing the school calendar. A distinct aspect is its orientation towards the community. Courses are given in the local language 102 and are taught by instructors, many working on a voluntary basis, all of whom have been locally recruited. A particularly relevant component for the rural poor is that children do not need uniforms and are provided with free textbooks, thereby decreasing central indirect costs of schooling. Having reached a total of about 85,000 children between 1996-2007, the scope of the School for Life program has been wide within its eight targeted districts. A study found that more than 90 percent of students graduated from the program, about four out of five students (81 percent) had attained third grade literacy and numeracy levels and 65 percent went on after completion to (re-)enter the formal schooling system. Graduates who transition into the formal system tend to outperform their counterparts in both English and math (UNESCO 2010). 262. At the district level, the prevalence of a net enrollment rate in primary below 60 percent fell from ten districts in 2006/07 to only one in 2008/09, illustrating strong enrollment growth among the educationally most deprived districts. In 2008/09, the only district in Ghana with primary net enrollment below 60 percent, as Map 4-2 indicates, was the Wa East district in the Upper West region (59.4 percent). Among the 17 districts where less than three fourths of the children in the appropriate age group had access to primary schooling, more than half are found within the Northern, Upper West and Upper East regions. At the other end of the spectrum, the number of districts with net enrollment in primary as high as 90 percent or higher more than doubled over this time span, increasing from 27 to 61 districts. Of these 61 districts, 57 are located outside of the three Northern regions, underling the North’s geographic isolation, higher incidence of poverty, child labor demand and thus lower participation in schooling as well as the common unwillingness of trained teachers to accept postings in remote communities. Map 4-2: Primary NER in 2006/07 and 2008/09, by District Source: Based on EMIS. 103 263. Primary Net Admissions Rates (NAR) is lowest in Northern Ghana. As Map 4-3 indicates, eight of the eleven districts with the lowest primary net admission (50% or below) were found within the three Northern regions – i.e. the Upper West, Upper East and Northern regions. The Wa district and Wa East district in the Upper West had the lowest primary net admission rates, averaging only 41.2 and 37.5 percent, respectively. Given that the aggregate primary NAR in Ghana reached 72.1 percent in 2008/09, geographic disparities are pervasive. While 31 districts within 7 regions had a primary net admission rate greater than 90 percent, 55 districts throughout all 10 regions had a primary NAR less than 70 percent. This suggests that though inter-regional disparities in primary NAR are large, they become much more accute when disaggregated at the district level. Map 4-3: Primary Net Admission Rate in 2008/09 Source: Based on EMIS. 264. Districts along the coast tend to have the highest concentration of students who complete primary, while those in more Northern and rural areas tend to have the lowest. In fact, in six districts of the Northern region less than 70 percent of primary students completed. Nearly two thirds of all districts with a primary completion rate of seventy percent or lower are found within or bordering the Upper West, Upper East or Northern regions. The Karaga district (Northern region) had the lowest primary completion rate at 48.4 percent, followed by the Afram Plains district (Eastern region) at 52.3 percent and the Gushiegu district (Northern region) at 54.6 percent. To meet Universal Primary Completion (UPC) by 2015, it will be essential to target efforts to the most disadvantaged and poorest performing districts, particularly in the off-road rural communities and impoverished areas of the North. 104 Map 4-4: Primary Completion Rate in 2008/09 Source: Based on EMIS. 265. Inter-regional disparities in primary quality and learning. The parity index in NEA results can be calculated by dividing the lowest scoring region in English and math, in this case the Northern region, by the region performing highest, i.e. Greater Accra. Thus, the parity index of NEA scores, 0.70 in English and 0.69 in math, illustrate inequities across geographic lines in learning achievement within these core subjects. In explaining these regional disparities in learning one must look at factors contributing to the higher concentration of trained teachers in Greater Accra and to lack of desire to be posted in “bush schools� within off-road rural settings, considering migration patterns to urban centers, but also that distances to schools in urban areas tend to be shorter and thus attendance rates higher. When asked, why children were not in primary school, about 5 percent stated that the distance to school was simply too far, according to data from the Core Welfare Indicators Questionnaire Survey (GSS 2003). Many districts suffer from a chronic shortage of (trained-) teachers, with some of these districts being located in the Northern region where on average only 27 percent of primary teachers possessed formal qualifications. 105 Figure 4-9: Average NEA English scores by region Average NEA English Test Scores by Region 60.0 40.0 20.0 English 2005 0.0 English 2007 Source: Wodon and Joseph 2010 based on NEA. 266. Relationship between primary P(T)TR and completion. At the regional level, 2008/09 data does not directly support the assumption that higher PTRs in primary impact completion rates in an adverse way. When looking at the pupil per trained teacher ratio in primary, there seems to be a slight, but not very significant relationship. This is to some extent explained by lower levels of efficiency and equity in the distribution of trained teachers in comparison to all teachers in general. In primary, the Western region had the highest pupil per trained teacher ratio of 145.5, but was able to reach a completion rate of 92 percent, about 6 percent higher than the national average. The Eastern region had the lowest pupil per trained teacher ratio of 54.9, while only reaching a completion rate of 83 percent, about 3 percent lower than the national average. This unexpected association appears even stronger in junior high school. Figure 4-10: Primary P(T)TR and Completion Rate, by District in 2008/09 220 Pupil (Trained) Teacher Ratio 200 180 160 140 120 100 PTR 80 60 PTTR 40 20 0 0.40 0.50 0.60 0.70 0.80 0.90 1.00 Primary Completion Rate Source: Based on data from EMIS. 267. Using the parity index as a potential policy tool. With the intention of closing geographic disparities, the Government could use parity indexes as a means to measure the inter-regional inequity between the poorest and highest performers, and then channel resources to the given input that has the lowest parity and target measures to those corresponding regions and districts with the poorest performance on the respective indicator. For instance, table 4-10 indicates that among the two key indicators illustrated, the parity index is lowest – thus the regional disparity largest – in terms of the pupil per 106 trained teacher ratio. (The flat PTR line shows almost no relationship between PTR and completion rates. The downward slope of the PTTR line shows a small relationship, and hence a lower parity index. Thus, measures aimed at reducing the primary PTTR would aid in closing some of the largest geographic disparities in Ghana’s education system. In addressing this issue, newly trained teachers would need to be deployed particularly in outlier districts with high PTTRs within, for example, the Western (145.5), Northern (140.3), Upper East (113.5) and Upper West (106.6) regions. 268. Students in the North tend to be about 3 years over-age. One cannot stress enough the importance of children beginning the primary cycle at the official age and keeping repetition rates to a minimum, in order to raise the chances of completion. However, Ghana’s education system faces challenges both in timely enrollment and transition. Children in schools within the three Northern regions tend to be about three years older than the official age in a particular grade, while in the Greater Accra and Ashanti regions the average actual age of students is about two years older than the appropriate age. Such high age-grade delay places pressures on schools in terms of higher costs, increased strains on the pedagogy process due to differences in students’ learning capabilities, preferences and maturity levels, reduced learning results, higher rates of drop out and increases in the likelihood of early entry into the labor market, discussed in greater detail in chapter 2 on access and enrollment. Figure 4-11: Comparison of Mean Age-grade Delay, by Region Mean Age-Grade 'Delay' by Region 4 mean 'delay' in years 3 2 1 0 i o n t t l n n a lta nt ra as es er af r er cr te ha vo nt re rw ah ac st rth es ce as ea pe no pe er g w on up at up br e gr Source: Akyeampong 2009 based on GLSS 5. 269. In general, primary-age boys living in rural areas within the Northern or Upper West regions of the country tend to have the highest probability of repeating or dropping out. The largest share of children tend to repeat the first grade, while no clear patterns seem to exist in the frequency of dropout at different levels of primary education. 270. Inequitable teacher deployment in pre-tertiary. Nationally, there was a slight decline from 33.5 to 32.4 pupils per teacher (PTR) between 2006/07 and 2008/09. However, at the district level there was a total of 40 or more students per public teacher in 11 districts in 2006/07, rising to 25 districts by 2008/09, mostly lying within the Upper West, Upper East and Northern regions, due mainly to the rapid expansion of admissions. Although high teacher intake growth has been taking place, rising geographical inequities 107 could undermine the impact of the sector’s growth. In 2008/09, out of the 33 districts with a primary PTR of 30 or below, 25 of these districts are found within three regions – the Volta, Eastern, and Ashanti regions. 271. Over time and geography, inequities in PTRs continue to reflect to a large degree Ghana’s underlying economic inequalities. The rural and socioeconomically most marginalized districts and regions tend by and large to have the highest PTRs and especially PTTRs in the country. Data suggests that regional variations in PTRs exist across the country and across all pre-tertiary education levels. Overcoming such disparities and teachers’ unwillingness to work in such areas is achievable through targeted measures that (i) recruit, train and deploy teachers locally, (ii) provide teachers with monetary and in-kind incentives to relocate and retain them in hard-to-place communities and, amongst others, (iii) enforce legislation that obliges all new teachers to be posted in such communities the first few years. Map 4-5: Primary PTRs in 2006/07 and 2008/09, by Region Source: Based on EMIS. Note: PTRs represent public schools. 272. Severe discrepancies in trained teacher distribution. Not only do large trans- regional disparities in the share of trained primary teachers exist, but also intra-regional. Among districts within the same region, significant inefficiencies in trained teacher distribution emerge primarily due to the urban-rural divide. In the Western region there is a variation of over 60 percent in the share of trained teachers among its districts, meaning that the difference ranges from the Shama-Ahanta East District where 69.3 percent of teachers are qualified to the Bia district with 8.2 percent qualified. 108 Figure 4-12: Portion of Primary Teachers with training, by Region/District, in 2008/09 100.0% 90.0% Ho district (Volta) 87.2% 80.0% Akwapim North district (Eastern) 76.9% 70.0% 60.0% 50.0% 40.0% 30.0% 20.0% Karaga district (Northern) 13.8% 10.0% Bia district (Western) 8.2% 0.0% EN A O TI ST A T N AL RN N LT AF ES CR ER AN ER EA TR TE VO AH W AC ST H TH ER ES AS ER EA G R C W PP O N PP O N U BR U Source: Based on EMIS. 273. Many schools need major repairs and lack clean water. In the Upper West and Upper East regions, where poor infrastructure and remoteness are characteristically critical barriers, nearly 30 percent of public primary classrooms need repairs, while several districts in the other regions classroom infrastructure is even worse, with 30 percent or more of classrooms needing repairs. Thus, despite large education deprivation in the Upper West and Upper East, classrooms and schools seem to be in overall better condition than in other regions. The Volta region has many districts with 40 percent or more of public primary classrooms in need of major repairs, in spite of the region having one of highest percentages of trained primary teachers and having high learning achievement results, for instance, on NEA. 274. In terms of public primary schools with access to drinking water, districts with 40 percent or less of schools having access tend to be more highly concentrated in the Northern areas, while those with 80 percent or more access to potable water tend to be located in the Southern regions, although some exceptions exist. Thus, while 14 of the 24 districts with at least 80 percent access to potable water were located in the Western and Ashanti regions, 13 of the 23 districts with schools that have 40 percent or less access to drinking water were located within the Upper West, Northern and Brong Ahafo regions. Given that schools in only 24 districts have on average 80 percent or more access to potable water, clean water scarcity may continue to lead to a higher prevalence of disease, such as guinea worm infection, and deter teachers from coming to such communities. 109 Map 4-6: Classrooms Needing Major Repairs and Schools with Potable Water, 2008/09 Note: Both maps refer to public primary schools. Source: Based on EMIS. 275. Regional variations in pupil per (trained) teacher ratio. Only analyzing primary pupil teacher ratios, the lowest being in the Eastern region (28.6) and the highest in the Upper East (47.6), doesn’t always offer an accurate picture of the teacher work force. When examining PTTRs one can see much larger regional disparities. In primary schools, the average classroom in the Eastern region had about 55 students per trained teacher, however, the primary PTTR was nearly three times higher in the Northern (140.3) and Western (145.5) regions. To date, the Ministry of Education has not been able to develop a solid policy scheme to overcome institutional inefficiencies of mobilizing trained teachers, especially to communities of Northern Ghana, but also to the most rural areas of the Western region. In terms of junior high school, while the number of students per teacher ranged from 15.2 in the Eastern region to 24.3 in the Upper East, the geographic gap in PTTRs varied more widely from 22.4 in the Eastern region to 43.4 in the Western region. Distributing trained teacher more equitably in Ghana has evolved into a central policy challenge, as discussed in the chapter on quality. 110 Figure 4-13: PTTR in Primary and JHS, by Region in 2008/09 160 140 145.5 Primary JHS 140.3 120 113.5 106.6 100 80 82.7 84.1 66.8 71.1 60 54.9 59.1 40 38.7 43.4 29.2 32.1 35.1 22.4 24.5 25.8 24.7 27.3 20 0 n lta ra st ti t o l rn rn tra es an er f Ea ha c te te Vo W Ac en th h s es A As or Ea er C er W ng pp N pp o U U Br Source: Based on EMIS. 276. JHS access and completion is lowest in the North. Geography matters at all levels in the access to and quality of education. Variations in enrollment and participation at the regional level are such that about 70 percent of children are not enrolled in junior high school in some parts of the country, more specifically in the Northern region. The nature of small communities in rural settings does not necessarily support high enrollment patterns particularly due to living settlements that are small, sparsely populated and extensively spread out. The goal of attaining universal basic education by 2015 confronts serious challenges in the rural and Northern areas of Ghana. The parity index in net JHS enrollment is 0.38, implying a nearly threefold gap between net enrollments in the Central region at 70 percent and the Upper East at 27 percent. 277. The Central region has not only been most successful in bringing the appropriate age children into junior high schools. It was also able to ensure that 93 percent of them completed the basic education system in 2008/09. However, in Northern regions the probability that a JHS student completes the three-year course of study is significantly lower than the national average; in 2008/09 only 62 percent in the Upper East, followed by the Upper West with 60 percent and 56 percent in the Northern region. 278. Extensive inter-regional variations in JHS learning. Given that the parity index of BECE pass rates was 0.41 in English and 0.53 in math, the junior high system appears to be plagued with considerable geographical inequities in learning results. On BECE English exams, it was more than twice as likely for a junior high school student to have passed, if he resided in Greater Accra (81% pass rate) than in the Northern region (33% pass rate). As to BECE math assessments, the regional divide is most extensive between the Ashanti region with the highest pass rate of 79 percent and the Northern and Volta regions with the lowest pass rates of 43 and 42 percent, respectively. This illustrates that the education administrators of the Ashanti region place a stronger emphasis on math curriculum at the JHS level, while Greater Accra is by far the highest performer in English. However, at the primary level the Greater Accra region clearly outperforms every other region in both English and math, based on the NEA results of 6th graders. 279. The Northern regions are not only the poorest performers on NEA English and math exams, but as a regional average they also had the overall weakest pass rates on 111 BECE assessments. It is important to rethink potential targeted measures to improve the learning results of students residing in the Northern regions. One can’t sufficiently stress the point that geography is a strong (maybe the strongest) determinant of learning results throughout the basic system. However, the inter-regional divide tends to conceal other underlying variables, among others, the fact that trained teachers are concentrated in cities, where social goods and services tend to be more accessible, work productivity is higher and walking distances to schools tend to be much shorter. Figure 4-14: Junior High P(T)TR and Completion Rate, by Region in 2008/09 50 The 3 Regions in the North 45 Western Pupil (Trained) Teacher Ratio Upper East 40 Northern 35 Upper West Brong Ahafo 30 Central Volta Greater Accra 25 Eastern Ashanti Northern Upper East 20 Brong Ahafo Western Greater Accra Upper West Central 15 Volta Eastern Ashanti 10 PTR 5 PTTR 0 0.50 0.60 0.70 0.80 0.90 1.00 Completion Rate Source: Based on data from EMIS. 280. Very large district-level disparities in JHS net enrollment. In aggregate terms, data on net enrollment in JHS suggest that a 3.2 percent decline took place between 2006/07 (51%) and 2008/09 (47.8%). Net enrollment trends in JHS, representing some of the largest inequities across districts in basic education, are not positive and the gap appears to be widening. Deterioration in net access has translated into a jump from 19 districts with net enrollment of 33 percent or lower (2006/07) to a total of 37 districts (2008/09). Eighteen 18 districts regressed to a junior high NER of 33 percent or below; they were located in the regions of the Upper West (3), Upper East (3), Brong-Ahafo (4), Ashanti (4), Eastern (3) and Volta (1). 281. While the number of districts with very low net enrollment in JHS increased, the total amount of districts with a NER of 70 percent or higher rose from 9 to 14 districts over this time span, all five of which were located in the Central and Western regions. Geographically, there seems to a strong comparative advantage in net access for junior high school students located in or near the coastal areas. While the Awutu-Efutu-Senya district (Central) and the Ga East district (Greater Accra) nearly achieved universal net enrollment in JHS at 99 and 96.8 percent, respectively. The Wa East District (Upper East) and the Sene district (Brong-Ahafo) have been least successful in widening net access, reaching only 12.2 and 14.2 percent, respectively, of the appropriate age population. 112 Map 4-7: JHS Net Enrollment in 2006/07 and 2008/09 Source: Based on EMIS. 282. In Ghana, about 65 percent of junior high school teachers have received formal qualifications. At this lower secondary level, while a total of 56 districts had a share of qualified teachers above 75 percent, there were 75 districts with a 50-75 percent qualified teacher work force and in 7 districts less than 50 percent of teachers had formal qualifications. Thus, the Western region, where four of the seven districts with less than 50 percent of qualified teachers are found, is home to the largest share of unqualified JHS teachers in Ghana. Map 4-8: Share of Trained Public JHS Teachers in 2008/09 Source: Based on EMIS. 113 283. Urban-rural divide. As Table 4-2 indicates, net enrollment of children in primary varies widely, depending upon if a child resides in a rural area (69.8) or in a urban area (80.3). In secondary schooling, disparities in enrollment are typically larger, with about half of the official secondary-school-age population enrolled in urban areas (51.9) compared to about one third in rural areas (34.5). Table 4-2: Gross and Net Enrollment in Primary and Secondary in 2008, by Urban/Rural Location Gross enrollment rate Net enrollment rate Male Female Total GPI Male Female Total GPI PRIMARY SCHOOL Urban 110.2 108.9 109.5 0.99 81 79.5 80.3 0.98 Rural 107.1 105.9 106.6 0.99 68.8 70.8 69.8 1.03 SECONDARY SCHOOL Urban 72.8 63.6 67.8 0.87 53 50.9 51.9 0.96 Rural 51.1 44.6 48.1 0.87 34.3 34.9 34.5 1.02 Source: Ghana Demographic and Health Survey, 2009. Note: Secondary school includes both junior and senior high school. 284. Overall SHS access, quality and resources are lowest in the North. As is the case in many other sub-Saharan countries, SHS enrollment in Ghana is both low and highly inequitable. In the Central region, roughly half of all youth are enrolled in senior high school (51.1%), but only about one in every five youth have access in the Upper East region (20.3%). As most Senior High schools are equipped with boarding facilities and it is common to be enrolled in school outside the region of residence, GER and GAR are therefore imperfect indicators of how well the population from a region is served by the schools within that region. As a vital actor, the private sector helps to close the national and inter-regional gap in enrollment, extending service delivery disproportionately to harder-to-reach areas. At the SHS level, only 26.6 percent of private schools are in urban areas and the rest, 73.4 percent, serve the rural population. Public high schools are more evenly divided with 48.6 percent in urban locations. Not surprisingly the distribution of SH schools shows low school numbers in the three Northern regions. In Greater Accra, the relatively low number of public SHS schools are compensated by a high number of private ones. 114 Figure 4-15: Regional distribution of S.H. Schools by type, 2008 Regional Distribution of SH Schools 100 90 80 70 60 Public 50 40 Private 30 20 10 0 Brong Ahafo Greater Northern Volta Eastern Central Ashanti Upper East Western Upper West Accra Source: EMIS Report 2008/09. 285. Across the country, the education sector is characterized by considerable variations in SHS learning results and in the percentage of SHS trained teachers. Viewed nationally, the Greater Accra region had the highest average English pass rate of 92.9 percent on WASSCE assessments and, at the same time, had the highest share of trained SHS teachers, i.e. 94.1 percent, with the vast majority of teachers in the region being employed in the public sector. The direct impact of such essential inputs on achievement outcomes also tends to be consistent at the other end of the performance spectrum. The Northern region, classified as having the most severe difficulties in attracting trained public teachers in senior high school (only reaching 75.2 percent), had the lowest pass rates on WASSCE exams in both English (71.6%) and Math (37.7%) in 2008/09. As to basic pedagogy infrastructure, existing disparities across Ghana are not trivial, particularly between the Central region, where nearly every senior high school student has their own desk (.94) and seat (.92), and the Upper West region, where only about half of SHS students are taught in classrooms with their own desk (.46) and seat (.47). Targeting Policies, Demand Side Interventions Deprived District Policy 286. As has been consistently shown, there are significant geographical disparities in Ghana, mostly in line with other social and economic factors. Since 1999, the Government of Ghana has classified roughly one third of the districts (originally it was 40 and gradually it increased to 68 as the district mapping was redrawn) as deprived based on various education outcome and resource indicators including GER in primary, gender parity, seats and core textbooks per pupil, share of schools needing major repairs, BECE pass rates both in English and math, per student budget in primary, PTR in primary and the share of qualified primary teachers. The plan was that by allocating 115 proportionately more funds to these disadvantaged districts, the Government would effectively address education deprivation. However, it is important to emphasize that the identification of deprived districts is based strictly on education indicators and not on indicators external to the sector i.e. poverty or economic production. This caused some level of overlap but also some incongruence between “educationally� deprived and, for instance “socially� deprived districts as it is illustrated below. Map 4-9: Government’s Identification of Deprived Districts and District Level Poverty Count Source: EMIS on Deprived Districts, Wodon 2009 on poverty count based on GLSS 2005/06 and CWIQ 2003. 287. Several factors co-determine geographical deprivation. Most of these districts are remote rural areas with limited economic opportunities and limited access to social services. This leads to persistent poverty, which in turn diminishes demand for education. Low interest, low-level access and low level supply are creating a self-perpetuating general deprivation. The rationale for Government intervention has been that providing extra support to improve supply will increase demand, and better education helps break the cycle of poverty in which families and communities are trapped. 288. Have Government policies been successful in dealing with geographical education deprivation? At first look the Government policies appear to have achieved some success in some districts. In 10 years some 12 districts out of the then 53 districts designated as deprived, have managed to improve their educational indicators, and move out of their deprived status. However, it remains unclear whether this was the result of improved supply. In the meantime, 15 districts that had originally been classified as non- deprived fell into the deprived category. 116 Table 4-3: Existing Deprived Districts and Those Calculated Using 2008/09 Data Deprived Districts using 2008-09 data and old index Not Deprived Total Deprived deprived Districts Not 70 15 85 using existing deprived index Deprived 12 41 53 Total 82 56 138 Source: Wodon 2009. 289. More importantly, the logic of Government intervention has been to target additional resources to the deprived districts. Two important findings indicate that the Government has not been successful in this effort. Most of the deprived districts are in the three Northern regions and actual per student expenditures are below standard and the median per student expenditures have remained below the median per student expenditures of non-deprived districts. Only recently has the Per Child Recurrent Expenditure (PCE) in deprived districts increased in the primary and JH cycles, but even now they remain lower than in non deprived districts. Figure 4-16: Visualizing the disparities: Deprived districts 2006-2008 290. Similarly, the key input that assures both enrollment and better learning outcomes is the trained teacher and here, too, per teacher salaries, a good proxy for qualification and outcome, are between 10 and 40 percent lower in the three regions than in better funded regions like Greater Accra or Ashanti. 291. Overall, the Government has not been successful in effectively addressing the geographical disparities; in fact per student expenditures and median salaries show that the social and economic disparities may have been further exacerbated by substandard education expenditures and by the supply of untrained and inexperienced teachers. 292. Evidence in this chapter points to the fact that the central obstacle confronting Ghana in its efforts to attain education for all is how it can open and maintain attendance 117 among very poor households in remote, isolated and deprived communities. Educational deprivation appears to be principally caused by low income, vulnerability, rural locality and poverty. However, the single most pressing characteristic of the Ghanaian context, impacting nearly every education indicator, seems to be the geographic divide, i.e. between the South and the three Northern regions. The central challenge here is to reduce the indirect and opportunity costs to families, of sending their children to school. Struggling to incorporate the remaining pockets of the hard-to-reach population into the school system, policies that focus on both supply and demand driven aspects may be more appropriate. Since rural and impoverished children are at the centre of low school participation, it is crucial to better target them. To this effect, increasing the capitation grant funding and closing in on its target population could increase educational demand. Capitation Grant Policy 293. The Capitation Grant (CG) was introduced in 2004 as a result of the 2003 Education Strategic Plan in response to the elimination of school fees. The CG was initially financed by Education Sector Project (World Bank) and targeted to the 40, then 53 Deprived Districts. The initial allocations were 3000 (old) Cedis (equivalent with 3Ghc) per student at primary and JHS levels. In 2005, the CG was scaled up to the whole country and was paid from HIPC funds. The law eliminating the fees and introducing the CG nationwide had major impact, especially on primary admission, which jumped that year. Initially, the impact on the Deprived Districts and, in general, low enrollment areas were more significant than in other areas, as the economic burdens of the school fees were especially hard on the poor. 294. By 2008, however, the initial enthusiasm subsided, growth in enrollment and in gender parity leveled out and by this time, dropout and even some learning outcomes had started to decline as the first post-CG generations started to reach 3rd and then 4th grade. This was the result of the large influx of poor students whose parents started to feel negative economic consequences, and overage students, who have had some difficulty staying in school. There has been some slowdown in initial admissions and inflation has also started to erode the value of the CG. In 2008, the Government increased the CG to 4.5Ghc. Demand-side Interventions: Programs Targeting the Poor 295. LEAP as best targeted program in Ghana. A benefit incidence conducted by the World Bank reveals that among a number of subsidies and programs in various sectors, LEAP (Livelihood Empowerment Against Poverty) is the best targeted social program in Ghana. 57.5 percent of the benefits of LEAP reach the poor (Wodon 2009). LEAP is a program carried out by MESW to provide a Conditional Cash Transfer (CCT) to households living in extreme poverty that is contingent not only on school enrollment, but also regular health visits. In 2009, LEAP allocated about 7.5 million GHC, resulting in a transfer per household of 8-12 GHC (about US$5-8 per month). Ghana School Feeding Program reaching 21.3% of the poor. 296. It is estimated that the Ghana School Feeding Program (GSFP), providing school lunches to students in participating public primary schools, does not have as large of a 118 pro-poor impact as LEAP and its targeting mechanism may be in need of remodeling. Only 21.3 percent of the poor are benefiting from the program, in spite of the GSFP using a wide range of prerequisites to target recipient schools, among them road, telecommunication and potable water inaccessibility, unavailability of a power grid, health facilities located outside a 15km radius, areas affected by conflict or prone to floods, as well as low enrollment rates and poor school infrastructure. According to a World Bank study, if the GSFP that aims to reach the poor improved its pro-poor targeting along the scheme designed for the free school uniforms, the antipoverty impact could be more than twice as high (Wodon 2009). 297. Administrative data on the benefits provided to 170 districts were transformed in order to obtain data for 138 districts for which we have poverty data. The total 2009 outlays to the districts indicated in the administrative data available to us amount to 26.3 million GH ¢. This is below the level of spending for 2008, at 33.0 million GH ¢. The poor, who represent 28.5 percent of the population, benefit from only slightly more than a fifth of the school lunches. The profile by quintile shows that the targeting performance for the top quintiles are twice as high as for those for the bottom quintiles (Table 4-4). Table 4-4: Targeting performance of school lunches using district level allocations Share of population Total Program benefiting Q1 Q2 Q3 Q4 Q5 Poor Expenditure School lunch expenditure 2.77 3,449.6 4,618.7 5,235.1 6,182.0 6,789.4 5,597.1 26,274.9 Share of benefits 13.1% 17.6% 19.9% 23.5% 25.8% 21.3% Source: Wodon and Joseph 2010. 298. Targeting Performance of School Lunches based on Child Needs Index as per the Core Welfare Indicators Questionaire (CWIQ). Then, instead of applying district level poverty rates, we estimated an index-based household level child needs using CWIQ. The following variables are included to estimate the household level index of child needs: proportion of children of the primary school age who are not enrolled, share of children below age five who are stunted, share of children who are underweight, ability of the household to cope with shocks and manage future shocks, household level food insecurity, and household calorie intake. When examined the targeting performance of school lunch expenditure in 2008 at the district level based on this index, we found that the targeting performance was slightly better than before. Table 4-5: Share of the benefits that go to each quintile using CWIQ and admin. data and education needs in CWIQ Program Q1 Q2 Q3 Q4 Q5 Poor Total School lunch expenditure 5,220.50 4,875.40 4,980.30 5,829.00 5,369.70 5,597.10 26,274.90 Share of benefits 19.87% 18.56% 18.95% 22.18% 20.44% 21.30% 100% Source: Wodon and Joseph 2010. 299. It should be pointed out that on a needs basis, the targeting performance of school lunches as estimated using district level data would be even weaker than suggested in Table 4-6. This is because the poorest quintiles of the population tend to have more 119 children, and among the poor, children are more likely to go to public schools than private schools. No adjustments have been made here for differences in needs. 300. Estimating deprivation using school level data. Using school level data from EMIS, a school level deprivation index is calculated using the following variables: text book to pupil ratio, whether the school has classrooms that need major repairs, percentage of qualified teachers, student teacher ratio, pass rate in primary, and girls’ enrollment rate. Table 4-6: Share of School Lunch Expenditure by Deprivation Index 1 National Q1 Q2 Q3 Q4 Q5 Poor Total enrollment 2,606,512 344,942 436,174 512,147 577,893 735,356 522736 Total students receiving school meals 372,372 26,484 48,832 63,116 74,583 159,357 45791 Total students receiving capitation grants 2,273,722 289,647 377,870 455,176 508,159 642,870 443828 Percentage of students receiving school meals 100 7.11 13.11 16.95 20.03 42.80 12.30 Percentage of students receiving capitation grants 100 12.74 16.62 20.02 22.35 28.27 19.52 Source: Wodon and Joseph 2010. Conclusion: Equity and Disparity in Education Services 301. If access is inequitable, it diminishes the performance and compromises the social and economic benefits. That is why the Ghanaian government has been committed to equitable access. A number of policies support this goal, including, among others, the elimination of fees, the provision of capitation grant, textbooks, school feeding and the delineation of deprived districts. But, by the same token, inequitable services, disparities in access and in educational inputs could not only compromise the effect of these policies but also result in greater social inequity and fragmentation. 302. There are three levels of potential and desirable policy actions in the context of disparities: a) Avoiding negative discrimination: At a minimum, the allocation of resources should not discriminate along social, cultural or geographical grounds. If x amount is available for education per student, the Government needs to assure that x amount is allocated so that all students see the benefit of the inputs equal to this amount. b) Establishing specific policy incentives to improve supply of services: At a higher level, the Government resources should be allocated in a way that it addresses specific difficulties in education service delivery. If, for instance, enrollment and attendance is more challenging in some areas, typically in the rural and poor Northern regions, the Government conceives policies that address this specific educational problem. For instance, it is important to top up teacher salaries for those who are willing to teach in remote areas; accommodations or subsidized transportation can be provided to these teachers. 120 c) Creating positive discrimination to stimulate demand: At a highest level, the Government may develop policies that address underlying social and economic problems by stimulating demand and thus access to education. These policies may include school feeding, cash transfers or any type of extra provisions that relieve the burden on families for sending children to school. It is critical to the success of these measures how they are targeted and monitored. B. Efficiency 303. In this section, we seek answers to questions on the degree of sectoral efficiency and in particular how efficiently physical, human, and financial resources are allocated throughout the country to maximize the benefits. This gives rise to the question: to what extent are allocations driven by Government policy or, alternatively, by spontaneous, unintended factors? School sizes 304. Basic education is largely delivered in small and medium sized schools. Enrollment in the average KG is 90, in the average Primary school 225, and in the average JHS 138. A key challenge is that most small schools in remote rural and poor areas where the school size often affords only one teacher who may or may not be qualified. Affording one qualified teacher to each school would lower pupil teacher ratios and inflate salary expenditures. Figure 4-17 illustrates the effect of increased enrollment in junior high schools, a phenomenon that affected disproportionately very small schools, the proportion of which decreased significantly between 2005/06 and 2008/09., thus increasing slightly the efficiency of JHS in terms of school size. Figure 4-17: Effect of increased enrollment in junior high schools 121 305. Senior High Schools tend to be large -- half of SH schools had more than 850 students in 2008/09 -- and are becoming larger. The distribution of SHS enrollment has changed significantly in the past 4 years, shifting from a high concentration of medium schools to a high concentration of medium to large schools. Figure 4-18: Senior High Schools Enrollment distribution 306. 94% of Kindergartens have three or fewer classrooms, almost 50% have just two and the large majority of rural schools have only one KG class. 88% of Primary schools have six or fewer classrooms and 47% have exactly six, a lot of them built based on central construction standards.. Similarly, 94% of Junior high schools have six or fewer classrooms and 63% only have three. Only 20, 40 and 46% of the variation in number of classrooms between schools can be explained by school level enrollment in KG, Primary and JHS respectively. This is not surprising given the complexity in funding of school construction. 307. Class size in basic education is relatively large, especially in KG, but the size has not increased significantly, indicating that new classrooms were constructed in line with 122 the increased enrollments. In four years efficiency13 improved in primary schools but declined in KG and JHS levels. At SHS level, the average classes are significantly bigger, indicating overcrowding and the need to build new SHS facilities. Table 4-7: Efficiency in classroom distribution in the Basic Cycle, 2004/05 - 2008/09 2005 2006 2007 2009 Kindergarten Median number of pupils per classroom 44.67 43.00 44.25 43.50 Efficiency in distribution 0.213 0.284 0.300 0.204 N 5341 6943 7496 7974 Primary School Median number of pupils per classroom 34.83 34.42 35.17 35.17 Efficiency in distribution 0.32 0.45 0.42 0.40 N 9335 10222 10489 10428 Junior High School Median number of pupils per classroom 31.33 31.33 33.33 33.83 Efficiency in distribution 0.52 0.57 0.53 0.46 N 5101 5642 5790 5957 Senior High Schools Pupil – Classroom 44.6 44.5 43.4 46.5 Efficiency 0.65 0.71 0.73 0.69 N 338 394 398 403 Source: Author's calculation from EMIS 2004/05 to 2008/09 Census data. Unadjusted. Note: Schools with Pupil - Classroom ratio >200 excluded N indicates the number of observations used to compute the coefficients Note: Data is not adjusted for non responses Figure 4-19: Pupils per Classroom and Teacher by school size in Primary, JHS, 2008 60 45 40 50 35 40 30 pupils per classroom 25 pupils per classroom 30 pupils per teacher 20 pupils per teacher 20 15 10 10 5 0 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 1 2 3 4 5 6 7 8 9 10 11 12 13 number of classrooms in the school number of classrooms in the school Source: Author's calculation from EMIS 2008/09 Census data. Unadjusted. Textbooks 308. Between 2005/06 and 2008/09 the average number of textbooks per pupil in primary schools declined more in deprived districts than in non-deprived. There appears to be no such trend in JH School. Too few schools however report correctly on the number of textbooks in deprived district to calculate sensible estimates of the efficiency in textbooks distribution. 13 Measured by the extent changes in the number of classrooms followed the changes in the number of students. 123 Figure 4-20: Efficiency in core textbooks distribution in Primary and JH, 2004/05 - 2008/09 2005 2006 2007 2008 Primary School Average number of Core textbooks per child 1.54 1.64 1.44 1.45 Efficiency in distribution 0.69 0.68 0.72 0.63 N 6193 6250 6446 6598 Junior High School Average number of Core textbooks per child 2.06 2.16 1.96 1.93 Efficiency in distribution 0.60 0.62 0.56 0.47 N 6193 6250 6446 6598 Senior High School Core Textbooks - Pupil 1.15 3.21 3.28 3.11 Efficiency 0.45 0.45 0.44 0.41 N 358 422 425 432 Source: Calculation based on EMIS 2004/05-2008/09. Unadjusted N indicates the number of observations used to compute the coefficients Note: Data is not adjusted for non responses 309. Core subjects in SHS are Mathematics, English, Sciences and Social studies. The number of textbooks in schools increased threefold between 2005/06 and 2008/09 passing from 1 to 3 per student. While this is a significant improvement, it means that the average student is missing 1 out of the 4 core textbooks. Furthermore, averages hide great disparities in textbook distribution. Only about 45% of the number of textbooks encountered in schools can be explained by enrollment (while a large number of students remain with no textbooks at all). 310. A large number of textbooks were procured and distributed in 2009. These textbooks reached schools after the 2008/09 census had taken place and are therefore not captured in this analysis. If efficiently distributed these textbooks would be sufficient to bring the Pupil-Core textbook ratio to 1: 4 at the senior high level. 124 Figure 4-21: Students - Core textbooks scatter plot in the Secondary Cycle 2005/06 - 2008/09 Source: Author's calculation from EMIS 2004/-5 to 2008/09 Census data. Unadjusted Efficiency in teacher deployment 311. National aggregates of PTRs hide significant disparities between schools, districts and regions. Both in the Primary and JH cycles the highest PTR is more than 30% higher than the lowest. Higher PTRs are found in the North and in the Greater Accra region, in the latter many due to migration patterns. Table 4-8: Primary Pupil Teacher Ratio at the Regional Level 2003/04 2004/05 2005/06 2006/0714 2007/08 2008/09 Ashanti 32.5 32.9 35.9 33 33 32 Brong Ahafo 30 30.4 34.5 30 32 32 Central 34 35.7 39.3 36 37 36 Eastern 29.3 29.7 33.3 30 30 30 Greater Accra 36.8 36.6 39.6 39 40 40 Northern 38.6 40.1 40.5 34 33 33 Upper East 59 57.4 53.3 43 46 48 Upper West 46 49 45.6 34 39 42 Volta 30 30.5 34.1 29 29 30 Western 34.1 36.1 38.9 34 36 37 National 34 34.8 37.8 33 34 34 Source: EMIS. 312. How efficiently are teachers deployed? Efficiency in teacher deployment generally decreased between 2005/06 and 2008/09. In 2008/09 only 29, 46 and 55% of teacher presence could be explained by enrollment in Kindergarten, Primary and Junior 14 It is unclear what accounts for the big drop (i.e. improvement) in 2006/07, in the whole country? 125 High School, respectively. In the primary subsector, the number of teachers whose presence in school was related to the number of pupils was 46% in 2008/09, down from 55% in 2005/06. In other words “randomness� in teacher deployment increased from 45% to 54% between 2005/06 and 2008/09. Teachers are even less efficiently distributed in deprived districts, where in 2008/09 “randomness� reached 62%. 313. Efficiency in teacher deployment has been decreasing in the Junior High subsector, too. Whilst 63% of teacher presence was explained by enrollment in 2005/06, only 55% of teacher presence could be explained by enrollment in 2008/09. Efficiency in teacher presence in deprived districts fell by a further 10 percentage points in the last year. In effect, this means that for the last four years, efficiency of teacher deployment was always lower at all levels in deprived than in non-deprived schools. Evidently, the Government efforts to assure better education services to deprived districts have not been effective. Figure 4-22: Efficiency in teacher distribution in Primary and JH by deprivation status, 2004/05-2008/09 Primary School - Effciency in Teacher Junior High School - Effciency in Deployment Teacher Deployment 0.70 0.80 0.60 0.50 0.60 0.40 0.40 0.30 0.20 0.20 0.10 0.00 0.00 2005 2006 2007 2009 2005 2006 2007 2009 Deprived Non Deprived National Deprived Non Deprived National Source: Calculation from EMIS 2004/-5 to 2008/09 Census data. Unadjusted. Figure 4-23 below shows the variations between PTEs by schools at primary level, showing how widely teacher allocations scattered around a linear regression line that is set at the national average of about 35 pupil teacher ratio. 126 Figure 4-23: Students - Teachers scatter plot in Primary, 2008/09 314. As is the case in primary schools, the district-level distribution of junior high school teachers appears equitable, with allocations to districts generally being efficient while allocations to schools within districts being inefficient. The figures below illustrate this phenomenon at the JHS level. Figure 4-24: Teacher Allocation in Junior High Schools in 2008/09, by District (left) and by Schools (right) 100000 y = 16.899x + 79.108 R² = 0.9786 Accra district (Greater Accra) 80000 Number of Pupils Kumasi district (Ashanti) 60000 40000 20000 0 0 1000 2000 3000 4000 5000 6000 Number of Teachers Source :EMIS 127 Figure 4-25: Pupils - Teachers- Classrooms in SH, 2005/06 - 2008/09 315. In contrast to primary and JHS schools, teacher distribution in SH schools is relatively efficient. Close to 80% of the distribution of teachers across schools can be explained by variations in enrollment. Efficiency has declined slightly since academic year 2005/06 due to increased enrollment against unchanged number of teachers in small schools. Figure 4-26: Students - Teachers scatter plot in the secondary cycle 2005/06 - 2008/09 Source: Author's calculation from EMIS 2004/-5 to 2008/09 Census data. Unadjusted Efficiency in Allocating Expenditures 316. Child expenditures are plotted in Figure 4-27, where every line is a district. The vertical red line cutting across the others demarcates the bottom third of the per capita expenditure. Any district on the left of the red line falls within the bottom third. It is important to note how (i) the three regions in the North fall almost entirely below the red line, and (ii), in all other regions there is a wide variation in district PCE. 128 Figure 4-27: Visualizing the disparities – The districts in 2008 317. Figure 4-28 helps in summarizing these two important features of the raw data. Noticeably higher per capita spending tends to go with higher disparities: in regions where the median PCE is higher there still exist districts with very low spending. Figure 4-28: Visualizing the disparities – Between and within regions in 2008 318. In blue, Figure 4-28 plots the median district in every region15. It is important to note the difference between the regions and in particular how much lower the median PCE is in the three regions in the North. In Red, the Figure plots the Inter-Quartile Range (IQR), which is a measure of disparity between the districts in a region16. 319. In order to give a dimension of the impact of low per child expenditures, Table 4-9 reports the percentage of children affected by sub-standard PCE, i.e. funding that falls within the bottom third of the per capita expenditures in the country. Regions where more than 30% of the children suffer from sub-standard spending are highlighted. A very large amount of under spending is recorded in the three regions in the North. More than 60% of children in the Northern region attend primary school in districts where the PCE is below 15 The median district is the district below which lie 50% of the districts in the region and is therefore a useful representation of the level of per capita expenditure in the region. 16 Specifically the IQR measures the difference between the top and the bottom 25% of the districts. 129 the bottom third of the nation. The proportion of Junior High School students suffering from under spending in the Northern region is even higher, affecting 83% of the total number of students in the region. 320. In the Senior High cycle, Greater Accra Northern and Upper East are the only regions with less the 30% of the districts falling in the bottom third of the nation. This is probably due to the low levels of enrollment in such regions, where NER is below 15%. Table 4-9: Percentage of children in districts with sub-standard PCE in 2008 KG Primary JHS SHS ASHANTI 28% 29% 23% 58% BRONG AHAFO 0% 35% 45% 49% EASTERN 0% 15% 0% 31% GREATER ACCRA 17% 7% 0% 18% NORTHERN 49% 65% 83% 26% UPPER EAST 59% 66% 63% 30% UPPER WEST 93% 45% 59% 53% VOLTA 15% 0% 12% 36% WESTERN 24% 18% 5% 40% 321. The conclusion is that a key source of deprivation for the deprived districts is the allocation of public expenditures itself. These expenditures, instead of compensating for the deprivation, exacerbate them by allocating fewer resources per child to the regions where the majority of deprived districts are located. 322. Having provided a cross sectional snapshot of the disparities across regions in PCE, this subsection will present the evolution of PCE from 2006 to 2008. The figures reported present large year-to-year fluctuations. It has to be kept in mind that as one district crosses the established threshold, all children is such district are counted above the threshold. It is therefore normal to see large fluctuations in the percentages. 323. Per capita expenditure in KG are very low: the national average is GHc47, with variations ranging from GHc5 to GHc220. Table 4-10: Percentage of children in districts with substandard PCE at KG level 2006 2007 2008 ASHANTI 13% 40% 28% BRONG AHAFO 0% 16% 0% EASTERN 10% 5% 0% GREATER ACCRA 39% 0% 17% NORTHERN 38% 63% 49% UPPER EAST 73% 63% 59% UPPER WEST N/A 82% 93% VOLTA 15% 20% 15% WESTERN 21% 19% 24% 130 324. The percentage of children affected by substandard funding in primary school, despite recent improvements, remains very high in the North of the country. More than 60% of the students in the Northern and Upper East regions attend school in district with substandard PCE. A success story worth pointing out is the Volta region, where no district fell in the bottom third in 2008, although 36% of children had substandard funding in 2006. Table 4-11: Percentage of children in districts with substandard PCE at Primary level 2006 2007 2008 ASHANTI 37% 65% 29% BRONG AHAFO 30% 27% 35% EASTERN 17% 16% 15% GREATER ACCRA 35% 9% 7% NORTHERN 76% 91% 65% UPPER EAST 100% 100% 66% UPPER WEST N/A 78% 45% VOLTA 36% 21% 0% WESTERN 62% 64% 18% 325. The picture in Junior high school, reported in Table 4-12 shows little change in the dynamics of disparities. . Low PCE affects large number of children in the North of the country with no hint of improvement. The proportion of children studying under substandard PCE saw a substantial increase in the Northern region between 2006 and 2008. A success story worth pointing out is the Western region, where from more than 50% in 2006, the number of students affected by under spending in 2008 fell to fall%. Those districts who appeared to have been underfunded in 2006 remained so later. Table 4-12: Percentage of children in districts with substandard PCE at JHS level 2006 2007 2008 ASHANTI 26% 89%17 23% BRONG AHAFO 34% 63% 45% EASTERN 4% 20% 0% GREATER ACCRA 26% 0% 0% NORTHERN 55% 80% 83% UPPER EAST 66% 80% 63% UPPER WEST N/A 50% 59% VOLTA 15% 18% 12% WESTERN 53% 32% 5% 17 This outlier seems to be related to low data reliability 131 326. There has been a general improvement in funding of SHS over the years. Particularly so in the Greater Accra, Northern and Upper East regions where now less than 30% of students study in districts with substandard PCE. It is however important to note that in the rest of the country, where NERs are above 15%, PCEs remain low. A noticeable exception is the Upper West region, where NER is only 8% yet 53% of students are below the bottom third of per capita expenditure. In simpler words, 53% of the very few students in the Upper West region receive substandard PCE. Table 4-13: Percentage of children in districts with substandard PCE at SHS NER 2006 2007 2008 2008/09 ASHANTI 72% 95% 58% 23% BRONG AHAFO 59% 89% 49% 15% EASTERN 100% 44% 31% 26% GREATER ACCRA 100% 3% 18% 13% NORTHERN 100% 97% 26% 11% UPPER EAST 65% 59% 30% 8% UPPER WEST N/A 83% 53% 8% VOLTA 51% 71% 36% 16% WESTERN 100% 70% 40% 15% 327. Figure 4-29 depicts the PCE in every district classified according to their deprivation status. Every blue bar in the picture depicts a district, while the red line cutting across demarcates the bottom third. From the pictorial representation, it emerges that: 328. Deprived districts have on average lower PCEs than non deprived district. In 2008, deprived districts spent about GHc50 less on every primary student and GHc60 less on every JH student than non deprived districts did. There are two potential explanations to this finding: either the targeting is purely nominal and no substantial resources have been allocated to the deprived districts, or only donor funds have been allocated to the deprived districts crowding out government spending. 329. There are a number of deprived districts well above the minimum acceptable PCE as well as a number of non-deprived districts below it. In 2008 primary PCE was above the minimum standard in 44% of deprived districts and below standard in 19% of non- deprived. Similarly PCE on JH students was above standard in 50% of deprived districts and below standard in 23% of non-deprived. 132 Figure 4-29: Visualizing disparities - Deprived Districts in 2008 330. The secondary cycle presents a very different set of challenges. Per capita expenditure in this cycle is significantly higher than in the basic cycle, and the definition of minimum standard adopted in this chapter might be ill suited to such levels of PCE. Nevertheless, the analysis reveals strong regional disparities, with regions with the highest enrollment level struggling to secure sufficient funding for their students. A special case is presented by the Upper West region, where despite the low levels of enrollment funding remains low. 331. Geographic disparities. There is a degree of disparity in per unit expenditures across regions. Per unit expenditure18 in Greater Accra is almost 30% higher than the national level; in the Northern region it is 10% lower. Teachers in deprived districts receive 16% less then teachers in non-deprived districts as reported in Table 4-14. Table 4-14: Per unit expenditure by region and deprived, 2009 Annual Per unit % difference from expenditure the national average ASHANTI 3573 6% BRONG AHAFO 3216 -5% CENTRAL 3478 3% EASTERN 3486 3% GREATER ACCRA 4305 28% NORTHERN 3042 -10% UPPER EAST 3129 -7% UPPER WEST 3257 -4% VOLTA 3513 4% WESTERN 3143 -7% NATIONAL 3376 non deprived 3550 5% deprived 3009 -11% Source 2009 Payroll. 18 Government recurrent expenditures per student 133 Efficiency in schools: Absenteeism and Time-on-task 332. High teacher absenteeism is one of the key inefficiencies in the education sector. It is a major problem in rural schools; poor teacher morale and poor work environment. The average teacher absentee rate for Ghana, as a survey by the Center for Democratic Development (2008) indicates, was 27 percent. Teachers in Tunisia and Morocco, by comparison, miss 11.6 days and 13.4 days a year, respectively, The common reasons for the high absentee rate in Ghana are (i) sickness, visits to health facilities, (ii) collection of salary at a bank some distance away, (iii) frequent funeral attendance, (iv) religious practices (for instance, Friday prayers among Muslim teachers), and (v) rural teachers engaging in activities related to farming (CDD 2008; World Bank 2004a; Abadzi 2007:26). But there are underlying reasons contributing to high absenteeism: (i) lack of supervision, (ii) long distances to school (schools far from bus/lorry stations as well as from healthcare facilities), especially in rural areas, (iii) scare resources, such as toilets and potable water, at schools. 333. Another study that examined 120 schools in rural Ghana reported that on average teachers attend schools four days a week, implying that 5.5 instructional hours are lost each week. The day that researchers visited the schools, nearly twenty percent of the teachers were absent (EARC 2003). In 2003, an evaluation conducted by the World Bank in Ghana established that in basic education almost 13 percent of the teacher workforce was absent in the month prior to the survey due to reasons other than sickness, a stark deterioration from about 4 percent reported in 1988 (World Bank 2004a). Survey results from the same study indicate that the incidence of absenteeism is worsening, with 15 percent of schools faced with the problem in 1988 rising to 39 percent by 2003 (Table 4-15). . In other words, 85 percent of schools were not affected by teacher absenteeism in 1988, but this share had dropped to 61 percent by 2003, with 13 percent of schools having more than one third of teachers absent in the last month for reasons other than sickness. As data in table 4-15 confirm, the phenomenon of teacher absenteeism is endemic, but affects rural schools more than urban, and public schools more than private. This is explained in part by rural teachers having to travel further distances to schools as well as to collect pay, and being occupied in seasonal farm activities more so than their urban counterparts (World Bank 2004a). The data also suggest that absenteeism is more common in schools suffering from poor infrastructure, located in poorer districts, and providing service delivery to children from households with lower levels of income (UNESCO 2008). 334. Since approximately 28.2 percent of the government’s entire budget is allotted to teacher salaries and teachers are absent between 20-27 percent of the time, as various studies suggest, one can conclude that roughly 5.6-7.6 percent of the government’s total budget is lost to teacher absenteeism. 134 Table 4-15: Share of schools affected by teacher absenteeism in 1988 and 2003, (%) 1988 2003 Disaggregated data, 2003 Frequency Urban Rural Public Private None .85 .61 .61 .61 .56 .80 Up to a third of schools .11 .26 .27 .23 .29 .14 Between one to two thirds of schools .3 .9 .9 .10 .11 .4 More than two thirds of schools .1 .4 .2 .7 .5 .1 Number of observations 518 706 451 255 568 138 Source: Based on data from World Bank 2004a. 335. A USAID study (2003) that randomly sampled 720 classrooms in 120 mainly rural Ghanaian schools found that 22.3 percent of teachers were absent on the day the survey was conducted. An additional approximately 25 percent of teachers – 157 out of the total 668 teachers – were present but engaged in other activities not involving teaching. Nine percent were present but idle. Over a quarter of teachers (28.8 percent) who were present at school had arrived late, on average by about half an hour, and 1.5 percent left early. 336. While teachers holding a university diploma in basic education were least absent, those holding a certified A level were absent almost half the time (47 percent). Also, teachers aged 50 and above tend to be absent less frequently. Common reasons explaining these poor teacher attendance patterns are: study leave with pay, attendance at workshops, preparation of events, teacher exodus to universities. 337. Improvements to reduce the problem of absenteeism include providing on-or- near-site salary collection, especially as 28 percent of teachers reported not being able to collect their salaries on time (World Bank 2004a). Also helpful would be ensuring access to local health clinics and basic medical services. The CDD study (2008), shows that enhancing effective supervision as well as strengthening evaluation have proven effective measures in considerably increasing teacher attendance, in particular, in public schools. The Brookings Institution (2007) underlines that the degree of parental involvement can greatly influence the absence patterns of teachers at a minimal financial cost. One measure would be to enlist PTAs in helping with this task. In principle, nearly all schools have a PTA, i.e. 99 percent of public and 95 of private basic schools (World Bank 2004a). Studies have also shown that rigorous and frequent school visits by circuit supervisors, though uncommon, play an essential management role (see chapter on management). Improving the teacher supervision structures and offering teachers incentives tend to directly decrease teacher absenteeism and lateness (Akyeampong et al. 2007). Where favorable working conditions and low pupil teacher ratios in schools exist, teachers are more likely to have regular attendance. 338. Various studies indicate that time on task is estimated to be between 33 to 39 percent. In Ghana, the phenomenon of low attendance rates is compounded by the problem of low instructional time in school, i.e. low time on task, which appears to be widespread. A simple observation to school activities can provide a good picture of the causes of low time on task. Frequent breaks, when teachers are separated from students, slow transition to instruction, down time spent disciplining students, collecting home- work, etc, all create an illusion of children learning while in school. The following table 135 is based on classroom observations on time use in Ghanaian schools by grade in primary schools. Table 4-16: Ghana’s Average Percentages of Time Use, by Primary Grade Grade 1 2 3 4 5 6 Interactive 71.1 58.6 50 57.5 50.2 62.4 Reading 14.4 10.4 8.8 16.3 4.1 6.7 Instruction 16.7 18.8 16.3 11.3 17 22.4 Discussion 20.6 21.4 22.5 26.3 24.2 27.6 Practice Drill 16.7 7.3 2.5 3.8 4.1 5.2 Passive 4.4 10.5 8.8 17.5 14.3 9.6 Seatwork 3.9 7.4 2.5 11.3 11.6 7.1 Copying 0.6 3.1 6.3 6.3 2.8 2.6 Total Engaged Time 75.5 71.1 58.8 75 64.5 72 Management 23.3 29 38.8 25 33.1 26.1 Student Off Task 19.7 24.7 26.5 21.3 23.8 16.6 Source: Abadzi 2007. 339. An alternative means of measuring instruction time spend on learning activities is to ask teachers themselves. A teacher questionnaire by the World Bank found that time- on-task reflects one-third of class time using a narrow definition and 72 percent based on a broader definition, although large deviations from the average exist. Regression results indicate that the inputs that are the strongest determinants of higher time-on-task are, in order of importance: visits from circuit supervisors, head teacher supervision, and teacher training (World Bank 2004a). Table 4-17: Time-on-task for classroom activities Mean Minimum Maximum Total class time (minutes) 51 25 120 Of which (percent) Preparing for class 21 0 68 Disciplining students 7 0 48 All students engaged copying etc 33 0 73 Dealing with student 1-to-1 13 0 68 Addressing whole class 26 0 80 Memo: time on task Broad definition 72 29 100 Narrow definition 33 0 73 Source: World Bank 2004a. 340. In addition to wastage during the instructional time, the length of instruction is also a factor. In an international comparison, only 76 days were devoted to learning tasks in Ghana of the 197 days officially available, whereas, by comparison, for Tunisian students a total of 148.1 days of the 190 available were dedicated to learning. In Morocco, 145 of 204 days in the school year were allocated to engaging students in learning. Consequently, time on task for students in both Tunisia (77.9%) and Morocco (71.1%) was nearly double that of the time their Ghanaian counterparts spent learning 136 (38.7%), as seen in Table 4-18. Sustaining internationally competitive learning and quality outcomes is not feasible in Ghana’s school system, when students spend merely 39 percent of their time learning. On international assessment tests, Tunisia for instance scores substantially higher than Ghana (TIMSS 2009). Figure 4-30: Expected and Actual Number of Days Students were Engaged in Learning 250 197194 190185 200195 204203 200 171 175 176 148 145 150 126 109 100 76 50 0 GHANA Tunisia Pernambuco (Brazil) Morocco Days in School Year Days after Closure No. Days Schools Operated Days Students Engaged in Learning Source: Based on Abadzi 2007. 341. Together, teacher absenteeism, poor ToT and short duration of school year can result the loss of as much as 50-60 percent of teaching time, clearly making this as a key constraint to learning. Table 4-18: Instructional Time Use in 4 Countries Ghana Tunisia Brazil* Morocco School Year (Days after Closure) 193.83 184.85 195.21 202.62 Teacher Absence (Days) 43.01 11.55 12.76 13.36 Teacher Delays 39.75 1.27 5.5 6.94 Early Class Dismissals 2.43 1.22 2.31 6.68 Time on Task (Days) 76.3 148.1 125.9 145.1 School Year, % of Time on Task 38.7% 77.9% 63.0% 71.1% Students Absence (Days) 9.04 3.35 7.82 4.30 Student Delays (Number of Times) 10.61 2.63 5.64 5.19 Note: *Data for Brazil is for the region Pernambuco. Source: Abadzi 2007:23. 342. Learning is compromised by high level of student truancy. In 1988, about 59 percent of primary and lower secondary students reported that they attended school on average four hours or less per day in the week prior to taking the household survey. The increase of school time in 2003, illustrated in Table 4-19, is primarily a result of the policy reform increasing the four-hour school day to five hours but it also reflects additional classes and longer hours in private schools (World Bank 2004a). 137 Table 4-19: Hours spent in primary and junior high school in past week, percent 1988 2003 public private 20 or less 58.7 16.3 23.1 20-30 39.1 55.3 30.1 More than 30 2.2 28.4 46.9 Total 100 100 100 No. of observations 3464 1158 286 Source: World Bank (2004a) based on GLSS data. C. Management Accountability and Social Accountability 343. The previous chapters have shown that there are impressive overall outcomes in education, but there are also great variations and disparities both in attainment and in learning. Moreover, resource allocations show similar disparities explaining to some extent the variations in outcomes. Overall, there are strong indications of inequity and inefficiency, which underline the need for strengthening management and information systems, better structuring the division of roles and responsibilities, and decentralizing some functions including resource management and service delivery. 344. Effective education management and administration has to assure that the outcomes, including enrollment, learning and participation are maximized, resource use is efficient and schools, teachers and administrators are held accountable. Accountability also assures that individuals, households and communities, even the poorest ones, increase their expectations and their participation in education. The objective of this section is to see how the flow of funds and management of resources, the division of roles and responsibilities, accountability measures, information systems, planning and policy-making all contribute to the effectiveness of education service delivery. Fragmented System of Flow of Funds 345. Education financing is fragmented among a number of sources and among an even larger number of flows of funds. The authority to allocate the key resources is divided among four agencies: The Ministry of Finance and Economic Planning (MOFEP) is responsible for setting the overall budget and determining Civil Service Agency (CSA) the remunerations; the Ghana Education Trust (GET) Fund is responsible for investments; Ghana Education Service (GES) is responsible for allocating recurrent expenditure and to set teacher numbers; the Ministry of Education (MOE) is responsible for allocating donor funds (totaling 6% of direct education expenditures, although donor contributions to the state budget bring donor support considerably higher) and proposing the annual budget to the Government. This fragmentation of budgetary responsibilities is the most fundamental reason why planned and executed budgets differ. 346. The large majority of the budget is channeled through the budget of the Ministry of Education. However, this budget only reflects teachers, new school construction and some inputs such as textbooks. A still significant portion of public resources benefiting the education sector is channeled through the budgets of other ministries, departments and agencies of the Government, among others, the Social Investment Fund and the Ghana Education Trust Fund. Education funding is also derived from the budgets of the 138 MoFEP and the Ministry of Local Government. Some of the resources from the “MP Fund,� allocated to Members of Parliament to promote development in their constituencies, also support school projects, scholarships and the like. 347. A significant proportion of (non-sectoral) Heavily Indebted Poor Country (HIPC) funds are used on education. HIPC funds are transferred directly to the Metropolitan, Municipal and District Assemblies (MMDAs) with involvement of Local Government (MoLG), which screens and monitors the funds. There are various non-sector-specific district support programs as well and several off-budget funding arrangements (sourced by NGO-, Community Based Organizations (CBO)- and donor-funded projects). 348. Size of the funding flows for basic education. Figure 4-31 below provides an overview of the various main funding streams to Basic Education. It should be noted that while Basic Education expenditures are directly classified under the budget lines for: Pre- school, Primary and Junior Secondary levels, they usually do not, but should include: a proportion of the administration budget for schools and regional services related to Basic Education (these expenditures are not broken down in the public accounts), off-budget donor support, e.g. from NGOs, the educational share of non-sectoral donor funds, funds for Basic Education funded from the HIPC initiative through MMDAs (MLGRD screening of proposal), and other sources such as some of the expenditure on special education, of which most of the expenditure are for Basic Education. Figure 4-31: Education expenditure by source, 2007 Expenditure by Source MDRI, 0.30% GoG EFA Catalytic Fund, 0.80% HIPC, 3.20% Donor GETFund, 12.90% IGF GETFund IGF, 8.70% HIPC EFA Catalytic Fund Donor, 6.00% MDRI GoG, 68.10% Source: RECOUP. 349. The systems vary greatly across the sources, ranging from: funding routed directly from the central ministry to the contractors (e.g. school books), GET-funds directly to the contractors (e.g. vehicles and equipment), direct payment from CAGD for services (teachers’ salaries), funds controlled through the MoFEP/MOE and passed through the DEOs (GOG funded services and investments), funds channeled through the MOES/GES 139 headquarters to the MMDAs, where the MMDAs function more or less as “post offices� (some construction funds and the HIPC funds), funds from the CAGD directly to the MMDAs Treasury (administration – new flow), funds from various sources directly to the MMDAs for local priorities (non-sectoral grants such as DACF, part of the HIPC- MMDA non-sectoral funds) and funds from the GES routed through the DEOs’ accounts to the schools’ accounts (capitation grants). 350. The most important funding flow is the GOG-funded salaries, followed by a number of significant fund flows from the GET-Fund, HIPC and DACF. Salaries are paid entirely from the GOG PE budget component, but the other budget components, particularly the investments, have several sources. 351. A good example of the complexity of financing is classroom construction, which is funded by a great number of sources including: GOG investment budget, Donor support (on-budget), GET-Fund, HIPC- MoE budget funds, HIPC-MMDA funds, DACF non-sectoral funds, various targeted education programs such as the Pilot Programmatic Scheme being implemented with WB funding support, and, non-sectoral district support programs e.g. the CIDA-supported DWAP. Provision of textbooks is another area of diverse financing with numerous funding sources (GOG, GET-Funds, HIPC and donor funds) and transfer schemes, although most of the textbooks are purchased directly by MOES. 352. The general picture of the funding of Basic Education shows a complex flow of funds and involvement of many authorities (GES, MoF, Regional Authorities and MMDAs). The existing system is one of fragmented planning, budgeting and accounting, which hinders the establishment of a comprehensive and consolidated overview of all inputs and the outputs measured against these inputs. The current funding system also reduces the possibilities for efficient programming and monitoring of the various schemes. The flow charts below (without all the details) provide a general overview of release path of the core GoG budgetary items (Item 1 – Personnel Emoluments, Item 2 – Administration, Item 3 – Services, Item 4 – Investment), and of the Capitation Grant. In all charts a bold arrow indicates a flow of fund and a dashed arrow indicates the issue of a warrant/authorization or just information flow. Figure 4-32: Item 1, GoG Core Budget- PE Personnel Emoluments MOF informs CAGD through instructions to the Bank of Ghana- BOG Teachers’ Bank Accounts Source: Steffensen 2006. 140 353. Upon instruction from MoF administration is transferred from the Controller and Accountant General Department (CAGD), directly to the MDA/MMDAs. Reliability and timeliness of funds is reported to have improved since 2006 when funds started being transferred directly and not through MoE/GES. 354. Teacher salaries are paid directly to commercial banks. Teachers in remote rural areas often experience delays in receiving their salaries through the banks. Newly qualified teachers also experience long delays – sometimes a whole year -- in getting their names onto the payroll system. Late payment of salaries is one of the major causes of teacher absenteeism and lateness especially at the end of the month when teachers take time off to chase salaries. Figure 4-33: Item 2, GoG Core Budget Administration. GOG - Administration (cost of running an office) MOF CAGD through instructions to BOG MDA/MMDA Treasury Accounts DEO Source: Steffensen 2006. 355. Services were, at the time of the study transferred from CAGD to the DEOs through the MoE/GES. Funds were reported to be very unpredictable and significantly delayed. After 2006, however, plans were in place to route Service funds directly through the DEOs, similarly to Administration funding. No study of funding flows has taken place since 2006 and it is therefore not possible to tell whether an improved path has effectively been established. Figure 4-34: Item 3, Core GoG Budget – Services GOG - Services MOF CAGD instructions to the MOES/GES HQ BOG Source:Steffensen 2006. Checks to the DEO Accounts 141 356. Investment funds are released from CADG to the DEOs, while informing the regional authorities of the process. Figure 4-35: Item 4, GoG Core Budget – Investment. GOG - Investments MOF CAGD through instructions to REC the BOG DEO Contractors Source: Steffensen 2006. 357. The Capitation Grant is sourced from the HIPC fund and the Social Impact Mitigation Levy (SIML) and released to the schools from the CAGD through the GES and DEO accounts. The PER reports funds to arrive quite reliably to the DEO accounts, but the transfer between the DEO and the school is poorly documented and generally unreliable. The results are cases of significant discrepancies in the actual per child grant received by the schools. Figure 4-36: Capitation Grant Capitation Grant CAGD through BOG GES HQ Account DEO Special Account School Accounts for Grants Source: Steffensen 2006. 358. Investment expenditures are separated from sector planning. Ghana Education Trust Fund is responsible for allocation of expenditures that totaled about 13 percent of the overall Government education budget for the fiscal year 2007. However, decisions are made independently by the GET Fund Board. It is unclear what proportion of the GET 142 Fund allocations go for what level or target which geographical location or how the decisions at the onset are linked with the priorities of the Education Sector Strategy. On the other hand, in theory, annual budgets (developed by the MOE based on GES estimates) would be expected to account for the recurrent implications of investments (including items 1 to 3). Chart 1: Statutory Funds Vehicle/Equipment to Schools GETFund e) Payment to the Contractors f) Source: Steffensen 2006. 359. Delays, Predictability and Budget Cuts. There are serious delays in several funding flows, particularly for the service component of the budget due to various systemic and capacity related reasons. There is inadequate capacity at the MMDA level to account and report in a timely manner on the use of funds. Moreover, these problems are worsened by the multiple funding flows, planning systems, accounting procedures, reporting systems, auditing requirements and bank accounts. 360. With the exception of salaries, fund flow is complex and often leads to a fragmented and unpredictable planning and budgeting in the MMDAs. This has a spill- over effect on service delivery at institutions/schools, which have to face lack of funding for operations and maintenance particularly at the beginning of the financial year. Capitation grants are supposed to address some of these problems and have improved the funding for operations of most of the schools. Roles and Responsibilities in Planning and in Service Delivery 361. The Ministry of Education (MOE) is the Government ministry responsible for policy and strategy for education in Ghana. The Ghana Education Service (GES) is the body responsible for overall implementation of that policy. Within GES, key departments include the departments of Basic and Secondary Education; Planning, Budget, Monitoring and Evaluation; Finance; Human Resources; and Teacher Education. GES is supported by a range of different agencies, including the West Africa Examinations Council and the Ghana Education Staff Development Institute. 362. The MOE organizes annually an Education Sector Review (ESR) and reports to interested stakeholders - including donors - on progress in and performance of the sector. In preparation for the ESR, MOE produces an annual performance report and this is discussed in detail at the review. In 2007,2008 and 2009, Regional Education Sector 143 Annual Reviews (RESAR) were conducted in all the ten regions in preparation for the annual review. 363. MOE represents the sector in strategic (Government and Development Partners-- DP) dialogue whereas GES is responsible for service delivery including deployment of teachers, allocation of textbooks, supervision of schools and teachers. Too many vertical levels blur responsibility; for example, in the context of leakages in the allocation of the capitation grant. 364. GES has no mandate to manage the payroll. It does not have effective tools to assure that teachers who are assigned to one school or even to one district take the assignments or remain there. Funds and resources to the local level flow through a variety of channels, limiting the capacity of agencies to pursue a coherent spending strategy. 365. In the meantime, GES has the mandate to assure that schools have the appropriate number of teachers. When teachers leave their posts for one reason or another (such as study leave or relocation) GES has the responsibility of substitution. As enrollments increased in the last decade, the difference between the planned and executed payroll has increased from about 10 percent to over 35 percent. 366. As the flow charts in the previous section illustrate, funding for basic education has a complex financial arrangement encompassing a number of authorities at different levels (GES, MoF, Regional Authorities and MMDAs). 367. Presently, the District Education Department system is de-concentrated rather than decentralized, inasmuch as District officials, including the Director are appointed by GES headquarters and report through regional offices to the center. The DEDs were created in 1995 when the Ghana Education Service (GES) reviewed its management structures at headquarters, regional and district levels in order to bring authority and responsibility for service closer to communities. Yet the predominant role that the central authorities continue to play limits the capacity of locally elected officials to oversee the sector and confront problems. 368. Under the new legislation District Education Directors are expected to report to the Metropolitan, Municipal or District Assemblies (MMDAs). This would strengthen local accountability; however, it is unclear how the standards for curriculum and teaching and other regulations set by the center would be enforced. 369. District Education Oversight Committees (DEOC), were introduced in 1993 to enhance broader stakeholders participation in the management, supervision and general governance of schools by the communities. These comprise stakeholders in education at the district level and represent the Ghana Education Service Council at the Community level. 370. In a school survey conducted by Akyeampong and Asante, headteachers were asked about the general behavior of their staff. Whilst rural school heads seem generally happy with the behavior of their teachers, urban heads complained that teachers were increasingly becoming rude and difficult to manage. They attributed this to general apathy, especially among older teachers. It may be that older teachers having experienced the frustrations in teaching for longer are becoming less concerned about maintaining their sense of professionalism. The school survey also showed that 144 dismissals rarely occur, which could mean that the majority of teachers are professionally well-behaved but given the previously described rate of absenteeism, more likely, the general laxities in the Ghanaian occupational culture makes enforcement of teacher regulations difficult. Most primary headteachers, as one union official pointed out, “lack the authority that goes with being a head�, and therefore are unable to initiate disciplinary measures against teachers. 371. School work, educational plans, classroom work and teaching are supposed to be overseen by District Circuit Supervisors and by the Headteachers. District Circuit Supervisors are responsible for overseeing school operations, education service delivery on behalf of the District, whereas the ultimate responsibility for service delivery rests with the head-teacher. However, supervision of teachers by circuit supervisors and head teachers is uneven, but has a significantly positive impact. About one in twenty teachers (5 percent) reported not having a head teacher regularly review their lesson plans. 44 percent of teachers do not have direct contact with circuit supervisors, even though the majority of schools are visited by them. 372. Many circuit supervisors don’t observe and assess teachers during instruction time, but rather visit schools to simply verify student and teacher attendance figures (World Bank 2004a). Although attendance is a fundamental prerequisite for learning, such an input-based approach to supervision will have little if any impact on achievement outcomes. According to the CDD (2008) study, less than 6 out of 10 teachers (57 percent) said circuit supervisors visit their schools at least once a month. Over time, visits from circuit supervisors have become slightly more frequent: when the number of visits averaged about six per year in 1988, this number increased to about nine per year in primary and slightly lower in junior secondary by 2003 (World Bank 2004a). An analysis of EMIS data indicates that in 2008 circuit supervisor visited schools more than once a year in only 77 percent of schools. Table 4-20: Frequency of head-teacher and circuit supervisor activities, 2003 (percent) Sits in Looks at a Looks at Discusses Discusses on class sample of lesson plans lesson plans career students’ work development Head- Circuit Head- Circuit Head- Circuit Head- Circuit Head- Circuit teacher Superv. teacher Superv. teacher Superv. teacher Superv. teacher Superv. Never 10.6 58.6 8.2 56.4 2.3 48.7 10.9 58.3 42.8 72.6 Less than weekly 28.1 38.4 39.4 41.6 2.4 49.2 43.4 40.3 48.7a 22.0a At least weekly 49.4 2.5 46.4 1.7 95.4 2.2 45.7 1.5 8.5a 5.5a Daily 11.9 0.4 6.0 0.2 .. .. .. .. .. .. Total 100 100 100 100 100 100 100 100 100 100 Schools sampled 3013 3120 3009 3119 3011 3121 3007 3097 3005 3114 Note: Schools sampled also include a small share of schools that are part of USAID and DFID programs (i.e. Quality Improvement in Schools [QUIPS] and Whole School Development [WSD]), which had virtually no impact on variations of the frequency of visits. a. Indicates per month and not weekly. Source: World Bank 2004a. 373. In Ghana there exist several mechanisms to promote community and stakeholder involvement and oversight in local schools. The Parent Teacher Associations (PTAs) are the oldest and most widespread, thought they have limited powers. School Management Committees and School Performance Assessment Meeting were established more 145 recently, but have had trouble gaining a real foothold and tend to operate sporadically. The most influential form of community involvement typically comes from local traditional authorities – headed by the local chief. When the traditional authority is interested and active, it can provide powerful oversight. 374. In basic education, virtually every school in Ghana has a PTA: 99 percent of public and 95 of private schools (World Bank 2004a). When parents are actively involved in their children’s schools through productive and effective PTAs, parental oversight can have a direct impact, for instance, on teacher attendance patterns (CDD, 2008; Brookings Institution, 2007). However, a study by Transparency International (2009a) found that parental knowledge of local school structures varies, with 97 percent of parents knowing about a PTA but only 44 percent about another community oversight structure, a School Management Committee. Survey results have shown that the degree of community support for PTAs is strongly linked to its level of income such that financial support for PTAs in schools located within wealthier areas is on average tenfold that of schools in the poorest areas (World Bank 2004a). Although funding cannot determine the quality of PTAs, a minimal finance threshold exists in order to function properly. 375. As part of the 1993 decentralization, School Management Committees (SMCs), were introduced and charged with the responsibility to rekindle the community spirit in improving education, especially at the basic education level, and to analyze their own schooling problems and adopt and internalize their strategies towards the improvement of teaching and learning. As with PTAs, the presence of SMCs will not necessarily have a positive impact on schools unless they are run efficiently. Among primary and junior secondary schools surveyed, four out of five (80 percent) had SMCs set up and functioning, with the figure reaching 90 percent within primary schools. A study found that school management committees met in the month previous to the survey within only about 50 percent of schools at the basic level (World Bank 2004a). The School Management Committees practice some level of community oversight, but their focus is limited to overseeing the use of the capitation grant and does not extend to educational issues, including whether teachers are present and whether students spend their school time on task. 376. The School Performance Assessment Meeting (or SPAM), which 92 percent of public primary schools have but only 6 percent of surveyed households attend, is a type of supervision mechanism that provides a forum to agree on action plans (World Bank 2004a). Decentralization 377. Since as early as 1951, local councils have had responsibility for providing and maintaining educational facilities while central government provided teachers’ salaries. This was reinforced in 1961 when Government “reaffirmed local authority control and management of primary education� (CREATE, 2007) and further reinforced by the Local Government Act of 1988 which made each district or community responsible for providing and maintaining basic school infrastructure. In 1993 devolution was legislated 146 and 22 Government departments, including the Ghana Education Service were decentralized19. 378. The 2007 reforms specify that “Metropolitan, Municipal and District Assemblies shall be responsible for the infrastructure, supervision and monitoring of Basic and Senior High Schools�. Currently, Ghana has 10 Regions and 138 districts (though these were increased to 154 in 2009). 379. Presently, the Metropolitan, Municipal and District Assemblies (MMDAs) are not fully autonomous in most expenditure decisions, especially in the areas of teacher management, remuneration (salaries and allowances), investment budget (classroom, school buildings, teachers' houses etc.) and textbooks provision (Steffensen 2006). Table 4-21: Level of MMDA Discretion Expenditure Degree of Discretion C\omments Salaries No discretion Handled by the central payroll system and hiring/firing is performed centrally Administration and No discretion Disbursement controlled centrally. Service components GOG funded Limited discretion Most investments are distributed centrally and heavily controlled. Investments within MMDAs have limited influence on the priorities. In theory, Education MMDAs are in charge of school construction, but in reality they do not fully control this function, which is managed by various funding (GES/GET/Donors etc.). Investments from Limited discretion Controlled by the center – MOES/GES – MMDAs function as “post HIPC for Education – offices� Investments from Some discretion within four MLGRD has a say in the screening process HIPC for MMDAs core sectors (Education, Sanitation, Health and Water) to make cross sectoral priorities. DACF Significant discretion for the DACF is a non-sectoral grant, but there are earmarked budget lines, largest share of the funds which limits the discretion of MMDAs to some extent. GETFund Limited discretion Controlled by the GETFund and MOES/GES, vertical lines of reporting and accountability. Capitation grants Discretion for the schools The new capitation grant provides the schools with some flexibility to cover the most urgent operational costs. Allocations are made fixed as per pupil enrolled and it does not involve the DEOs and/or the MMDAs. Donor funds Some discretion This varies greatly from source to source. The DWAP and the PPS provide the MMDA with a significant flexibility within a set menu (funds to finance work-plans, excluding salaries), others are involved in direct purchases of classrooms and text books. Source: Steffensen 2006. Education Information Systems, Monitoring 380. Education Management System has Limited Scope and Relevance 381. The Education Management Information System (EMIS) has been built up throughout the decade. The capability was strengthened following the ESP 2003 and since then, an annual performance review exercise has been carried out based largely on EMIS data. But, the EMIS to date is basically an annual school census, usually carried 19 Arguably, this should be considered deconcentration rather than decentralization, given that District Office staff is appointed by the Center, reports to the Region and uses centrally defined operational guidelines and standards. 147 out at the end of the calendar year or early of the following year. All data in the EMIS relies on school responses to a detailed questionnaire and the definitions in the sector performance reports are dependent on the data collection methodology. For instance, primary completion rate in Ghana is equal with GER at 6th grade. Given the reliance on school responses, and considering the lack of IT at many schools, data collection is not computerized. 382. The reliance on an annual school-based census is not unique, but it limits the EMIS reliability and timeliness for decision making. A key concern with this EMIS design is that it maintains a fragmentation between types and sources of data and that, in turn, limits its usefulness for decision makers. This report has indicated some of the key challenges in the education sector, but very few of them can be addressed through the EMIS data collection: 383. Children and student based information: The report has shown so far that key concerns with outcomes include out-of-school children, overage students and migrations. Even if statistics were collected on these, EMIS is not linked up with such information and it doesn’t collect data on the age of children in schools. 384. The private sector has become a key provider of services at all levels, especially at the lowest levels of basic education. Yet, EMIS has a limited reach to private schools. Learning outcomes, test results: We have seen the challenges in competence, proficiency measured by tests. EMIS does not integrate these results. 385. Teacher qualification, teacher payroll. Information about these are part of the payroll data and the two are not integrated. Expenditures, sources and use of funds: EMIS is not integrated with the diverse sources of financial information. Schools have no information about expenditures and financing. 386. A key concern of EMIS has been that it was a hierarchical and highly centralized data collection, serving essentially the annual sector performance review process. For the last three years, however, the annual reviews included regional reviews based on district level information. These reviews have shown the importance of performance review at local level and also the limitations of the centralized EMIS. Information at the regional reviews, based on district feedback has often used data that was incongruent with the centralized EMIS-based statistics. Also, the limited response from private schools has created a dilemma for the MOE: whether to adjust the data to estimate the likely trends in non-responding schools or leave key statistics, such as enrollment and admission rates, and gender parity unadjusted and distorted. Limited Information about the Quality of Education Services 387. The report has shown some of the most critical factors in determining equity and efficiency in education. These included the inefficient system of teacher deployment, disparities in qualified teachers and in per child expenditures as well as wastage in services caused by absenteeism, and low ToT. These likely are the key reasons for some of the major challenges in education service delivery. Yet, information about teachers, funding, and attendance is sporadic and is not available at local levels where social accountability is critical. 148 Data is not collected in time for key management decisions 388. Some of the key decisions include annual budget planning, investment decisions (GET fund), deployment or redeployment of teachers. Information is not adequately available for users at local levels, especially for the MMDAs, which appear to have some level of authority to use non-regular resources to improve education services. Conclusion: Efficiency and Accountability in Education Management and Administration 389. Resources will have to be better managed if the government wants to achieve further improvements in access to and quality of education services, given the already very high overall level of public expenditures. The Government has been considering some strategic changes at the local, school-level and central levels but has not fully implemented them. a) Local authorities need stronger capacities for decision making, supervision and resource allocation to more effectively address the diverse social and economic conditions of education. These include more consistent promotion of innovative practices, strengthening the supervision by district level authorities and strengthening the role of local communities through improved accountability. b) School management needs to be professionalized through training, strengthened authority and adjustment of the career patterns in education to make sure that those become headteachers who are trained and qualified for the post rather than those who are the most effective or senior teachers (teaching and school management are two different professions). c) District authorities need to be given authority to hire, fire and deploy teachers and more resources to better supervise their attendance and performance. d) The central Government needs to build capacities to develop and monitor policies and incentives, assess learning, monitor performance and guide those responsible for service delivery as it devolves management to district and school levels. 149 5. HIGHER EDUCATION20 390. Higher education (HE) in Ghana began when the then British Colonial Administration upon the recommendation of the Asquith Commission on Higher Education established the University College of the Gold Coast in 1948. When the country attained independence in March 1957, it became known as the University College of Ghana and until it gained a full university status as University of Ghana (UG) in 1961 was affiliated to the University of London which supervised its academic programs and awarded external degrees to its graduates (UG, 2007; Manuh et al., 2007). 391. As of 2009, the National Accreditation Board (NAB) had accredited 126 public and private institutions as HEIs. There are 8 public universities. Private institutions, most of which are affiliated to other autonomous institutions since only 4 have been granted a charter to award their own degrees, are accredited for a period ranging from two to five years. The growth in the number of public institutions is largely attributable to the upgrading of the 38 Teacher Training Colleges which were under the Teacher Education Division of the Ministry of Education, into Colleges of Education in September, 2007 and placed under the National Council for Tertiary Education (NCTE). All the Colleges of Education are affiliated to the University of Cape Coast, which awards the Diploma in Education to graduates (NAB, 2009). Hitherto, students who enrolled in the Teacher Training Colleges were awarded the Three-Year Post-Secondary Teacher’s Certificate ‘A’. Table 5-1: List of Accredited Tertiary Education Institutions as of February, 2009 Type of Institution Total Number Public Universities 6 Other Public Tertiary Institutions 2 Chartered* Private Tertiary Institutions 4 Private Tertiary Institutions 45 Polytechnics (all public) 10 Public Teacher Training Colleges** 38 Private Teacher Training Colleges 1 Public Nursing Training Colleges 14 Private Nursing Training Colleges 4 Total 124 * Granted Presidential authority to award qualifications. ** Now Colleges of Education Source: NAB 2009. 392. Ghana has made considerable progress in widening access to and promoting the subject diversity of tertiary education. In a constantly changing global market, this improvement should be accompanied by a persistent reanalyzing of curriculum relevance. As of 2010, there were a total of six public universities, encompassing three older and well established ones – University of Ghana, University of Cape Coast, Kwame Nkrumah University of Science and Technology – and three more recently established ones – 20 This section on higher education is based on a background study by James Anyan. 150 University of Education at Winneba, University of Development Studies and University of Mines and Technology. As to private universities, the total number is uncertain, even thought they are required to register with the National Accreditation Board. Thus, adequate data on private tertiary institutions, which are vital for national development, are currently lacking. In terms of the public polytechnics, all 10 of these institutions acquired university accreditation per 1992 legislation (PNDC Law 321). Other public institutions also exist, such as institutes of professional studies and languages (e.g. focusing on accounting, translating, journalism, media communications). 393. Tertiary enrollment has grown thirteen-fold since 1990. The estimate of the transition rate from senior high school to the first year of tertiary is about 10 percent (Boissiere 2008). Since the early 1990’s tertiary education has expanded at an exponential rate. Before 1992 when only three institutions (UG, KNUST and UCC) in the country were regarded as institutions of higher learning; total enrollment stood at 14,500. Following the upgrading of the polytechnics in 1993 to HEIs, total enrollment in the tertiary sector more than doubled within a year from 14,500 in 1993 to 30,000 in 1994. By the 1997/98 academic year, the sector had witnessed a further spike in enrollment of about 33% from the 1994 figure to 40,000. Within a decade (1997/98-2007/08) total enrollment in the public HEIs alone (excluding the colleges of education) has more than tripled from 40,000 to 132,604 (Chambas, 1998: 3; NCTE, 2008b: 1). This represents a 232% increase in enrollment within a decade, or a thirteen-fold increase since 1992. A glance at three year trends in tertiary enrollment between 1990 and 1995 illustrates that access to tertiary has experienced a tremendous development: 9,997 students in 1990, 15,500 in 1993, 24,500 in 1996, 46,500 in 1999, 86,570 in 2002, and 121,390 in 2005 (Adu-Orivel 2006). In 1995, a total of only 171 students attended tertiary institutions in Ghana, owing to a year-long strike among teachers. Since the academic year 2003/04, enrollment growth averaged 10,000 new students per year. A 104% increase occurred in the public universities while the polytechnics recorded a 69% increase over the period 2001/02-2007/08. With regard to polytechnics, in Ghana none existed in 1990; currently, there are over 18,000 students enrolled in ten public polytechnics. Of all tertiary students, 71 percent are enrolled in universities, 26 percent in polytechnics and 3 percent in professional institutions. 394. Such rapid expansion has placed substantial pressures on the supply side, in particular, in terms of new faculty members and adequate infrastructure. Adding to the stresses of the enrollment influx and resource demands, governance structures in tertiary institutions have not adapted and basically have remained unchanged over the past years (Boissiere 2008). As a consequence of the growing enrollment numbers, there are also increasingly more unemployed graduates, some of which eventually seek job opportunities abroad (ESPR 2009:15). The principal determinants of the rapid enrollment growth were, firstly, tertiary institutions capacity to levy fees, in turn, allowing for ‘natural expansion’ and consequently more revenue and, secondly, the GETFund provided resources to expand infrastructure (Boissiere 2008). Moreover, the analysis conducted by the Country Economic Memorandum using household and employment surveys demonstrates that in Ghana the rate of return on tertiary education is considerably high (Boissiere 2008), thus justifying past and further growth in the sector. 151 395. A burgeoning private sector. The emergence and the increasing number of private HEIs in Ghana in the last decade is in sync with similar developments in the sub-region. The number of accredited private HEIs reached 54 in 2009. The private higher education institutions are playing a significant role by easing the social demand for HE. As of December 2008, there were 36 unaccredited institutions listed on NAB’s website as compared to 54 accredited ones. While 26.3 percent of all tertiary students (about 24,300) were enrolled in private institutions in 2003/04, 20.5 percent of students (about 28,700) were enrolled in such institutions in 2006/07, which corresponds to a 5.8 percent decrease as a share of the total over these years. In other words, while private enrollment is in fact rising, public enrollments are expanding at a much faster pace and represent a growing share of total tertiary students. 396. Polytechnics. The polytechnics in Ghana operated as second-cycle institutions under the Ministry of Education until they were upgraded to tertiary level status in 1993 and brought under the supervision of the NCTE following the promulgation of the Polytechnics Law (Provisional National Defence Council [PNDC] Law 321 in 1992 (Ministry of Education, Science and Sports, 2008). Since the law, they are more oriented to technical and vocational education and the training of middle level human resources for national development, and mandated to award the Higher National Diploma (HND) to graduates who have successfully completed the three-year polytechnic education. 397. It is worth mentioning that the NAB has accredited five of the ten polytechnics: Cape Coast, Takoradi, Accra, Ho and Sunyani to award the more advanced B-Tech. degree in specified programs (NAB, 2009). The emerging academic drift in Ghanaian HE, as evidenced by the increasing tendency of the polytechnics to be like the universities, is a threat to differentiation and defeats the purpose of establishing them to widen the scope of student choice and to develop middle-level human resources for socio-economic development. As the NCTE (2008a:10) puts it,�...diversity in tertiary education could be hampered by duplication of academic programs in public universities and the increasing concentration of polytechnics on business programs rather than applied science and technology�. 152 Figure 5-1: Enrollment in Public Universities and Polytechnics 2001/02-2007/08 Figure : Enrollment in Public Universities and Polytechnics 2001/02 - 2007/08 140,000 120,000 100,000 Students 80,000 60,000 40,000 20,000 0 2001/ 2002/ 2003/ 2004/ 2005/ 2006/ 2007/ 02 03 04 05 06 07 08 Univ. 46,184 53,895 63,576 73,408 84,078 88,445 93,973 Poly. 20,44 23,11 24,35 24,98 24,66 28,69 34,44 Total 66,626 77,012 87,929 98,391 108,74 117,14 128,42 Source: Based on NCTE 2008b. 398. The phenomenal growth in student numbers in Ghana, is not a deviant case when viewed from the context of similar developments in HE both at the regional (SSA) and global levels. As far as tertiary enrollment growth is concerned, SSA has been leading the world during the last four decades. Although the region enrolled less than 200,000 students in 1970, enrollment increased more than 20-fold to approximately 4 million by 2007 (UNESCO-UIS, 2009a:1). ISSUES/CHALLENGES 399. Tertiary Gross Enrollment Ratio remains rather low. Despite the huge increases in student numbers over the years, gross enrolment in tertiary education was 6 percent as of 2007. This figure however shows that participation in HE has doubled since 2002 (3%) according to UNESCO-UIS (2009c: 2). The 6% GER is similar to the average for SSA, although the global average stands at 26% based on UNESCO data. Countries whose participation rates exceed the regional average include South Africa (15.4%), Mauritius (14.0%), Nigeria (10.2%), Cape Verde (8.9%), Côte d’Ivoire (7.9%), Senegal (7.7%) and Cameroon (7.2%). Other countries such as the Central African Republic (1.1%), Chad (1.2%), Mozambique (1.5%), United Republic of Tanzania (1.5%) and Burundi (1.9%) have rates considerably below the average for SSA (UNESCO-UIS, 2009a:1). According to the Government’s ‘Education Sector Performance Report’ (2010a), the tertiary gross enrolment rate increased to 9.7 percent by the academic year 2008/09. Considering the fact that Ghana’s GER for the secondary level grew from 38% in 2002 to 49% in 2007 and the Net Enrollment Rate (NER) from 32% to 45% for the same period, it is evident that pressure will continue to mount on HEIs, especially the public ones for admissions. 400. Globally, the tertiary gross enrollment averaged around 17 percent while the average for developed countries was around 67 percent in 2006. This illustrates how far Ghana still has to go to equitably widen tertiary access. 153 Figure 5-2: Overview of Primary, Secondary and Tertiary GERs, 2006 Primary GER 06’ Total Secondary GER, 06’ Tertiary GER 06’ 120 112 107 106 98 96 96 95 100 80 76 60 54 50 49 40 32 32 32 20 12 10 5 3 6 5 5 0 India Botswana Kenya GHANA Nigeria Benin SSA Note: Variations between EMIS and UNESCO data may arise due to differing methodologies applied. Source: Based on UNESCO, 2008. 401. Access and equity. Access to HE, by and large, remains one of the greatest challenges of the HE system in Ghana. As noted earlier, the country’s secondary level NER of 45% for 2007 compared with the tertiary GER of 6% means that all things being equal, the number of secondary graduates seeking admissions to the tertiary level far outweighs the capacity of the HEIs to absorb them. Despite the huge increase in student enrollment in recent times, available data indicate that access has not necessarily broadened and opportunities for HE have been lopsided and unevenly distributed. Analyses of student enrollment with respect to students’ socio-economic status (SES); gender, places of origin, previous secondary schools, family income among others prove this point. Indeed, “it is at the entry point of tertiary education, that the compound effects of inequalities in access to and completion of basic education, and progression through secondary education become most visible� (UNESCO, 2008: 90). The problem of inequities in the distribution of opportunities in education goes beyond HE. But HE sums up the disparities at the lower levels of education. 402. A few secondary schools dominate tertiary intake. The secondary school background of an applicant to the public universities in Ghana greatly influences their admission, and to which program of study. In a study of demography, social background and admissions into tertiary institutions in Ghana, Addae-Mensah (2000) found that over 70% of the country’s future doctors, scientists, engineers, architects, pharmacists, agriculturists, and other professionals emerged from just about 10% of the schools, with almost 50% emerging from less than 4% or only 18 of Ghana’s 504 Senior Secondary Schools. In fact, the study also revealed that 70% of admissions in the Humanities and 91% in the Sciences including Medicine and Agriculture for the 1999/00 academic year at UG were accounted for by the top 50 secondary schools, constituting 9.9% of secondary schools in Ghana. The situation at KNUST was no different. A recent survey of 1,500 students in the country’s five oldest public universities confirms that there has been little variation in this pattern of admissions as it emerged that 50% of students had come from only 29 schools in the country (Manuh et al., 2007). 403. Geographic Distribution of Tertiary Enrollment. Alongside the influence of students’ secondary schools on the admission process, geography has also been found to be a strong determinant of who gets into a particular HEI in Ghana. Residence impacts 154 who gets a slot in Ghana’s HE. The results of the survey cited above revealed that 75% of students in the public universities came from five of the ten administrative regions in Ghana; namely Ashanti, Eastern, Central, Volta and Greater Accra. The Eastern Region is ranked the highest with over 18%. The other five regions had enrollments ranging between 4-6%. Again, approximately 70% of students in the five oldest universities reside in only three regions – Ashanti, Eastern and Greater Accra -- implying that students from these regions stand a better chance of making it to the university relative to those from the rest of the regions. Not only do these regions (Ashanti, Greater Accra and Eastern) have the highest population densities in Ghana respectively, they are also home to most of the well-endowed SHS in the country. Figure 5-3: Region of Origin of Undergraduates Students at UG, 2002/3 and 2007/08 Figure: Region of Origin of Undergraduate Students at UG, 2002/03 and 2007/08 25 20 Students (%) 15 10 2002/03 5 2007/08 0 Source: Based on data from the Planning Management and Information Systems Directorate (PMISD) of UG, 2008. 404. While the regions that are historically privileged in the distribution of students extended their domination, the disadvantaged ones grew worse off during the period under review with the exception of the Brong Ahafo which saw a very slight increase. Whereas in 2002/03 Ashanti, Central, Greater Accra, Eastern and Volta had almost 82% share of undergraduate students at UG, in 2007/08 the five regions’ share of students had increased to 85%. ,During the same period the other five regions saw their share decrease from 18% 2002/03 to 15% in the 2007/08 academic year. 405. Gender-Based Equity: Women still under-represented. Despite government policy aimed at achieving gender parity in Ghanaian HE, this target is far from being achieved. Moreover, the portion of female students has barely budged in recent years. Recognizing that the gender imbalance in favour of males starts well before HE, Government has created a Girl Child Education Division in both the Ministry of Education and the Ghana Education Service to tackle the problem from both the basic and the secondary levels. More action however, is required to achieve the target of 50:50 enrollments which is 155 currently the world average and has been exceeded by other SSA countries like Mauritius, Lesotho and South Africa. 406. In 2003/04, 31.8 percent of students in Ghana’s tertiary sector were female, compared to and 33 percent in 2007/08, signifying that the gender gap narrowed minimally over the past few years. The 2007/08 gender parity equaled 0.49 -- about two males for every female enrolled. According to data from the World Development Indicators (World Bank), gender parity has subsequently improved to 0.54 at the tertiary level. Table 5-2: Male/Female Tertiary Enrollment 2003/04 2004/05 2005/06 2006/07 2007/08 2008/09 Tertiary 92,530 103,222 121,390 139,768 132,604 167,346 Male 63,076 69,179 79,475 91,827 88,836 Female 29,454 34,043 41,915 47,941 43,768 Gender Parity index .47 .49 .53 .52 .49 Source: EMIS. 407. Socio-economic equity. The influence of one’s socio-economic background such as family income and occupation of parents on access, participation and success in HE is well documented. In South Africa for example, 70% of students who drop out after the first year in the university come from low income families. In the Ghanaian context, a survey conducted by the NCTE in 2002 on the socio-economic status of students in HE showed that enrollment is skewed in favour of students with white-collar parentage over those with blue-collar parents. HE in Ghana is disproportionately ‘consumed’ by the richest 20% of the population. Male students from the highest income quintile (Q5) are more than seven times more likely to enter and successfully complete HE than those from the poorest quintile (Q1). The situation is even more precarious for the female category where students come from only the richest 40% of the population. 156 Figure 5-4: Distribution of Tertiary Gross Enrollment Ratios by household Wealth and Gender, 2006 Figure: Distribution of Tertiary Students by Household Wealth and Gender, 2006 8 Gross Attendance (%) 6 4 2 0 Q1 Q2 Q3 Q4 Q5 (Poorest) (Richest) Male 0.11 0.90 1.23 2.63 7.84 Female 0.00 0.00 0.00 1.81 7.31 Source: Based on data from Ghana Multiple Indicator Cluster Surveys (MICS) 2006, World Bank-EdStat. 408. Responses to Access and Equity Problems. It is noteworthy that KNUST pioneered a quota-based admission system which allows students from deprived schools categorised by the Ghana Education Service (GES) as Less Endowed Schools (LESs) to enroll as long as they have met the basic entry qualification, even if they do not achieve the competitive ‘cut-off’ points required for entry into the various programs. This initiative has been replicated by the UG and UCC. The latter has reserved 200 admission slots for this category of students (Morley et al. 2007). In the 2007/08 academic year for example, a total of 243 students comprising 183 males and 60 females were offered admissions at KNUST through the LES representing 3.5% of total admissions (6,868) for the year (Adarkwa, 2007). Apart from the quota system for the deprived schools, the public universities have also adopted affirmative action for female enrollment. Female applicants to the UG are admitted one point lower than their male counterparts while the UDS admits all female applicants with the basic entry qualifications into all programs. At the UCC, the Joint Admissions Board according to Morley et al. (2007) is poised to keep female enrollment at levels not below 35% of total enrollment. A ‘dual-track’ tuition policy (fee-paying) window is also available for applicants who possess the basic entry qualifications but cannot get into the various academic programs due to the competitive ‘cut-off’ points to enrol in the public universities. 409. Relevance: the changing role and new demands on HEIs. HE is expected to justify the public investments it receives by helping to raise the appropriate human capital to deal with the socio-economic challenges confronting countries and to position them to effectively survive, compete and succeed in the knowledge economy. 410. However, the 60:40 enrollment policy goal for programs in Science & Technology and Arts & Humanities respectively, for the public HEIs is far from being realised. In fact, enrollment in Science & Technology in the public universities averaged 36% during the 2001/02-2007/08 period, while Arts & Humanities accounted for 64% during the period under review. 157 Figure 5-5: Trends in Enrollments in Public Universities by Broad Fields of Study, 2001/02-2007/08 Figure: Trends in Enrolments in Public Universities by Broad Fields of Study, 2001/02-2007/08 Arts/Hum. Sci./Tech. 80,000 60,000 Students 40,000 20,000 0 Source: Data based on NCTE (2008b:iv). 411. Graduate Output and the Labor Markets. Tertiary education attainment rate for the 25-64 age groups has increased from 2.6% in 1991 to 7.6% in 2005 yet unemployment among well-educated urban youth (6.1% in 2005 according to one very narrow measure) is a persistent problem despite the country’s appreciable economic growth rate and poverty reduction compared to other SSA countries in the past 15 years, a stable labor force participation of 89% (age 25-64) and a reduction in the general unemployment rate from 2.7% in 1991 to 2.3% in 2005, according to one very narrow measure (World Bank 2009). Table 5-3: Graduate Output by Broad Fields of Study, 2001 & 2007 (ISCED 5-6) 2001 2007 Field of Study Total MF % F %F Total % F %F Education 1,122 8 288 26 2,917 13 1,219 42 Hum.& Arts 3,984 28 1,255 32 6,921* 30 2,737 40 SSBL 4,147 29 1,145 28 5,748 25 1,900 33 Science 1,399 10 346 25 2,418 10 705 29 EMC 2,248 16 254 11 3,116 13 313 10 Agriculture 480 3 104 22 1,059 5 247 23 Health & Welfare 591 4 163 28 628 3 223 36 Services 492 3 357 73 496 2 415 84 Total 14,463 100 3,912 x 23,303 100 7,759 Notes: SSBL=Social Science, Business and Law. EMC= Engineering, Manufacturing and Construction. *Figure may include some Social Science graduates. Percentages have been rounded up and may exceed or be lower than 100 in some cases. Sources: Data from UNESCO Institute for Statistics (UIS) database for 2001, NCTE data for 2007 category. 412. As expected, graduates from Ghana’s HEIs increased by 61% between 2001 and 2007. In general, female graduate output nearly doubled (98%) during the period under review from 3,912 in 2001 to 7,759 in 2007. Whereas in 2001 Science and Technology fields produced 26% of graduates in 2001 compared to 74% for the other fields, by 2007 the share of the former had fallen to 23% and that of the latter increased to 77%. Clearly, however, employers are not just looking for ‘graduates’ but ‘value-added graduates’. 158 Over time, employers’ demand for graduates with computer and analytical skills, as well as other personal attributes has increased considerably, making the trend away from science and technologies studies even more troubling. 413. Internal efficiency: human resources. Here the quality of institutional management in HE is examined, with particular emphasis on resource mobilisation and allocation, the so-called productive efficiency. The ability of HEIs to mobilise and allocate human, material and financial resources in pursuance of the trinity of mission – teaching, research and community service in a manner that maximises output and reduces ‘production costs,’ and assures quality determines the level of efficiency. 414. Graduate studies, particularly, those which lead to the award of a PhD Degree, impact positively on the faculty strength and quality of any HEI. However, Ghanaian HEIs have not fared well in attracting, retaining and graduating individuals with PhDs to augment the stock and quality of academic staff for teaching and research activities. A fundamental problem with the low production of PhDs in Ghanaian Universities is attributable to poor ‘throughput’, which is an element of internal efficiency determined by the rate as well as the speed of completion of degrees, retention and survival rates (Salmi & Hauptman, 2006). Data available at UG’s School of Research and Graduate Studies show that of the 38 PhD candidates enrolled in the Arts and Sciences full-time in 2000, only 16 had graduated by 2007 although even part-time students had a maximum of seven years to complete the PhD (UG, 2007: 52). Table 5-4: Student/Academic Staff Ratios and Norms for Public Universities, 2003/04 and 2007/08 Academic Field of Student No. of Student/Staff NCTE Year Study Enrollment Staff Ratios Norms Science 18,726 755 25:1 12:1 Medicine 2,013 307 7:1 8:1 2003/04 Humanities 36,009 763 47:1 18:1 Pharmacy 602 34 18:1 12:1 Education 6,226 76 82:1 15:1 Total 63,575 1,935 33:1 Science 30,615 1,051 29:1 12:1 Medicine 4,138 342 12:1 8:1 2007/08 Humanities 41,833 776 54:1 18:1 Pharmacy 648 38 17:1 12:1 Education 16,739 276 61:1 15:1 Total 93,973 2,483 38:1 Sources: NCTE, 2004; 2008b. 415. In 2007/08, there were on average 38 students per teacher (STR) in public universities. The STRs for all the fields of study exceeded the norms of the NCTE for both 2003/04 and 2007/08 academic years, excepting Medicine whose STR of 7:1 in 2003/04 was slightly below the norm (8:1) for all the public universities. 159 Figure 5-6:Full-Time Teaching staff in Public Universities by Qualifications and Gender, 2006/07 Figure: Full-Time Teaching Staff in Public Universities by Qualification and Gender, 2006/07 2000 No. of Staff 1500 1000 Male 500 0 Female Source: Data based on NCTE, 2007. 416. Female teachers in Ghana’s universities are grossly underrepresented: 15% compared to 85% male representation. Moreover women tend to be concentrated in the lower ranks of teaching staffs. Approximately, a third of teaching staff in the country’s public universities had a Ph.D. in 2006/07. The recruitment of First Degree holders (BSc/B.A.) into teaching positions in the universities reveals the extent of academic staff shortages in the public universities. The Ghanaian Government has taken steps to improve remuneration to attract and retain teaching staff in the universities. Remuneration increased by 80% in real terms from 2001-2005 (Adu & Orivel, 2006). 417. Internal efficiency: financial resource mobilization and allocation. Public HE in Ghana is funded from four principal sources: annual budgetary allocations from the Government of Ghana (GoG), allocations from the Ghana Education Trust Fund (GETFund) for the development of infrastructure as well as research and staff, funds generated internally by the HEIs, and student fees. GoG allocations account for 65% of recurrent expenditure of the public HEIs with the remaining 35% borne by funds from private sources. Through the so-called “Akosombo Accord� in 1997, it was agreed that government or the taxpayer takes 70% of the costs of HE, while the remaining 30% is borne equally from three sources: students (tuition fees), institutions (IGFs) and private donations (Manuh et al., 2007). Consequently, students in the country’s public HEIs were required to pay Academic Facility User Fees (AFUF) and Residential Facility User Fees (RFUF) in addition to other ancillary fees for such services as examination, transcripts and technology service, and as a penalty for late registration. Tuition remains free for students whose grades are competitive enough to get them into the government- subsidised admission slots, the so-called regular students. 418. Per student costs. Per student cost of HE in Ghana , is 297% of Gross Domestic Product per capita (GDP p.c.) compared to 19% and 55% for primary and secondary (senior high school) education, correspondingly. Thus, per student costs at the tertiary level are about 16 times the costs of primary and five times that of secondary. The SSA average is 314% of GDP p.c. (World Bank; MOE, 2008). 160 Table 5-5: Countries with the Highest Share of GDP p.c. per Tertiary Student (2005-2007) GDP Ratio of Ratio of per Tertiary Tertiary Tertiary capita, NER to Pry. to Sec. tertiary Lesotho 1141.5 3.6 45.6 22.9 Ethiopia 785.5 2.7 62.7 87.9 Botswana 449.6 5.1 27.9 10.9 Rwanda 372.8 2.6 36.6 10.6 Niger 371.4 1.0 13.0 8.1 Burundi 363.1 1.9 18.2 4.7 Chad 348.2 1.2 49.3 11.9 Swaziland 337.6 4.2 21.3 8.4 Central African Rep. 305.2 1.1 40.8 - Ghana 297.0 - 15.6 5.4 Burkina Faso 236.5 3.0 6.6 10.1 Source: World Bank- EdStats, 2009; World Bank, 2008 data. 419. Suggestions and recommendations: access and equity. Available evidence, thus far, indicates that demand for HE in Ghana will continue to rise. One indication of this is that with a secondary NER of 45% and tertiary GER of 6%, applications for admissions can be expected to outstrip supply by the HEIs. Measures to diversify the sector and take the pressure away from the university sector should be strengthened. The polytechnic sub-sector could be made attractive for students as a viable and alternative option to university. Steps should be taken to halt the academic drift of the polytechnics by ensuring that the sub-sector focuses on its core mission of training technicians and professionals at an intermediate level of HE. 420. The lifelong learning option could be explored further, developed, and made more appealing by government and the HEIs to allow non-traditional students whose qualifications may not get them into HE as traditional students, to access HE. Professional experience, age and similar attributes should be given greater weights in such admissions. Distance education initiatives by the HEIs go in the right direction; government should encourage the HEIs by investing in modern educational technologies and study centers. Cross-border distance education should be encouraged and handled in a manner that does not compromise quality or discourage investments. In particular, investments that foster foreign collaboration with Ghanaian HEIs should be encouraged, since such initiatives stand to give students, teaching and research staff international experience and exposure. 421. Differentiation of the sector is key to increasing access, widening participation and reducing government spending in Ghana’s HE. In spite of the increase in enrollment in Ghana’s HE, the 6% GER is still not the optimum for a country aspiring to middle- income status as evidence from other SSA countries shows. Requisite human capital and skills development is a sine qua non for the attainment of such a status and to make the country competitive in the knowledge economy. Diversifying the sector, ensuring that 161 academic programs respond to socio-economic needs will make graduates not just job- seekers but job-creators, which should help reduce the problem of graduate unemployment. 422. In terms of efforts to help incorporate marginalized groups, setting quotas would be an important step to enable the Government to do more in the area of policy-making and monitoring to ensure that efforts are properly coordinated and harmonized. The Government should revisit the policy of ensuring that each district has at least one well- endowed senior high school to address the gross underrepresentation of graduates of the majority of the senior high schools in the HEIs. Again, improving information dissemination about opportunities in HE, fee-rebates, scholarships etc., especially to rural students who in most cases have had no relatives who have enrolled in or completed HE, would go a long way to demystify HE and reduce debt-aversion for students with low socio-economic status. Public subsidies for HE should also be targeted at such minority groups since benefit incidence analysis has revealed that HE in Ghana is disproportionately being ‘consumed’ by the rich and urban dwellers. 423. Suggestions and recommendations: relevance. The Government policy goal of 60:40 enrollments for Science & Technology, compared to Humanities & Arts (and other fields) is gravely off-track. Enrollments from 2001/02 to 2007/08 averaged 36% and 64% correspondingly. Worse, the number of graduates in the Sciences took a down turn during the same period. This has far-reaching implications for Government’s vision of “[...] aspiring to build a knowledge-based economy within the next generation� (MOEYS, 2004: 3). Here again, Government should look beyond HE to address this challenge. The teaching and learning of science-based courses, especially at the secondary level should be given a boost to attract and sustain the interests of students. At the tertiary level, interventions should include the creation of science-based programs other than Medicine which would lead to employment opportunities and thereby be attractive to many students. It is not uncommon to find science students abandoning the field for programs in Business and Management at both the undergraduate and postgraduate levels because such fields are deemed to hold better job prospects for graduates in Ghana. It is important to note that training human resources for today’s labor markets may not be the best long- term option due to the volatile nature of the labor markets; hence the need for both job- specific and broad liberal arts education accompanied by the development of universal skills and aptitudes that will make the graduate adaptable in the labor market. 424. Suggestions and recommendations: internal efficiency. At the heart of the issue of high STRs and an ageing faculty in the public HEIs is the low throughput from graduate studies and the cumulative effect of poor teacher remuneration and other inferior conditions of service in the past. The best and brightest first degree holders from the universities tend to enroll in graduate studies abroad as that is typically perceived as superior to graduate study in Ghana. Inadequate scholarships and other funding opportunities for graduate studies also account for this. Neither can departments with few faculty members holding PhDs effectively mount PhD programs nor supervise PhD candidates. Urgent steps must be taken at both the system and institutional levels to improve quality and restore confidence in graduate education and research training in the country’s HEIs. Some rationalization of non-academic staffing in both the public universities and polytechnics is required to keep them within the limits set by the NCTE. 162 The underrepresentation of females within academic staffs in the HEIs is an outcome of the gender imbalance in female enrollment in Ghana’s HE. The majority of female academics are Master’s Degree holders and at the rank of Lecturer. As female enrollment in the HEIs improves it will have a positive influence on the stock of female academic staff in the HEIs. Conclusions: Improving access to and relevance of higher education 425. Ghana needs policies to deal with social pressures to increase higher education enrollment and economic pressures to improve relevance; without these, most social and economic expectations, and also the Government’s strategic agenda will remain unfulfilled. Differentiation of the sector is also key to increasing access, widening participation, increasing relevance, and reducing government spending. It’s also clear that effective policies require a change in the way higher education is financed and governed, how institutions and services are managed, and how quality is assured. Already almost half of the higher education revenues come from non-state resources; it is imperative that both state and non-state resources are used more strategically. A few policy options may include the following: a) Government financing should be more closely linked to performance, norms, and strategic priorities. The financing of inputs including staff and other costs based simply on enrollment, and the separation of investments from recurrent expenditures limit the Government’s ability to stimulate improved performance or progress along the desired strategic directions. b) By the same token, Government financing should also be more targeted to meet specific social, economic or other strategic objectives. c) Institutions need to have more management authority, responsibility and capability to run higher education effectively and efficiently. There are well tested instruments that can help the improvement of institutional performance including institutional planning, performance contracts, but also partnerships with other HEIs or the private sector. d) Independent quality assurance, accreditation and performance monitoring should be established and strengthened, with the participation of both the academic community and the Government. These processes should involve members of the international academic community, the business community, social partners and civil society. 163 6. EDUCATION STRATEGIC PLANNING, POLICY OPTIONS The Need to Revise the Education Strategic Plan (ESP) 2003 426. The 2003 Education Strategic Plan was established to project the funding and policy requirements of reaching the EFA and MDG sector goals. The policy goals of the ESP were the following: 1. Increase access 2. Improve quality 3. Extend and improve TVET 4. Promote health/ sanitation in institutions 5. Improve planning and management 6. Expand science and technology education 7. Improve academic quality and research 8. Expand pre-school education 9. Promote programs to prevent HIV&AIDS 10. Provide equal opportunities 427. The ESP accordingly helped establish some key milestones, policies, funding and operational frameworks and monitoring arrangements. Five years later, it was clear that the Plan had to be fundamentally revised for the following reasons: 428. Financial gap: From the very beginning, there was a funding gap within the plan. A 2005 funding projection indicated that by 2010, the gap would grow up to 80 percent of the basic education budget. 429. Changing External Factors: Over the last ten years, economic growth has accelerated, thanks to solid governance and positive changes in global markets. This brought about higher state revenues, but also caused inflation and radical increases in the wage bill. The latter affected heavily the sector given that 70 percent of the entire system and over 90 percent of basic education expenditures are used to pay salaries; 40 percent of the entire civil service is teachers. So, both the budget and the payroll have radically expanded especially over the last five years. 430. Competing priorities: Also, from the onset, competing education policy priorities were defined in ESP and in the White Paper, also dated in 2004. The latter has put much stronger emphasis on post-basic education. 431. Emerging new challenges: By 2008, it had become clear that improvements in access had slowed down, learning results shown low quality and there were persistent equity and efficiency problems in the sector, including those that create major obstacle to achieving the goals defined in the ESP. 432. The Government, in strong cooperation with the Development Partners launched in 2009, the new Education Strategic Planning process for 2010-2020. The new ESP was to be completed by the Fall of 2010. It is unclear what the status of the ESP is, to what extent the new targets and benchmarks were endorsed and to what extent budgeting is 164 affected by the plan. However, it is very likely that the plan will provide a framework for monitoring the sector performance for the coming decade. Positive Trends in the ESP 2010-2020 433. The ESP drew up the following lessons (ESP 2010-2020 p.12) a) Improvements in 1st and 2nd cycle enrolments and gender parities, primary completion, transition to JHS and initiatives in complementary education are achievements that must be maintained and built upon. b) The improvements in equity and access (above) now need to be reflected in comparable improvements in delivery of quality education. c) Effective means of measuring and subsequently improving education quality are required - necessitating regular national assessments of literacy and numeracy rates at P6 level, with greater emphasis on qualitative performance reporting. d) Enrolment growth in the 1st cycle has placed a strain on the provision of trained teachers for kindergarten, primary and JH schools, leading in turn to reduced capacity for delivering quality education. e) (…) non-formal education (including various forms of NGO-supported complementary education that have demonstrably improved literacy and enabled out-of-school children to join mainstream classes) is good value for money and should be supported in the ESP. f) Retention of pupils in JHS needs to be improved in the next decade. g) PTRs in JHS are very low, leading to inefficient teaching and high costs – staff rationalization and deployment in this subsector need to be addressed. h) In terms of examination success and skills development, the rate of returns at second cycle level, particularly in SHS, is poor. i) The issues of boarding and student accommodation at 2nd and 3rd cycle levels, with their high levels of public expenditure and corresponding inequities, need to be addressed. j) Escalation in tertiary enrolment has skewed public expenditure (away from basic education), distorted government training targets for science and technology, and contributed to graduate unemployment. k) Official and unofficial leave is expensive and takes people (particularly teachers) away from their place of work, invariably at public expense, at a cost for the delivery of quality in the workplace. l) Decentralization, while an undoubted opportunity, will place strain on management capacity and public resources during the first five years of the ESP. 165 434. There seems to be a strong concurrence between these lessons drawn in the ESP and the findings of the present report, especially in some key areas, including on access where there is a slow-down in aggregate terms, in quality, which needs significant overhaul, in efficiency, relevance -- especially at post-basic level -- and in management. 435. A key advantage of the new ESP is that it recognizes the importance of efficiency gains to finance the long-term development agenda. Accordingly, the ESP targets the following efficiency objectives: a) elimination of inequitable subsidies b) reduction in non-productive staff in District and Regional Offices and in tertiary institutions c) reduction in absenteeism d) effective staff utilization e) effective reporting on use of financial and other resources 436. As part of the effective staff utilization, the ESP recommends increasing the PTRs (with the exception of KG where the PTRs are already extremely high). Table 6-1 below shows the expected changes in PTR by levels: Table 6-1: Expected changes in PTR by levels 2007 2010 2012 2015 2020 (baseline) KG 74 67 63 56 45 Primary 34 37 38 41 45 Junior Sec 24 27 28 31 35 Senior High 19 22 23 26 30 437. The new ESP also intends to strengthen management through gradually decentralizing some areas but also building stronger central institutions, including a National Inspectorate Board and strengthened, multilevel monitoring and evaluation as well as a better EMIS system. Remaining Challenges 438. While it is unquestionable that the ESP is comprehensive and that it addresses some of the most pressing issues in access, quality, efficiency and appropriate institutional/management issues, an effective policy agenda needs to cover some further issues that can create stubborn obstacles for achieving the long-term goals. This report will call the attention to five of these issues: Intrinsic inefficiencies 439. Some of the inefficiencies stem from uneven distribution of student population. The median size of basic schools is low and it is very difficult to provide quality education to all especially in small rural schools in poor communities without some innovative policies and practices. These may include new pedagogical approaches, like 166 multi-grade teaching, decentralizing key decisions on service delivery and resource deployment to local levels, reliance on private and non-formal delivery and creating school clusters and using new technologies in delivery and in teacher training. Such ideas will have to be properly priced and incentivized. Stubborn disparities 440. Despite the commitment by subsequent Governments, resources are not targeted to deprived areas; instead such areas are short-changed as qualified teachers are not deployed effectively and other resources are leaked before they reach the classroom. Even if the Government increases PTR to a level that appears to be sustainable on an aggregate level, the disparities in PTR and in PTTR are so significant that the uneven service delivery dwarfs any efficiency gains. In the last five years, the efficiency of resource allocations has declined in all areas. The Government needs to strengthen its monitoring capacity, policy arsenal and resolve to address these disparities. Social inequalities requiring better targeting 441. Deprivation can be clearly located through educational outcomes and in the geographic mapping of the country, but deprivation is fundamentally originated in underlying social and economic position of families and communities. Better educational services can overcome the social inequalities but only if these services are better targeted. These may include better targeting of school feeding, programmatic schemes to compensate the educational handicap of poor communities (i.e. to assure improved access to KG education) or scaled up version of LEAP. But such measures require much more effective targeting to the poor. Coherent financing 442. While more and more resources are used for education, those who could use the resources effectively are not empowered to do so. Instead, resources remain fragmented, each serving a variety of agendas and clients. Typically, the key source of investment, the GET fund is not linked to the Sector’s long-term strategy nor to the decisions of those who are ultimately responsible for service delivery. Education financing needs to be decentralized at a level, where effective decisions can be made, efficiency gains can be achieved. Institution building for service delivery 443. The ESP recognizes the challenges with decentralization and correctly identifies some of the central capacities to be built. However, ultimately education will be delivered at the level of districts and schools. These levels need to have mechanisms, tools and capacities to allocate resources, supervise staff performance, measure and monitor outcomes and report on the results to the stakeholders. Without these capacities not only decentralization will fail, but also education service delivery will stagnate. Programmatic financing 444. Addressing persistent and emerging new challenges requires well informed and well targeted policies as well as programmatic financing that assigns clearly 167 responsibilities to agencies in charge of service delivery. Such a programmatic financing was adopted by the Ministry of Education in 2008, but this was never implemented during the budget preparation process in any of the subsequent budget years. 168 169 References Abadzi, Helen. 2007. Absenteeism and Beyond: Instructional Time Loss and Consequences. Washington DC: World Bank. Adarkwa, K. K. (2007, September). Matriculation address. Delivered at the 41st Matriculation Ceremony of KNUST. Retrieved March 9, 2009 from, http://www.knust.edu.gh/news/speeches/speech7Speech%20for%2041st%20Matricul ation.pdf Addae-Mensah, I. (2000). Education in Ghana: A tool for social mobility or social stratification? Delivered at the J.B. Danquah Memorial Lectures. Accra: Ghana Academy of Arts and Sciences. Adu, K. & Orivel, F. (2006). Tertiary education financing strategy in Ghana. Report to the National Council for Tertiary Education. Accra, Ghana. Akyeampong, Kwame, Pryor, John and Ampiah, Joseph Ghartey. 2006. A vision of successful schooling: Ghanaian teachers' understandings of learning, teaching and assessment. Comparative Education, 42:2, 155-176. Akyeampong, Kwame et al. 2007. Access to Basic Education in Ghana: The Evidence and the Issues. Country Analytic Report. CREATE. June 2007. Akyeampong, Kwame. 2008. Public Private Partnership in the Provision of Basic Education for Poor and Disadvantaged Groups in Ghana and Rwanda: Possibilities and Constraints. Brighton, UK, University of Sussex, Sussex School of Education, Centre for International Education, unpublished paper. Akyeampong, Kwame. 2009. Access and Educational Equity in Ghana. Brighton: Centre for International Education, University of Sussex, commissioned by the World Bank for this Country Status Report, unpublished. Akyeampong, Kwame. 2009a. Revisiting Free Compulsory Universal Basic Education (FCUBE) in Ghana. Routledge, Comparative Education, 45:2, 175-195. Amaral, A. & Magalhặes, A. (2007). Implementation of higher education policies: A Portuguese example. In A. Gornitzka, M. Kogan & A. Amaral (Eds.), Reform and change in higher education: Analysing policy implementation, pp. 117-134. Dordrecht: Springer. Ampiah, Joseph Ghartey and Adu-Yeboah, Christine. 2009. Mapping the incidence of school dropouts: a case study of communities in Northern Ghana. Comparative Education, 45:2, 219-232. Anyan, James. 2010. Higher Education in Ghana: A Diagnostic Review. Accra, Ghana: commissioned by the World Bank for this Country Status Report, unpublished. Annor, J. 2002. Implementing government policy for supporting technology use by persons with disability. Center on Disabilities Technology and Persons with Disabilities Conference 2002, California State University. Anthony, J. 2009. Access to education for students with autism in Ghana: Implications for EFA. Background paper for EFA Global Monitoring Report 2010. Atchoarena, David and Andre Marcel Delluc. 2001. Revisiting Technical and Vocational Education in Sub-Saharan Africa: An Update on Trends, Innovations, and Challenges. IIEP/Prg.DA/1,320. Paris: International Institute for Educational Planning. 170 Bennell, P. and Akyeampong, K. 2006. Is There a Teacher Motivation Crisis in sub- Saharan Africa and South Asia? Key Findings and Recommendations of an International Research Project, Knowledge and Skills for Development. Brighton, UK, Knowledge and Skills for Development. Brookings Institution. 2007. Review of Trends in Public Spending for Education and Health in Ghana (2002-2006). Transparency and Accountability Project. Bruneforth, M. 2006. Interpreting the distribution of out-of-school children by past and expected future school enrollment. Background paper for EFA Global Monitoring Report 2007. CDD (Center for Democratic Development). 2008. TRACKING LEAKAGE OF PUBLIC RESOURCE IN THE EDUCATION SECTOR: A pilot investigation of teacher absence in public primary schools in Ghana. In collaboration with: the Transparency and Accountability Project/ Brookings Institution. Ghana. Chambas, I. M. (1998). Higher education in the twenty-first century: Vision and action. Volume V- plenary .Ghana. Paper presented at the World Conference on Higher Education. Paris: UNESCO. Daily graphic. 2009. Educational woes. Accra: Daily Graphic Press. 2nd November, 2009, cover article. De Walque, D., 2004: How does the impact of an HIV/AIDS information campaign vary with educational attainment? Evidence from rural Uganda. Working Paper March. Washington DC: World Bank, Development Research Group. Dunne, Máiréad, Leach, F., Chilisa, B., Maundeni, T., Tabulawa, R., Kutor, N., Forde, L. and Asamoah, A. 2005. Gendered School Experiences: The Impact on Retention and Achievement in Botswana and Ghana. Education Series Research Report No. 56. London, DfID. EARC. 2003. Teacher Time on Task. United States Agency for International Development, Grant No. 641 G-00-03-0055 unpublished report. Educational Assessment and Research Center, 35 Dabebu Rd. Osu, Ghana. Education Policy and Data Center. 2008. 2008 series of education projections to 2015 and 2025. Background paper for EFA Global Monitoring Report 2009. EER. 2009. Ghana: Education Expenditure Review; Component of External Review of Public Finance Management (ERPFM). …. Education Management Information System (EMIS). EMIS school census database and administrative WAEC data. Accra: Ministry of Education. Fentiman, A., Hall, A. & Bundy, D. 2001. Health and cultural factors associated with environment in basic education: a study in rural Ghana. Social Science and Medicine, 52, pp. 429-439. Fobih, D., Akyeampong, K. A. and Koomson, A. 1999. Ghana Primary School Development Project: Final Evaluation of Project Performance. Accra: Ministry of Education. Foster, Philip J.1965. Education and Social Change in Ghana. Chicago: University of Chicago Press. Fredriksen, B. 2007. School Grants: One Efficient Instrument to Address Key Barriers to Attaining Education for All. Capacity Development Workshop 'Country Leadership and Implementation for Results in the EFA FTI Partnership', Cape Town, South Africa, 16–19 July. 171 Freedom House. 2009. Freedom in the World 2009. Washington DC: Freedom House. Gershberg, A. I. and Maikish, A. 2008. Targeting education funding to the poor: universal primary education, education decentralization and local level outcomes in Ghana. Background paper for EFA Global Monitoring Report 2009. Ghana Demographic and Health Survey. 2009. Ghana Demographic and Health Survey 2008. Accra, Ghana: Ghana Statistical Service, Ghana Health Service, and ICF Macro. Ghana Education Service. 2008. INSET National Guidelines. Second Edition, GES, Accra, Ghana. Ghana Education Service, 2007. INSET Sourcebook: District Guidelines and Operational Manual for District Level INSET. GES, Accra, Ghana Ghana Education Service and University of Cape Coast (Institute of Education) 2007. Two Year Diploma in Basic Education Brochure. Teacher Education Department, GES, Accra, Ghana Ghana Education Service and Ghana National Association of Teachers. 2000. Conditions and Scheme of Service and the Professional Code of Conduct for Teachers. GESC and GNAT, Accra, Ghana Ghana National Association of Teachers, 2000. Manual on GNAT. GNAT, Accra, Ghana Ghana National Commission on Children (GNCC). 2000. Ghana's children: the child's perspective. Accra: GNCC. Ghana Statistical Service (GSS). 2003. Core Welfare Indicators Questionnaire. Accra: Ghana Statistical Service. Government of Ghana (GoG). 2004. White Paper on the Report of the Education Review Committee. Accra: Government of Ghana. Government of Ghana. 2009. Underserved Areas in Ghana, A study on Teacher Placement. Accra: Government of Ghana. Harttgen, K., Klasen, S. and Misselhorn, M. 2008. Education for All? Measuring pro- poor educational outcomes in developing countries. Background paper for EFA Global Monitoring Report 2009. Hashim, I. 2005. Exploring the Linkages between Children’s Independent Migration and Education: Evidence from Ghana. Working Paper T-12, Brighton: Sussex Centre for Migration Research. Hayford, B. K. (ed.).1988. Social Studies for JSS. Accra: Curriculum Research and Development Division (CRDD). Hedges J. 2000. The importance of posting in becoming a teacher in Ghana MUSTER discussion paper No. 13. University of Sussex, UK JICA, 2008. Final Report on Project to support the operationalization of the In-Service Training Policy. JICA, Accra, Republic of Ghana. J-PAL and IPA. 2009. An [Emergency] Primary School Quality Initiative. Jameel Poverty Action Lab (JPAL) and Innovations for Poverty Action (IPA). Leach, F. 2006. Researching gender violence in schools: methodological and ethical considerations. World Development, Vol. 34, No. 6, pp. 1129–47. Lewin, Keith M. 2009. Access to education in sub-Saharan Africa: patterns, problems and possibilities. Comparative Education: 45:2, 151-174. 172 Manuh, T., S. Gariba & J. Budu (2007). Change & transformation in Ghana’s publicly funded universities: A study of experiences, lessons & opportunities. Oxford: James Curry Ltd., Accra: Woeli Publishing Services. Ministry of Education and Sports/EMIS. 2002. Policy Analysis Brief (1). Accra: Ministry of Education. Ministry of Education, Youth and Sports. 2004. White Paper on the Report of Education Reform Review Committee. Accra: Ministry of Education, Youth and Sports. Ministry of Education, Youth and Sports. 2006. Preliminary Education Sector Performance Report 2006. Accra: Ministry of Education, Youth and Sports. Ministry of Education. 2008. Policy, Budget, Monitoring and Evaluation Unit. March, 2008. Accra: MOESS. Ministry of Education. 2008a. Education Sector Performance Report. July 2008. Accra: MOESS. Ministry of Education. 2008b. Trends in International Mathematics and Science Study: TIMSS 2007 – Ghana Report. Accra: Adwinsa publications. Ministry of Education Science and Sport, 2008. Basic Teaching and Community Practices for the Teacher: A compendium of useful information for the practicing teacher. MOESS, Accra, Ghana. Ministry of Education. 2009. Education Sector Performance Report. Accra: Ministry of Education. Ministry of Education. 2010. Basic Education Comprehensive Assessment System: Report Summary 2009 National Education Assessment Primary 3 and Primary 6, English and Mathematics. Accra: Ministry of Education, Assessment Services Unit, June. Ministry of Education. 2010a. Draft Education Sector Performance Report. Accra: Ministry of Education, unpublished version for internal circulation and development partner discussion. Monk, Courtney, Justin Sandefur, and Francis Teal. 2007. Apprenticeship in Ghana. Centre for the Study of African Economies, Department of Economics, University of Oxford, (processed). Montgomery, M. Kouamé, A and Oliver, R. 1995. The Tradeoff between Number of Children and Child Schooling: Evidence from Cote d'Ivoire and Ghana: Ghana Living Standards Measurement Study. Washington DC: World Bank. Morley, L., F. Leach, R. Luggs, J. Opare, E. Bhalalusesa, L.D. Forde & R. Mwaipopo (2007). Widening participation in higher education in Ghana and Tanzania: Developing an equity scorecard. An ESRC/DfID Poverty Reduction Programme funded Research Project. Murphy, Patrick. 2009. Teachers in Ghana; Summary and Observations. Washington DC: World Bank, unpublished. National Education Reform Implementation Committee. 2007. Education Reform 2007 At a Glance. NERIC Secretariat. Accra, Ghana National Accreditation Board. (2009, February). Accredited tertiary institutions as at February 2009. Retrieved April 27, 2009, from http://www.nab.gov.gh/nabsite/List%20of%20accreditation%20Tertiary%20Instit ution%20as%20at%20February%202009.pdf 173 National Council for Tertiary Education [NCTE]. (2007). 2006/2007 statistical digest of publicly funded universities, IPS and GIJ. Accra: NCTE. NCTE (2008a). Tertiary education policy: Investing in human capital; securing our future. Accra: NCTE. NCTE (2008b). Summary of basic statistics on public tertiary education institutions 2007/2008. Accra: NCTE. NEA (National Education Assessment). 2008. 2007 NATIONAL EDUCATION ASSESSMENT. Accra, Ghana: Ministry Of Education, Science and Sports. Nii Moi Thompson and Leslie Casely-Hayford (RECOUP). 2008. The Financing and Outcomes of Education in Ghana. RECOUP Working Paper Series, Cambridge, UK. Nti, Benjamin. 2007. Report on Monitoring of Teaching and Learning at Debibi Circuit. July 2007. Accra, Ghana. Obeng, C. 2007. Teacher’s views on the teaching of children with disabilities in Ghanaian classrooms. International Journal of Special Education, 22 (1), pp.96-102. Oduro, A. D. 2000. Basic Education in Ghana in the Post-Reform Period. Accra: Centre for Policy Analysis. OECD. 2008. A profile of Immigrant Populations in the 21st Century: Data from OECD Countries. Paris: Organization for Economic Co-operation and Development. Payson Center for International Development. 2008. Second Annual Report. Oversight of Public and Private Initiatives to Eliminate the Worst Forms of Child Labor in the Cocoa Sector in Côte d'Ivoire and Ghana. New Orleans: Tulane University Law School, Payson Center for International Development. Perry C. 2007. MoESS/GES Staff Supply/Demand Forecasting Model, Final Report. Minisry of Education Science and Sports, Accra, Ghana Salmi, J.& Hauptman, A. M. (2006). Innovations in tertiary education financing: A comparative evaluation of allocation mechanisms. Education Working Paper Series, No. 4, Washington D.C: The World Bank. Steffensen, Jesper. 2006. Study on Improving Basic Education through a more Transparent, Equitable and Better Financial and Performance Management: Review of the Funding Flows in Basic Education. Nordic Consulting Group, June. Teacher Education Division, Ghana Education Service. 2000. Teacher Trainees Handbook: 3 year diploma in basic education for teacher training colleges. GES, Accra, Ghana Teal, Francis. 2001. Education, incomes, poverty and inequality in Ghana in the 1990s. Working paper WPS/2001.21. Oxford: CSAE, University of Oxford. te Velde, D. W. et al. 2009. The Global Financial Crisis and Developing Countries: Synthesis of the Findings of 10 Country Case Studies. London, UK Overseas Development Institute, Investment and Growth Program. ODI Working Papers, 306. Thurman, M. S. 2003. A study on Children with Disabilities and Special Education in Ghana: Key Findings. Washington DC: USAID. TIMSS. 2009. Trends in International Mathematics and Science Study. Washington DC: IES, http://nces.ed.gov/timss/index.asp Tooley, James. 2005. Private Schools for the Poor. Education Next, Winter 2005, pp.22- 32. Transparency International. 2009. Corruption Perceptions Index 2009. Berlin: Transparency International Press. 174 Transparency International. 2009a. Africa Education Watch. Berlin: Transparency International Press. UNDP. 2009. Human Development Report (HDR) 2009. New York: United Nations Development Program, in: http://hdr.undp.org/en/statistics/data/ [1.12.2009]. UNESCO. 2006. EFA Global Monitoring Report 2007: Strong foundations Early childhood care and education. Paris: UNESCO Publishing, in: http://unesdoc.unesco.org/images/0014/001477/147794e.pdf [10.08.2009]. UNESCO. 2007. EFA Global Monitoring Report 2008: Education for All by 2015; Will we make it? Paris: UNESCO and Oxford publishers, in: http://www.unesco.org/education/gmr2008/press/Full-report.pdf [11.08.2009]. UNESCO. 2008. EFA Global Monitoring Report 2009; Overcoming Inequality: Why Governance Matters. Paris: UNESCO, http://unesdoc.unesco.org/images/0017/001776/177683e.pdf UNESCO. 2010. EFA Global Monitoring Report; Reaching the Marginalized. Paris: UNESCO; and, Oxford: Oxford University Press. UNESCO-UIS (2009a). Trends in tertiary education: Sub-Saharan Africa. UIS Fact Sheet, July 2009, No.01. Retrieved September 2, 2009 from http://www.uis.unesco.org/template/pdf/ged/2009/Fact_Sheet_2009_SSA.pdf UNESCO-UIS (2009c). UIS statistics in brief: Education in Ghana. Retrieved September 2, 2009, from http://stats.uis.unesco.org/unesco/TableViewer/document.aspx?ReportId=289&IF_La nguage=eng&BR_Country=2880&BR_Region=40540 United Nations Population Division. 2009. World Population Prospects: The 2008 Revision. Population Division of the Department of Economic and Social Affairs of the United Nations Secretariat, http://esa.un.org/unpp. University of Education, Winneba, Ghana. September 06. Ghana: The Country Analytic Review, CREATE Country Analytic Reviews, Sussex, UK University of Ghana (2007). Report of the Visitation Panel to the University of Ghana. Legon: University of Ghana. USAID. 2003. Teacher Time on Task. Conducted by Educational Assessment and Research Center, Ghana. Sponsored by USAID grant no. 641-G-00-03-0055. USAID. 2009. Basic Education Quality in Ghana; Basic Education in Ghana: Progress and Problems. June 2009. Washington DC: USAID. Wodon, Quentin. 2009. Improving the Targeting of Social Programs in Ghana. October 2009. Development Dialogue on Values and Ethics. Washington DC: World Bank. Wodon, Quentin; George Joseph. 2010. School Level Analysis, School Lunch Analysis and NEA Analysis in Ghana. Accra, Ghana: commissioned by the World Bank for this Country Status Report, unpublished. World Bank. 1989. Basic education for self-employment and rural development. Washington, DC: World Bank. World Bank, 2002: Opening Doors; Education and the World Bank. Washington, DC: World Bank. World Bank. 2004. World Bank Project Assessment Report (Report No.: 29581). July 2004. Washington DC: World Bank. World Bank. 2004a. Books, Buildings and Learning Outcomes: An impact evaluation of World Bank support to basic education in Ghana. Washington, DC: World Bank. 175 World Bank. 2008. Ghana: Job Creation and Skills Development – Draft Report. Volume 1. Washington DC: World Bank. World Bank. 2009. Gross national income 2008, Atlas method. Washington DC: World Bank. http://siteresources.worldbank.org/DATASTATISTICS/Resources/GNI.pdf World Bank. 2009a. Ghana job creation and skills development. Volume 1: Main Report. Document of the World Bank. World Bank Ghana Survey Data. 2009. Ghana Survey Data. Chris Reinstadtler. Washington DC: World Bank, unpublished. World Health Organization (WHO). 2008. World health statistics 2008. Geneva: WHO Press. 176 Annex 1: Political and Administrative System Ghana at a Glance Geography. Ghana is a small country located on the coast of West Africa. It borders the Gulf of Guinea to the South, Côte d’Ivoire to the West, Burkina Faso to the North and Togo to the East, three French-speaking countries (see map below). It covers an area of 238,540 square kilometres (92,085 square miles) and its coastline stretches along 540 kilometres. Ghana’s Lake Volta is created by a hydroelectric dam in the East and is the largest manmade lake in the world. Due to its proximity to the Equator, Ghana has a warm and tropical climate, supporting crops such as cocoa, maize, cassava, sorghum, millet, yam and rice. It is also endowed with natural resources such as gold, timber, diamond, bauxite and manganese. Commercial quantities of oil were discovered in 2007 containing up to 3 billion barrels, and explorations and findings continue. People. As of 2009, Ghana had a population of about 24 million people, consisting of over 100 ethnic groups, who speak over 250 languages. Although the vast majority of Ghanaians speak a minimum of one local language, English is the official language and the most common language in administration, schooling and commerce. As census data from 2000 indicates, about 69 percent of Ghanaians practice Christianity, 16 percent Islam and 15 percent traditional beliefs. Politics. Internationally, Ghana has been and continues to be recognized as one of the most peaceful countries in Africa. It was the first sub-Saharan country in colonial Africa to gain independence, declared on March 6th, 1957. In that year Ghana was founded as a sovereign parliamentary democracy, led by Kwame Nkrumah. In subsequent decades, power alternated between civilian and military rule of the government. With the adoption of the 1992 Constitution, the division of powers laid the political framework to help consolidate one of the most stable democracies on the continent. Administration. The administrative structure divides the country into 10 regions and 170 districts. Every district has its own District Assembly. Districts are subdivided into town, area and zonal councils, and unit committees at the lowest level. Accra, the executive, legislative and judicial capital city, is located within the Greater Accra region, and is home to over two million Ghanaians. Having won the 2008 presidential and parliamentary elections, the National Democratic Congress is the current ruling party. Economy. In Ghana, a country rich in natural resources, the economy depends strongly on exports of cocoa, gold, timber, manganese ore, diamonds, aluminium, bauxite, horticulture and tuna. Worldwide, Ghana is one of the leading producers of gold and the second largest cocoa producer. In 2008, the labor force in Ghana constituted about 11.5 million people, the majority of whom were employed in the agricultural sector, representing about 40 percent of GDP. Thus, many Ghanaians are heavily dependent on agriculture and are subsistence farmers. In 2006, more than a quarter of Ghanaians 177 (28.5 percent) lived below the national poverty line. Gross National Income (GNI) per capita reached US$670 in 2008 ($1,430 using PPP), below the sub-Saharan average of US$1,082. While GDP growth in 2008 reached 6.2 percent, Ghana’s GNI totalled US$15.7 billion in that year ($33.4 billion using PPP). In 2008, nearly two thirds of GDP represented imports of goods and services (63.2 percent), resulting in only about one third of GDP from exports (36.8 percent). In Ghana, official development assistance flows constituted US$1.2 billion in 2007 (US$49 per capita), net foreign direct investment averaged US$970 million in 2007, and remittances and employee compensation totalled US$128 million in 2008. The total debt serviced in 2007 was 3.1 percent of all exports of goods, services and income. Political and Administrative System The 1992 Constitution founded the Fourth Republic of Ghana. Since that year the country has undergone two peaceful democratic transitions, in 2000 and 2008. The constitution supports a hybrid presidential regime, with the president elected by an absolute majority. The two major political parties in Ghana are the New Patriotic Party (NPP) and the National Democratic Congress (NDC). As leader of the NDC, President Atta Mills has held office since January of 2009. Internationally, Ghana’s elections are widely recognised as credible and transparent. Based on a number of political indicators, Ghana is internationally classified as a free country (Freedom House 2009). With the high degree of political freedom its citizens enjoy, Ghana is a strong outlier in sub-Saharan Africa. Transparency International’s Corruption Perceptions Index (2009) ranked Ghana 69th out of 180 analyzed countries, thereby ranking better than about 62 percent of the countries examined. However, only 11 percent of seats in Ghana’s parliament are held by women (World Bank). In contrast, in many other sub-Saharan countries the share of women holding parliamentary seats is significantly higher, such as in Mozambique (35 percent), Angola (37 percent) and Rwanda (51 percent) (UNDP 2009). In Ghana, the parliament is elected every four years, there is an independent judiciary, and the media is known to be one of the most vibrant in West Africa. 178 Annex 2: Map Map: Ghana and Bordering Countries, with Main Cities and Regions Source: Ghana, Political Map 1995. All Annex graphs, figures and tables are in the separate document titled “CSR Annex� in the folder: “Ghana Education CSR 2009� Table: Selected Economic and Infrastructure Indicators in Ghana Indicator Value Value (Macro-)Economic GNI per capita, Atlas method (current US$), 2008 670 GNI per capita, PPP (current international $), 2008 1430 GNI, Atlas method (current US$), 2008 15.7 billion GNI, PPP (current international $), 2008 33.4 billion GDP (current US$), 2008 16.1 billion GDP growth (annual %), 2008 6.2 Inflation, GDP deflator (annual %), 2008 18.0 Imports of goods and services (% of GDP), 2008 63.2 Exports of goods and services (% of GDP), 2008 36.8 Merchandise trade (% of GDP), 2008 99.5 Services, etc., value added (% of GDP), 2008 41.7 Official development assistance and official aid (current US$), 2007 1.2 billion Foreign direct investment, net inflows (BoP, current US$), 2007 970,380,000 Workers' remittances and compensation of employees, received (current US$), 2008 128,000,000 179 Time required to start a business (days), 2008 34 Infrastructure Access Roads, paved (% of total roads), 2005 14.9 Electric power consumption (kWh per capita), 2006 311.9 Energy use (kg of oil equivalent per capita), 2006 424.4 Other indicators Renewable internal freshwater resources per capita (cubic meters), 2007 1324.8 Military expenditure (% of GDP), 2008 0.7 Source: World Bank, World Development Indicators database. Note: Value indicated is for the latest year with available data. i The scale of reported improvements between 2007 and 2009 raises questions about the consistency in the methodology of the subsequent tests and/or the reliability of data collection and reporting. 180