THE PRICE OF EXCLUSION: DISABILITY AND EDUCATION THE CHALLENGE OF INCLUSIVE EDUCATION IN SUB-SAHARAN AFRICA QUENTIN WODON, CHATA MALE, CLAUDIO MONTENEGRO, AND ADA NAYIHOUBA DECEMBER 2018 THE PRICE OF EXCLUSION: DISABILITY AND EDUCATION THE CHALLENGE OF INCLUSIVE EDUCATION IN SUB-SAHARAN AFRICA QUENTIN WODON, CHATA MALE, CLAUDIO MONTENEGRO, AND ADA NAYIHOUBA Awareness of the need for inclusive education systems BACKGROUND TO THIS is increasing. The United Nations Convention on the Rights of Persons with Disabilities has been ratified by SERIES 177 countries. It calls for full integration of persons with disabilities in societies. Several targets in the Sustainable More than one billion people may experience some form Development Goals are related to disabilities, including of disability. Of those, up to one in five may experience with respect to inclusive education. However, at the significant disabilities. Individuals with disabilities have country level, and especially in low and middle-income on average worse socioeconomic outcomes than those countries, resources are often lacking to effectively without disabilities. They often have poorer health, lower promote inclusive education. levels of employment and earnings, and higher poverty rates. In developing countries, the prevalence of disability This note is part of a series on The Price of Exclusion: and its impacts on a wide range of development outcomes Disability and Education prepared as part of broader work are typically larger. program on children with disabilities that benefited from funding from USAID. Previous global work for this series Children with disabilities are especially at a disadvantage in benefited from support from the Global Partnership terms of school enrollment, educational attainment, and for Education. The series documents gaps in education learning. This is especially the case in low income countries outcomes between children with and without disabilities. and sub-Saharan Africa, the region on which this note It also showcases examples of programs and policies and focuses where disability gaps in educational attainment lessons from the literature on how to improve inclusion in are increasing and affordability and other constraints lead education systems. many children with disabilities to never enroll in school or drop out prematurely. They also often learn less while in school. Ensuring that these children have the same opportunities as other children is a challenge, but also an opportunity, in that inclusive education may bring benefits to all children, and not only those with disabilities. 1 | THE PRICE OF EXCLUSION: DISABILITY AND EDUCATION THE CHALLENGE OF INCLUSIVE EDUCATION IN SUB-SAHARAN AFRICA | NOVEMBER 2018 KEY MESSAGES ›› Large gains in secondary completion rates have also been achieved for boys and girls without disabilities, The challenge of achieving inclusive education for sub- but gains are again smaller for children with disabilities, Saharan Africa is massive, but conversely the opportunities leading to disability gaps in completion rates of that inclusive education could provide are major as well. This close to 11 points for boys and 7 points for girls. note provides an analysis of gaps in educational opportunities for children with disabilities in sub-Saharan Africa. Five he last indicator considered for the analysis of trends ›› T main outcomes are considered: whether children ever enroll over time is literacy. In part, as a result of differentiated in school, are literate, complete their primary education, trends in educational attainment by disability status, complete their secondary education, and perform well in the disability gaps for literacy also grew over time, school. In addition, the note measures the coverage of a small reaching 13 points for boys and 12 points for girls. number of programs in schools towards inclusive education, and it measures the returns to investing in the education of hen looking through regression analysis at the ›› W children with disabilities. These returns are high. Therefore, marginal effects of exclusion associated with disabilities, apart from the fact that all children have a basic right to findings are similar to the results from simple statistical education, the economic case for investing in the education comparisons. This suggests that statistical differences of children with disabilities is strong as well. Key findings are in educational outcomes are indeed due to exclusion as follows. related to disabilities as opposed to other (observable) characteristics of children with disabilities. ENROLLMENT IN SCHOOL, EDUCATIONAL ATTAINMENT, AND LITERACY ›› After controlling for other factors affecting educational outcomes, the average reductions at the margin for ›› The Sustainable Development Goals (SDGs) children with disabilities in the probabilities of ever call for ensuring inclusive and quality education enrolling in school, completing primary schooling, for all and promoting lifelong learning (Goal 4). completing secondary schooling, and being literate are The SDGs explicitly mention equal access to all estimated at 8 points, 10 points, 7 points, and 14 points levels of education and vocational training for the respectively (the first three effects are statistically vulnerable, including persons with disabilities. significant, but the fourth for literacy is not). Unfortunately, gaps in education outcomes between children with and without disabilities are large, he effects on education outcomes of exclusion related ›› T and they appear to have increased over time. to disabilities are often larger than the effects of other child or household characteristics. For example, or children without disabilities, completion rates at ›› F the effect of a disability is often larger than that of the primary level have increased substantially over the the quintile of wealth of the child’s household. last few decades. Smaller gains have been observed for children with disabilities. As a result, the gap in PERFORMANCE IN SCHOOL, SCREENING primary completion rates between children with and FOR DISABILITIES, SCHOOL FACILITIES, AND without disabilities has increased over time, reaching 13 TEACHER TRAINING points for boys and 10 points for girls according to the latest publicly available census data for 11 countries. mong children who are in primary school, children ›› A self-reporting hearing or seeing difficulties tend to do any children with disabilities are never enrolled in ›› M worse on standardized mathematics and reading tests in school. Among children aged 12, the likelihood of having all but one of ten countries that participated in the latest ever enrolled in school was 10 percentage points lower PASEC assessment implemented in primary schools. for children with disabilities versus children without ›› The negative impact associated with hearing and seeing disabilities at the time of the latest available census difficulties is confirmed in regression analysis after data. As for primary education, the disability gap in ever controlling for a wide range of other factors that affect enrolling has increased over time. student performance on PASEC. The loss associated with hearing and seeing difficulties is at up to three DECEMBER 2018 | THE PRICE OF EXCLUSION: DISABILITY AND EDUCATION THE CHALLENGE OF INCLUSIVE EDUCATION IN SUB-SAHARAN AFRICA | 2 percent of mean performance, which is larger than the effects of many other variables in the regression analysis. INTRODUCTION ›› PASEC data also suggest that the lack of infrastructure Children with disabilities are at a disadvantage in terms of for pupils with special needs is a reason for children to school enrollment, educational attainment, and learning. drop out. While other reasons related among others to Many never enroll in school or drop out prematurely. cost, poverty, health and nutritional status, and early They also often learn less while in school because of their marriages or pregnancies are mentioned more often disabilities and the lack of inclusive education policies and by teachers as major reasons for children dropping out, programs. In many ways, exclusion from education systems the lack of adequate infrastructure comes up as well. due to disabilities contributes to multidimensional poverty. ›› While slightly more than one in four teacher Children with disabilities should have the same opportunities mentions that medical check-ups for students as other children. To make the case for investments towards are in place in their school, less than five percent inclusive education in sub-Saharan Africa, this note measures mention that students benefit from hearing or gaps in educational outcomes between children with and eye tests. This suggests that disability screening without disabilities, including in terms of whether children among students remains very rare in schools. ever enroll in school, complete their primary education, complete their secondary education, are literate, and ›› PASEC data finally suggest that less than one in ten perform well in school. The note also measures the returns to teachers benefit from in-service training aiming to investing in the education of children with disabilities. These promote inclusive education. Among a dozen categories returns are high. Apart from the fact that all children have a of in-service training, this is the category with the basic right to education, the economic case for investing in lowest coverage rate among teachers across countries. the education of children with disabilities is strong. ›› These estimates are symptomatic of a much larger issue related to the lack of comprehensive programs There are interesting experiences in sub-Saharan Africa on and policies towards inclusive education not only in how to make schools inclusive, including initiatives in the sub-Saharan Africa, but also in the developing world grassroots level. However, country governments, bilateral and and even some developed countries more generally. multilateral development agencies, and other stakeholders Interventions often tend to be piecemeal as opposed to are only starting to place a larger emphasis on ensuring comprehensive, and they often have low coverage rates. access to education for children with disabilities. Reaching marginal groups, and especially children with disabilities, is RETURNS TO EDUCATION FOR INDIVIDUALS often a key priority of Ministries of Education, but this has WITH DISABILITIES often been less the case for children with disabilities than for other groups of vulnerable children. Unfortunately, adequate ›› Overall, the analysis demonstrates that children data and the capacity to analyze existing data to suggest with disabilities are being left behind by efforts to and implement appropriate policies are lacking. To indicate improve education opportunities for all. The rising the need for better data and capacity building related to gaps between children with and without disabilities education for children with disabilities are, findings from a in sub-Saharan Africa call for stronger policies and Global Partnership for Education (GPE, 2018) study are interventions to achieve the target of inclusive education instructive. The analysis was conducted in 2017 and is based adopted under the Sustainable Development Goals. on data available at the time for 51 countries benefitting from GPE investments, most of which are in sub-Saharan Africa. ›› Analysis of earnings data from household surveys For this note, two findings stand out (some progress may suggests that the returns to education for individuals have been achieved since the analysis was conducted): with disabilities are large and similar in magnitude to those observed for individuals without disabilities. ›› Commitment to children with disabilities but lack Investing in the education of children with disabilities of specific strategies and targets: Some 38 GPE is therefore not only the right thing to do, it is developing country partners have signed and ratified the also a smart investment with high returns. Convention on the Rights of Persons with Disabilities 3 | THE PRICE OF EXCLUSION: DISABILITY AND EDUCATION THE CHALLENGE OF INCLUSIVE EDUCATION IN SUB-SAHARAN AFRICA | NOVEMBER 2018 and all countries but three either state or guarantee Washington Group’s Short Set of Questions on Disability (as within their Constitution the right to primary education explained in more details below). As part of its engagement for all children, including those with disabilities. Some in the Summit, the World Bank signed the Charter for 33 countries have a national disability law or policy, but Change along with DfID and other Summit participants. The at the time of drafting the report only three countries global commitments laid out in the Charter reinforce the had an established policy on inclusive education, World Bank Group Commitments on Disability-Inclusive and six countries have drafted one. There is a long Development. There is clear momentum going forward way to go to make inclusive education operational. towards disability inclusive education. The goal should not be to only concentrate on access but to ensure learning of all ack of disability-disaggregated data: Ministries ›› L by supporting the social inclusion of children with disabilities identify the need for robust, reliable data regarding and systemic change. This is also why in this note the focus is the education of children with disabilities as a high not only on access to school and educational attainment, but priority. Today, only 29 countries include an estimated also on learning, and some of the interventions that could be percentage or number of children with disabilities implemented in school towards inclusive education. enrolled at any level in the school system. Only 12 countries have data disaggregated by disability To make the case for investments in the education of children domain (such as mobility, cognition, sight, hearing, with disabilities, a simple theory of change suggests that and communication). The lack of data on disability three components are needed: (1) showing that there is a is the most commonly cited barrier for policy. The need for interventions by documenting gaps in education second-most cited barrier identified in Education outcomes between children with and without disabilities; Sector Plans is a widely held negative attitude toward (2) arguing that better educational outcomes for children people with disabilities. Lack of infrastructure, learning with disabilities can make a major difference not only for material, and strategies on inclusive education, them but also for society as a whole – including in economic as well as the lack of financial resources, the lack term; and (3) demonstrating that successful interventions of inter-ministerial coordination and economic are feasible and affordable. Building on Male and Wodon barriers are also cited as barriers to education. (2017), this note contributes to the first two tasks. Its aim is simple: by showing how large disability gaps in education outcomes remain in sub-Saharan Africa, and by showing that Ministries of Education clearly have a particular stake investments in education for children with disabilities can in ensuring that children with disabilities have access to have high returns, the note contributes to enabling Ministries education since ensuring that children with disabilities are of Education to prepare country diagnostics based on this able to go to school could dramatically improve educational theory of change (country profiles will be prepared separately attainment and learning in low income as well as other apart from this note focusing on the region as a whole). countries. Supporting and providing technical assistance to help countries invest in the education of children with The broader analytical framework that informs this work and disabilities is also a major new priority for the World Bank and the series of notes more generally is provided in Figure 1. At other donors. In July 2018, the World Bank announced ten the bottom are the data sources used for various components corporate commitments at the Global Disability Summit co- of the work. In the middle are the areas of focus of the work, hosted by the United Kingdom Department for International with the aim of generating a series of reports and notes. Development (DfID), the Government of Kenya, and The overall aim of the work is to advocate for investments the International Disability Alliance. These commitments in inclusive education and provide guidance to operational include (1) Ensuring that all World Bank-financed education teams on how to implement appropriate programs and programs and projects are disability-inclusive by 2025; policies. This note is but a first attempt at measuring gaps and (2) Scaling up disability data collection and use, guided and marginal impacts and illustrating some of the potential by global standards and best practices, such as using the benefits from investments. DECEMBER 2018 | THE PRICE OF EXCLUSION: DISABILITY AND EDUCATION THE CHALLENGE OF INCLUSIVE EDUCATION IN SUB-SAHARAN AFRICA | 4 Figure 1: Analytical Framework Advocacy for investments Guidance for operational teams Africa-wide diagnostic reports and notes Measuring gaps and marginal impacts of various Making the case for interventions types of disabilities on outcomes Benefits (lifetime Program Enrollment and earnings/human costs and Learning Gaps capital wealth and potential Attainment Gaps other benefits) impacts Estimation of returns Case studies Student Census Household Other with labor force surveys and surveys Others Assessments Data Surveys Datasets and other impacts using with partners DHS/MICS Source: Authors. The note is structured as follows. The first sections document The analysis is based on censuses, household surveys, and disability gaps in educational attainment and literacy based student assessments (see Box 1 on the advantages and limits on census data. The focus shifts thereafter to gaps in learning of various data sources). While census data especially are for students in school based on student assessment data known to underestimate the extent of disabilities and may (PASEC, which stands for Programme d’analyse des systèmes thereby capture for the most part severe disabilities, they éducatifs de la Confemen) for ten Francophone countries. The are nevertheless useful – given their large size –, including to last part of the note estimates the returns to education for look at trends over time in educational outcomes for children children with disabilities. These returns are similar to those with and without disabilities. Census data are also useful for other children, helping to make the case that investments to measure the impact at the margin of exclusion related in the education of children with disabilities make economic to disabilities on education outcomes for various types of sense apart from being the right thing to do. A brief disabilities, something that is more difficult to do with surveys conclusion follows. given limited sample sizes. 5 | THE PRICE OF EXCLUSION: DISABILITY AND EDUCATION THE CHALLENGE OF INCLUSIVE EDUCATION IN SUB-SAHARAN AFRICA | NOVEMBER 2018 BOX 1: ANALYZING DISABILITY AND EDUCATION OUTCOMES WITH VARIOUS DATA SOURCES This note relies in part on census data. The use of censuses for work on disability and education has some advantages. The number of observations is large even when only a subsample is available for the analysis. This makes it feasible to measure trends over time in education outcomes and the impact at the margin (controlling for other factors) of exclusion related to disabilities. Marginal impacts can be estimated for all children with disabilities as well as by type of disability. By contrast, the sample size of household surveys is often insufficient to perform a similar analysis. But there are also limitations when using census data for such analysis. First, the data tend to underestimate the prevalence of disability. Second, despite efforts to improve questionnaire, the only information that is typically available relates to whether a child has a disability or not, and not whether the disability is severe or mild. The Washington Group on Disability Statistics has developed ways to improve census data on disabilities through six questions related to core functional domains (these questions were not yet available in most of the census data used for this note). The functional domains are: seeing, hearing, learning, walking, cognition (remembering or concentrating), self-care (washing all over or dressing), and communicating. For each question, four responses are suggested – the individual has no difficulty, some difficulty, a lot of difficulty, or no ability at all to perform the task. These questions help measure functionality and thereby detect disabilities that may be less severe or apparent. By contrast, when a single question is asked, the result is typically a substantially lower rate of disability identifying only those with the most severe disabilities in the household as having a disability. Because mot censuses have yet to include these questions, readers should be aware that only severe disabilities tend to appear in census datasets used here. In addition, in most countries censuses are implemented only once every ten years, so the data may not account for the latest developments in education systems. The timing of disabilities is also typically not observed (as is the case for most surveys). Especially for older individuals, the disability may have been observed after the individual has left school. Comparisons of educational attainment for individuals with and without disabilities may be less precise in identifying the role of disabilities in affecting educational attainment for older individuals. This is one of the reasons why the regression analysis in this note is performed on younger age groups than the statistical comparisons provided for all age groups. Because censuses tend to identify severe disabilities that are often observed at or soon after birth, the risk of bias may however not be too large. Despite these limitations, censuses remain a useful source of data for measuring the impact of exclusion related to disabilities on education outcomes by type of disability, especially over time. In addition, for a few countries, the analysis of educational attainment and (subjectively declared) literacy conducted with census data in this note is complemented by additional analysis using household surveys. In addition, the note also relies on household surveys to measure the returns to education for children with disabilities, and on student assessment data for assessing the performance of students with disabilities while in school. The student assessment data is from PASEC and covers 10 Francophone countries. It should be noted that efforts are underway to improve data on children with disabilities. One effort of note is being undertaken by UNICEF, especially for young children. For these children, the Washington Group questions may not be appropriate. A new module under the MICS (Multiple Indicators Cluster Surveys) is being implemented to better measure child functioning. This new module – the Child Functioning Module, covers children between 2 and 17 years of age and assesses functional difficulties in different domains. DECEMBER 2018 | THE PRICE OF EXCLUSION: DISABILITY AND EDUCATION THE CHALLENGE OF INCLUSIVE EDUCATION IN SUB-SAHARAN AFRICA | 6 PRIMARY EDUCATION and avoid jumps in the data when too few observations are available for individuals with disability). For children COMPLETION without disabilities, completion rates at the primary level increased substantially. The gain is at 3.6 percentage points The analysis of the first four educational outcomes is based for boys and 13.3 percentage points for girls over the time for the most part on census data. Figure 2 provides primary span separating the youngest and oldest groups. Girls completion rates for individuals aged 16 to 35 based on have essentially caught up with boys in terms of primary the latest publicly available census data for 11 countries completion. Smaller gains are observed for children with (Burkina Faso, Ethiopia, Ghana, Kenya, Liberia, Mali, Malawi, disabilities, at -0.9 percentage points for boys and 9.5 points Mozambique, South Africa, South Sudan, and Zambia). for girls. For both boys and girls that have disabilities, there Completion rates for children aged 12-15 are not shown seems to be a plateau in terms of the completion rates for because they tend to be too low versus expected lifetime primary school, probably in part because few countries have completion rates. This is because some children enter developed strategies to reach these children. primary school late or repeat grades, which leads them to complete primary school well beyond the normal completion As a result of these trends, the absolute gap in primary age. Note that since the data from the censuses is not the completion rates between children with and without most recent in comparison to data from household surveys disabilities has increased over time. This is visualized in Figure of education management information systems, changes 3. A few decades ago, as fewer children had the opportunity that may have taken place in recent years are not accounted to complete primary school, differences in completion rates for. Yet it is unlikely under current circumstances that such by disability status were at only four to six percentage points changes would reverse the long-term trends observed in the for boys and girls alike. At the time of the latest available census data. censuses, gaps are at 12.8 points for boys and 10.1 points for girls on average across countries. In short, completion rates have improved for children with disabilities, but more slowly Disability gaps in educational than has been the case for children without disabilities, so attainment for older individuals disability gaps in primary completion have widened. could be underestimated. This is because older individuals may have suffered from a disability For children without disabilities, after leaving school. This would completion rates at the tend to reduce the measures primary level have increased of disability gaps observed for substantially. Smaller gains have older individuals. To reduce the been observed for children risk of bias, comparisons are with disabilities. As a result, made until age 35 because until the absolute gap in primary that age, disability rates tend completion rates between not to increase too much. children with and without disabilities has increased over time. Four groups are considered: boys with no disability, girls with no disability, boys with one or more disabilities, and girls with one or more disabilities. The statistics in Figure 1 are average completion rates for individuals of the corresponding age bracket a at the time of the census (three-year moving averages are used to better capture underlying trends 7 | THE PRICE OF EXCLUSION: DISABILITY AND EDUCATION THE CHALLENGE OF INCLUSIVE EDUCATION IN SUB-SAHARAN AFRICA | NOVEMBER 2018 Figure 2: Primary Completion Rates by Age and Group (%) 60 55 Completion Rates by Group (%) 50 45 40 35 30 25 35 34 33 32 31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 Age Men, No Dis. Men, Dis. Women, No Dis. Women, Dis. Source: Authors. Sample: 11 censuses. Figure 3: Disability Gap for Primary Completion (%) 15 14 13 Disability Gap by Sex (%) 12 11 10 9 8 7 6 5 35 34 33 32 31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 Age Men Women Source: Authors. Sample: 11 censuses. EVER ENROLLING IN SCHOOL The gaps in primary school completion between children with enrollment at the primary level. Figures 4 and 5 provide data and without disabilities may be due in part to the fact that on the share of children who ever enrolled in school. Statistics children with disabilities are at a higher risk of dropping out are computed for individuals aged 12 to 35. Statistics for before completing the cycle than children without disabilities. individuals who are younger but in age of primary school are However, many children with disabilities never even get to not shown because some of them could still enroll at a later enroll in school. Indeed, the disadvantages faced by these age. This is less likely for children aged 12 and over if they children start in their early childhood years including for have not yet enrolled by then. DECEMBER 2018 | THE PRICE OF EXCLUSION: DISABILITY AND EDUCATION THE CHALLENGE OF INCLUSIVE EDUCATION IN SUB-SAHARAN AFRICA | 8 Patterns observed in Figures 4 and 5 for ever enrolling in school are similar to those observed in Figures 2 and 3 for Many children with disabilities primary completion. For children without disabilities, the are never enrolled in school. At likelihood to ever enroll in school increased by 12.6 points for boys and 26.6 points for girls over the time separating the time of the latest census the youngest and oldest age groups. Girls caught up with data, among children aged boys, as is the case for primary completion. Smaller gains are 12, the likelihood of having observed over the period for children with disabilities, at 6.5 ever enrolled in school was 10 points for boys and 20.3 points for girls. Even if girls with percentage points lower for disabilities have caught up with boys with disabilities, absolute gaps in the likelihood of ever enrolling in school between children with disabilities versus children with and without disabilities have grown over time. children without disabilities. The This is shown in Figure 5. At the time of the census, the disability gap has increased over absolute gaps associated with disabilities were at 10.4 points time. for boys and 9.8 points for girls. Figure 4: Shares of Children Ever Enrolled by Age and Group (%) 80 75 70 65 Shares by Group (%) 60 55 50 45 40 35 34 33 32 31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 Age Men, No Dis. Men, Dis. Women, No Dis. Women, Dis. Source: Authors. Sample: 11 censuses. Figure 5: Disability Gap for Share of Children Ever Enrolled (%) 11 10 Disability Gap by Sex (%) 9 8 7 6 5 4 3 35 34 33 32 31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 Age Men Women Source: Authors. Sample: 11 censuses. 9 | THE PRICE OF EXCLUSION: DISABILITY AND EDUCATION THE CHALLENGE OF INCLUSIVE EDUCATION IN SUB-SAHARAN AFRICA | NOVEMBER 2018 SECONDARY EDUCATION COMPLETION The analysis is repeated for secondary education completion in Figures 6 and 7. One again, similar patterns emerge, albeit with lower completion rates than for primary education. Large gains for boys and Estimates are provided for individuals aged 20 and above girls without disabilities were to allow for a few additional years versus the typical age of also achieved in secondary on-time completion of 18. For children without disabilities, completion rates, but gains were completion rates at the secondary level increased by 3.9 once again smaller for those percentage points for boys and 7.5 percentage points for girls over the period separating the youngest and oldest groups with disabilities. This has led to (note that since we look at older individuals, this period of disability gaps in completion time is shorter than for primary completion). Girls made a lot rates of 11 points for boys and 7 of progress but still lag slightly behind boys, probably in part points for girls. due to the risks of child marriage and early childbearing in many countries, both of which lead girls to drop out of school at the secondary level (Wodon et al., 2018) The gains for children with disabilities were once again smaller, at 0.7 points for boys and 5.5 points for girls. While girls with disabilities have caught up with boys with disabilities, gaps in secondary completion rates have widened between children with and without disabilities. They reached 11.1 points for boys and 7.0 points for girls aged 20. As for other indicators, children with disabilities did not benefit as much as other children from the large gains in educational attainment achieved by countries. Figure 6: Secondary Completion Rates by Age and Group (%) 40 35 Completion Rate by Group (%) 30 25 20 15 35 34 33 32 31 30 29 28 27 26 25 24 23 22 21 20 Age Men, No Dis. Men, Dis. Women, No Dis. Women, Dis. Source: Authors. Sample: 11 censuses. DECEMBER 2018 | THE PRICE OF EXCLUSION: DISABILITY AND EDUCATION THE CHALLENGE OF INCLUSIVE EDUCATION IN SUB-SAHARAN AFRICA | 10 Figure 7: Disability Gap for Secondary Completion (%) 12 10 Disability Gap by Sex (%) 8 6 4 2 0 35 34 33 32 31 30 29 28 27 26 25 24 23 22 21 20 Age Men Women Source: Authors. Sample: 11 censuses. LITERACY The last indicator considered for the analysis of trends over time is literacy, as declared subjectively by census respondents for their household members. While subjective The last indicator considered perceptions of literacy may not necessarily indicate than an for the analysis of trends over individual is indeed literate as a reading and comprehension time is literacy. In part as a test would, the data are still useful to conduct a tentative assessment of differences in literacy rates between result of differentiated trends individuals with and without disabilities. Figures 8 and 9 in educational attainment by provide the trends in literacy over time by age groups, disability status, the disability considering individuals from 12 to 35 years of age. For gaps for literacy also grew over children without disabilities, the likelihood of literacy time substantially, reaching 13 increased by 3.3 points for boys and 20.6 points for girls over the period separating the youngest and oldest age groups. points for boys and 12 points for As expected, girls caught up with boys. The gains for children girls. with disabilities were small for boys at -3.4 points, while they reached 13.7 points for girls. As shown in Figure 9, the disability gaps grew over time substantially, reaching 13.0 points for boys and 11.9 points for girls. 11 | THE PRICE OF EXCLUSION: DISABILITY AND EDUCATION THE CHALLENGE OF INCLUSIVE EDUCATION IN SUB-SAHARAN AFRICA | NOVEMBER 2018 Figure 8: Literacy Rates by Age and Group (%) 80 75 Completion Rate by Group (%) 70 65 60 55 50 45 40 35 35 34 33 32 31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 Age Men, No Dis. Men, Dis. Women, No Dis. Women, Dis. Source: Authors. Sample: 11 censuses. Figure 9: Disability Gap for Literacy (%) 14 13 12 Disability Gap in Literacy Rate (%) 11 10 9 8 7 6 5 4 35 34 33 32 31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 Age Men Women Source: Authors. Sample: 11 censuses. DECEMBER 2018 | THE PRICE OF EXCLUSION: DISABILITY AND EDUCATION THE CHALLENGE OF INCLUSIVE EDUCATION IN SUB-SAHARAN AFRICA | 12 MARGINAL IMPACTS FOR ENROLLMENT, ATTAINMENT, AND LITERACY The data presented in the previous sections suggest large children with disabilities versus children with no disabilities and growing gaps in educational attainment and literacy among 11 years old. Similar gaps are observed for primary and between children with and without disabilities over time. Table secondary school completion. For literacy, the gaps are even 1 summarizes the key estimates for the youngest appropriate larger, at 14 percentage points for boys and 12 points for girls age cohorts for each of the four indicators. The last row in of the appropriate age group. These are the average values of the Table displays the disability gaps for the various indicators the disability gaps observed across the 1 countries for which for both boys and girls. For example, the probability of never estimations were carried using census data. going to school at all is about 10 percentage points higher for Table 1: Disability Gaps for the Most Recent Appropriate Age Cohorts (%) Sample: 11 censuses. Ever Enrolled Primary Completed Secondary Completed Literacy (11 Years Old) (16 Years Old) (20 Years Old) (12 Years Old) Boys Girls Boys Girls Boys Girls Boys Girls No disability 76.7 75.3 50.0 46.2 36.1 28.6 65.7 65.3 Disability 66.3 65.4 37.2 36.2 25.0 21.6 52.7 53.5 Disability gap 10.3 9.9 12.8 10.1 11.1 7.0 13.0 11.9 Source: Authors’ estimation using census data. Are the current gaps the result of exclusion associated with While census data have limits in terms of the variables that disabilities, or do they result from other characteristics can be used as controls, a number of important variables of children that could be correlated with disabilities? For known to affect educational outcomes are available in the example, are the disability gaps mentioned in Table 1 the data. The regression analysis includes the following variables result of exclusion related to disabilities, or could they be as controls (with minor variations between countries): the sex related to other characteristics of children that could lead of the child; whether the child has a birth certificate; whether to a higher risk of disability, such as extreme poverty or the child has a disability and the type of disability observed stunting in early childhood? To assess the likely impact of (in addition, an interaction effect is included to assess if a exclusion related to disability on educational attainment and disability has a differential impact for boys or girls); whether literacy after controlling for other factors that may affect the child is an orphan on the side of the mother, the father, education outcomes, regression analysis is needed. The or both; whether the child resides in an urban or rural area; term “impact” is used for simplicity (see Box 2 on what is the geographic area in which the child resides (these areas meant by “impact”). The regression analysis for the marginal clearly differ between countries); the mother tongue of impacts of exclusion related to disabilities considers the same the child; the quintile of wealth of the households to which four education outcomes, but with slightly different age the child belongs; the religion of the child; the size of the groups mostly for sample size reasons: (1) whether a child household in which the child resides; the sex of the household ever enrolled in school (the sample for the regression analysis head; a number of characteristics for the household heal consists of children ages 6 to 11 to account for conditions at (age according to 10 years intervals, education level, and the time of the implementation of the census); (2) whether type of work); the same characteristics for the spouse of the a child completes primary education (sample of children 15 household head; and the leave-out-mean of the dependent to 18 years old); (3) whether a child completes secondary variable. This leave-out-mean variable is computed among all education (sample of youth ages 19-22); and (4) whether other children in the area where a child lives and is meant to a child is considered by the parents as literate (sample of capture local conditions that affect education outcomes for children ages 15-18). For each regression, the sample of children in disaggregated areas where children live. children included is a bit older than the normal threshold to complete a level of schooling or be literate to account for children who may start school late or repeat grades. 13 | THE PRICE OF EXCLUSION: DISABILITY AND EDUCATION THE CHALLENGE OF INCLUSIVE EDUCATION IN SUB-SAHARAN AFRICA | NOVEMBER 2018 BOX 2: WHAT DO WE MEAN BY “IMPACTS” OF EXCLUSION RELATED TO DISABILITIES? The aim of this section is to estimate the impact of exclusion associated with a disability on education outcomes for children with disabilities in comparison to children without a disability but with otherwise similar characteristics. The term “impact” is used for simplicity, but one must be careful about not necessarily inferring causality. Estimates of impacts are obtained through regression analysis to control for other variables that may affect education outcomes. What is measured are statistical associations, and not necessarily impacts as could be observed for example with randomized control trials. Since a disability cannot be randomized, we must rely on regression analysis to estimate likely impacts, but there is always a risk of bias in the measures of the impacts on outcomes of exclusion associated with a disability. At the same time, the fact that we observe strong effects that are robust to various specifications does suggest, as expected, that exclusion related to disability is often strong. Table 2 provides a summary of the results for the impacts associated with exclusion related to disabilities. The interpretation of the marginal impacts is in percentage The reductions at the margin points. The marginal impacts in Table 2 are almost always statistically significant as well as large (when a coefficient for children with disabilities is not statistically significant, the value is not provided and in the probabilities of ever instead a notation NS is mentioned in the Table). After enrolling in school, completing controlling for other factors that may affect outcomes, the primary schooling, completing reductions at the margin for children with disabilities in the secondary schooling, and being probabilities of ever enrolling in school, completing primary schooling, completing secondary schooling, and being literate are estimated at 8.3 literate are estimated at 8.3 percentage points (this is the points, 10.2 points, 6.8 points, interpretation of the coefficient value of -0.083 in table and 14.0 points respectively. 2), 10.2 points, 6.8 points, and 14.0 points, respectively, in comparison to children with no disability. These estimates are of a similar order of magnitude to the statistical measures mentioned in the previous section, suggesting that most of the differences between children with and without disabilities are indeed associated with the disability status of the children as opposed to other variables that could be correlated with that status. There are differences in the impacts associated with different types of disabilities. For example, children with cognitive, psychosocial or multiple disabilities often fare worse than children with physical disabilities. This is also shown in Table 2 which provides the marginal impacts in the regression analyses when different types of disabilities are considered separately, as opposed to considering all types of disabilities together. DECEMBER 2018 | THE PRICE OF EXCLUSION: DISABILITY AND EDUCATION THE CHALLENGE OF INCLUSIVE EDUCATION IN SUB-SAHARAN AFRICA | 14 Table 2: Marginal Impacts of Exclusion by Type of Disability on Educational Attainment and Literacy Sample: 11 censuses. Ever Enrolled Primary Completed Secondary Completed Literacy (6-11 Years Old) (15-18 Years Old) (19-22 Years Old) (15-18 Years Old) All types of disabilities combined -0.083 -0.102 -0.068 NS By type of disability Seeing -0.048 NS NS -0.052 Hearing -0.049 -0.102 -0.084 -0.104 Speech -0.177 -0.153 -0.152 -0.266 Mental -0.217 -0.223 -0.111 -0.313 Physical -0.055 -0.101 -0.057 -0.064 Multiple -0.235 -0.168 -0.142 -0.281 Other -0.042 -0.075 -0.098 -0.059 Source: Author’s estimations. Note: All coefficients are statistically significant at the ten percent level except those noted NS for Not Statistically Significant (at the ten percent level). To provide perspective, it is useful to compare the marginal impacts on education outcomes of exclusion related to The marginal effects on disabilities with the impact of other factors affecting those outcomes (this is not shown in Table 2 but based on the full education outcomes of regression results available from the authors). Girls continue exclusion related to disability to have lower educational outcomes than boys, but the are often larger than the effect magnitude of the gender gaps is much smaller than for the of other child or household disability gaps. Being an orphan is also associated with a characteristics. For example, the lower likelihood of completing various levels of schooling as well as being literate, but with smaller effects than exclusion marginal effect of a disability related to disability. There are also wealth effects at work, is often larger than that of with children from better off households more likely to do the quintile of wealth of the well. When comparing children from the lowest quintile of households in which a child wealth to those from the top quintile of wealth, marginal effects can be large, but again, socio-economic differences resides. tend to have smaller effects at the margin that exclusion related to disability. Finally, there are also marginal effects on education outcomes associated with the education level of the household head and the location of the household, but these impacts are smaller again than those associated with disability even if differences between urban and rural areas can be substantial. 15 | THE PRICE OF EXCLUSION: DISABILITY AND EDUCATION THE CHALLENGE OF INCLUSIVE EDUCATION IN SUB-SAHARAN AFRICA | NOVEMBER 2018 STUDENT PERFORMANCE IN SCHOOL Apart from being less likely to go to school, complete their Table 3 provides estimates of the average scores of students primary or secondary education, and being literate, children with and without hearing and seeing difficulties on PASEC with disabilities tend to perform less well on standardized for both mathematics and French. Data are available for the student assessments when they manage to remain in school. ten countries: Benin, Burkina Faso, Burundi, Cameroon, While student assessments in sub-Saharan Africa rarely Chad, Cote d’Ivoire, Congo, Niger, Senegal, and Togo. provide a clear identification of children with disabilities, Note that what is measured is the mastery of core literacy questions in the 2014 PASEC assessment can be used as and numeracy skills, not the mastery of specific aspects of can be used as proxies for disabilities. Students in grade 6 are the curriculum in each country (for more details on what asked whether they have difficulties seeing or hearing while PASEC measures, see Box 3). A score of 500 indicates that in the classroom. These are imperfect proxies for disability a student is performing at the average level in the sample because in large classrooms other factors apart from for the 10 countries. The estimates suggest that there are disabilities (for example being far away from the teacher at systematic differences in test scores for both mathematics the back of a large classroom) may lead to difficulties seeing and reading/language between children with and without or hearing. The share of students declaring having difficulties hearing and seeing difficulties. The average differences hearing or seeing is too large to represent only children with are large for the whole sample with students with hearing/ disabilities, and other types of disabilities are not represented seeing difficulties having substantially lower scores than by these two simple questions. Still, there is a clear link with those without difficulties (Chad is the only exception among disabilities. For example, essentially all children wearing eye the ten countries with a surprising reverse effect). The glasses declare having seeing difficulties. Therefore, in a differences in performance tend to be larger for children with tentative way, these two variables are useful for a tentative hearing than seeing difficulties. Perhaps this may be because assessment of the impact of difficulties/disabilities on learning hearing impairment may in some cases be temporary, outcomes. for example when a child has an ear infection, while visual impairment is typically not temporary (although eye glasses can help). DECEMBER 2018 | THE PRICE OF EXCLUSION: DISABILITY AND EDUCATION THE CHALLENGE OF INCLUSIVE EDUCATION IN SUB-SAHARAN AFRICA | 16 Table 3: Average Student Performance on PASEC in Grade 6 (Test Score – Mean at 500) Sample: 10 PASEC countries Hearing Difficulties Seeing Difficulties No Yes Difference No Yes Difference Mathematics Benin 502.9 464.3 38.6 503.5 473.2 30.3 Burkina Faso 544.2 524.1 20.1 541.2 535.5 5.7 Burundi 597.3 571.4 25.9 598.8 584 14.8 Cameroun 494.3 469.1 25.2 495.3 468.6 26.7 Congo (Republic) 485.0 470.7 14.3 485 474.7 10.3 Cote d'Ivoire 481.0 457.5 23.5 483 455.2 27.8 Niger 409.9 397.7 12.2 409.1 403.4 5.7 Senegal 561.7 476.9 84.8 560.3 507.8 52.5 Chad 449.7 464.3 -14.6 448 472.1 -24.1 Togo 523.2 500.2 23.0 519 527 -8.0 Total 508.7 482.8 25.9 505.9 498.6 7.3 Reading Benin 530.3 486.9 43.4 531.8 493.5 38.3 Burkina Faso 538.4 509.0 29.4 535.4 520.8 14.6 Burundi 528.3 509.9 18.4 530 517.8 12.2 Cameroun 524.1 495.5 28.6 523 501.1 21.9 Congo (Republic) 506.0 504.5 1.5 507.6 495.3 12.3 Cote d'Ivoire 524.8 491.5 33.3 525.8 493.4 32.4 Niger 407.5 397.5 10.0 406.8 401.8 5.0 Senegal 564.8 472.4 92.4 563 506.4 56.6 Chad 433.0 433.1 -0.1 429.6 452.1 -22.5 Togo 499.7 481.8 17.9 495.8 506.1 -10.3 Total 511.3 486.2 25.1 509.7 497.4 12.3 Source: Author’s estimations. Note: NS = Not statistically significant at the ten percent level. 17 | THE PRICE OF EXCLUSION: DISABILITY AND EDUCATION THE CHALLENGE OF INCLUSIVE EDUCATION IN SUB-SAHARAN AFRICA | NOVEMBER 2018 BOX 3: INTERPRETING PASEC SCORES The 2014 PASEC assessment was implemented in grades 2 and 6. The grade 2 test for reading/language assesses three main areas: listening comprehension, familiarization with written language and reading-decoding, and reading comprehension. For mathematics, the grade 2 test measures pupils’ core competencies in two main areas: arithmetic, and geometry, space and measurement. For grade 6, the test aims to evaluate student’s ability to understand, learn and adapt their knowledge to situations encountered in daily life. The language test assesses pupils’ comprehension of informative texts and documents, including the ability to extract information from literary texts. For mathematics, the test considers the ability of students to assimilate concepts and apply them in diverse situations. Scores for both tests are scaled so that the international average is 500 points and the standard deviation is 100 points when weighting all countries equally. This implies that two in three pupils are in a range of 400 points to 600 points, and most students are in the 250 to 750 range. An average score of 500 does not mean however that a student is doing well. As an example, in 2014, more than 70 percent of students did not achieve sufficient language competency in grade 2. For mathematics, almost half of the students in grade 2 did not reach the sufficient mathematics competency threshold. To assess how well students are doing versus how they should be doing, students can be categorized by proficiency levels. Pupils below Level 1 are not able to correctly answer a majority of the most basic test questions; these pupils do not display the competencies measured by the test. Students are then categorized according to three other levels (levels 1 to 3 for mathematics, levels 1 to 4 in reading) with increasing degrees of mastery. A sufficient level of proficiency versus the test’s standards is achieved by students performing at levels 3 and 4 for reading, and levels 2 and 3 for mathematics. Source: CONFEMEN (2015). Another way to look at the data is to group students by levels of proficiency. As noted in Box 3, students below level 1 do not display the competencies measured by the test, while students in levels 1 to 3 for mathematics, and 1 to 4 for reading display increasing levels of competencies. To achieve sufficient proficiency, a student must be at level 3 or 4 in reading, and level 2 or 3 in mathematics. As shown in Table 4, many students do not achieve sufficient levels of proficiency, and students with hearing or seeing difficulties tend to do more poorly in terms of the degree of competency that they display. On average across the 10 countries, a larger share of students with hearing or seeing difficulties tends to be clustered in low proficiency levels as compared to students without hearing or seeing difficulties. DECEMBER 2018 | THE PRICE OF EXCLUSION: DISABILITY AND EDUCATION THE CHALLENGE OF INCLUSIVE EDUCATION IN SUB-SAHARAN AFRICA | 18 Table 4: Share of Students by Proficiency Level in PASEC, 10 Countries, Grade 6 (%) Sample: 10 PASEC countries Hearing Difficulties Seeing Difficulties No Yes Difference No Yes Difference Mathematics Level 0 24.2 30.4 -6.2 24.7 26.8 -2.1 Level 1 31.1 37.6 -6.5 32.2 32.5 -0.3 Level 2 28.9 24.3 4.6 28.1 28.2 -0.1 Level 3 15.9 7.7 8.2 15.0 12.5 2.5 All 100.0 100.0 - 100.0 100.0 - Reading Level 0 6.2 7.6 -1.4 6.7 5.9 0.8 Level 1 19.2 23.7 -4.5 19.7 20.8 -1.1 Level 2 26.8 33.5 -6.7 26.5 33.1 -6.6 Level 3 28.2 25.3 2.9 27.6 27.9 -0.3 Level 4 19.6 9.9 9.7 19.6 12.3 7.3 All 100.0 100.0 - 100.0 100.0 - Source: Author’s estimations. student performance. The impacts are slightly larger for MARGINAL IMPACTS FOR hearing than for seeing difficulties, but they are broadly similar in magnitude for boys and girls and for mathematics TEST SCORES IN SCHOOL and reading within each type of difficulties. As done for educational attainment and literacy, regression To provide some perspective, the impacts in Table 5 are larger analysis can be used to test whether controlling for other in magnitude than the benefits from (i) going to a preschool; factors, hearing and seeing difficulties lead to lower (ii) having textbooks at home; (iii) having a teacher involved performance on PASEC. The analysis controls for a in tutoring; (iv) having small class sizes; and (v) being in a wide range of student, household, teacher and school larger school. Impacts for variables that have impacts of a characteristics that may affect test scores. The results are similar magnitude to those observed for hearing and seeing provided in Table 5. Hearing and seeing difficulties are difficulties include those from (i) having homework to do; associated with substantial reductions in performance even and (ii) the socio-economic quintile of household well-being after controlling for other factors. Recall that the average to which the child belongs. Impacts that are slightly larger score for students is set at 500. Therefore, a reduction of include those from (i) female teachers; (ii) electricity in 15 points is equivalent a loss of about 3 percent versus the schools (which may denote a well-developed area effect); (iii) average score after controlling for other factors affecting urban location; and (iv) private school provision. 19 | THE PRICE OF EXCLUSION: DISABILITY AND EDUCATION THE CHALLENGE OF INCLUSIVE EDUCATION IN SUB-SAHARAN AFRICA | NOVEMBER 2018 Table 5: Marginal Impact of Exclusion from Hearing/Seeing Difficulties on Performance, Grade 6 Sample: 10 PASEC countries Mathematics Reading All Boys Girls All Boys Girls Hearing difficulties -14.4 -14.6 -15.2 -15.2 -15.5 -15.4 Seeing difficulties -9.0 -6.0 -8.9 -10.5 -8.4 -12.0 Source: Author’s estimations. can be provided by teachers, and each teacher can provide TEACHER PERCEPTIONS ON up to five responses. One of the responses relates to the accessibility of the school for children with disabilities, which REASONS FOR DROPPING in most likelihood relates to mobility issues. Tabulations of the most common teacher responses for both grade 2 and grade OUT OF SCHOOL 6 are provided in Table 6. While the lack of infrastructure It should also be noted that exclusion related to disabilities for pupils with special needs is not the most frequently cited may lead some children with disabilities who enrolled in school response (not surprisingly given that most students do not to drop out of school prematurely. One of the questions asked to teachers in PASEC related to their perceptions of have disabilities), it is mentioned quite often as one of the why students are dropping out. A total of eighteen responses reasons leading some students to drop out. Table 6: Teacher Perceptions on Reasons for Students Dropping Out of Primary School (%) Sample: 10 PASEC countries Teacher Responses Teacher Responses in Grade 2 (all countries) in Grade 6 (all countries) Pupil's health and nutritional status 65.7 58.1 Child labor 33.8 30.8 Pupil's lack of motivation for learning 55.0 54.2 Families' lack of interest in school 68.4 68.5 Excessive distance between home and school 32.2 24.0 Lack of school canteen 32.2 30.2 Cost of schooling 23.0 19.2 Household poverty 66.9 70.3 Early marriage 18.5 28.8 Pregnancy 13.4 21.1 Teachers' poor educational mentoring 4.5 3.8 Teachers' lack of motivation 4.2 4.7 Teacher absenteeism 5.8 4.6 School's bad reputation 3.5 3.8 Overcrowded classrooms 8.2 5.8 Lack of security at school 4.7 2.6 Negative school climate 2.7 2.3 Lack of infrastructure for pupils with special needs 16.1 20.6 Source: Author’s estimations. DECEMBER 2018 | THE PRICE OF EXCLUSION: DISABILITY AND EDUCATION THE CHALLENGE OF INCLUSIVE EDUCATION IN SUB-SAHARAN AFRICA | 20 LACK OF PROGRAMS FOR teachers who answer that this was the case for their students. Only slightly more than a fourth of teachers mention the INCLUSIVE EDUCATION existence of medical check-ups in both grades. For hearing and eye tests, the proportions are at less than five percent There is no doubt that achieving inclusive education is a major on average across countries. Without such diagnostics tests, challenge in sub-Saharan Africa, as it is in other regions of it is often very difficult for teachers to be able to support the the world. While this note does not provide by any means learning efforts of children who may have disabilities, because a comprehensive assessment of where the region stands, a the teachers may not know who these children are given that few statistics again from PASEC are useful to illustrate the class sizes of 50 students tend to be more the norm than the challenges that remain. A question is asked in PASEC to exception in many countries. The lack of such tests also does teachers in both grade 2 and grade 6 on whether students not enable the schools to use referral services for hearing aids have benefitted from a medical check-up, a hearing test, or glasses. and an eye test in the school. Table 7 provides the share of Table 7: Share of Teachers Reporting Medical Checkups and Eye/Hearing Tests, Grades 2 and 6 (%) Sample: 10 PASEC countries. Grade 2 Grade 6 Medical Medical Eye Test Hearing Test Eye Test Hearing Test Check-Up Check-Up Benin 5.4 0.9 0.9 4.9 1.6 0.4 Burkina Faso 19.1 5.6 1.8 18.3 5.3 4.0 Burundi 41.4 2.8 1.9 23.2 5.0 4.0 Cameroun 48.1 6.0 1.3 39.6 4.5 2.6 Congo (Republic) 14.5 7.4 4.5 9.6 2.3 0.4 Cote d'Ivoire 29.1 1.8 1.4 38.4 18.0 10.5 Niger 32.0 3.9 0.0 26.1 2.3 1.1 Senegal 50.1 2.2 7.9 50.7 10.3 8.1 Chad 18.6 8.3 5.2 9.9 4.4 1.8 Togo 19.0 3.9 1.0 16.8 2.4 2.3 Total 28.8 3.8 2.3 27.0 7.6 4.7 Source: Authors’ estimation using census data. Another interesting question asked to teachers in both on how they were trained, one of which is training towards grades is whether they have received in-service training inclusive education (disability or other). As shown in Table in various areas. Teachers need both pre-service and in- 8, less than one in ten teachers in both grade 2 and grade service training with respect to inclusive education, but the 6 declares having received such training, and this is the PASEC data provide information only on in-service training. category of training that is the least common among the This information is very instructive about the relatively low eleven types of training identified. Without such training, it priority assigned to inclusive education. Eleven categories is unlikely that teachers will be able to adequately support of training are included in potential responses by teachers children with disabilities. 21 | THE PRICE OF EXCLUSION: DISABILITY AND EDUCATION THE CHALLENGE OF INCLUSIVE EDUCATION IN SUB-SAHARAN AFRICA | NOVEMBER 2018 Table 8: Share of Teachers Reporting Receiving Various Types of Training, Grade 2 and Grade 6 (%) Sample: 10 PASEC countries Teacher Responses Teacher Responses in Grade 2 (all countries) in Grade 6 (all countries) Teaching of English 84.0 82.1 Teaching of mathematics 70.2 72.9 Teaching of other subjects (history, science, ICT, etc.) 56.2 62.1 Skills-based approach 61.2 71.5 Situation-based learning approach 18.6 23.8 Multi-grade class management 17.9 22.9 Large group-centered teaching 22.5 23.7 Adaptive dimension 16.2 17.2 Teaching by objectives 37.4 36.3 Child-friendly school 10.3 11.4 Inclusive education (disability or other) 8.2 7.6 Source: Author’s estimations. The data provided in Tables 7 and 8 are symptomatic of of children with disabilities; (iv) the availability of appropriate a much larger issue related to the lack of comprehensive human resources to support inclusive education, covering programs and policies towards inclusive education not only in children with disabilities; (v) the existence of curricula and sub-Saharan Africa, but also in the developing world and even support from teachers and school administrators that are some developed countries more generally. While discussing inclusive of children with disabilities; and (vi) the existence what such programs and policies should entail is beyond of an EMIS that is inclusive of children with disabilities. the scope of this note, some guidance is available from a More generally, IIEP-UNESCO and UNICEF are currently diagnostic tool that UNICEF (2018) developed for service working on a framework for inclusive education that delivery and system strengthening. The tool considers among identifies multiple areas of service delivery as well as enabling others whether the following conditions are being met: (i) factors that should be considered to achieve inclusive the existence of appropriate law/policy explicitly mentioning education. But clearly, the simple statistics provided in tables the rights of children with disabilities to receive an education; 7 and 8 suggest that there is ample scope for improving the (ii) the existence of a physical environment in schools that is coverage of programs and interventions in this area (as well as appropriate for children with disabilities; (iii); the availability the quality of those programs and interventions). of materials and communication that support the inclusion DECEMBER 2018 | THE PRICE OF EXCLUSION: DISABILITY AND EDUCATION THE CHALLENGE OF INCLUSIVE EDUCATION IN SUB-SAHARAN AFRICA | 22 LABOR MARKET GAINS FROM EDUCATION The analysis so far suggests that disability gaps in educational To test for robustness, two types of models are estimated attainment, literacy, and performance on standardized tests (OLS and Heckman selection model). Models are estimated are large, and that schools often do not have programs in for the whole sample of workers with earnings, and then only place to detect disabilities, or train teachers to be able to for those workers with a disability to be able to compare the promote inclusive education. Investing in the education returns to education for both samples. A number of other of children with disabilities is the right thing to do from a tests have been performed that suggest that the results rights perspective, but it is also a smart investment. Indeed, tend to be fairly robust to the estimation approach. Table 9 apart from a wide range of other benefits from educational provides the main results for four countries: Burkina Faso, attainment, the labor market returns to education for The Gambia, Rwanda and Senegal [additional countries to be individuals with disabilities are typically large and of a similar added for final version]. Having a disability is associated with order of magnitude to the returns observed for other lower earnings after controlling for education, experience, individuals. There is a large body of literature on the potential sex, and location. In Burkina Faso for example, workers with impact of educational attainment on earnings. The benefits a disability earn on average 16 percent to 18 percent less are typically measured through regression analysis whereby than other workers with similar characteristics (this is the the potential effect on earnings of educational attainment interpretation of the coefficients -0.163 in the first column and experience is estimated. In some models, the focus is the with the OLS model and -0.186 with the Heckman model). correlation between years of schooling and earnings, and the But for both the full sample of workers with earnings and implicit gain associated with each additional year of schooling. for workers with disabilities, the marginal gains in earnings Other models look at the potential impact on earnings associated with higher educational attainment tend to be large. of different levels of schooling, such as having a primary, secondary, or tertiary education. For example, again in Burkina Faso, workers with primary education tend to make 50 percent more than those with no Apart from educational attainment, whether measured education at all, and the gain is similar when looking only at through years of schooling or in levels, the models may also workers with a disability. When testing formally for equality control for other variables that may affect earnings. Apart in the gains from educational attainment between the from education and the level of experience of the workers two groups, in most cases the tests suggest no statistically (and its squared value), the models used here include a significant differences. This pattern is stable across countries, dummy variable to identify workers with disabilities when with the exception of The Gambia where the gains from estimations are conducted for all workers together, as higher educational attainment for workers with disabilities are well as variables to capture the sex of the worker and his/ often not statistically significant and, in some cases, negative, her location. Models with education in levels typically possibly in part due to small sample sizes that lead to less provide better estimates of the gains from various levels of robust coefficients. Overall though, the analysis suggests educational attainment than those based on the number that the economic benefits from investing in the education of of years of education, so levels of educational attainment children with disabilities are likely to be large when they reach are used here. Based on standardized variables available in adulthood, as is the case for investments in the education of the World Bank’s I2D2 database, all individuals with some children without a disability. primary education or primary education completed but no education at a higher level are combined in a single category for primary education. The same is done for secondary and tertiary education. In other words, no distinction is made as to whether individuals have completed or not a specific cycle of study. 23 | THE PRICE OF EXCLUSION: DISABILITY AND EDUCATION THE CHALLENGE OF INCLUSIVE EDUCATION IN SUB-SAHARAN AFRICA | NOVEMBER 2018 Table 9: Marginal Gain in Earnings from Higher Educational Attainment and Impact of Disability (%) Sample: Four Countries Full Sample with Workers with Full Sample with Workers with Earnings Disabilities Earnings Disabilities OLS Heckman OLS Heckman OLS Heckman OLS Heckman Burkina Faso The Gambia Primary 0.516 0.520 0.425 0.556 0.543 NS -0.877 NS Secondary 0.803 0.995 1.340 1.318 0.714 NS -0.671 -0.588 Tertiary 1.108 1.150 1.466 1.549 1.239 0.724 - - Disability -0.163 -0.186 NA NA -0.315 NS NA NA Rewanda The Senegal Primary 0.382 0.348 0.468 0.442 0.098 NS 0.266 NS Secondary 1.429 1.362 1.587 1.606 0.454 0.573 0.621 0.954 Tertiary 2.627 2.550 2.273 2.270 1.323 1.412 1.328 1.574 Disability -0.166 -0.159 NA NA -0.064 NS NA NA Source: Authors’ estimation. Note: NA = Not applicable. NS = Not statistically significant at the ten percent level. DECEMBER 2018 | THE PRICE OF EXCLUSION: DISABILITY AND EDUCATION THE CHALLENGE OF INCLUSIVE EDUCATION IN SUB-SAHARAN AFRICA | 24 CONCLUSION In essence, children with disabilities are being left behind by efforts to improve education opportunities for all children As noted in the introduction, to make the case for in sub-Saharan Africa. Questions in student assessments investments in the education of children with disabilities, a about school policies related to inclusive education suggest simple theory of change suggests that three components that little is done currently for such inclusion. This represents may be needed: (1) showing that there is a need for a potentially large loss for countries because analysis of interventions by documenting gaps in education outcomes earnings in household surveys shows that the returns to the between children with and without disabilities; (2) arguing education for children with disabilities are as large as those that better educational outcomes for children with disabilities for children without disabilities. There is a clear economic can make a major difference not only for them but also for rationale in addition to a rights-based approach for investing society as a whole – including in economic term; and (3) in the education of children with disabilities in sub-Saharan demonstrating that successful interventions are feasible and Africa as elsewhere. affordable. This note contributes to the first two components of this theory of change. The note has provided a simple The Sustainable Development Goals call for building and analysis of education outcomes for children with and without upgrading education facilities that are child, disability and disabilities using a range of datasets. Children with disabilities gender sensitive and also provide safe, inclusive and effective have been shown to be at a substantial disadvantage for learning environments. While progress continues to be ever enrolling in school, completing primary or secondary achieved to reduce gender and wealth gaps in education, education, or being literate. The gaps between children progress is much too slow for children with disabilities. with and without disabilities have been increasing over time. Ensuring that children with disabilities in sub-Saharan Africa In addition, students with hearing or seeing difficulties in have opportunities to go to school and learn while in school the classroom tend to perform less well when in school in should be a top priority to end the persistent crisis of so comparison to students without such difficulties. Finally, many of these children remaining out-of-school or dropping the lack of adequate school infrastructure for children with out prematurely, and not learning nearly enough while in disabilities is an important reason for children dropping out of school, with often dramatic consequences for their prospects primary school according to responses provided by teachers later in life. on the factors that lead to such drop-outs. 25 | THE PRICE OF EXCLUSION: DISABILITY AND EDUCATION THE CHALLENGE OF INCLUSIVE EDUCATION IN SUB-SAHARAN AFRICA | NOVEMBER 2018 REFERENCES CONFEMEN. 2015. PASEC 2014: Education System Performance in Francophone sub-Saharan Africa - Competencies and Learning Factors in Primary Education. Dakar: Conférence des ministres de l’Éducation des États et gouvernements de la Francophonie. Global Partnership for Education (GPE). 2018. Disability and Inclusive Education - A Stocktake of Education Sector Plans and GPE-funded Grants. Washington, DC: Global Partnership for Education. Male, C., and Q. Wodon. 2017. Disability Gaps in Educational Attainment and Literacy. The Price of Exclusion: Disability and Education Notes Series. Washington, DC: The World Bank. UNICEF. 2018. 2018-2021 Strategic Plan: Goal Area 2 SMQ Guidance for Country Offices. New York: UNICEF. Wodon, Q., C. Montenegro, H. Nguyen, and A. Onagoruwa. 2018. Missed Opportunities: The High Cost of Not Educating Girls. The Cost of Not Educating Girls Notes Series. Washington, DC: The World Bank. Recommended citation for this note: Wodon, Q., C, Male, C. Montenegro, and A. Nayihouba (2018). The Challenge of Inclusive Education in sub-Saharan Africa, The Price of Exclusion: Disability and Education Notes Series, Washington, DC: The World Bank. The authors are with the World Bank Group. This series of notes was launched by the Education Global Practices at the World Bank in collaboration with the Social, Urban, Rural, and Resilience Global Practice which houses the World Bank’s disability inclusion team. Comments on this note from Hanna Katriina Alasuutari, Charlotte Vuyiswa McClain-Nhlapo and Ruchi Kulbir Singh are gratefully acknowledged. The series is part of a broader work program that benefitted initially from the support of the Global Partnership for Education. This specific publication benefitted from a grant from the Disability-inclusive Education in Africa Program Trust Fund and was made possible through support provided by the Office of Sustainable Development, Bureau for Africa, U.S. Agency for International Development, under the terms of Award No. AID-AFR-IO-17-0001. The opinions expressed in this publication are those of the author(s) and do not necessarily reflect the views of the U.S. Agency for International Development. The findings, interpretations, and conclusions expressed in this note are entirely those of the authors and should not be attributed in any manner to the World Bank, its affiliated organizations or to members of its Board of Executive Directors or the countries they represent. The World Bank does not guarantee the accuracy of the data included in this work. Information and illustrations contained in this report may be freely reproduced, published or otherwise used for noncommercial purposes without permission from the World Bank. However, the World Bank requests that the original study be cited as the source. © 2018 The World Bank, Washington, DC 20433. DECEMBER 2018 | THE PRICE OF EXCLUSION: DISABILITY AND EDUCATION THE CHALLENGE OF INCLUSIVE EDUCATION IN SUB-SAHARAN AFRICA | 26 THE PRICE OF EXCLUSION: DISABILITY AND EDUCATION THE CHALLENGE OF INCLUSIVE EDUCATION IN SUB-SAHARAN AFRICA