Marginalization in education From research to policy Kevin Watkins Senior Fellow Brookings Institution
Why it matters Highly predictable failure rooted in structural disadvantage - clearly remediable injustice linked to poverty, gender, ethnicity, disability, location and wider markers for disadvantage A brake on progress towards the EFA goals on access + learning Intrinsic importance and links from education to wider inequalities and to wider debates
Current approaches OECD data rich longitudinal studies, disaggregated learning outcomes, and school vs. pupil characteristics Education Gini and composite indices (e.g. HDI) years in school World Bank pre-determined circumstance and inequality of opportunity Group-based analysis Overlapping dimensions of marginalization- UNESCO Global Monitoring Report Learning disparities regional assessments/egra
The education poverty threshold (age 17-22) Philippines Turkey Vietnam Egypt Kenya Congo India Nigeria Yemen Nepal Pakistan Morocco Senegal Chad Burkina Faso Share of the population with less than 4 and less than 2 years of education The gender effect: Girls from the poorest households who are in education poverty The wealth effect: People from the poorest households who are in education poverty Education poverty People with less than 4 years of education In Yemen, the poorest 20% of households have an education poverty incidence double the national average Extreme education poverty People with less than 2 years of education And, for girls from the poorest 20% of households, the proportion triples. 100% 90% 80% 70% 60% 50% 40% 30% 20% 10% 0%
Overlapping disadvantages and education poverty Extreme education poverty % with less than 2 years of education (age 17-22) Nigeria, poor, Hausa, girls 8% 17% 25% Kenya, rural, Somali, girls Ghana 20% 31% 57% 73% 84% Pakistan, rural, Sindhi, girls 97% 96%, northern region, rural, girls India, poor, Uttar Pradesh, girls In Kenya, 96% of rural Somali girls (aged 17-22) have less than 2 years of education. The current primary net attendance rate for Somali girls is only 30%. Country average Group average
Average number of years of schooling Education marginalization inequalities within countries 14 12 Ukraine The case of Nigeria 10 8 Cuba Bolivia Indonesia Honduras 9.7 years Richest 20% 10 years Rural Urban 10.3 years Rich, rural boys Rich, urban boys Rich, rural girls Boys Girls 6 Nigeria 6.7 years Cameroon Bangladesh Urban 6.4 years Poor, urban boys 4 2 Education poverty Chad Extreme education poverty Poorest 20% 3.5 years Rural 3.3 years Poor, rural girls 2.6 years 0 C. A. R. Rural Hausa 0.5 years Poor, rural Hausa girls 0.3 years
SACMEQ wealth-based learning disparities 70% 60% 50% Proportion of students scoring at the lowest level (percentage) 52.9% 55.7% 62.6% 40% 33.7% 35.8% 38.9% 30% 27.2% 20% 10% 10.5% 13.4% 0% South Africa Mozambique Zambia National Average Highest 25 % SES Lowest 25% SES Source: SACMEQ 2007
Percentage of Students Ghana: Levels of Achievement in Math (P6) Not Reached Min. Competency Reached Min. Competency Reached Proficiency 100% 90% 80% 70% 60% 50% 40% 30% 20% 10% 0% Total Achievement Source: Ghana National Education Assessment 2012 Male Female Highest (greater Accra) Median (Central) Region Lowest (Northern)
Three core research areas Opportunities and entitlements Accessibility Financing to mitigate disadvantage/incentivize education Links to wider strategies for combating marginalization Household and pupil characteristics Poverty and nutrition Parental literacy Attitudes The learning environment Teacher motivation, competencies and attitudes Infrastructure Curriculum, language, textbooks
Student percentage Factors related to learning outcomes (P3 and P6) in Ghana 100% 90% 80% 70% 60% 50% 40% 30% 20% 10% 0% Reached Minimum Competency Reached Proficiency Facilities (P6)* Textbooks (P3) School registers (P6-Maths) Teacher gender and training (P6) Multi-grade classrooms (P3) *Water, electricity, and and girls' toilets
Pupail Mean ORF Score Nigeria: EGRA Results in Sokoto State 5 4.5 4 3.5 3 2.5 2 1.5 1 0.5 0 Mean Oral Reading Fluency (ORF) Scores by Home Characteristics of Pupils Neither One or No Yes No Yes No Yes more Literate parents Read to at home Quiet place to Eats before school read at home
Kenya s unequal distribution of out- of-school children (47 counties) Share out of school primary aged children 11.0% 10.0% Mandera 9.0% 8.0% 7.0% 6.0% 5.0% 4.0% 3.0% Turkana Wajir Garissa West Pokot Narok 2.0% Samburu Kajiado Marsabit 1.0% Isiolo Lamu Tana River 0.0% 0.0% 1.0% 2.0% 3.0% 4.0% 5.0% 6.0% 7.0% 8.0% 9.0% 10.0% 11.0% Share of primary school age children Source: EMIS/Census 2009
Turkana Wajir Garissa Samburu Mandera Marsabit Tana River West Pokot Isiolo Baringo Narok Kilifi Kwale Kajiado Lamu Laikipia Mombasa Busia Kakamega Migori Trans Nzoia Nandi Homa Bay Kisumu Kitui Meru Uasin Gishu Nakuru Siaya Bungoma Nairobi Taita Taveta Kisii Vihiga Elgeyo Marakwet Nyamira Kericho Tharaka Nithi Bomet Makueni Kiambu Nyandarua Machakos Embu Kirinyaga Nyeri Murang'a Derived share of FPE spending as a proportion of school age population: 47 counties (2009) 1.40 1.20 1.00 0.80 0.60 0.40 0.20 0.00 *Calculated on the basis of school enrollment data and census data on school age population
Research agenda discussion Need for more policy-oriented research on who is being left behind and why. Identifying what works Focused research on specific groups to inform national strategies eg UCW on child labor The role of state and non-state providers Beyond the school linkages to wider sources of marginalization Strengthening disaggregated data and re-coupling quantitative with qualitative analysis