Global OOSCreport Full Web - 217 PDF

FIXING THE BROKEN PROMISE
OF EDUCATION FOR ALL

Findings from the Global Initiative
on Out-of-School Children
FIXING THE BROKEN PROMISE
OF EDUCATION FOR ALL
Findings from the Global Initiative
on Out-of-School Children
UNESCO Institute for Statistics
The UNESCO Institute for Statistics (UIS) is the statistical office of UNESCO and is the UN depository for global statistics in
the fields of education, science and technology, culture, and communication. The Institute is the official source of data used
to monitor progress towards Education for All and related targets of the Millennium Development Goals.
The UIS education database is the most comprehensive in the world covering a wide range of indicators—from girls’ enrolment
in primary school to the mobility of university students. It is updated three times each year based on results of the UIS annual
education survey, which is conducted in more than 200 countries and territories. Established in 1999, the Institute serves
Member States and the UN system, as well as inter-governmental and non-governmental organizations, research institutes,
universities and citizens interested in high-quality data. Our central goal is to improve the opportunities and living conditions of
children and adults around the world by producing the data needed for effective policies and interventions.
United Nations Children’s Fund (UNICEF)

UNICEF promotes the rights and wellbeing of every child, in everything we do. Together with our partners, we work in 190
countries and territories to translate that commitment into practical action, focusing special effort on reaching the most
vulnerable and excluded children, to the benefit of all children, everywhere.
UNICEF is a vital part of international efforts to achieve the Millennium Development Goal of universal primary education
by 2015. UNICEF provides policy and programming expertise on the ground for the Out-of-School Children Initiative. It is
involved in advancing education in countries around the world and is a global leader in advocating for equity in education
and improvements in learning outcomes. Among its contributions is serving as the secretariat for the United Nations Girls’
Education Initiative (UNGEI). UNICEF is also on the forefront of promoting inclusion of Education for All goals in the post-2015
development agenda.
Published in 2015 by:
UNESCO Institute for Statistics

P.O. Box 6128, Succursale Centre-Ville
Montreal, Quebec H3C 3J7
Canada
Tel: +1 514 343 6880

Email: [email protected]
http://www.uis.unesco.org
© UNESCO-UIS 2015
ISBN: 978-92-9189-161-0
Ref: UIS/2015/ED/SD/7
DOI: http://dx.doi.org/10.15220/978-92-9189-161-0-en
Typesetting: Em Dash Design / www.emdashdesign.ca

Photo credits: Frontcover: © International Labour Organization (ILO)/Crozet M. Interior: Shutterstock (p. 6); © ILO/Ferry Latief
2012 (p. 12); © Ollivier Girard/Center for International Forestry Research (CIFOR) (p. 16); © International Labour Organization
(ILO) (p. 38); © ILO/Huynh Ha (p. 88); © DVIDSHUB/flickr (p. 106); © Adam Cohn/flickr (p. 120); © ILO/Luu Van Tien (p. 130)
This publication is available in Open Access under the Attribution-ShareAlike 3.0 IGO (CC-BY-SA 3.0 IGO) license (http://creativecommons.org/
licenses/by-sa/3.0/igo/). By using the content of this publication, the users accept to be bound by the terms of use of the UNESCO Open Access
Repository (http://www.unesco.org/open-access/terms-use-ccbysa-en).
The designations employed and the presentation of material throughout this publication do not imply the expression of any opinion whatsoever on the
part of UNESCO concerning the legal status of any country, territory, city or area or of its authorities or concerning the delimitation of its frontiers or
boundaries.
The ideas and opinions expressed in this publication are those of the authors; they are not necessarily those of UNESCO and do not commit the
Organization.
Download the data in this report

This report uses DataLinks, which allow you to download Excel files with the data presented in most of the figures and statistical
tables. In the electronic version of this report, click on the link beside the DataLink icon. For the print copy, type the link starting with
http://dx.doi.org into your Internet browser.
This report should be cited as:
UNESCO Institute for Statistics (UIS) and UNICEF (2015). Fixing the Broken Promise of Education for All: Findings from the Global
Initiative on Out-of-School Children. Montreal: UIS. http://dx.doi.org/10.15220/978-92-9189-161-0-en
Table of contents
Foreword �� 7
Acknowledgements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
Abbreviations. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10
Chapter 1. Introduction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
Chapter 2. Data on out-of-school children and adolescents . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17

Sheena Bell and Friedrich Huebler (UIS)
2.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
2.2 Latest data on out-of-school children and adolescents and trends since 2000. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18
2.3 The individual and household characteristics of out-of-school children of primary and
lower secondary school age. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27
2.4 How to improve the accuracy of estimates of out-of-school children . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30
2.5 How to better identify the characteristics of children out of school. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35
Chapter 3. Barriers and policy solutions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39

Angela Hawke (Independent), Sheena Bell (UIS), Hiroyuki Hattori (UNICEF), Friedrich Huebler (UIS), Albert
Motivans (UIS) and Mark Waltham (UNICEF)
3.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39
3.2 System-wide barriers, system-wide solutions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40
3.3 One-half of the world’s out-of-school children: Children caught up in conflict. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45
3.4 Why gender matters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 56
3.5 Child labourers and school participation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 65
3.6 The language barrier. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 70
3.7 Excluded and uncounted: Children with disabilities. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 76
Chapter 4. Financing needs for out-of-school children. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 89

Milan Thomas and Nicholas Burnett (Results for Development Institute)
4.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 89
4.2 An evolving understanding of financing needs for universal basic education. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 90
4.3 A model for estimating financing needs focused on out-of-school children and adolescents. . . . . . . . . . . . . 91
4.4 Illustration: An equity-based approach to assessing the cost of enrolling out-of-school children in the
Democratic Republic of the Congo. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 93
4.5 Innovations in costing and simulating targeted interventions for out-of-school children . . . . . . . . . . . . . . . . . . . 98
4.6 Illustration: Comparing the impact of targeted interventions to enrol out-of-school girls in the
Balochistan province, Pakistan. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 99
4.7 Reaching the marginalised may cost more, but better data and innovative tools can help
us spend smarter . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 99
Findings from the Global Initiative on Out-of-School Children 3

Chapter 5. Conclusions and recommendations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 101
5.1 The broken promise of education. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 101
5.2 Fixing the broken promise. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 102
References. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 107
Appendix I. UNICEF regional classification. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 121
Appendix II. Definitions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 123
Appendix III. Reader’s guide. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 127
Appendix IV. Statistical tables. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 131
List of boxes
Box 1.1 The Global Initiative on Out-of-School Children (OOSCI) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14
Box 1.2 Explore the data. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15
Box 2.1 When is a child considered to be ‘out of school’?. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21
Box 2.2 Schooling trajectories in Pakistan. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31
Box 2.3 Selecting and calculating estimates on out-of-school children in India . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34
Box 3.1 Follow the money: The funding gap. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48
Box 3.2 A lost generation? The children of the Syrian Arab Republic. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49
Box 3.3 The economic argument. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 58
Box 3.4 Measuring the impact of child marriage on education . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 61
Box 3.5 Out-of-school boys and boys who fall behind. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 62
Box 3.6 Defining child labour. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 66
Box 3.7 Grade 4 Portuguese lesson for Changana speakers, Mozambique. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 71
Box 3.8 Education in the United Nations Convention on the Rights of Persons with Disabilities . . . . . . . . . . . . . . . 75
Box 3.9 Impairments and disabilities . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 78
Box 3.10 Towards inclusive education in Serbia. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 84
Box 3.11 A way forward on the data. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 85
Box 3.12 Making the invisible child visible in Cambodia. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 86
Box 4.1 Formulae for estimating the cost of enrolling out-of-school children . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 92
List of tables
Table 2.1 Out-of-school children of primary school age, 2000 and 2012. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20
Table 2.2 Out-of-school adolescents of lower secondary school age, 2000 and 2012 . . . . . . . . . . . . . . . . . . . . . . . . . . . 25
Table 2.3 Selected countries with a large difference between the male and female out-of-school rates,
latest data available, 2010-2013. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29
Table 3.1 Effective implementation of multilingual education (MLE). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 74
Table 3.2 Approaches to schooling children with disabilities. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 79
Table 3.3 A checklist for the effective implementation of inclusive education. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 82
Table 4.1 Out-of-school children and adolescents of primary and lower secondary age in the
Democratic Republic of the Congo, 2012. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 94
Table 4.2 Public Expansion cost in the Democratic Republic of the Congo . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 95
Table 4.3 Household Expansion cost in the Democratic Republic of the Congo. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 95
Table 4.4 Examples of Targeted Intervention costs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 96
4 Fixing the Broken Promise of Education for All

Table 4.5 Total estimated annual cost of enrolling 1.3 million out-of-school children and adolescents
in the Democratic Republic of the Congo. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 97
Table 4.6 Comparison of a system-wide and targeted intervention to improve the enrolment of girls,
Balochistan province, Pakistan . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 99
List of figures
Figure 2.1 Global out-of-school rate for children of primary and lower secondary school age, 2000-2012 . . . . . 18
Figure 2.2 Out-of-school children of primary school age by region and sex, 2000-2012 . . . . . . . . . . . . . . . . . . . . . . . . . . 19
Figure 2.3 Out-of-school adolescents of lower secondary school age by region and sex, 2000-2012 . . . . . . . . . . 20
Figure 2.4 Selected countries with more than 0.5 million out-of-school children of primary school age,
by region, 2012 or most recent year available . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22
Figure 2.5 School exposure of out-of-school children of primary school age by region, 2012 . . . . . . . . . . . . . . . . . . . . 23
Figure 2.6 Distribution of children who leave school before completing primary education,
by age group, selected countries, 2007-2012. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25
Figure 2.7 Evolution of the primary and lower secondary school-age population from 2000 to 2012,
by region. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27
Figure 2.8 Out-of-school rate by region, age group and sex, 2012. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28
Figure 2.9 Percentage of children of primary school age in school, by sex, location, household wealth
and education of the household head, various years . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30
Figure 2.10 Educational attainment of adolescents of upper secondary school age (13-16 years) in
the richest and poorest household wealth quintiles of Pakistan, 2012-2013. . . . . . . . . . . . . . . . . . . . . . . . . . . 31
Figure 3.1 Gaps in data, analysis and policymaking on out-of-school children. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 44
Figure 3.2 Percentage of primary school-age children out of school, both sexes, 2012. . . . . . . . . . . . . . . . . . . . . . . . . . . 46
Figure 3.3 Consolidated appeal requests and funding for education received by conflict-affected
countries, 2013 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48
Figure 3.4 Percentage of primary and lower secondary school-age children in school, by household
wealth quintile, location and sex . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 57
Figure 3.5 Children involved in child labour and its impact on school attendance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 67

Foreword
Education represents the hopes, dreams and aspirations of children, families, communities and nations around
the world—the most reliable route out of poverty and a critical pathway towards healthier, more productive
citizens and stronger societies. Not surprisingly, when people are asked to list their priorities, education tops
survey after survey, poll after poll.
There is consensus at virtually every level, from the poorest family in the most remote village to the global policy
leaders who are shaping the world’s future development goals: education matters. This consensus has been
translated into concrete action, propelling millions of children once denied an education into the classroom.
In the 15 years since the launch of the Millennium Development Goals—which set the target for every child to
complete a full course of primary education by 2015—the latest data show that the number of primary school-
age out-of-school children has dropped by 42%, and for girls by 47%, despite rapid population growth.
Why, then, are there still 58 million children, roughly between the ages of 6 and 11, out of school globally? Each
and every one of these children is a stark reminder of the broken promise to achieve universal primary education
by the original deadline of 2015.
Fixing the Broken Promise of Education for All, a report produced by the UNESCO Institute for Statistics and
UNICEF, could not be more timely. As the international community renews its commitment to advance every
child’s right to education, it explores why global progress has stalled since the early 2000s, when millions of
additional children poured into the world’s classrooms, and provides the data and analysis needed to move
forward and reach every child excluded from education.
With its rich combination of data and analysis, this report provides a nuanced assessment of why some children
never make it into the classroom at all, why some children start going to school far later than others, and why
some children are more likely than their peers to drop out before they complete their schooling. It reminds us—if
any reminder were needed—of the critical need for good data to inform the educational policies that can reduce
the barriers that continue to stand between children and their fundamental right to an education.
This report sets out some of those policies and strategies. They include a deeper focus on improving the quality
of education so that children will be more likely to go to school and stay in school if the education on offer is
fit for purpose. And, given the alarmingly high number of adolescents out of school—63 million worldwide in
2012—it advocates for universal secondary education, drawing from and building on the lessons learned since
2000 on universal primary education.
Finally, this report shows the children behind the numbers. The boy who pushes a cart each day in a Kyrgyzstan
bazaar to help feed his family. The girl pulled out of school in Yemen and married off against her will when still a
child. The child in Sri Lanka, humiliated at school for lacking proper shoes, who drops out altogether rather than

be demoted to a lower grade. The Namibian child with an undiagnosed hearing impairment who struggles at
school. The Syrian refugee child turned away from one over-burdened school after another.
As the international community renews and expands its commitments as part of the post-2015 development
agenda, we must focus on these children, and the millions of others struggling to realise their right to an
education—and to fulfil their dreams for a better future. By working together and promoting greater investment,
we can and must dismantle the barriers that stand in their way, one by one—and in doing so, deliver on our
global promise of education for every child.
Irina Bokova Anthony Lake

UNESCO Director-General UNICEF Executive Director

Acknowledgements
This report would not have been possible without the commitment and efforts of the countries which
participated in the Global Initiative on Out-of-School Children. It draws upon numerous national studies prepared
as part of the Initiative, each one of which provides much-needed information on out-of-school children at the
national and sub-national levels. The report team would like to thank the governments of these countries and all
research partners on the ground for their support throughout this venture, as well as colleagues in UNICEF and
the UNESCO Institute for Statistics (UIS) regional and country offices worldwide, who provided vital support and
encouragement.
We are grateful to the following experts for submitting invaluable background papers and other contributions
used for this report: Carol Benson (Independent) on language issues; Natasha Graham (Imperial College)
on the situation of children with disabilities; Lorenzo Guarcello and Furio Rosati (Understanding Children’s
Work) on child labour; Hiroyuki Hattori (UNICEF) on data analysis; Patricia Justino (Institute of Development
Studies) on children in conflict; Janet Lennox (UNICEF) on barriers and policies in Latin America; Robert Prouty
(Independent) on system-wide policies; Shailendra Sigdel (UIS) on data in India; Nelly Stromquist (University of
Maryland) on gender; Annababette Wils (Independent) on Simulations for Equity in Education; Quentin Wodon
(World Bank) on data analysis; and Mari Yasunaga (UNESCO) on non-formal education.
We would also like to thank the following peer reviewers for their careful revision of the report:
External reviewers: Manos Antoninis (EFA Global Monitoring Report) and Frank van Cappelle (Independent)
UNICEF: Jo Bourne, Claudia Cappa, Joost Kooijmans, Changu Mannathoko, Sreerupa Mitra, Francesca Moneti,
Nicole Petrowski, Aarti Saihjee and Morgan Strecker
UIS: Alison Kennedy, Elise Legault, Patrick Montjourides, Pascale Ratovondrahona and Shailendra Sigdel
REPORT TEAM
Research and writing: Sheena Bell, Friedrich Huebler, Albert Motivans (UIS), Hiroyuki Hattori, Mark Waltham
(UNICEF) and Angela Hawke (Independent)
Editor: Angela Hawke (Independent)
Report coordinator: Sheena Bell (UIS)
Production and promotion: Katja Frostell, Amy Otchet (UIS) and Rudina Vojvoda (UNICEF)

Abbreviations
AIDS Acquired immunodeficiency syndrome

BEPS Balochistan Education Sector Plan
CT Cash transfer
CEE/CIS Central and Eastern Europe and the Commonwealth of Independent States
CDD Centre for Disability in Development
CPIA Country Policy and Institutional Assessment
CULP Centre for Unfolding Learning Potentials
CRPD Convention on the Rights of Persons with Disabilities
DFID Department for International Development
DHS Demographic and Health Survey
DDR Disarmament, demobilization and reintegration
EADE-RDC L’enquête nationale sur la situation des enfants et adolescents en dehors de l’école en République
Démocratique du Congo
ECD Early childhood development
EFA Education for All
EPDC Education and Policy Data Center
EMIS Education Management Information System
FGM/C Female genital mutilation/cutting
GDP Gross domestic product
GPE Global Partnership for Education
HIV Human immunodeficiency virus
ICF International Classification of Functioning, Disability and Health
ICF-CY International Classification of Functioning, Disability and Health for Children and Youth
ILO International Labour Organization
IPEC International Programme on the Elimination of Child Labour
ISCED International Standard Classification of Education
MICS Multiple Indicator Cluster Survey
MOE Ministry of Education
MOEYS Ministry of Education, Youth and Sport
MEPSP Ministry of Primary, Secondary and Professional Education
MLE Multilingual education
NGO Non-governmental organization
NSSO National Sample Survey Office
NFE Non-formal education
NFE-MIS Non-Formal Education Management Information Systems
OCHA United Nations Office for the Coordination of Humanitarian Affairs
OOSCI Global Initiative on Out-of-School Children
RALS Rapid Assessment of Learning Spaces

SECOPE Service de contrôle et de la paie des enseignants
SIMPOC Statistical Information and Monitoring Programme on Child Labour
SEE Simulations for Equity in Education
SRI-IMRB Social and Rural Research Institute of IMRB International
UCW Understanding Children’s Work
UIS UNESCO Institute for Statistics
UN United Nations
UNESCO United Nations Educational, Scientific and Cultural Organization
UNHCR United Nations High Commissioner for Refugees
UNICEF United Nations Children’s Fund
UNRWA United Nations Relief and Works Agency for Palestine Refugees in the Near East
WHO World Health Organization

Chapter 1
Introduction
Since 2000, the progress made on access to As a result, the promise made to children in 2000—
primary education—a fundamental human right— that they would all be able to complete a full course
has been nothing short of remarkable. Spurred of primary schooling by 2015—has been broken.
by the Millennium Development Goals and the
Education for All (EFA) goals, governments worldwide There are also alarming gaps in the enrolment of
have expanded their education systems, built children of lower secondary school age (typically
more schools and deployed more teachers—often between 12 and 15 years). Lower secondary
abolishing school fees at the same time—in an education, considered compulsory in most countries,
attempt to ensure that all children complete primary is crucial to further develop the foundational skills
education. As a result, the number of out-of-school needed for decent work and a productive life. Yet
children of primary school age fell by 42% between 63 million young adolescents were out of school
2000 and 2012. worldwide in 2012. Although the numbers in South
Asia have fallen by nearly one-third since 2000,
This is a notable achievement for the developing the region still has the largest population of out-of-
world. It is not, however, any justification for school adolescents at 26 million. Another 22 million
complacency. Despite the progress that has been adolescents are out of school in sub-Saharan Africa
made, 58 million children of primary school age and their numbers will likely grow (UIS and EFA GMR,
(typically between 6 and 11 years) are out of school 2014a).
worldwide (UIS and EFA GMR, 2014a). If current
trends continue, around 43% of these children—or This report delves into a rich new body of data and
15 million girls and 10 million boys—will probably analysis from the Global Initiative on Out-of-School
never set foot in a classroom. Most of the 30 million Children, which confirms that the task of achieving
out-of-school children in sub-Saharan Africa will education for all is far from over (see Box 1.1). The
never go to school if current trends continue. government-backed national studies carried out
under the Initiative have marshalled a wide range of
The progress made has not been equitable: it is the data sources for innovative analyses, revealing crucial
most disadvantaged children who are still left behind. information on the magnitude of the problem, and on
What’s more, progress has stalled: while access to who the out-of-school children are and where they
education expanded considerably at the beginning live. The studies have used the data as a cornerstone
of the 2000s, there has been little or no change in the to identify context-appropriate policies to overcome
global number of out-of-school children since 2007. the specific barriers to education in their country. For
The global primary out-of-school rate has stagnated many countries, participation in the Global Initiative
at around 9% for the past seven years. has provided an unparalleled opportunity to bring

Box 1.1 The Global Initiative on Out-of-School Children (OOSCI)
The Global Initiative on Out-of-School Children (OOSCI) was launched in 2010 by UNICEF and the UNESCO
Institute for Statistics (UIS) to help participating countries develop evidence-based strategies to reduce
the number of out-of-school children and adolescents. OOSCI looks beyond the goal of universal primary
education and examines exclusion at the pre-primary, primary and lower secondary levels of education. The
Initiative works closely with national and local governments, as well as civil society partners, to focus on
three core objectives:
■■ Develop detailed profiles of out-of-school children and children in school who are at risk of dropping out;
■■ Assess the underlying barriers that prevent those children from completing basic education; and
■■ Recommend innovative policies and strategies that can bring them into school and keep them there.
Twenty-six countries participated in OOSCI in its first phase, and many more governments have joined since.
Participating countries produce in-depth studies that focus on the data, barriers and policies for children
excluded from education. These studies span the pre-primary, primary and lower secondary school levels to
include children who are out of school and those at risk of dropping out. The approach further distinguishes
between out-of-school children who have been to school but dropped out, those whose entry to school is
likely to be delayed, and those who are unlikely to ever attend.
OOSCI studies examine the data and provide concrete recommendations tailored to the barriers to education
that are of most relevance to the local context. This evidence-based and equity-focused approach enables
governments to make targeted changes in their policies and strategies to eliminate these barriers and
increase the number of children in school. By providing much-needed evidence and recommendations on
out-of-school data and policy, OOSCI aims to build political commitment and action to generate a real and
sustained decrease in the number of out-of-school children and adolescents worldwide.
together experts in statistics and policy to shine in combination, over time, with their cumulative
a light on excluded children, who remain largely impact delaying or curtailing children’s education,
voiceless and invisible in government interventions. or preventing their enrolment entirely.
This report draws on this experience to reveal—and The report highlights five important barriers to
attempt to fill—crucial gaps on data, analysis and education and the children affected. First, one-half
policy, aiming to revitalise the momentum on reaching of the world’s out-of-school children live in conflict-
out-of-school children at a critical moment, as new affected countries. Second, entrenched gender roles
international development goals and targets are continue to influence whether or not a child starts
being set. and stays in school. Third, a household’s reliance on
child labour often competes with that family’s hopes
Drawing on the data and analysis of national and for education. Fourth, too many children are side-
regional OOSCI studies, this report provides a lined by education that is delivered in a language
nuanced assessment of system-wide barriers that they neither speak nor understand. And finally, the
keep children out of the classroom. Overcoming considerable barriers that prevent children with
these impediments, which relate to the availability, disabilities from claiming their right to an education
affordability and quality of schools, is crucial to the are only reinforced by a lack of data on their numbers
achievement of education for all. Time and time and their needs.
again, poverty and rural location have been linked
to persistent obstacles to education, despite the The report takes us beyond ‘one size fits all’ solutions
fact that a multitude of countries have built more to these barriers by making a clear distinction
schools and abolished school fees. The problems between two types of countries: those that face
arise from the way in which various barriers work an increasingly narrow set of challenges to achieve

Box 1.2 Explore the data
The hardest to reach children are still out of school. They are poor, rural and often girls. But the situation
is different in every country. The UIS interactive data explorer illustrates the multiple and overlapping
barriers to education in the countries that participated in the Global Initiative on Out-of-School Children.
Learn more about the circumstances that unfairly exclude these invisible and voiceless children.
http://on.unesco.org/oosci-global
More data are available in the UNESCO eAtlas on Out-of-School Children: http://on.unesco.org/oosc-map
education for all and that must, therefore, focus of the children most likely to be excluded, which are
more intensely on interventions targeted towards analysed in greater detail in Chapter 3. Chapter 2
their ‘hard-to-reach’ children; and countries that still also highlights the challenge presented by the rapid
account for a disproportionate percentage of the growth of the school-age population in sub-Saharan
world’s out-of-school children, where system-wide Africa. Finally, this chapter describes some of the
reforms are urgently needed. The report argues that challenges in accurately measuring which children,
most countries must adopt a mixture of system-wide and how many, are in and out of school. It offers
and targeted responses if they are to guarantee recommendations to improve data, showcasing
universal basic education—that is completion of both efforts by the Global Initiative participant countries,
primary and lower secondary education. such as India and the Democratic Republic of the
Congo.
To put it simply, ‘build it and they will come’ optimism
will not pull the world’s 121 million out-of-school Chapter 3 begins with an analysis of system-wide
children and adolescents into education without barriers and responses to out-of-school children,
mobilising policymakers to enact the specific before exploring targeted responses to the obstacles
interventions to address the specific barriers that to education faced by five main groups: children
they face. caught up in conflict, girls (and in some cases, boys),
child labourers, children who do not speak the
The report argues for a new and stronger political language of instruction, and children with disabilities.
commitment to the education of every child, backed In each case, the report attempts to summarise the
by the necessary resources, both human and supply- and demand-side barriers to their education
financial. As the report shows, there is no doubt that and the possible policy responses.
governments and practitioners will have to find far
greater resources to ensure that all children are in the The costs of universal primary education—and
classroom and learning. But it is a worthy investment, ways to assess them in any given country—are
given the long-term benefits for the social and outlined in Chapter 4. Finally, Chapter 5 provides a
economic well-being of every nation. summary of the report’s conclusions and sets out key
recommendations for policymakers.
ABOUT THIS REPORT
Chapter 2 of this report presents the latest global

and regional data on out-of-school children and
adolescents. It analyses trends over time to highlight
different dimensions of this challenge for children
who may never enter school at all, children who enter
school later than their peers, and children who drop
out. The data reveal crucial information on the profiles

Chapter 2
Data on out-of-school children and

adolescents
2.1 INTRODUCTION
Progress on the rate and number
The data are clear: despite substantial gains in of out-of-school children has
school enrolment over the past 15 years, the world stalled since 2007
has missed the goal of universal primary education
by 2015 and there has been virtually no progress 9% of primary school-age children
in reducing the global rate and number of out-of- and 17% of adolescents of lower
school children since 2007. It is increasingly apparent secondary school age are excluded
that business-as-usual approaches have failed to from education
reach 58 million children of primary school age who
continue to be denied their right to education.
Children excluded from education often face
While primary education has long been viewed multiple and overlapping disadvantages, as outlined
as essential for a child’s full development, lower in Chapter 3. If we are to reach them, we need a
secondary education is also increasingly recognised more complete picture of who they are, where they
as the foundation for the acquisition of the skills are and why they are out of school. The evidence
needed for a healthy and productive life and base must draw on a wide range of data sources:
access to decent work. There are now 63 million gathering information about households and
adolescents of lower secondary school age who are schools, and from parents, teachers and children
out of school—5 million more than children of primary themselves. This chapter presents the most recent
school age, even though there are twice as many data from the UIS on the school participation of
primary school-age children worldwide. children and adolescents of primary and lower
secondary school age in order to take stock of
This chapter shows that the stagnation seen in recent global progress since 2000. Such comparative data
years is, in part, the consequence of rapid population are important because they alert the international
growth in sub-Saharan Africa. In most regions, the community to worrying trends, reinforce calls to
school-age population has fallen or remained stable stop the abuse of children’s right to education,
since 2000. In sub-Saharan Africa, however, the support the monitoring of development challenges
soaring school-age population makes it more difficult and cases of exclusion, and provide the basis
to reduce the number of out-of-school children and for requests for international aid. The chapter
adolescents. Nevertheless, countries in the region also describes some of the challenges in
have managed to enrol millions of additional children accurately measuring the number of children in
in primary and lower secondary education over the and out of school and discusses how the data
past two decades. could be improved.

2.2 LATEST DATA ON OUT-OF-SCHOOL 2007, with the primary out-of-school rate stuck at
CHILDREN AND ADOLESCENTS AND around 9% and the lower secondary out-of-school
TRENDS SINCE 2000 rate at around 17%. The gap between the out-of-
school rates of girls and boys has narrowed steadily
Exclusion from education in 2012 since 2000, but even this trend has slowed in recent
mm 58 million children of primary school age are out years.
of school. Of these children:
>> 23% attended school in the past but left; The initial decrease and subsequent stagnation of
>> 34% are likely to enter school in the future; the out-of-school rate is reflected in the evolution of
and the number of out-of-school children between 2000
>> 43% are likely to never enter school. and 2012 (see Figure 2.2). The number of out-of-
mm 63 million adolescents of lower secondary school children of primary school age has hovered
school age are out of school. just below the 60 million mark since 2007. There
has even been a slight rise in the number of out-
of-school children since 2010, but it is too early to
As well as missing the goal of universal primary conclude whether this is the start of a true reversal
education by 2015, the world is far from delivering of the previous steady decline since 2000 or just a
universal lower secondary education. As shown temporary increase. The number of out-of-school
in Figure 2.1, out-of-school rates for children of adolescents of lower secondary school age remains
primary and lower secondary school age fell between on a declining trend overall, but there were still
2000 and 2007, but this progress has stalled since 63 million of them in 2012 (see Figure 2.3).
Figure 2.1 Global out-of-school rate for children of primary and lower secondary school age, 2000-2012
%
30
25
Lower secondary out-of-school rate

20
Female
TOTAL
Male
15
10 Female
Primary out-of-school rate TOTAL
Male
0
2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012
Source: UNESCO Institute for Statistics, August 2014 DataLink: http://dx.doi.org/10.15220/2014/ed/sd/7/f2.1

Out-of-school children of primary school age Independent States (CEE/CIS) (0.6 million). By contrast,
the number of out-of-school children increased slightly
Table 2.1 presents a breakdown of the rate and between 2000 and 2012 in Latin America and the
number of out-of-school children of primary school Caribbean and in Western Europe, North America and
age (typically between 6 and 11 years old) by Australasia, by a combined total of 1.1 million.
region and sex in 2000 and 2012. The numbers
demonstrate considerable progress in the expansion A closer look at national data in the UIS database
of access to primary education. In 2000, 100 million shows that much of the global progress since 2000
children of primary school age, 15% of the children has been driven by a small number of countries
in this age group, were out of school. By 2012, that (see the UNESCO eAtlas of Out-of-School Children
number had fallen by 42 million. at http://on.unesco.org/oosc-map). In India alone,
the number of out-of-school children decreased by
The biggest decrease in the number of out-of-school nearly 16 million between 2000 and 2011, the latest
children was seen in South Asia, where their numbers year with data for that country. Pakistan and the
fell by 23 million between 2000 and 2012 (see also Islamic Republic of Iran have managed to reduce
Figure 2.2). There were also decreases in Eastern their numbers of out-of-school children by 3.4
and Southern Africa (8.3 million), Middle East and million and 1.2 million, respectively, since 2000. The
North Africa (4.1 million), East Asia and the Pacific number of out-of-school children fell by 0.5 million to
(4.1 million), West and Central Africa (3.0 million) and 1 million in eight countries: Algeria, Burundi, Ghana,
Central and Eastern Europe and the Commonwealth of Morocco, Mozambique, Nepal, Yemen and Zambia.
Figure 2.2 Out-of-school children of primary school age by region and sex, 2000-2012
Sub−Saharan Africa South Asia Rest of world
World in 2000
99.7 million
100
12.8 million
80 10.5 million
World in 2012
21.6 million Female 57.8 million
Out-of-school children
60
Male 7.6 million

11.1 million
Female 7.7 million
40 4.8 million
Male
5.1 million
23.7 million
Female
18.2 million
20
20.0 million
Male 14.5 million
0
2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012
Note: The data refer to the regional classification used by UNICEF.


Figure 2.3 Out-of-school adolescents of lower secondary school age by region and sex, 2000-2012
Sub−Saharan Africa South Asia Rest of world

World in 2000
96.9 million
100
18.0 million
80
World in 2012
19.6 million Female
Out-of-school adolescents
62.9 million
60 7.2 million
Male
7.4 million
21.3 million
Female 12.6 million
40
16.1 million Male 13.7 million
20
12.4 million Female 11.8 million
9.5 million Male 10.2 million

0
2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012
Note: The data refer to the regional classification used by UNICEF.

Table 2.1 Out-of-school children of primary school age, 2000 and 2012
2000 2012
% Number (in millions) % Number (in millions)

Region MF M F MF M F MF M F MF M F
W. EUROPE/N. AM./AUSTRALASIA 1.9 1.9 1.9 1.2 0.6 0.6 3.5 3.6 3.4 2.2 1.2 1.1
LATIN AMERICA AND CARIBBEAN 6.2 5.0 7.3 3.6 1.5 2.1 6.3 6.5 6.1 3.8 2.0 1.8
CEE/CIS 6.6 5.4 7.8 1.6 0.7 0.9 4.7 4.7 4.8 1.0 0.5 0.5
EAST ASIA AND PACIFIC 5.4 5.2 5.7 11.0 5.4 5.5 4.6 4.6 4.6 6.9 3.6 3.2
SOUTH ASIA 20.1 13.1 27.6 32.7 11.1 21.6 5.8 5.7 5.9 9.8 5.1 4.8
MIDDLE EAST AND NORTH AFRICA 18.2 14.7 21.8 8.4 3.5 4.9 9.3 7.6 11.1 4.3 1.8 2.5
EASTERN AND SOUTHERN AFRICA 35.1 33.2 37.0 19.3 9.2 10.2 15.1 13.6 16.6 11.0 5.0 6.0
WEST AND CENTRAL AFRICA 43.3 37.6 49.2 21.9 9.6 12.2 27.1 23.1 31.2 18.8 8.1 10.7
WORLD 15.0 12.1 17.9 99.7 41.6 58.1 8.9 8.1 9.7 57.8 27.3 30.5
Notes: The data refer to the regional classification used by UNICEF. The category ‘Western Europe, North America and Australasia’ is not an official UNICEF
region, but it is used in this report to group all countries not belonging to other UNICEF regions. It includes countries in which UNICEF does not operate. They
are primarily high- and upper-middle-income countries located in Australasia, Europe and North America. The list of countries is available in Annex I.
Source: UNESCO Institute for Statistics, August 2014

The combined decreases from these 11 countries
account for more than one-half of the global 31 million of the 58 million
decrease in the number of out-of-school children— primary school-age children out
nearly 26 million—since 2000. of school are girls
In relative terms, 42 countries with data were able

to more than halve their numbers of primary school- school age worldwide—8% of all boys and 10%
age out-of-school children between 2000 and 2012, of all girls—were still out of school in 2012. The
including Algeria, Burundi, Cambodia, El Salvador, majority, 31 million of the 58 million out-of-school
Ghana, Guatemala, India, the Islamic Republic of children, were girls.
Iran, Lao People’s Democratic Republic, Morocco,
Mozambique, Nepal, Nicaragua, Rwanda, Viet Nam, One-third of all out-of-school children of primary
Yemen and Zambia, all of which had more than school age lived in West and Central Africa, the
100,000 out-of-school children at the turn of the region with the highest out-of-school rate. Here,
millennium. more than one in four children (31% of all girls and
23% of all boys) were not in school, far more than
However, despite such impressive progress in in any other region. In Eastern and Southern Africa
many countries, about 9% of all children of primary and in South Asia, another 11 million and 10 million
Box 2.1 When is a child considered to be ‘out of school’?
International statistics on out-of-school children from the UIS are used to monitor EFA and related
international goals. Indicator estimates are based on enrolment data from administrative records,
collected by the UIS from more than 200 countries and territories through its annual survey on education
statistics. Any children of primary or lower secondary school age who are not enrolled in primary or
secondary education are considered to be out of school. This includes a small number of children in
pre-primary education and in non-formal education (NFE).1 Children of primary school age who are
enrolled in pre-primary education are counted as out of school, because the educational content of
pre-primary education and the pedagogical qualifications of its teaching staff are not equivalent to the
standards required for primary education. Children in NFE programmes are also considered to be out of
school, because the nature of these programmes is not, in general, equivalent to that of formal primary
and lower secondary education.
Household surveys, such as the Demographic and Health Surveys (DHS) and the Multiple Indicator Cluster
Surveys (MICS), collect data on school attendance rather than enrolment.2 In these surveys, children who
did not attend school at any time during the reference school year are considered to be out of school.
Household survey data on attendance complement administrative records on enrolment and provide
important information on the characteristics of out-of-school children and their households that cannot
be obtained from enrolment data in the UIS database (see Section 2.3).
National and regional studies conducted as part of the Global Initiative on Out-of-School Children by
UNICEF and the UIS use data on enrolment and attendance from both administrative and household
survey sources. To ensure cross-national comparability of the data, national education programmes are
classified in accordance with the International Standard Classification of Education (ISCED).3
1 The International Standard Classification of Education (ISCED) 2011 defines non-formal education as “education that is institutionalised, intentional
and planned by an education provider. The defining characteristic of non-formal education is that it is an addition, alternative and/or complement to
formal education within the process of the lifelong learning of individuals. It is often provided to guarantee the right of access to education for all. It
caters to people of all ages but does not necessarily apply a continuous pathway-structure; it may be short in duration and/or low-intensity, and it is
typically provided in the form of short courses, workshops or seminars. Non-formal education mostly leads to qualifications that are not recognised
as formal or equivalent to formal qualifications by the relevant national or sub-national education authorities or to no qualifications at all. Non-formal
education can cover programmes contributing to education for out-of-school children and adult and youth literacy, as well as programmes on life
skills, work skills, and social or cultural development” (UIS, 2012a).
2 More information on these survey programmes is available at http://dhsprogram.com and http://www.childinfo.org/mics.html
3 The most recent version of ISCED was adopted in 2011 (UIS, 2012a), but the out-of-school studies published so far as part of the UNICEF-UIS
Initiative are based on data classified in accordance with ISCED 1997 (UIS, 2006).

Figure 2.4 Selected countries with more than 0.5 million out-of-school children of primary
school age by region, 2012 or most recent year available
LATIN AMERICA
AND THE CARIBBEAN Colombia 0.6
EAST ASIA
AND THE PACIFIC
Indonesia 1.3
Bangladesh 0.6
SOUTH ASIA India 1.4
Pakistan 5.4
MIDDLE EAST AND Yemen 0.5

NORTH AFRICA Sudan 2.8
Angola 0.5
Eritrea 0.5
EASTERN AND South Africa 0.7
SOUTHERN Uganda 0.7
AFRICA
Mozambique 0.7
South Sudan 1.0
Ghana 0.5
Mali 0.6
WEST AND Chad 0.8
CENTRAL Burkina Faso 0.9
AFRICA Niger 1.0
Nigeria 8.7
0 1 2 3 4 5 6 7 8 9
In millions
Notes: Data for Ghana refer to 2013; data for Angola, Chad, India, South Sudan, Sudan and Uganda refer to 2011; data for Bangladesh
and Nigeria refer to 2010.
children, respectively, were out of school. In relative estimate of the number of out-of-school children
terms, however, South Asia fares much better than for the war-torn Democratic Republic of the Congo
the sub-Saharan regions because 94% of its primary is 4.9 million back in 1999. For Ethiopia, the latest
school-age children are in school, compared to 85% available estimate refers to 2006, when 3.9 million
of children in Eastern and Southern Africa and 73% children were not in school. For Kenya, no data have
in West and Central Africa. Out-of-school rates are been available since 2009, when the number of out-
lowest in South Asia and in Latin America and the of-school children was 1.1 million. Reasons for this
Caribbean (6%), in CEE/CIS and East Asia and the lack of data are described in Section 2.4.
Pacific (5%) and in Western Europe, North America
and Australasia (4%). For many countries without reliable administrative
data, household surveys can give an indication of the
Figure 2.4 lists selected countries with more than extent of exclusion from education. In the Democratic
half a million out-of-school children of primary school Republic of the Congo, for example, four DHS and
age. Among them, India, Indonesia, Niger, Nigeria, MICS surveys were carried out between 2001 and
Pakistan, South Sudan and Sudan had 1 million or 2014. Drawing on these data and a national household
more out-of-school children in 2012 (or the most survey on out-of-school children, the authors of the
recent year for which data are available). When national study for the Global Initiative on Out-of-School
reviewing these numbers, it is important to keep in Children estimated that 3.5 million children of primary
mind that there are no reliable estimates available for school age were out of school in 2012, more than in
recent years for some countries with large numbers all but two countries in Figure 2.4 (UNICEF and UIS,
of excluded children. For example, the most recent 2013d). Appendix IV provides the latest out-of-school

Figure 2.5 School exposure of out-of-school children of primary school age by region, 2012
Likely to enter Unlikely to ever Million children

Left school enter school out of school
school in the future
WEST AND 14 26 60 18.8

CENTRAL AFRICA
EASTERN AND
SOUTHERN AFRICA
17 58 25 11.0
SOUTH ASIA 28 15 57 9.8
EAST ASIA
AND THE PACIFIC
59 32 9 6.9
MIDDLE EAST
AND NORTH AFRICA
19 37 44 4.3
LATIN AMERICA
AND THE CARIBBEAN
20 44 36 3.8
W. EUROPE, N. AMERICA
AND AUSTRALASIA
17 64 19 2.2
CEE/CIS 10 52 38 1.0
WORLD 23 34 43 57.8
0 20 40 60 80 100% 0 In millions
Notes: The data refer to the regional classification used by UNICEF. The category ‘Western Europe, North America and Australasia’ is not an
official UNICEF region, but it is used in this report to group all countries not belonging to other UNICEF regions. It includes countries in which
UNICEF does not operate. They are primarily high- and upper-middle-income countries located in Australasia, Europe and North America.
children statistics from administrative and household Taken as a whole, the data suggest that most of
survey sources for all countries. sub-Saharan Africa’s out-of-school children are
unlikely to ever enter school. However, the patterns
Twenty-five million primary school-age children of school exposure in West and Central Africa differ
will probably never enter school markedly from those in Eastern and Southern Africa.
The former region has the highest concentration
Children out of school can be divided into two broad of school exclusion, similar to South Asia, where
groups: those who have attended school in the past three in five out-of-school children will probably
but dropped out and those who have never attended never enter a classroom. By contrast, most out-
school. The second group can be further sub-divided of-school children in Eastern and Southern Africa
into children who will attend school at some point in the are expected to start school in the future, a pattern
future and those who will never attend. By examining shared with CEE/CIS and Western Europe, North
the pattern of the age at which children enter and America and Australasia. Similarly, most out-of-
leave school it is possible to estimate the distribution school children in the Middle East and North Africa
of out-of-school children across these three groups. and Latin America and the Caribbean are expected
Figure 2.5 shows that more than two-fifths, or close to receive formal education at some point, although
to 25 million, of the world’s 58 million primary school- around 40% of them will probably never go to
age out-of-school children are unlikely to ever enter a school. East Asia and the Pacific is the only region
classroom. Of the remaining 33 million out-of-school where most primary school-age out-of-school
children, 13 million have left school and 20 million are children have dropped out, rather than having never
expected to be late entrants to school in the future. attended at all.

completion between 2007 and 2012 were, in theory, in
About 15 million girls and 10 million the age group for secondary education or were even
boys of primary school age are not older. Primary school-age children account for the
expected to ever attend school. This majority of early primary school leavers in only three
group of children poses the most countries in Figure 2.6: Ghana, Nigeria and Timor-
serious challenges to policymakers Leste. However, the true scale of over-age school
leaving is even greater than suggested in Figure 2.6,
because the analysis only considers children who were
Classifying out-of-school children by past and older than primary school age at the time they left
possible future school attendance yields important school. In fact, many children who drop out of primary
insights for policymakers. For countries like Bolivia, school are still within the primary school age range but
where most out-of-school children are likely to attend were too old for the grade they last attended.
school in the future (albeit when they are older than
the official age of entry into primary education), Out-of-school adolescents of lower secondary
the goal is to ensure earlier, on-time entry into the school age
education system (UNICEF and UIS, 2011).
In 2012, 63 million adolescents of lower secondary
However, it is children who are not expected to school age (typically between 12 and 15 years) were
gain access to schooling—roughly 15 million girls not in primary or secondary school (see Table 2.2).
and 10 million boys according to the latest available The out-of-school rate in this age group was 17%
data—who pose the most serious challenges to for girls, 16% for boys and 17% for girls and boys
policymakers. For countries with large populations of combined. More than 40% of all out-of-school
out-of-school children with a scant likelihood of going adolescents live in South Asia and more than one-
to primary school, such as Burkina Faso, Nigeria third in sub-Saharan Africa.
and Pakistan, accelerated learning programmes or
other forms of remedial NFE can be vital to provide The global number of out-of-school adolescents is
schooling to children who would otherwise be similar to the global number of out-of-school children,
excluded entirely from education. even though there were 1.7 times more children of
primary school age in 2012 (650 million) than lower
Most children who drop out of primary school secondary school-age adolescents (374 million).
early are over-age While adolescents are far fewer in number, they are
nearly twice as likely to be out of school as children
The classification of children by their past and of primary school age (17% compared to 9%). As
possible future exposure to education is only a partial mentioned, children who are over-age for their level
indicator of early school leaving because it only or grade are more likely to drop out of school. At
covers children of primary school age. An important the same time, the opportunity cost of education
share of primary school pupils who drop out are over- increases with age as vulnerable families weigh the
age by several years, because they have entered benefits of keeping older children in school against
school late or have had to repeat school grades. the need for income, with children often left with no
choice but to work instead of going to school (see
Figure 2.6 shows that in 20 of 23 countries with Section 3.4 on child labour).
recent household survey data, early primary school
leavers are more likely to be significantly over-age for Similar to trends for out-of-school children of
their level of education, often by three years or more. primary school age, the number and rate of
When these children leave school, they are counted out-of-school adolescents have fallen significantly
as out-of-school adolescents, as discussed in the since 2000, when 97 million adolescents—25%
next section. For example, around 80% of pupils in of the entire age cohort—were not in primary or
Haiti and Madagascar who left primary school before secondary school.

Figure 2.6 Distribution of children who leave school before completing primary education,
by age group, selected countries, 2007-2012
Primary school age 1-2 years older 3 or more years older
%
100
12
18
32 30 33
34 10
80 39 37 36 36 35 36
47
54 54 52
23
63 63
Early primary school leavers
65
71
75
60 80 21 15
20 17
22 20
98 34 24
33
28
21
40 18 78
17 17
14 59
17
52
20 11 47 49
43 45
20
14 39 42
33 34 34
8 29 30 30 32
23
18 19
14
11 12
0 1
ZA AL
A
G IA
O
IA
N AQ
A
A
R
E
RA LIC
BH AR
N
BO IA
I
AS I
E
LO N
IA
AZ AM
E
D
AU AW
AG AIT
VI
N
BI
AN
BI
,D
RE O
ER
ST
N
IP
TA
IB
EG
AN
ER
IR
O
LI
C
TO
O
TA
ER UB
M
M
SW T N
AM
O
M AL
LE
H
IG
LE
U
AN ER
N
IL
AN LIB
IN
G
RI
N
SE
M
R-
E
N
PR
IC M
O
VI
O
IN CA
C
AD
M
D
TI
M
SI
E
M
M
TO
DO
O
SA
Note: The figure refers to children and adolescents who were in primary school in the previous year, are not in school in the current year
(when the household survey was conducted), and whose highest completed grade was lower than the last grade of primary education.
Source: UIS calculations based on the Demographic and Health Surveys (DHS) and Multiple Indicator Cluster Surveys (MICS), 2007-2012
DataLink: http://dx.doi.org/10.15220/2014/ed/sd/7/f2.6
Table 2.2 Out-of-school adolescents of lower secondary school age, 2000 and 2012
2000 2012
% Number (million) % Number (million)
Region MF M F MF M F MF M F MF M F
W. EUROPE/N. AM./AUSTRALASIA 2.8 3.2 2.4 1.1 0.7 0.5 2.7 2.6 2.8 1.0 0.5 0.5
LATIN AMERICA AND CARIBBEAN 10.1 9.9 10.2 3.6 1.8 1.8 7.5 7.7 7.3 2.8 1.5 1.3
CEE/CIS 11.6 11.2 12.1 3.8 1.9 1.9 5.2 5.1 5.5 1.2 0.6 0.6
EAST ASIA AND PACIFIC 20.9 22.3 19.4 24.5 13.5 11.0 8.4 8.6 8.2 7.4 4.0 3.4
SOUTH ASIA 39.9 33.2 47.0 37.3 16.1 21.3 26.4 26.3 26.4 26.3 13.7 12.6
MIDDLE EAST AND NORTH AFRICA 11.7 9.4 14.1 2.9 1.2 1.7
EASTERN AND SOUTHERN AFRICA 42.8 36.3 49.3 10.5 4.4 6.0 26.9 24.2 29.5 8.5 3.8 4.6
WEST AND CENTRAL AFRICA 46.5 40.7 52.5 10.8 4.8 6.0 39.7 37.0 42.4 12.5 5.9 6.6
WORLD 24.7 22.5 27.0 96.9 45.2 51.6 16.8 16.2 17.5 62.9 31.3 31.6
Notes: The data refer to the regional classification used by UNICEF. The category Western Europe, North America and Australasia refers to primarily high- and
high-middle-income countries in which UNICEF does not operate. Data for Eastern and Southern Africa refer to 2011. No regional figures are available for the
Middle East and North Africa for 2000 because of insufficient data coverage.

The global reduction to 63 million out-of-school were found in Western Europe, North America and
adolescents is largely the result of progress in East Australasia (3%) and in the CEE/CIS region (5%).
Asia and the Pacific, where their number fell by
more than two-thirds from 25 million to 7 million The effect of population growth in
between 2000 and 2012, and South Asia, where sub-Saharan Africa
their number fell by 11 million over the same period,
from 37 million to 26 million. Enrolment rates across sub-Saharan Africa are
improving but not enough to keep up with the increase
The progress in East Asia and the Pacific is linked in the school-age population. In 2012, sub-Saharan
closely to the situation in China and its estimated Africa was home to 57% of the global population of
reduction in the number of out-of-school children out-of-school children of primary school age, up from
and adolescents since 2000.4 Another country in 44% in 2000, and its share of the global population of
the region with a large drop in its out-of-school out-of-school adolescents increased from 23% in 2000
population is Indonesia, where the number of out- to 35% in 2012. Between 2000 and 2012, the primary
of-school adolescents fell from 3.5 million in 2000 school-age population in sub-Saharan Africa grew
to 1.7 million in 2012. from 110 million to 148 million and the lower secondary
school-age population from 49 million to 66 million.
In addition to these success stories, 29 countries Sub-Saharan Africa—and this is true for both Eastern
for which data are available managed to reduce the and Southern Africa and West and Central Africa—is
number of out-of-school adolescents by more than the only region that has been confronted with such
one-half between 2000 and 2012. Among them are a rapidly-growing population. Countries in this region
eight countries that had more than 100,000 out- face a double challenge: not only do they have to
of-school adolescents in 2000: Ecuador, Ghana, provide educational facilities for the children who are
Indonesia, South Africa, Tajikistan, Turkey, Ukraine out of school today, they must also accommodate
and Venezuela. the ever-growing numbers of children who will reach
school-going age in the coming years.
West and Central Africa was the only region that
bucked the global trend, with an increase in the Figure 2.7 displays the evolution of the region’s
number of out-of-school adolescents from 11 combined primary and lower secondary school-age
million to nearly 13 million between 2000 and 2012, population from 2000 to 2012, with the starting point
although the lower secondary out-of-school rate for the population in 2000 set at 100. In both Eastern
fell from 47% to 40% over the same period. The and Southern Africa and West and Central Africa the
increase in the number of out-of-school adolescents population of primary and lower secondary school
in the region is a direct consequence of high age grew by more than one-third between 2000 and
population growth. 2012. For every 100 school-age children in sub-
Saharan Africa in 2000, there were 134 school-age
This particular region also had the highest lower children in 2012. Projections by the UN Population
secondary out-of-school rate in 2012, followed by Division show no slowdown in this trend before 2050,
Eastern and Southern Africa (27%) and South Asia when the primary and lower secondary school-age
(26%). In the Middle East and North Africa, 12% of all population in sub-Saharan Africa is estimated to be
adolescents of lower secondary school age were not more than 2.5 times as large as it was in 2000. By
in school. Similar to the primary out-of-school rate, contrast, in the East Asia and the Pacific and CEE/
the lowest percentages of out-of-school adolescents CIS regions, the school-age populations fell by
more than 20% since 2000. Given these population
4 In 1997, 17 million primary school-age children and an unknown number
of lower secondary school-age adolescents were out of school in China. trends, it is remarkable that the countries in sub-
The UIS has no publishable data for China for recent years, mainly due
to uncertainty about national population figures. However, the UIS has Saharan Africa have managed to reduce the number
produced estimates that are used to calculate regional rates and numbers
of out-of-school children and adolescents in East Asia and the Pacific. of primary school-age children out of school and

Figure 2.7 Evolution of the primary and lower secondary school-age population from 2000 to 2012,
by region
School-age population in region School-age population in other regions
WESTERN EUROPE,
EAST ASIA AND THE PACIFIC CEE/CIS
NORTH AMERICA AND AUSTRALASIA
140
120
100
School-age population relative to 2000 (2000=100)
80
MIDDLE EAST AND NORTH AFRICA LATIN AMERICA AND THE CARIBBEAN SOUTH ASIA
140
120
100
80
EASTERN AND SOUTHERN AFRICA WEST AND CENTRAL AFRICA WORLD
140
120
100
80
2000 2002 2004 2006 2008 2010 2012 2000 2002 2004 2006 2008 2010 2012 2000 2002 2004 2006 2008 2010 2012
Notes: The data refer to the regional classification used by UNICEF. The category Western Europe, North America and Australasia refers to
primarily high- and upper-middle-income countries in which UNICEF does not operate. Regions are sorted by the school-age population in
2012 relative to 2000, from the largest decrease to the largest increase.
Source: UNESCO Institute for Statistics, August 2014, calculated from population estimates by the UN Population Division, 2013
avoid an increase in the number of lower secondary on Out-of-School Children and other evidence to
school-age children out of school in recent years. look more closely at the barriers that keep the most
disadvantaged children out of school: children affected
2.3 THE INDIVIDUAL AND HOUSEHOLD by armed conflict, child labourers, children whose
CHARACTERISTICS OF OUT-OF- home language differs from the language used at
SCHOOL CHILDREN OF PRIMARY AND school, and children with disabilities. The role of
LOWER SECONDARY SCHOOL AGE gender as a determinant of exclusion from education
is also discussed in greater detail in Section 3.3.
The analysis so far has focused the rate and number
of out-of-school children globally and how trends Administrative data on gender disparities
have evolved since 2000. However, in order to target
policies and strategies to bring out-of-school children The analysis of gender disparities in out-of-school
into school, robust information is needed on who rates is crucial for the wider analysis of statistics on
they are and where they live. Chapter 3 draws on out-of-school children. UIS data confirm that there
country studies carried out under the Global Initiative has been considerable progress in reducing gender

disparities since 2000, despite persistent gender At the regional level, gender disparities are greatest
gaps in some countries and regions. Globally, the in West and Central Africa, Eastern and Southern
gender gap in the out-of-school rate fell from about Africa, and the Middle East and North Africa. In these
6 percentage points to about 2 percentage points regions, the gap between the male and female out-
for children of primary school age between 2000 and of-school rates ranges from 3 percentage points for
2012, and from about 4 percentage points to about 1 primary school-age children in Eastern and Southern
percentage point for adolescents of lower secondary Africa to 8 percentage points in West and Central
school age. South Asia began that period with the Africa. Gender disparities tend to be greater in
largest gap between female and male out-of-school regions with higher out-of-school rates, but the case
rates (15 percentage points for primary school- of South Asia—where 26% of both girls and boys
age children and 14 percentage points for lower of lower secondary school age are not in school—
secondary school-age adolescents), but by 2012, shows that high out-of-school rates are not always
there was hardly any difference between the rates for associated with gender disparity.
girls and boys.
While gaps in enrolment between girls and boys
Figure 2.8 displays regional out-of-school rates of have decreased over the past two decades, girls
children of primary and lower secondary school age are still more likely to face persistent barriers to their
in 2012. Among primary school-age children, 10% of education than boys in many countries. Table 2.3
girls and 8% of boys worldwide were out of school, presents countries with recent data where girls face
while the rate among lower secondary school-age the greatest disparities. At the national level, the
adolescents was 18% for girls and 16% for boys. female out-of-school rate for children of primary
Figure 2.8 Out-of-school rate by region, age group and sex, 2012
Both sexes Male Female
W. EUROPE, Primary age

N. AMERICA AND
AUSTRALASIA Lower sec. age
EAST ASIA Primary age

AND THE PACIFIC Lower sec. age
Primary age
CEE/CIS
Lower sec. age
Primary age
SOUTH ASIA
Lower sec. age
LATIN AMERICA AND Primary age

THE CARIBBEAN Lower sec. age
MIDDLE EAST AND Primary age

NORTH AFRICA Lower sec. age
EASTERN AND Primary age

SOUTHERN AFRICA Lower sec. age
WEST AND Primary age

CENTRAL AFRICA Lower sec. age
Primary age
WORLD
Lower sec. age
0 10 20 30 40%
Notes: The data refer to the regional classification used by UNICEF. The category Western Europe, North America and Australasia refers to
primarily high- and upper-middle-income countries in which UNICEF does not operate. Regions are sorted by the percentage of out-of-school
children of primary school age of both sexes combined.

Table 2.3 Selected countries with a large difference between the male and female out-of-school rates, latest
data available, 2010-2013
Female out-of-school rate 10 percentage points or more Male out-of-school rate 10 percentage points or more
greater than male out-of-school rate greater than female out-of-school rate
Primary school age Lower secondary school age Primary school age Lower secondary school age
Angola Central African Republic Antigua and Barbuda
Cameroon Guinea Bangladesh
Central African Republic Mali
Chad Mozambique
Guinea Yemen
Niger
Nigeria
South Sudan
Yemen
school age was at least 10 percentage points households are on average more likely to be out of
greater than the male out-of-school rate in ten school than boys, urban children and children from
countries, nearly all of them in sub-Saharan Africa. wealthier households (UIS, 2010; 2011a; 2012b).6
For adolescents of lower secondary school age, the Furthermore, the impact of personal and household
female out-of-school rate was at least 10 percentage characteristics on school attendance tends to be
points greater than the male out-of-school rate in five cumulative, so that for example, girls from poor, rural
countries. By contrast, there are only two countries households often have far lower attendance rates
(Antigua and Barbuda and Bangladesh) where the than boys from rich, urban households. (Explore the
out-of-school rate for boys of lower secondary school data for countries participating in the Global Initiative
age was 10 percentage points or more greater than with the UIS online tool at http://on.unesco.org/
that for girls.5 As with Figure 2.4, it should be noted oosci-global)
that the list in Table 2.3 is incomplete because of a
lack of reliable data for many countries. Figure 2.9 displays data from 63 nationally-
representative household surveys carried out
Household survey data on exclusion from between 2008 and 2012. Hattori (2014) calculated
education average out-of-school rates across the 63 countries
and found that 14% of all children of primary school
Administrative data in the UIS database can be age were out of school.7 There was hardly any
disaggregated by sex but provide no information on difference between the out-of-school rates of girls
the other individual and household characteristics and boys (14% and 13%, respectively). On the other
of children who are excluded from education. For hand, there were clear links between the out-of-
this, we must turn to household survey data, which school rate and the location of a household (urban
also complement administrative data by providing or rural), household wealth and the level of education
additional information on the possible extent of of the household head. Children from the poorest
exclusion from education, as discussed in Section 2.4. household quintile had the highest average out-of-
school rate, 22%, compared to an out-of-school
Analysis of data from international household rate of less than 6% among children from the richest
survey programmes, such as DHS and MICS, households. Higher educational attainment of the
consistently demonstrates that, where disparities household head and living in an urban area were
exist, girls, rural children and children from poor
6 Administrative records typically contain data on enrolment, while
household surveys typically collect data on attendance.
5 The data for Antigua and Barbuda should be interpreted with caution 7 Each of the 63 countries had an equal weight in the calculation of the
because the country has a very small school-age population, which can average out-of-school rate in Figure 2.9. The relative size of each country’s
exaggerate gender disparities for certain indicators. The national number population of primary school age was not considered. The focus of the
of primary school-age out-of-school children was estimated to be less analysis is, therefore, on individual countries rather than the national or
than 1,600 in 2012. combined number of children in and out of school.

Figure 2.9 Percentage of children of primary school age in school, by sex, location, household
wealth and education of the household head, various years
Mean values (unweighted) per group National values
TOTAL
14
MALE
13
FEMALE
14
URBAN
8
RURAL
16
RICHEST QUINTILE
6
SECOND RICHEST QUINTILE
9
MIDDLE QUINTILE
13
SECOND POOREST QUINTILE
16
POOREST QUINTILE
22
HOUSEHOLD HEAD SEC. OR HIGHER ED.

6
HOUSEHOLD HEAD PRIMARY ED.
12
HOUSEHOLD HEAD LESS THAN PRIM. ED.
21
0 10 20 30 40 50 60 70%
Source: Hattori, 2014 DataLink: http://dx.doi.org/10.15220/2014/ed/sd/7/f2.9
associated with lower out-of-school rates among household wealth had only a weak effect on school
children of primary school age. attendance had already achieved high attendance
rates (above 95%). There was a positive association
Hattori (2014) confirmed these findings with a between school attendance and having a household
regression analysis of the determinants of school head with a high level of education (secondary
attendance among children of primary and lower education or higher) in 54 of the 63 countries
secondary school age.8 Among the factors that analysed, evidence of the inter-generational impact
were considered, household wealth was by far the of education. The link between school attendance
most statistically significant determinant of school and other factors considered in the analysis (age,
attendance in 56 of the 63 countries studied. In sex and location of the household) was, however,
Guinea, for example, children from the wealthiest mixed. Older children, boys and children from urban
household quintile were nearly 40 percentage households were, on average, more likely to attend
points more likely to attend school than those from school, but this was not the case in every country.
the poorest quintile, when other factors were held
constant. In many other countries in sub-Saharan 2.4 HOW TO IMPROVE THE ACCURACY
Africa and South Asia, children from the richest OF ESTIMATES OF OUT-OF-SCHOOL
households were at least 20 percentage points CHILDREN
more likely to be in school than children from the
poorest households. The majority of countries where UIS estimates of the rate and number of out-of-
school children are used to monitor progress towards
8 Analysis of data from household surveys, such as DHS and MICS, can international education goals, acting as barometers
only consider demand-side determinants of school attendance. Supply-
side factors, such as the distance to the nearest school or the quality of to identify and compare key trends at the global
education on offer, cannot be examined because DHS and MICS surveys
do not collect data on the supply side of the education system. and regional levels. At the same time, countries also

Box 2.2 Schooling trajectories in Pakistan
Household survey data can be used to examine the dynamics by which children progress through all of the
different levels of the education system.9
In Pakistan, for example, a country with one of the world’s largest out-of-school populations, adolescents
aged 13 to 16 years (the official age for upper secondary education) who entered primary school on
time and did not repeat any grades should, in theory, have completed their lower secondary education.
However, data from a DHS survey carried out in 2012 and 2013 show that Pakistan’s reality is very
different. Only 79% of those aged 13 to 16 had entered primary school and only 63% had completed
primary education. Another 10% were still in primary education at the time of the survey and may
eventually complete that level.
Most children in the sample who completed primary education continued their education at the lower
secondary level (55%), but only 26% had completed lower secondary education by the time they were
13 to 16 years old. This is, in part, because 26% of this age group were still attending lower secondary
education at the time of the survey. Only 3% had dropped out of lower secondary education without
completing that level.
These results can be further broken down by household wealth. Figure 2.10 compares the schooling
trajectories of upper secondary school-age adolescents from the richest and poorest household quintiles in
Pakistan. In the richest households, nearly all of those aged 13 to 16 (96%) had entered primary education,
only one-half (49%) had completed lower secondary education, and around one-third (35%) were still
in lower secondary school at the time of the survey (as seen in the difference between the values for
‘attending or completed lower secondary education’ and ‘completed lower secondary education’ in Figure
2.10). In the poorest quintile, only 47% had entered primary school and only 5% had completed lower
secondary education by the time they reached upper secondary school age.
Figure 2.10 Educational attainment of adolescents of upper secondary school age (13-16 years)
in the richest and poorest household wealth quintiles of Pakistan, 2012-2013
RICHEST QUINTILE: Both sexes Male Female
POOREST QUINTILE: Both sexes Male Female
%
100
80
60
40
20
0
EP ACP CP ELS ACLS CLS
Notes: EP = entered primary education; ACP = attending or completed primary education; CP = completed primary education; ELS = entered
lower secondary education; ACLS = attending or completed lower secondary education; CLS = completed lower secondary education.
Source: UIS calculations based on the Pakistan Demographic and Health Survey (DHS) 2012-2013
9 Similar analysis was carried out by Nguyen and Wodon (2014a) and by the authors of the regional OOSCI report for West and Central Africa (UNICEF
and UIS, 2014c).

Figure 2.10 also shows a striking gender gap in the schooling trajectories of adolescents aged 13 to
16 years from the poorest household quintile: 60% of boys had entered primary school, compared to
only 30% of girls. However, once girls enter school, they are less likely to drop out—even if they are
poor—and the gap between the entrance and completion rates of girls and boys shrinks at higher levels
of education. By contrast, girls and boys from the richest households are equally likely to enter primary
education and progress similarly through the education system, although boys are somewhat more likely
to drop out along the way.
This analysis of schooling pathways yields important insights into the points within the education system
where children are likely to discontinue their education.10 It also demonstrates that enrolment rates or
out-of-school rates alone are not enough to obtain a full picture of exclusion from primary and secondary
education. The fact that 10% of children of upper secondary school age in Pakistan were still in primary
school in 2013 indicates widespread late entry into the education system. Under such circumstances,
the out-of-school rate of primary school-age children on its own can give a false impression of the
extent of exclusion from primary education. One way to address this problem is to examine the past and
possible future school attendance of out-of-school children, as presented in Figure 2.5 on the school
exposure of out-of-school children.
10 It should be noted that the experience of older cohorts will not necessarily be repeated by children who enter a country’s education system today.
In Pakistan and other countries, younger cohorts are more likely to have entered and completed a given level of education than members of older
cohorts, and this trend is likely to continue.
produce national and sub-national out-of-school Population censuses are usually carried out every
estimates to inform their own decision-making. The 10 years but have not been conducted in the
challenge for the calculation of more precise figures, Democratic Republic of the Congo and Somalia
whether for international monitoring or national in a quarter of a century. Population estimates
policymaking, lies in improving the accuracy of data for years between censuses rely on projections
on population, enrolment and attendance, refining based on inter-census data on births, deaths and
consistency between population and enrolment data, migration, which are not always reliable. Generally
and developing new indicators to measure exclusion speaking, the further a year is from the last full
from education. census, the more likely it is that the population
estimate is inaccurate. It is important, therefore, to
We need better population estimates to continue to improve the population estimates that
calculate the number of children in each country are generated between censuses. Accuracy can
also be undermined when censuses underestimate
The accuracy of estimates of the school-age hard-to-reach marginalised groups, such as nomadic
population has a direct and major impact on estimates populations, undocumented residents and those in
of out-of-school children from administrative data. slums and refugee camps (Carr-Hill, 2013). Improving
When the school-age population is underestimated the coverage of census data collection would also
or overestimated, the out-of-school rate follows suit. improve household survey-based estimates, because
Improving the consistency between population and the census provides the survey sampling frame.
enrolment data begins with ensuring that both are as
accurate and complete as possible. Improvements are also needed in the consistency
between data on age from population censuses
It is a fact that population data for many countries and enrolment records. In most countries, the
lack precision. Doubts about population estimates government mandates that children must enrol in
for countries with large populations, such as China school if they are of primary school age on or before
and Ethiopia, are the reason for the lack of recent a certain date. The most accurate estimates of out-
UIS estimates of the rate and number of out-of- of-school children would refer to the ages of children
school children and adolescents in these countries. at this specific date. If age data for students and

the population are collected at a later time, children
who were too young to enter primary school by the To accurately measure the number
date mandated by law but had their birthday by the of out-of-school children, we need
time the population data are collected are incorrectly to improve the accuracy of data
counted as part of the school-age population and, on population, enrolment and
therefore, as being out of school. Given the crucial attendance; refine consistency
role population data play in calculating out-of-school between population and enrolment
rates and numbers, lines of communication should data; and develop new indicators to
be established between the agencies responsible measure exclusion from education
for population estimates and enrolment data to
understand and resolve inconsistencies.
Household surveys are used for multiple purposes,
Household survey data face a similar issue. including demography and health planning, and
Household surveys can be conducted at any time education is not always a major component.
and often take place more than six months after the Because the data are collected in households,
school year has started. This discrepancy can have these surveys do include information on out-of-
a substantial impact on out-of-school estimates. school children and their individual and household
For example, the large gap (11 to 15 months after characteristics. However, such surveys are generally
the start of the reference school year) in the data sample-based and the reliability of their results,
collection of the Kenya 2008-2009 DHS survey including their estimates of the percentage and
resulted in an inflated primary out-of-school rate of number of children in and out of school, can be
26%, which halved to 13% after the data on age affected by sampling and non-sampling errors.
were adjusted to the start of the school year during
indicator calculation. The solution is to ensure that Both sources of data, administrative records and
surveys collect birth date information for all children household surveys, are used to take a snapshot of
to determine their exact age at the start of the school school attendance, which is a dynamic phenomenon.
year. While most surveys, such as DHS, do not do so Children are constantly entering and leaving school
at present, recent MICS surveys have included this or moving from one school to another. Yet, statistical
information, which greatly improves the accuracy of tools differ in fundamental ways: who collects the
the data on age that underpin the estimates of out- data, as well as how, when and for what purpose.
of-school rates. As the example of India in Box 2.3 demonstrates,
different sources of data can yield very different
We need better enrolment and attendance data estimates of the number of children and adolescents
to know how many children are in school or out of school.
out of school
There is no one best data source and each has its
Barring a few national exceptions, there is no regular advantages, which may make it more appropriate
data collection that targets out-of-school children for a given purpose over other data sources. In order
specifically. The global and regional numbers to assess the extent of exclusion from education,
presented in this report are based on administrative statistics on school attendance from all sources
data captured through school censuses—official must be considered. A thorough examination of
government data used for planning. Because the data quality and suitability can help countries decide
data are collected in schools, they do not provide which data to use for their policies. The following
information on out-of-school children. As we have recommendations apply to all sources of data on
seen, population data are needed to calculate out-of- enrolment or attendance to improve the accuracy
school estimates, and errors can result where there is of estimates of out-of-school children.
a mismatch between population and enrolment data.

Box 2.3 Selecting and calculating estimates on out-of-school children in India
Data collection and analysis are uniquely challenging tasks in India, with its large and diverse population.
Although India’s primary and lower secondary school-age out-of-school rates are low compared to the
rest of South Asia, a difference of one percentage point in the country’s out-of-school rate can mean a
difference of more than one million in the number of out-of-school children. India’s participation in the
Global Initiative on Out-of-School Children was motivated, in large part, by a need to understand the
reasons for different estimates of the number of out-of-school children and to harmonise methodologies
(Sigdel, 2014).
A data inventory at the outset of the national study revealed the availability of a whole range of
administrative and household data sources to generate statistics and develop profiles of out-of-school
children. However, the out-of-school rates calculated from these sources were not consistent. In-depth
investigation by the national report team revealed two main sources of discrepancy.
First, the sample design and timing of the survey matters. With the 2011 census data not yet available at
the time of the analysis, all recent data sources were sample surveys. The sampling frameworks of some
surveys were designed specifically to collect data on education, while others were not. In addition, the
data collection of each of the two relevant National Sample Survey Office (NSSO) surveys (2007-2008 and
2009-2010) took place during two consecutive school years. By contrast, the 2009-2010 Social and Rural
Research Institute of IMRB International (SRI-IMRB) Survey of Out-of-School Children conducted by the
Ministry of Human Resource Development was designed to collect data on this topic and data collection
was aligned with the academic year. As a result, estimates of out-of-school children from these surveys did
not match.
Second, surveys used different definitions of out-of-school children. Despite the fact that the NSSO 2007-
2008 and SRI-IMRB 2009-2010 surveys used the same sampling framework, the out-of-school rates
from the SRI-IMRB survey were much lower. It found that 4% of children of primary school age and 5%
of children of lower secondary school age were out of school, while the NSSO reported that 11% and
8% of primary and lower secondary school-age children were out of school, respectively. The differences
stemmed in part from what each survey considered as being ‘in school’. SRI-IMRB considered school-
age children, even when enrolled, who had not attended any kind of formal school during the two months
preceding the survey as being out of school; children in formal pre-primary education were counted as
in school. By contrast, the NSSO survey counted as out-of-school children those who were enrolled but
had not attended primary or secondary school at any time in the reference academic year; children in pre-
primary education were considered out of school.
Overall, the report team from India concluded that there is no one perfect estimate of the rate and
number of out-of-school children in the country: the data sources revealed a range of estimates of the
true value. Nonetheless, by identifying the causes of discrepancies and by adopting standard definitions
and methodologies, it is possible to arrive at better estimates. The Government of India is now working
towards the establishment of a clear, national definition of school drop-out.
First, efforts should be made to ensure that school over providers of non-formal education and may
censuses collect data on all schools, public and not even know they exist. Household surveys, like
private, to avoid the underreporting of enrolment. For DHS and MICS, do not collect data specifically on
example, many children identified as out of school in non-formal education because of the difficulties in
national statistics may, in fact, attend schools in the identifying and classifying such forms of education.
non-formal sector, such as community, NGO-run or While several countries have developed Non-Formal
unregistered schools. The large diversity of education Education Management Information Systems (NFE-
programmes and providers, coupled with a lack of MIS) in the past decade, there has been increasing
standards, pose great challenges for the collection of effort to develop more comprehensive Education
accurate data. Ministries of education, which manage Management Information Systems (EMIS) that
school censuses, typically have no regulatory control integrate both formal and non-formal education

in a number of countries, such as Cambodia, the goals. The out-of-school rate is derived from the
Democratic Republic of the Congo and Tanzania percentage of children enrolled in or attending
(Yasunaga, 2014). primary or secondary education. However, these
indicators are not perfect because they measure
At the same time, some countries face the challenge enrolment or attendance of children at the official
of inflated enrolment. This can occur when schools age for the respective level of education. Take the
report children who never attend as enrolled or where example of a country where all children enter and
children are registered at two or more schools but complete primary education but where some children
attend only one of them. A major cause of inflated enter primary school one year late. In this case, late
enrolment figures is the linking of funding to the entrants will be counted as out-of-school children
number of students. In these cases, schools have in national statistics, even though the country has
an incentive to overreport actual enrolment, and reached universal primary education. 910
as a result, the number of out-of-school children
may be underestimated. Monitoring and verification Indicators of primary completion that could measure
procedures of reported enrolment data can help to more accurately how close a country is to universal
check overstated enrolment counts. primary education, regardless of the age at which
children complete that level, are more difficult
Second, there must be a clear and consistently to calculate because of limited data availability.
applied definition of drop-out at the national level. If Graduation is not a well-defined concept at the
children who have left school remain enrolled in the primary level and the UIS calculates, therefore, a proxy
school register, the number of out-of-school children measure of primary school completion—the gross
will be underestimated. A first step is to establish intake ratio to the last grade of primary education—
a standard list of the authorised and unauthorised which is not as straightforward to interpret as the
reasons for school absence. Then, clear guidelines out-of-school rate and related indicators.11 The
can be established to deregister students who have advantages and disadvantages of current indicators
been absent without an authorised reason (such as are an important issue to consider as the international
suspension or illness) for a given amount of time. community defines new and improved indicators for
post-2015 goals to succeed the EFA goals and the
Finally, improvements are needed in the recording of Millennium Development Goals.
data on children’s age. In countries with weak birth
registration systems in particular, the age of children 2.5 HOW TO BETTER IDENTIFY THE
can be misreported by teachers, administrators or CHARACTERISTICS OF CHILDREN
households members, who may assume the children OUT OF SCHOOL
are of primary school age when they may, in fact,
be older or (less commonly) younger. This affects Statistics on out-of-school children produced by
both administrative and household survey-based the UIS are intended for international comparison,
estimates. As demonstrated in a report by the UIS but to develop in-depth profiles of these children for
and UNICEF, this can result in an overestimate of targeted policies, more fine-grained, sub-national
enrolment rates for primary school-age children and information is needed.
an underestimate of enrolment rates for secondary
school-age children (UIS and UNICEF, 2005). The most important and practical way to identify the
profiles of out-of-school children is to delve further
We need better indicators to measure into the data sources that already exist. National-level
exclusion from education
11 The gross intake ratio to the last grade of primary education is calculated
by dividing the number of new entrants to the last grade of primary
education by the number of children at the official age of entry into the last
It is also necessary to review the indicators that grade. The ratio can exceed 100% if many over- or under-aged children
enter the last grade of primary education as a result of early or late entry
are used to assess progress towards international into primary school and grade repetition.

mobile or nomadic groups do not live in households
Household surveys are underutilised and are, therefore, generally excluded from survey
data sources on out-of-school data collection. Invisible to regular data collection,
children, providing rich information these children require specific data collection,
on their profiles similar to a special survey on out-of-school children
conducted by the Democratic Republic of the
Congo in 2012. This survey showed that one-half of
administrative records and household surveys are the the children living in shelters and on the street are
primary sources used to identify the characteristics orphans, largely concentrated in the conflict-affected
of children excluded from education. Administrative provinces of North and South Kivu and in the capital
data, with their relatively complete coverage of all Kinshasa (UNICEF and UIS, 2013d). The survey
students in all schools, have great potential to zoom also revealed that out-of-school rates vary between
in on the characteristics at the district or school level, children in shelters and those on the street. Primary
for example, to examine areas with high rates of school-age children in shelters have lower out-of-
over-age students or early school leaving. school rates than the national average (11% versus
26%), but for lower secondary school age, the
A major finding of the Global Initiative on Out- rates are the same (13%). In contrast, 96% of the
of-School Children is that household surveys 1,160 street children of primary school age studied
are underutilised data sources on out-of-school were out of school. Among the 970 lower secondary
children, providing rich information on the profiles of school-age adolescents sampled, 93% were not
these children by sex, location, household wealth, in school. The survey also found that most street
education of the parents, and other personal and children are boys, though the girls who are on the
household characteristics. However, such surveys streets face especially harsh conditions and are
are rarely managed by the ministry of education, often victims of forced prostitution. This gender
whose staff may not be aware of or trained in using dynamic is common in many countries (Salmon
this kind of data. Cross-sector collaboration can and Wodon, 2014). Such targeted research brings
facilitate access and analysis of this rich source of visibility to vulnerable groups who would otherwise
data for education policymaking. For future data be overlooked in regular data collections.
collection, ministry of education staff should be
consulted during household survey design to ensure In other cases, existing data can be greatly improved
that education questions are accurate and useful. to enhance the information on vulnerable children.
Surveys may underestimate the number of out-of-
At present, crucial information is either lacking or of school children among groups for whom exclusion
poor quality on many of the most marginalised out- from education is particularly acute. In regions with
of-school children. For these children, such as those security risks, no data may be collected at all—a
affected by armed conflict or with disabilities, the lack challenge faced by national out-of-school studies in
of quality data is a major barrier to the development Pakistan and other conflict-affected countries. While
of effective, evidence-based policies. it is often said that children with disabilities are likely
to comprise a significant proportion of out-of-school
Targeted data collection initiatives and analyses are children, precise and reliable data on their situation
needed to close data gaps. Out-of-school estimates are rare. Until recently, only data on the most visible
from household surveys are likely to underestimate or severe disabilities were reported. The availability
the out-of-school population because they often and quality of information on children with disabilities
omit—by design—many of the most vulnerable has been improved by a shift in focus towards their
groups of children (Carr-Hill, 2013). Homeless functioning, as well as mild and moderate disabilities
children, those in institutions (such as care homes, and impairments, which is more useful for designing
orphanages and hospitals), refugee camps, and policies and interventions. Yet, these advancements

have not reached all data collection systems. move in and out of the education system (UNICEF
Similarly, surveys often report on the languages and UIS, 2012h). It would, however, be a challenge to
spoken at home, but information is rarely collected implement national ID numbers or similarly advanced
on the language of instruction at school—information systems in less-developed countries, including many
that is crucial to understand the impact of language of the countries with large numbers of out-of-school
barriers on school attendance. The data gaps for children and adolescents.
these three vulnerable groups are discussed in more
detail in Chapter 3. The increasing complexity of education management
information systems requires updated procedures for
Lastly, while information on the characteristics data entry, processing and analysis to harness the
of out-of-school children comes mainly from true potential of these advancements. Such a system
household survey data, improvements in education must also go hand in hand with a comprehensive
data management systems mean that increasingly legal framework to ensure that firstly, confidentiality
administrative data can be used to identify out-of- of data is maintained and that, secondly, all children
school children. In countries with relatively robust can easily acquire the documentation necessary to
information systems, such as many countries in register in school.
the CEE/CIS region and Latin America and the
Caribbean, civil registry data are being linked with These issues are discussed in more detail in
school census data, using unique ID numbers for Chapter 3 as we turn to the system-wide barriers
each child. The development of such a national and solutions to the global challenge of out-of-school
identification system in Turkey, for example, has children and look more closely at who these children
allowed more accurate monitoring of children as they are and what is keeping them out of the classroom.

Chapter 3
Barriers and policy solutions
3.1 INTRODUCTION include social barriers, such as discrimination against

girls, financial barriers such as school fees, and
The faltering global progress towards the EFA goals practical barriers such as the sheer distance to the
and the education-related Millennium Development nearest school. Responses to improve access must
Goals has varied significantly across countries. Some address all barriers that keep—or push—children
have managed to considerably expand education out of school. These responses may affect the entire
access and are now focusing specifically on the most education system, with the abolition of school fees
disadvantaged children—the so-called ‘last 10%’— being a prime example, or may be targeted towards
who have not yet been reached. Other countries, children who face particular disadvantage, such as
however, face persistent and widespread challenges children with disabilities who need specific support
that continue to leave large numbers of children and equipment.
excluded.
This is followed by an examination of the specific
The chapter begins by analysing responses that barriers and targeted policies that have a particular
span entire education systems. These system-wide impact on some of the most marginalised groups
responses involve changes in government policy of out-of-school children as identified in Chapter 2:
designed to improve access to school, the quality of those affected by conflict, girls, child labourers,
the education on offer or its affordability. In general, children whose first language is not the language
such ‘universal’ responses will have some impact on of instruction and children with disabilities. Such
every child in the country. The chapter explores how barriers often work in combination to deny children
such system-wide responses might differ in countries an education, with severe challenges faced by, for
that still have many miles to go in their journey example, a refugee girl with disabilities or a boy from
towards universal completion of primary education an ethnic minority labouring to support his family.
and those that are in the final mile. On the latter, the
chapter describes ongoing system-level challenges In countries in the final mile, it is clear that there is an
and initiatives in regions such as Latin America urgent need for specially-targeted efforts to overcome
and the Caribbean and CEE/CIS countries, which the particular barriers that keep the hardest-to-reach
are approaching universal completion of primary children out of school. They will not be reached simply
education. by business-as-usual approaches that expand existing
education systems still further. Instead, there needs
The chapter then describes the main causes to be a shift towards greater equity in education,
of persistent exclusion from education and the moving away from systems that allocate resources
responses that are needed to ensure the full uniformly and towards systems that allocate resources
educational inclusion of every single child. The according to actual needs of marginalised children. In
barriers that deter children from going to school many countries, the shift to more equitable resource

allocation should be complemented by advocacy The chapter confirms that the barriers to universal
campaigns to reduce persistent cultural and social education are complex and interlinked, and that the
barriers, such as a bias against educating girls or ability to respond to them is constrained by a lack
prejudice against ethnic minorities. of data and coordination among stakeholders to
deliver a comprehensive response. The responsibility
The Global Initiative on Out-of-School Children for the information, capacity and scale-up that are
has shown that all countries must, to some extent, needed to respond to the system-wide and specific
adopt both system-wide reforms and targeted barriers faced by children lies not only with national
responses. Countries that are approaching universal ministries of education but also with any ministry that
access to education need to focus, in general, implements programmes for vulnerable children or
on targeted responses that address the specific collects data on their situation. As well as examining
barriers faced by the most marginalised children. specific policy barriers and solutions, this chapter
Countries that still have large populations of out- attempts to pull together what we know about the
of-school children, on the other hand, usually need barriers, what we do not know about them and how
to balance targeted interventions with broader to bridge the gap between the two.
system-wide reforms. In these countries, targeted
interventions alone cannot compensate for weak 3.2 SYSTEM-WIDE BARRIERS, SYSTEM-
education systems, and the emphasis has to be WIDE SOLUTIONS12
on investment to strengthen and expand these
systems, combined with a sharp focus on inclusion “There are many reasons [why children are out
and the quality of education. Indeed, in many of school]. First and foremost, financial reasons:
developing countries, resources are skewed in the most Congolese parents are poor, even those
opposite way, with more resources for the most who do work earn little and the salary is paid
advantaged: urban, more affluent areas receiving when the employer decides to give it. It’s difficult
more funding per student than poorer, rural areas. to manage; with the salary you pay the rent,
Particularly in sub-Saharan Africa, a first crucial step education and medical expenses, transportation
toward equity in educational spending is working and more. Parents say: I can’t pay for it all; I
towards achieving equality in resource allocation, have to choose one or two children [to attend
such that resources are distributed equally among school]… the others must wait.”
different parts of the country (UNESCO 2010b). Educator, Democratic Republic of the Congo
(UNICEF and UIS, 2013d)
The issue of teaching and learning quality cuts
across all countries and is seen increasingly as a vital Countries with the greatest distance to travel
component of efforts to achieve universal primary
education. Some have called for a move away from At the global level, a handful of countries account for
language that refers to ‘education for all’ to a language one-half of all out-of-school children. These are the
that supports more directly the concept of ‘learning countries that have the greatest distance to travel to
for all’, given that learning is the ultimate purpose of achieve universal primary enrolment and completion.
schooling. The rapid expansion of efforts to improve Many are characterised by instability and conflict,
learning quality is an opportunity to address learning as well as extreme poverty. Yet, despite needing
in ways that avoid the inequitable patterns that have proportionally more support, they are also the
characterised expansions in access—urban boys first, countries that have the greatest difficulty accessing
then urban girls, then rural boys and rural girls, with external expertise and financial support, largely due
the most marginalised children always the very last of to their adverse political environments. For example,
all. This chapter examines policies to improve access the Democratic Republic of the Congo and Nigeria,
and learning of children who are often marginalised
12 This section draws on “Analysis of System-Wide Issues in Latin America
from mainstream education. and the Caribbean”, written by Janet Lennox (UNICEF).

which are both vast countries that are home to many can only be addressed through additional investment
millions of out-of-school children, have received in classroom construction, teacher training and
disproportionately little external support. The same is the production of learning materials. But even a
true for smaller countries, such as the Central African substantial expansion in the numbers of classrooms
Republic and Chad. and teachers will not be enough to reach the most
disadvantaged children—those who are kept out
In virtually every region, a relatively small number of school by extreme poverty, discrimination or the
of countries account for a disproportionately large threat of violence.
percentage of children out of school (see Figure 2.4
and Appendix IV). Two countries, Sudan and Yemen, Non-formal education programmes can play a
account for three-quarters of the out-of-school crucial role in providing second-chance education
population in the Middle East and North Africa. for out-of-school children and expanding educational
Burkina Faso, Chad, Ghana, Mali, Niger, Nigeria and opportunities to areas beyond the reach of the
Senegal account for the same proportion in Western mainstream public school system. However, it
and Central Africa. Household survey data indicate is important that such educational opportunities
that Ethiopia, Kenya, Somalia, South Sudan and provide a recognised pathway into the formal system.
Tanzania account for a significant proportion of out- Equivalency programmes have been developed in
of-school children in Eastern and Southern Africa. countries with large out-of-school populations, such
In South Asia, Pakistan alone accounts for more as Afghanistan, Burkina Faso, Ethiopia, Ghana, Mali
than one-half of the out-of-school children in the and Zambia, as well as countries in South Asia and
region, while Afghanistan also has a large number Southeast Asia, to bridge formal and non-formal
and proportion of out-of-school children according to education by linking curriculum and developing
household survey data. frameworks to recognise outcomes of prior learning
(Yasunaga, 2014). In Ghana, for example, the
In these countries, as well as in those with smaller Complementary Basic Education Policy targets
populations, such as the Central African Republic, out-of-school children aged 8 to 14 years, based
Chad, Djibouti and Mali, exclusion is a broad-based on positive experiences of the flexible school model
phenomenon that cuts a wide swathe through the (UNICEF and UIS, 2012d).
school-aged population. The principal barrier in
these countries is that there simply are not enough In all of these countries, poverty plays a crucial
classrooms and teachers to enable all children to role in keeping children out of school. Not only are
go to school. For example, in the Central African children from poor families less likely to be offered
Republic, Chad and the Democratic Republic of an opportunity to go to school, but their parents
the Congo, there are on average more than 80 are far less likely to take advantage of schooling
students in Grade 1 classes, implying a massive opportunities when they are available. In West and
shortage of classrooms and teachers. Other Central Africa, parents have reported that household
essential materials are also in short supply: for every economic hardships were the main reason why
mathematics textbook, there are on average three their children were not attending school. Household
primary students in Djibouti and five students in survey data presented in Appendix IV reveal the
Chad (UIS Data Centre, 2014). To achieve universal extraordinary impacts of inequalities in household
primary education by 2015, 4 million teachers would income on access to schooling, with, for example,
be needed to staff new classrooms and replace differentials in the primary out-of-school rate of
attrition of the teaching workforce (UIS and EFA more than 40 percentage points between children
GMR, 2014b). While a shortage of schools is often in households in the top wealth quintile and those
felt most keenly in particular locations, such as rural in the bottom quintile in 12 countries, all of which
areas and urban slums, the capacity of a country’s are in sub-Saharan Africa, with the exception of
education system is clearly a system-wide issue and Pakistan and Yemen. In Eastern and Southern Africa,

13% that translates into an out-of-school rate for
The single most important barrier to child labourers of around 88%. Here again, instability
education in the countries that have fuels non-attendance with widespread violence in
the farthest to travel is usually the the Swat valley in 2007 leading to a mass exodus
high cost of education for families of internally-displaced people and a marked drop in
such as fees for schooling, informal school attendance.
fees paid to teachers or the loss of
income from a child’s labour Not surprisingly, the single most important barrier
to education in the countries that have the farthest
to travel is usually the high cost of education for
primary school-age children from the poorest families families: whether these are direct, such as fees for
in Kenya and Malawi are over six times more likely schooling, or ‘hidden’, such as informal fees paid to
to be out-of-school than children from the richest teachers or the loss of income from a child’s labour.
families. The correlation between poverty and school
non-participation is also evident in East Asia and the An increasing body of literature documents the
Pacific (Hattori, 2014). positive impact of strategies that address these
costs, with three responses identified as having a
An important insight is the way that poverty strong impact (whether applied universally or targeted
interacts with other factors of exclusion, such as towards particular children): the abolition of school
location and gender, to intensify disadvantage. fees, cash transfer programmes, and school feeding
According to the Global Initiative on Out-of-School programmes.
Children study covering West and Central Africa,
barriers to education include “economic hardships mm The abolition of school fees is a system-
related to family issues, child health problems, wide approach to addressing the cost of
cultural factors and a poor perception of the value education. While such programmes require
of education” together with “the direct and indirect careful management to ensure that quality is not
costs of education, lack of schools, teachers and compromised, they have played a critical role in
equipment, as well as bad teaching practices and enrolment gains in Eastern and Southern Africa
violence at school” (UNICEF and UIS, 2014c). (World Bank, 2009) and to a lesser extent in West
In many countries, it is the poorest children who and Central Africa (UNICEF and UIS, 2014c). This
receive the poorest quality education, served by approach is most effective in reducing the barriers
schools with overcrowded classrooms, insufficient faced by children from poorer families when other
teaching materials and textbooks, high teacher expenses, such as the cost of textbooks, are
absenteeism and poor quality facilities. abolished at the same time as tuition fees.
Pakistan illustrates the range of challenges faced mm Cash transfer programmes (either conditional
by policymakers who must determine what should or unconditional) are most often used as a
be prioritised. In this one country, for example, the targeted intervention to reduce the barriers to
challenges include the yawning disparity in school access for particular groups of disadvantaged
attendance by household wealth and the highest children. These programmes have been effective
urban-rural gap in South Asia, with children in rural in increasing enrolment and attendance in school,
Balochistan having signiﬁcantly higher rates of as well as contributing to reductions in child
exclusion than the national average. As shown in labour. Cash transfers have been implemented
the following sections, Pakistan must also contend successfully in Latin America and the Caribbean,13
with barriers faced by particular marginalised groups,
with its dismal attendance rates for girls throughout 13 A well-known example is Brazil’s Bolsa Familia (Family Grant), which
is conditional on the recipient family ensuring that their children attend
the basic education cycle and a child labour rate of school.

and in some Eastern and Southern African Countries facing these challenges are undertaking
countries like the Basic Education Assistance system-wide approaches to streamline schooling
Model in Zimbabwe. In Bangladesh (UNICEF and pathways of children and to enrol all children
UIS, 2014d), they have stimulated girls’ enrolment into primary school at the intended age and with
(in particular) in lower secondary school. Despite adequate school preparation. In Latin America and
these successes, the spread of cash transfer the Caribbean, one of the most common risk factors
programmes is often hampered by the challenges for not completing a full course of primary education
associated with the testing that is necessary to is over-age enrolment. Over-age students are at
determine whether or not families fall into the much higher risk of dropping out of school early.
eligible target group. The relatively large numbers of over-age children
in many countries is due to a combination of
mm School feeding programmes can also be significant proportions of children who start primary
either system-wide or targeted, and often have school late, stubbornly high grade-repetition rates
significant impacts. The largest such programme (especially in the early years of school), limited pre-
was implemented in India with 120 million school primary education to prepare children for entry into
children benefiting by 2006 and has been credited primary school, and the prevalence of temporary
with a significant positive effect on both school drop-out and re-enrolment. Taken together, these
enrolment and attendance rates. A systematic factors too often result in an age-grade gap from
review of the evidence over the past 20 years the very start of a child’s education—a gap that
finds consistent positive effects of school feeding can widen over time—and may contribute to early
on children’s enrolment and attendance, while drop-out.
its impact on academic achievement is less
conclusive (Jomaa et al., 2011). The cumulative impact of this age-grade gap
often materialises at the transition from primary to
Countries in the final mile secondary education, with the most disadvantaged
children at much greater risk of being pushed out of
Countries that are in the last mile of their journey to the education system early. What begins as exclusion
universal primary school completion face a different that affects a relatively small proportion of the primary
set of challenges. These countries have already school-age population grows to affect a significant
made very significant strides in providing primary percentage by adolescence. In violence-scarred
education to their children, and their efforts are now Honduras, for example, enrolment rates in 2011
often centred on expanded access to pre-primary stood at 89% for 6-year-olds and close to 100%
and secondary education. Despite these advances, for 8- to 11-year-olds, but these were followed by
however, the central challenge of realising the right a precipitous drop to 64% by age 14, meaning that
to education for every child remains unfulfilled. There one out of every three adolescents of this age group
are still some children of primary age who are out of was excluded. As is common in Latin America and
school because they are the most disadvantaged the Caribbean, boys in Honduras were more likely to
and hardest to reach, requiring more complex and be over-age for their grade and, therefore, at greater
often more costly policy responses. These children risk of early drop-out (UNICEF and Asociación Civil
might represent only a small percentage of the total Educación para Todos, 2011).
student population, but their numbers may still be
large. In Brazil, for example, an out-of-school rate of Such barriers to the completion of education are often
only 2.4% in 2009 represented over 730,000 children reinforced by irrelevant curricula, poor pedagogy that
of primary and lower secondary school age (UNICEF fails to impart basic skills to children in primary school,
and UIS, 2012a). and low expectations about the academic potential of

Figure 3.1 Gaps in data, analysis and policymaking on out-of-school children
All out-of-school children in the compulsory school-age population
Out-of-school children
Out-of-school children accurately recorded in government databases, including the
not recorded in any
ministry of education database government database
Out-of-school children in ministry of education database only (usually those who Out-of-school children
not recorded in the ministry
enrolled but left school before completion)
of education database

Out-of-school
children included in analysis and reporting

Out-of-school children supported through targeted
policies and strategies
Out-of-school children supported through targeted

Out-of-school children not receiving any support
policies and strategies

0% 100%
Source: UNICEF and UIS (2014f, forthcoming) DataLink: http://dx.doi.org/10.15220/2014/ed/sd/7/f3.1
certain children, including those from poorer families or One of the simplest and yet most fundamental
other disadvantaged groups. approaches to providing learning opportunities to
children is simply to ensure that these children do
In general, the response to these barriers involves not remain invisible. In another region that is well
system-wide reforms, particularly a commitment within the final mile, CEE/CIS, several system-wide
to expand, or even universalise, pre-primary and barriers, including a lack of information-sharing
secondary education. These are often combined with and coordination among the various ministries
efforts to clarify the regulatory framework to minimise whose programmes target vulnerable children,
over-age enrolment, improvements in teacher training have impeded the development of a more robust
and curriculum reform to improve the relevance of system to monitor the situation. However, efforts
education in children’s lives. These efforts are often to enrol the remaining out-of-school children
accompanied by communications initiatives to dispel (once again, the hardest to reach) and monitor
myths about some children, such as those with the students at greatest risk of dropping out are
disabilities, and their ability to benefit from education. now being supported by efforts to improve data
An example of the latter is the Cero Falta (Zero systems and interagency cooperation. Key data
Absence) campaign in Uruguay, where children, gaps and problems with the reliability of data have
classes and schools are invited to share their been identified in the national Global Initiative on
experiences in an annual competition, with selected Out-of-School Children studies in Kyrgyzstan,
entries awarded prizes and made into short videos. Romania, Tajikistan and Turkey, and have led to the
Another is the follow-up to Brazil’s Global Initiative on development of a Regional Framework for Monitoring
Out-of-School Children study organised around the Out-of-School Children (UNICEF and UIS, 2014f
slogan, Fora da escola não pode! (‘Out of school, forthcoming).
just won’t do!’), which showcases related multimedia
content, including a web-based documentary and Work in these countries has focused on the
a user-friendly website where people can learn development of a complete and accurate national
about the situation in their municipality and add their database of school-age children that can act as the
comments and ideas.14 foundation for evidence-informed policies, in addition
to mapping and streamlining the procedures and
14 www.foradaescolanaopode.org.br

interventions undertaken by multiple ministries and 3.3 ONE-HALF OF THE WORLD’S OUT-
other actors to ensure that all children complete OF-SCHOOL CHILDREN: CHILDREN
their basic education (see Figure 3.1). It requires the CAUGHT UP IN CONFLICT15
closing of the horizontal information gaps between
ministries by improving and formalising relationships “My father went to ten schools [to find a school
to share information on school-age children. place for his child]. Finally, I was registered.
They referred us to one school with all the Syrian
The cross-sector nature of the barriers and children. Now they want to move us to another
responses around out-of-school children means that school farther away. When we go to school,
reporting and policymaking for this group too often no one respects us. My father wanted to cry
lack the sustained, coordinated commitment needed because the principal doesn’t respect us.”
to reach all children excluded from education. Syrian refugee boy aged 14, in Irbid, Jordan
Ministries collect data on and design policies for the (UNICEF and UIS, 2014b)
children for whom they are responsible: ministries
of education for the children who enter school, and Few challenges have more devastating consequences
ministries of health, social protection and justice for for education than armed conflict. Its most immediate
only specific sub-groups of children. To guarantee and grave threats are obvious, from the loss of life
that no child falls through the cracks, there is a need to the deliberate targeting of schools, teachers and
to understand and streamline the way in which these pupils and the abandonment of schooling as families
authorities at the national and community levels flee for their survival. Less apparent is the creeping
intervene to ensure that all children enter school and erosion of vital educational resources—human as well
complete a full cycle of education. as financial—and the cumulative and life-long impact
on the children who miss months, or even years, of
In East Asia, Cambodia has demonstrated that it schooling.
is possible to conduct a relatively low-cost survey
of out-of-school children that illuminates the faces
behind the numbers—identifying individual children Most out-of-school children and
district by district and region by region, identifying the adolescents in conflict-affected
particular challenges they face, and engaging both countries are girls
schools and NGO partners to bring these children to
school. Cambodia has also pioneered the use of a
cross-sectoral survey instrument to identify children The scale of the problem is immense. Children in
with disabilities and to provide them with the services conflict-affected countries account for just 22%
they need, as described in Section 2.6. Viet Nam of primary school-age children, but one-half of all
also incorporated provincial level surveys in its Out- children who were denied an education in 2011—and
of-School Children study allowing a more fine-grained their share of the global out-of-school population is
approach to the varied contexts that exist, especially rising, up from 42% in 2008 (UNESCO, 2014a).
in the mountainous areas and for specific ethnic
groups. This has been linked to the development In 2011, the largest numbers of primary school-
of new policies around the language of instruction age out-of-school children in conflict-affected
(UNICEF and UIS, 2014h). areas—12.6 million—were found in sub-Saharan
Africa. A further 5.3 million children out of school as a
This chapter now turns the spotlight on to the result of conflict lived in South and West Asia and—
children who are out of school, examining the barriers prior to the Syria crisis (see Box 3.2)—4 million in the
they face to their education and highlighting a range
of policy solutions and responses that enable them to 15 This section draws on “Barriers to Education in Conflict-Affected Countries
and Policy Opportunities”, a background paper prepared for this report by
take their place in the classroom. Patricia Justino, Institute of Development Studies, United Kingdom.

Figure 3.2 Percentage of primary school-age children out of school, both sexes, 2012
Conflict-affected countries Countries not in conflict
ERITREA 66
LIBERIA 59
SOUTH SUDAN 59
SUDAN 49
DJIBOUTI 42
CÔTE D'IVOIRE 38
EQUATORIAL GUINEA 38
NIGER 36
CHAD 36
NIGERIA 34
BURKINA FASO 33
MAURITANIA 30
GUINEA-BISSAU 29
CENTRAL AFRICAN REPUBLIC 28
PAKISTAN 28
MALI 27
GAMBIA 26
GUYANA 25
GUINEA 25
NAURU 24
SENEGAL 21
0 10 20 30 40 50 60 70 %
Notes: Data for Djibouti refer to 2013; data for Chad, Liberia, South Sudan and Sudan refer to 2011; data for Guinea-Bissau and Nigeria refer
to 2010; data for Côte d'Ivoire refer to 2009. Conflict-affected countries are identified based on the list provided in the EFA Global Monitoring
Report 2013/2014 (UNESCO, 2014).
Source: UNESCO Institute for Statistics, August 2014 and UNESCO, 2014b DataLink: http://dx.doi.org/10.15220/2014/ed/sd/7/f3.2
of the Congo, Sierra Leone and Somalia. In the

Almost one-third of out-of-school Democratic Republic of the Congo, for example,
adolescents of lower secondary the eastern provinces with recurrent violence face
school age worldwide live in the greatest challenges: 40% of primary school-age
conflict-affected countries children in North Kivu are out of school, compared
with a national average of just under 27% (UNICEF
and UIS, 2013d).
Arab States (UNESCO, 2013). The picture is just as
grim for secondary schooling: one-third of the world’s In addition to the millions of out-of-school children
out-of-school adolescents of lower secondary school in countries affected by armed conflict, there are
age lived in conflict-affected countries in 2011. And millions more who live in countries plagued by the
most of those out of primary or secondary school in growing violence linked to organized crime, the
conflict-affected areas are girls (UNESCO, 2013). trafficking of drugs or people, and gang wars. This is
a particular issue in some countries of Latin America
Of the 21 countries that have out-of-school rates and the Caribbean, where Honduras, for example,
above 20%, 12 are conflict-affected according to has a homicide rate of more than 90 deaths for
available administrative data (see Figure 3.2). This every 100,000 people, three times higher than the
excludes other conflict-affected countries where rate in Afghanistan and Iraq—countries with ongoing
such administrative data are not available but where armed conflicts (UNODC, 2014). Continued vigilance
household surveys confirm similarly high rates, and monitoring are needed in non-conflict contexts
such as Afghanistan, the Democratic Republic which experience a high level of internal insecurity,

due to inter-tribal warfare, street violence or gender-
based violence. These countries can also benefit Only 65% of children in conflict-
from lessons learned in supporting education in affected countries reach the
conflict-affected countries. The impact on education final grade of primary school,
for children in the affected areas can be every bit as compared with 86% across other
severe as for children in countries facing all-out war developing countries
(UNESCO, 2011).
This failure means that children of primary school in the extreme, given the importance of education
age in fragile and conflict-affected situations16 are in preventing conflict, in creating a vital sense of
nearly three times more likely to be out of school than normalcy for children during conflict, and as an
children in other parts of the developing world (World essential part of post-conflict recovery in its wake.
Bank, 2011). Conflict means that children in school
are more likely to drop out, with only 65% of children The barriers to the education of children affected
in conflict-affected countries reaching the final grade by conflict are formidable, but evidence from the
of primary school, compared with 86% across other Global Initiative on Out-of-School Children and other
developing countries. Children who miss school sources reveals the potential for a more concerted
during episodes of armed violence tend not to go and comprehensive response, even at the height of
back (UNESCO, 2011). As a result, the countries in the violence.
which they live have some of the lowest literacy levels
in the world. The barriers
The loss of education deprives children of at least “As a result of the conflict, my family had to
some protection from the sexual exploitation, leave home and was pushed into poverty. The
physical attacks and recruitment into armed groups continuous displacement and being separated
that are grotesque features of warfare, and they lose from my family, relatives and friends affected my
the precious sense of ‘normalcy’ that education can mental ability and totally destroyed my eagerness
provide (Norwegian Refugee Council, 1999; ICWAC, for studies. Schools were also closed and their
2000). They miss the chance to acquire vital skills activities ceased.”
for the future and the long-term impact includes Boy from Kilinochchi, Sri Lanka
diminished employment prospects and earnings in (UNICEF and UIS, 2013c)
later life (Justino, 2011), which may aggravate the
risks of an outbreak or renewal of violent conflict Supply barriers
(Justino et al., 2013).
Armed conflict destroys or consumes the infrastructure
The scale of the response to the impact of armed and resources needed to keep the supply of education
conflict on education has been totally inadequate. As flowing. During the Gaza emergency that began in
millions of children elsewhere take their rightful place July 2014, for example, displaced people (among
in the classroom, the lack of progress for children them children) sheltering from the violence in school
in conflict-affected countries serves as a constant buildings were killed or injured when those schools
reminder of the failure of political will, effective policies came under fire (OCHA, 2014).
and adequate resourcing, as well as the logistics,
to tackle this problem. This failure is short-sighted Conflict scatters communities as people flee, with
the number of people living as refugees from war
16 This statistic relates to fragile and conflict-affected situations as outlined
in the 2011 World Development Report on Conflict, Security and and persecution in 2013 exceeding 50 million for the
Development: countries or territories that have a harmonised average
Country Policy and Institutional Assessment (CPIA) rating of 3.2 or less (or first time since World War II, and available evidence
no CPIA), and/or have or have had a UN and/or regional peace-keeping or
peace-building mission during the past three years. suggests that one-half of the world’s refugees are

Box 3.1 Follow the money: The funding gap
Following the money for education in conflict-affected countries often reveals the entrenched security
interests of powerful donor countries, with an emphasis on countries seen as strategic priorities, such as
Afghanistan, Iraq and Pakistan (UNESCO, 2011). The global education community has called for a modest
4% of humanitarian aid to be allocated to education, but the share of humanitarian aid for education is
actually falling, from 2.2% of aid in 2009 to just 1.4% in 2012. This represents the largest funding gap for any
humanitarian sector, a gaping hole of US$221 million in 2012 (UNESCO, 2013), and languishes far below the
4% target.
When looking in detail at the 19 consolidated humanitarian appeals made to the United Nations Office
for the Coordination of Humanitarian Affairs (OCHA) in 2013 (see Figure 3.3), only 4 of the 16 countries
with requests for education financing received funds equal to at least 4% of humanitarian aid: Somalia
(4%), the Syrian Arab Republic (4%), Sudan (6%) and the Central African Republic (8%). In the case of the
Central African Republic, however, 81% of the resources received for education were for school feeding
programmes.
Figure 3.3 Consolidated appeal requests and funding for education received by conflict-affected
countries, 2013
Unmet requirements Funded % of total humanitarian funding to education
% 6%
1%
100
In the Democratic Republic of the Congo, only 9% of requests for the
90 education sector were met. Of the total available funds from
80 the appeal, just 1% was for the education sector.
4%
4%
70
In million US$
60
50 3% 3% 8% 1%
1%
40
1%
30 2%
3% 0% 1% 0%
0%
20
10
0
SYRIAN ARAB SUDAN SOUTH SOMALIA PALESTINE CENTRAL CONGO, MALI YEMEN AFGHANISTAN BURKINA PHILIPPINES NIGER KENYA MAURITANIA CHAD
REPUBLIC SUDAN AFRICAN REP. DR FASO
Source: Office for the Coordination of Humanitarian Affairs, 2014 DataLink: http://dx.doi.org/10.15220/2014/ed/sd/7/f3.3
of state presence and local coherence—are often

One-half of the world’s more targeted for violence by armed groups intent on local
than 50 million refugees are control (O’Malley, 2007, 2010; UNESCO, 2011).
thought to be children
Schools in areas affected by violence may become
temporary shelters for those who have been uprooted
now children (UNHCR, 2014). It also skews the or may even be closed down. Add to this the
equity of education, with some children even more problems of recruiting or retaining teachers in areas
excluded than others in times of war. affected by conflict, the disruption of examinations
and education supplies, and the woeful lack of
Reports from several conflict-affected countries show funding for education in conflict-affected countries
that schools, teachers and students—visible symbols (see Box 3.1), and even schools that manage to stay

open will struggle to provide a reasonable education
(UNESCO, 2011). Such effects have been seen “Typically, children in conflict zones and
in a mass of conflict-affected countries, including broken-down states have been provided
Afghanistan, Colombia, the Democratic Republic of with food and shelter as refugees, but
the Congo, Iraq, Nepal, Niger, Pakistan, Palestine, few receive any education”
Somalia, Sudan, Thailand and Zimbabwe (O’Malley, Gordon Brown,
2007, 2010; UNESCO, 2011). UN Special Envoy on Education
The uprooting of entire communities will, inevitably,

derail the supply of education. While some schooling Primary education also remains inaccessible for
may be available in camps for displaced and refugee many of Kenya’s refugee children. The Eastern and
children, it is often disorganised, temporary, under- Southern Africa Global Initiative on Out-of-School
resourced, overcrowded and limited to primary Children study shows that while many refugee
education (UNHCR/OSCE, 2002; Watkins, 2013). children in Nairobi, for example, are born in Kenya,
These children may be unable to access local their lack of birth certificates hinders their enrolment
schools beyond the camps because of restrictions in public schools. Nairobi’s city council primary
on their movements, security fears or those schools schools require refugee parents and guardians to
being unable to cope with more children—a produce a proper registration document, such as
challenge facing Syrian refugee children in Lebanon a UNHCR-mandated certificate, in addition to their
(see Box 3.2). child’s birth certificate (UNICEF and UIS, 2014a).
Box 3.2 A lost generation? The children of the Syrian Arab Republic
The Syrian conflict has devastated its children’s education. While estimates on the precise impact differ, they
all confirm that a large proportion of Syrian children have missed out on education since the outbreak of
violence in 2011. Data from the Syrian Ministry of Education (MOE) show that enrolments in Grades 1 to 12
fell by more than one-third (35%) between the 2011/2012 and 2012/2013 school years. The MOE estimates
that nearly one-half of those children have left the country, while the remainder are still in the Syrian Arab
Republic but have dropped out of school. Another 1.3 million children (approximately) attend school
irregularly and are at risk of dropping out (UNICEF and UIS, 2014b).
Many children who have left the Syrian Arab Republic with their families are in Lebanon, where at least
300,000 are out of school. If the Syrian refugee population in Lebanon were a country, it would have one of
the world’s lowest primary school enrolment rates—lower than some of the worst-performing countries in
sub-Saharan Africa. The net enrolment rate among Syrian refugee children of primary and lower secondary
school age (aged 6 to 14 years) is around 12%—less than one-half of the level in South Sudan. For children
of upper secondary school age, probably below 5% are attending upper secondary education (Watkins,
2013). By contrast, in 2010 before the conflict began, the Syrian Arab Republic had enrolled virtually all of its
primary school-age children in school and was nearing universal enrolment of lower secondary school-age
adolescents at 90%.
Lebanon faces immense pressures on all basic services but particularly on education. Its schools have
thrown open their doors to the refugee children of the Syrian Arab Republic but are now stretched beyond
breaking point: absorbing every refugee child would be equivalent to New York taking in the entire school
populations of Washington D.C. and Chicago (Watkins, 2013).
This challenge cannot be resolved through short-term humanitarian appeals, which are already chronically
under-funded. Providing education for every Syrian refugee child requires a strong international response and
strengthened partnerships, backed by an international action plan of an estimated US$165 million per year
(Watkins, 2013) to get every refugee child into the classroom. Without such a response, there are growing
concerns at the prospect of a ‘lost generation’ of Syrian children.

The situation may be even worse for children who only intensifying their exclusion. Armed conflict can
are internally displaced within the border of their own also sweep away a key incentive for education—the
conflict-affected country and beyond the reach of chance of a decent job. To put it simply, it may no
international educational support for cross-border longer pay to send children to school where job
refugees or domestic educational services. opportunities have been shattered by the destruction
of industries, markets and infrastructure, eroding
While displacement can be temporary, as in the prospects of higher incomes for households
Timor-Leste (Justino et al., 2013), it often lasts for that invest in the education of their children (Santos,
decades (as in Colombia, the Democratic Republic 2014; Shemyakina, 2011; Chamarbagwala and
of the Congo, Palestine and Sudan), leaving whole Moran, 2009).
generations without access to education and its
important social structures (Watkins, 2013). More evidence is needed on how this loss of incentive
plays out across different conflict-affected contexts
‘Winners’ have been known to punish ‘losers’ by and populations. It seems certain, however, that
prioritising school enrolment for some groups or many parents have to make tough choices about
segregating schools along the lines of language the costs and benefits of education for their children
(Timor-Leste), race (South Africa), ethnicity (pre- in the face of deteriorating job prospects, continued
1994 Rwanda) and religion (Northern Ireland) (Bush insecurity and growing poverty. Most reports from
and Saltarelli, 2000; Shemyakina, 2011). Existing the Global Initiative on Out-of-School Children from
disparities linked to location and income are often conflict-affected countries confirm the link between
reinforced: the Colombia Global Initiative on Out- conflict, poverty and the loss of education, as seen
of-School Children study, for example, reports a in Colombia, the Democratic Republic of the Congo,
geographic divide between children in urban and Pakistan, Sri Lanka and Tajikistan (UNICEF and UIS,
rural areas, with the worst educational outcomes 2012c; 2013b; 2013c; 2013d; 2013e). Families
found among rural populations in areas at risk that have been plunged into extreme poverty and
of armed attacks. These effects only aggravate destitution by armed conflict may well remove their
economic disparities, with obvious consequences children from school to augment household income
for the likelihood of future conflict (UNICEF and UIS, (Justino et al., 2013; Rodriguez and Sanchez, 2009;
2012c, 2009; Watkins, 2013). UNICEF and UIS, 2014b, 2013a) or to fill the shoes
of an adult breadwinner lost to recruitment, death or
Demand barriers injury (Akresh and de Walque, 2008; Merrouche, 2006;
Shemyakina, 2011; Rodriguez and Sanchez, 2009).
In the worst scenarios, violent conflict undermines
demand for education by taking the lives of pupils No study has, to date, examined the links between
and teachers, instilling relentless fear and insecurity, bereavement during armed conflict and reduced
and by pulling children out of school and into active demand for education, although a link seems likely.
combat. What has been studied, however, is the impact of
armed conflict on child and maternal health and
Beyond the school gate, conflict reduces demand nutrition. Children exposed to high levels of violence
for education by exacerbating poverty and poor are often inches shorter than other children—a
health and by reducing the returns to education. sign of long-term undernutrition (Bundervoet et
Families that are already impoverished, including al. 2007; Alderman et al., 2006; Bundervoet and
the displaced and refugees, may be unable to cover Verwimp, 2005; Guerrero-Serdan, 2009). Similarly,
the costs of education, such as uniforms, school a combination of stress and limited access to health
fees, school lunches, books and other materials, services during pregnancy in conflict-affected regions
certificates, transportation and so on (Shemyakina, in Jammu and Kashmir (Parlow, 2012), Colombia
2011; UNICEF and UIS, 2014b), with armed conflict (Camacho, 2008) and Nepal (Valente, 2011) has

resulted in children being born with low birth weight. depressed and socially withdrawn (UNFPA, 2006),
Malnourished children are more likely to miss school, and those who return to the classroom are likely to
fall sick and struggle to concentrate in the classroom lag behind or abandon their studies prematurely.
and are, therefore, more likely to drop out of school.
Breaking the barriers
Exposure to violent conflict can result in deep
psychological trauma and stress among children, The reform of education systems is no easy
as well as greater family stress: factors that can limit task in environments characterised by shattered
not only their school attendance but also their school infrastructure and social institutions. The question
performance and future development. is what to prioritise and how? Should the focus be
on a good quality education as part of wider social
Demand for education is reduced by fear and reforms to prevent conflict, keeping children in school
insecurity. Parents may have valid concerns about during conflict, or the reform of education in the wake
keeping their children in school during armed conflict, of conflict? The answer is a combination of all three,
given that schools are often in the firing line and that backed by strenuous efforts to close the data gap on
teachers and pupils are often seen as legitimate education in conflict-affected areas.
targets for violence, rape and other sexual violence,
as well as forced recruitment. Nothing has illustrated
this problem more vividly than the kidnap of more
A three-pronged approach is needed:
than 200 girls from their school in northern Nigeria
reforms that go beyond education,
in early 2014 by Boko Haram, an armed group
measures to keep education going
opposed to ‘western’ education (OHCHR, 2014). The
during conflict, and reforms in the wake
consequence is, as one United Nations report puts
of conflict—all backed by better data
it: “a growing fear among children to attend school,
among teachers to give classes, and among parents
to send their children to school” (United Nations, While more resources are needed for education
2010). during armed conflict, the focus needs to broaden
beyond education itself to address what is happening
Fear can linger long after the violence ends, beyond the school gates—the economic and social
particularly in contexts where the original conflict exclusion that can ignite violence and, in turn, push
lasted for years and might reignite (Justino, 2012). children out of school.
Fear may also aggravate gender inequalities,
restricting the movements of girls, in particular, in the Reform has to go beyond education. The
wake of conflict (UNICEF and UIS, 2012b, 2013a, education sector alone cannot address the complex
2013b). and interlinked barriers to education presented by
armed conflict. Formal education systems must be
One of the most extreme demand-side barriers to part of wider reform efforts that transform economic
education is the recruitment of children (forced or and social institutions to address the inequalities,
voluntary). There is no agreed estimate of the number poverty, vulnerabilities and, indeed, the aspirations
of child combatants worldwide, given the severe of children and families that can work for or against
data challenges in conflict-affected areas, but it is conflict.
clear that children are recruited as soldiers, porters,
messengers, cooks and sexual slaves (UNHCR/ This requires education systems that are designed
OSCE, 2002; USAID, 2007). Their chance of any and implemented alongside wider economic and
education is, effectively, zero, and those who survive social policies. The most critical are economic
and escape are unlikely to return to school (Blattman interventions to reduce poverty, undernutrition and
and Annan, 2009). Many are left traumatised, the need for child labour (see Section 3.5).

religious, ethnic or language groups; and children
A non-formal education programme with disabilities.
in Iraq—the accelerated learning
programme—is now reaching more Keep education going. Education systems must
than 60,000 students a year, up from do whatever it takes to keep children learning—even
around 17,000 in 2007 at the height of conflict—while recognising that some
children are more likely to be out of school than
others. The response must also be flexible enough
Standard anti-poverty measures may have little to respond to fast-changing situations without losing
impact in countries affected by conflict, where acute sight of the need for long-term engagement, not only
economic distress requires a more robust response. during a conflict but also once the conflict has ended.
Evidence from the Global Initiative on Out-of-School
Children studies and wider research suggests that Sector-wide approaches are crucial, with the
the demand for schooling may be augmented maintenance of early childhood education providing a
by economic-support interventions, including pathway into—and reducing drop-out from—primary
social protection safety nets and cash transfer education, while continued support for secondary
programmes. education and teacher training will enhance the
benefits of completing a primary education.
Such approaches are relatively new among conflict-
affected populations, but results are encouraging: Common measures are seen across a number of the
the Familias en Acción cash transfer programme in Global Initiative on Out-of-School Children studies
Colombia, for example, is pulling children into school to address the loss of education among displaced
in conflict areas (Bozzoli et al., 2011) and reducing and refugee children, including the hiring of teachers
overall school absenteeism (UNICEF and UIS, drawn from displaced populations, advocacy and
2012c). While children in conflict areas still do less negotiations with host governments and schools
homework and miss more days of school (Bozzoli to let refugee children attend school, temporary
et al., 2011) than other children, a complementary documentation for those who have lost or do not
school feeding programme provides a simple possess key documentation, and the provision of
incentive to go to school (UNICEF and UIS, 2012c). non-formal catch-up programmes for youth.
Other examples include the Punjab Education
Voucher Scheme and the Benazir Income Support Non-formal education (NFE) needs special attention
Programme in Pakistan (UNICEF and UIS, 2013b) in conflict-affected countries, including system-
and the Samurshi Poverty Alleviation Programme in wide approaches to coordinate diverse providers
Sri Lanka (UNICEF and UIS, 2013c). in each phase of the emergency and the inclusion
of NFE in national plans of action from the earliest
Education reforms also need to ensure that the reconstruction phase (Yasunaga, 2014). Several
schooling on offer is effective, inclusive and equitable. NGOs and UN agencies in the Middle East and
Wider social policy reforms should be complemented North Africa region support non-formal alternative
by interventions to reduce the economic burden education programmes as a flexible response to
of schooling, such as the abolition of school fees education exclusion. Experiences from Iraq and
and the provision of subsidised uniforms and other Sudan show that such programmes play a key role
school materials (UNESCO, 2011). While such during protracted crises, where large numbers of
interventions have a universal impact on enrolment, children have missed years of schooling. In Iraq,
more specific approaches are needed to reach UNICEF has intensified its support to the accelerated
the children most likely to miss out on schooling— learning programmes, expanding the number of
whether in peacetime or in times of conflict: the very students benefiting from such programmes from
poorest; girls; child labourers; children from particular around 17,000 in the 2007/2008 school year to more

than 60,000 in 2010/2011 in close collaboration with Rwanda has made a remarkable recovery from
national authorities and local communities (UNICEF the impact of the 1994 Genocide, during which
and UIS, 2014b). schools had become sites of betrayal and massacre.
Indeed, the country’s recovery has been particularly
The second-chance Youth Education Programme marked in the education sector, which was seen
supported by the Norwegian Refugee Council spans as fundamental in tackling the historic inequity and
ten countries, targeting displaced, returnee and exclusion that fuelled the conflict, and went through
other vulnerable youth aged 14 years and older, a series of reforms to promote peaceful social
and combines literacy training with practical skills transformation. The key approach was to create
for employability. One barrier has been the lack of an inclusive education system that achieved good
trained teachers, with Sudan among only a handful learning outcomes for every student, with an emphasis
of countries where the government has seconded on the ability of the system to mitigate the sources of
teachers to the programme and paid their salaries violent conflict in Rwandan society. This was backed
(UNICEF and UIS, 2014b). by intense personal commitment, with ministers and
other leaders going from province to province to
At the most basic level, children need to be safe persuade parents to return children to their schools.
on the way to and from school and while in the As a result, the World Bank Country Study of 2004
classroom. Examples of physical protection include noted that the numbers of children enrolled in primary
guards to protect schools, provision of housing for school from 1994 to 1999 surpassed the number that
children close to schools, accompanying children would have been enrolled had the system expanded
to and from school, and transportation safety. at its pre-conflict rate (World Bank, 2013).
In Palestine, a system of so-called protective
presence groups has seen international volunteers
accompanying children to school when feasible,
The Back on Track Programme for
given the continuing volatility of the situation there.
conflict-affected countries restored
In the Gaza Access Restricted Areas, parents can
access to school for 6 million
call teachers to get advice on safe routes to school
children in 40 countries and
and an alert system has been introduced to support
territories between 2006 and 2010
children’s safety (UNICEF and UIS, 2014b).
When it is simply too dangerous to go to school, A multi-country response was the Back on Track
tried and tested alternatives have included the Programme on Education in Emergencies and Post-
organization of temporary schools in religious Crisis Transition, which operated on the premise that
buildings or at home, summer sessions and distance countries with strong internal capacities are less likely
learning programmes. to slip back into armed conflict or be overwhelmed
by the next disaster. This four-year programme
Reform education in the wake of conflict. launched in 2006 by UNICEF, the Government of the
Education is critical for the economic and social Netherlands and the European Commission aimed
recovery of households and countries affected by to help countries tackle the precarious conditions
conflict, but there are clear constraints to the reform that were preventing them from accessing global
of education systems in the wake of violent conflict. education funding. It helped countries to build
Countries may lack the financial capacity to rebuild their own capacity, reform their education sectors
schools—let alone reform education systems—while and deepen partnerships between governments,
trying to meet many other pressing needs, from communities and civil society. By the end of 2010,
housing to clean water. It can be done, however, as the programme had restored access to schools
Rwanda has demonstrated. and improved education quality for almost 6 million

children in 40 countries and territories (Back on Track violence. Educating and training young people to act
and UNICEF, 2011). as constructive and productive citizens may reduce
the risk of future conflict, while addressing family
The programme had a sharp focus on access and community vulnerabilities may help to break the
to education for displaced, refugee and returnee economic and social barriers that keep children out
children; girls and children from ethnic minorities, of school.
including concrete measures such as the provision
of temporary learning spaces; the construction and Filling the data gap
rehabilitation of schools; back-to-school campaigns;
and education and recreation kits, as well as support While data always matter for effective policymaking
for early childhood development (ECD) programmes. and efficiency in government spending, they are
As part of these efforts, a total of 4,700 schools were particularly crucial during conflicts. The building
either constructed or rehabilitated. UNICEF country of equitable and sustainable education systems
offices reported improved enrolment and retention to reduce the number of out-of-school children in
rates in the target locations: for example, 23 districts conflict-affected countries requires at least some
in the Acholi sub-region, northern Uganda reported good data and analysis, and there are ‘quick fix’
higher net intake and primary completion rates approaches that can, at the very least, provide vital
as a result of the programme (Back on Track and information on the scale of the problem.
UNICEF, 2011).
At present, however, the lack of timely evidence
Work is also needed to improve post-conflict on what is happening is, in itself, a barrier to the
disarmament, demobilisation and reintegration educational inclusion of children in conflict-affected
(DDR) programmes that aim to reintegrate former areas. The data that are available tend to describe
child combatants into civilian life. The effectiveness the impact of conflict after it has happened, when
of these programmes, which provide money, it is too late to address its most devastating effects
social and psychological support and employment (Montjourides, 2013).
training for ex-combatants, has been mixed, often
because the reasons why individuals join armed Despite recent improvements in data availability and
groups—particularly children—are poorly understood the collection of qualitative and quantitative data in
(Humphreys and Weinstein, 2008; Guichaoua, 2009). conflict-affected contexts (Justino et al., 2013), the
Clearly, armed groups expose children to extreme data remain sparse, scattered and non-comparable,
violence but may also provide some sense of while rigorous monitoring and evaluation of policy
‘belonging’ at a time of extreme instability (Blattman interventions in such contexts is a rarity. Household
and Annan, 2009; Humphreys and Weinstein, surveys, such as DHS, often miss areas affected by
2008)—considerations that are often absent from conflict and violence, as well as internally displaced
DDR programmes and from education programming and refugee populations. According to the Pakistan
in conflict-affected countries. Global Initiative on Out-of-School Children study,
the DHS survey excludes the Khyber Pakhtunkhwa
It is also crucial that education nurtures the province because of security concerns. Similarly,
aspirations that are so often undermined by armed the Global Initiative on Out-of-School Children
conflict by delivering a good quality education that Regional Report on Eastern and Southern Africa
prepares children for a peaceful and productive adult notes that population samples for Angola, Burundi
life. Education can support stability and economic and South Sudan were not representative as data
resilience when children and young people learn were only gathered from non-conflict areas. In all of
to support positive social change and when it these countries, therefore, the true impact of conflict
works in tandem with interventions to address the on education remains unclear (UNICEF and UIS,
vulnerabilities and aspirations of those affected by 2013b; 2014a).

Policies and programmes to meet the needs of
out-of-school children require a new approach to At present, available data tend
data, with the emphasis on speed and flexibility and to describe the impact of conflict
on tools that can be mobilised at different phases after it has happened
of the conflict to assess the specific and changing
situation of different groups of children and the
impact of current interventions. Rapid Assessment Uganda to collect real-time data, including teacher
of Learning Spaces (RALS) is a prime example, and student attendance and delivery of materials. It
allowing the evaluation of the changing educational provides districts with a tool to identify bottlenecks
needs of children in conflict-affected areas in order to at the school level, facilitates the tracking of
plan effectively for the reconstruction and reform of accountability for solving any issues arising from
education systems in the post-war environment. system reports and helps to improve planning for
education. At the same time, EduTrac complements
The RALS approach has been mobilised successfully the existing EMIS. EduTrac has also been used to
in Ethiopia and South Sudan, the world’s youngest monitor the schooling status of refugee children
state. Southern Sudan had made impressive from South Sudan who fled the country’s continuing
progress on education before its independence in conflicts. Drawing on its successful experience in
2011, doubling the number of pupils in school and Uganda, EduTrac has been introduced to Kandahar
halving the pupil-teacher ratio between 2006 and province in Afghanistan, where there are many
2009 and prioritising education infrastructure in the out-of-school children and monitoring is hampered
most disadvantaged states. This would have been by both conflict and remoteness. Teachers are
impossible without assessments conducted between now using EduTrac for real-time reporting and
2003 and 2005, but these assessments were limited getting feedback on the problems they identify. In
to the most secure provinces. other countries, such as Liberia and Sierra Leone,
similar progress has been made through the use of
In December 2005, the Ministry of Education, comprehensive school censuses or the establishment
Science and Technology, working with UNICEF, of an EMIS. While it is too early to determine the
undertook a RALS to fill in information gaps and precise links between education progress and such
prepare the ground for an annual education census approaches to data, it is clear that, without them,
and the development of a full EMIS. The information it would be impossible to estimate progress at all
collected, combined with data from geographic (Montjourides, 2013).
information systems, gave planners a global view of
the system as well as detailed information. They had Data from conflict-affected countries do not meet the
access to gender-disaggregated enrolment figures most basic data prerequisites in terms of resources,
by grade and level, numbers of teachers by level of coverage and methodologies. First, there are almost
training and gender, learning and teaching materials no resources for data collection during emergencies.
available to students and teachers, subjects taught, Second, coverage is patchy (at best), excluding large
and language(s) of instruction. parts of the population from the picture, particularly
the most vulnerable; refugees and internally
A yearly statistical booklet now details the state of the displaced children. Even when available, data cover
education system at all levels and the progress being only some parts of education, neglecting crucial
made. areas such as learning outcomes, early childhood
education and teachers. Third, methodologies
Another innovative approach to education data need to improve: lack of means does not have to
collection is the EduTrac system: a mobile phone- equate lack of rigour, and the comparability and
based data collection system developed by UNICEF comprehensiveness of indicators produced should
to support the Ministry of Education and Sports of follow the same standards that apply in countries

that are not affected by conflict (Montjourides, 2013). now embedded in national education strategies, with
UN agencies, such as UNICEF, UNHCR, UNRWA governments of most political orientations backing it
and the UIS, should partner to develop efficient and wholeheartedly.
flexible toolkits to collect and analyse education data
during emergency situations. The situation on the ground, however, is often more
complex—particularly when looking beyond school
There can be no doubt that there are serious enrolment to school completion. Perceptions and
constraints to data collection and analysis in conflict- traditions around gender can often combine with
affected countries. To address these constraints, the location, income and age to determine whether a
international community needs to be more creative child is in school or not, as shown in Figure 3.4,
and innovative than ever before. contrasting the situation of the poorest rural girls of
primary school age in Iraq and the poorest urban
3.4 WHY GENDER MATTERS17 boys of lower secondary school age in Bangladesh.
They often shape what the state is willing to provide
When her mother migrated to work overseas, and what families and communities are likely to
Lalanthi, a 14-year-old girl from Puttalam, Sri demand, and when all three perceive gender-based
Lanka, was left to care for her younger siblings. inequities as ‘normal’, some children—most often
Her father was a violent alcoholic and there was girls, but sometimes boys (as shown in Figure 3.4
nobody else who could help. She had to drop out and Box 3.5)—fall through the educational net.
of school.
(UNICEF and UIS, 2013c) While the world as a whole has progressed towards
gender parity in primary enrolment, girls still account
for 53% of out-of-school children of primary school
Globally, girls still account for 53% of age. The vast majority of girls do now complete
out-of-school children of primary school worldwide. Yet if they are out of
primary school age school, girls of primary age are more likely to never
enter school at all compared to out-of-school
boys, who are more likely to have some exposure
70% of countries are expected to to schooling (UIS and EFA GMR, 2014a). Girls’
reach the goal of gender parity in enrolment rates tend to fall when they reach lower
primary education by 2015, 15% will secondary school age, which coincides with puberty,
be far from that target, and 7% very and tumble even further when they reach upper
far from it—three-quarters of them in secondary school age.
sub-Saharan Africa
According to the 2014 EFA Global Monitoring
Report, many countries had not achieved gender
It seems so obvious that girls and boys should have parity by 2011, six years after the 2005 deadline.
the same opportunities when it comes to schooling. The proportion of 161 countries with gender parity
Indeed, gender parity in education has been seen at the primary level of education rose only marginally
as a crucial indicator of gender equality overall since between 1999 and 2011: from 57% to 63%. It is
2000 as part of the Millennium Development Goals projected that 70% of countries will have reached the
and was an intermediate goal to be achieved by goal by 2015—a decade late—and that another 9%
2005, well ahead of the other goals. The principle of of countries will be getting close. However, 15% of
gender parity has gained considerable traction and is countries will still be far from the target, and 7% will
be very far from it—three-quarters of them in sub-
17 This section draws on “Out-of-School Children: Why Gender Matters”,
a background paper prepared for this report by Nelly P. Stromquist, Saharan Africa (UNESCO, 2014b).
University of Maryland, United States.

Figure 3.4 Percentage of primary and lower secondary school-age children in school, by
household wealth quintile, location and sex
% IRAQ, 2011
100
Richest Urban Richest urban boys
Richest urban girls Richest rural boys
Primary school-age children in school
Average Rural Richest rural girls

90
Poorest urban boys
Urban
Poorest rural boys
80
Poorest urban girls
Poorest
Rural
70
Poorest rural girls
60
50
% BANGLADESH, 2011
Lower secondary school-age adolescents in school
100
Rural Richest rural boys

90
Richest Richest rural girls
Richest urban boys
Average Urban Richest urban girls
80
Poorest rural girls

Poorest urban girls
70
Rural
Poorest
60 Urban
Poorest rural boys
Poorest urban boys

50
Source: UIS calculations based on the Iraq Multiple Indicator Cluster Survey (MICS) 2011 and the Bangladesh Demographic and Health
Survey (DHS) 2011 DataLink: http://dx.doi.org/10.15220/2014/ed/sd/7/f3.4
Sub-Saharan Africa is now home to more than Some girls remain far more likely to be out of school
one-half of all out-of-school girls of primary school than others, with the poorest girls in rural areas
age (UIS and EFA GMR, 2014a) and has the lowest particularly disadvantaged and women from poor
regional proportion of countries—just 13 out of 49— households, in general, far less educated than any
which have reached gender parity to date. As shown other group. In Nigeria, wealthy urban women have
in Chapter 2, West and Central Africa has the world’s attained, on average, around ten years of education,
largest gender gap for both primary and lower in stark contrast to the poorest women from the
secondary school age groups. Hausa ethnic group in rural areas, who had just a
few months of schooling on average (UNESCO,

A number of countries—including some of the
Boys have higher repetition rates than poorest—have taken a lead on closing the gender
girls in primary and lower secondary gap in primary education, offering lessons for
education in nearly every country in countries that still have a long journey ahead of them.
Latin America and the Caribbean As shown in Chapter 2, the greatest improvements
have been seen in South Asia, where girls of primary
school age were twice as likely to be out of school
2010b). The situation for Hausa girls also reflects the as boys in 2000—a gap that had been closed by
way in which poverty and gender intersect with social 2012. Some countries in sub-Saharan Africa have
and cultural practices, such as purdah, and attitudes also made impressive progress from a low starting
to education (UNICEF, 2011a). point. Ethiopia, in particular, has emerged as a global
leader, increasing the number of children enrolled
Such gender gaps matter, with girls’ education both in primary education five-fold between 1994 and
an intrinsic right and a pathway to wider economic 2012 and introducing a special policy focus on girls’
and social objectives. It helps to break cycles of education (Nega, 2012).
poverty and poor health, with adolescent girls in
school less likely to marry early and against their The barriers
will; less likely to die in childbirth; less vulnerable
to diseases, including HIV and AIDS; more likely to Many barriers to gender parity in education—and
have healthy babies; more likely to send their own to girls’ education in particular—remain firmly
children to school; and more likely to acquire the entrenched. They range from broad institutional
information and skills that lead to increased earnings constraints, such as inadequate legislation and
(UNESCO, 2003) (see Box 3.3). Around one-half of policies on sexual violence, female genital mutilation/
the reductions in maternal and infant mortality over cutting (FGM/C) or child marriage, to the deliberate
the past four decades have been attributed to the targeting of girls’ education seen, for example, in
expansion of girls’ education, especially when they Afghanistan, Nigeria and Pakistan that can result in
finish primary school and complete at least lower their physical harm, as well as their removal from
secondary school (Gakidou et al., 2010). school.
Box 3.3 The economic argument

The moral case for girls’ education is overwhelming: it is a human right that helps to forge more equitable
societies. But there is also a compelling economic argument to be made.
■■ Countries with greater gender parity in primary and secondary education are more likely to have higher
economic growth. Based on World Bank research and data and UIS education statistics, Plan (2008)
estimated that the economic cost to 65 developing countries of failing to educate girls to the same
standard as boys was a staggering US$92 billion each year, just less than the US$103 billion annual aid
budget of the OECD countries in 2007.
■■ Girls who have even one year of education above the national average earn 10% to 20% more than the
average national income in later life (Psacharopoulos and Patrinos, 2002). Returns to female secondary
education range from 15% to 25% (Schultz, 2002).
■■ More productive farming as a result of increased female education is thought to have accounted for
43% of the decline in malnutrition between 1970 and 1995, according to a 63-country study (Smith and
Haddad, 2000). Better nutrition, in turn, boosts returns to educational investments, with children better
able to concentrate in class.

Within schools, the way in which girls experience that parents or other relatives had prevented them
their education and the cumulative nature of the from going to school because of concerns for their
disadvantages they face leave them vulnerable to the safety.
risk of drop-out. All too often, curriculum materials,
teaching and learning processes, assessment In some school environments, girls face considerable
systems and teachers’ expectations of their students peer pressure to experiment sexually, usually with very
are highly gendered and, when combined with a lack little understanding of the consequences (Stromquist,
of female teachers, present major barriers to girls’ 2014). For example, high drop-out levels have been
completion of their education and learning. detected among girls in the Bolivian Amazon region,
where 27% of adolescents leave school as a result of
Such barriers can become insurmountable when pregnancy (UNICEF and UIS, 2011).
compounded by conflict and natural disasters,
migration and displacement, HIV and AIDS, disability,
ethnicity, religion and caste. By one UNICEF estimate, only 49 of
155 countries have dedicated policies
Supply barriers to allow pregnant girls and young
mothers to continue their education
The lack of a nearby school is a problem for
any child—boy or girl—undermining punctuality,
attendance and learning, all precursors of school The lack of private and separate latrines and washing
drop-out. Girls, however, face particular risks linked facilities in under-resourced schools presents an often
to distance and the time it takes to go to and from unacceptable risk to a girl’s modesty, dignity and basic
school, including the danger of being assaulted. human rights. The absence of such basic facilities is
Fears of such assaults may explain why girls in rural a particular issue for girls who are menstruating and
areas are more likely than boys to attend school at has been cited in studies across Africa and Asia,
a later age, when they are better able to make their in particular, as a factor in girls abandoning their
own way to school. Distance to school is an issue education (Adukia, 2014; Herz and Sperling, 2004;
that also affects boys and that cuts across nations, Mooijman et al., 2005; Sommer, 2010).
from Bolivia to India and from Tajikistan to Togo.
The presence of female teachers helps to attract girls
Even where there is ready access to a local to school and improve their learning outcomes in
school, the lack of a safe and supportive school some contexts, and female teachers can be valuable
environment is another major barrier to girls’ role models for young girls, helping to ensure that
education. While schools are usually considered girls are not merely present in the classroom but
‘safe havens’, many girls experience intimidation are also active participants in lessons. But there is a
and abuse from teachers and other pupils and chronic shortage of female teachers in some regions,
sexual harassment is a major cause of female especially in secondary education. In sub-Saharan
drop-out (Stromquist, 2014). An estimated African countries with available data, female teachers
500 million to 1.5 billion children are thought to account for less than 40% of teachers in 43% of
experience violence every year, many of them countries at the primary level, in 72% of countries at
within schools, and millions more live in fear of the lower secondary level and in all countries at the
physical abuse under the guise of discipline: more upper secondary level (UNESCO, 2014b).
than 80% of students in some countries suffer
corporal punishment at school (Greene et al., Another supply barrier is the lack of flexible education
2013). According to the Global Initiative on Out- opportunities for girls who have dropped out of
of-School Children study in Tajikistan (UNICEF and school and who now want to return, particular girls
UIS, 2013e), 15% of girls out of school reported who are pregnant or young mothers. According to

FGM/C and child marriage—marriage before the
Each year of marriage before the age of 18—are thought to account, in part, for the
age of 18 in Africa reduces the low educational attainment of girls in African and
probability of secondary school Asian countries and high levels of illiteracy among
completion by 6.5 points women in these regions. The precise links between
child marriage and low educational achievement are
not yet clear (see Box 3.4), but the decisions not
UNICEF Country Office Annual Reports, only 49 to send a girl to school and to marry her off early
out of 155 countries appeared to have policies in are often made at the same time. In some African
place to allow pregnant girls and young mothers to populations, FGM/C is the first signal that a girl is
continue their education in 2013 (UNICEF, 2014d). marriageable and disrupts the attendance of girls
Even where such policies are in place, they may who are in school, and the prolonged absence it
have little impact on the ground if school principals, causes often leads to drop-out. For girls who have
teachers and fellow pupils reject girls who have given never enrolled in school at all, child marriage may be
birth. It is unlikely that legislation alone can entice seen as the only possible and appropriate option.
girls back into the classroom or keep them there, Whether girls are withdrawn from school or have
given the continuing stigma around pregnancies or never attended, child marriage marks the beginning
births outside marriage, and the social assumption of subordination to their husbands and the dangers
that education is over once a girl marries and gives of early pregnancy, as well as the end of education
birth. What is clear is that girls of school age who (Equality Now, 2014).
become pregnant tend to abandon schooling (Chae,
2013), as seen in Brazil, where over 70% of girls “I finished seventh grade and left school because
aged 10 to 17 years who have children are not in the of marriage. I didn’t want to get married, but my
classroom (UNICEF and UIS, 2012a). father forced me to. He told me that education
won’t do anything for me... I had no choice.”
Few of the most marginalised and excluded children Sultana, married at age 16 in Yemen
have access to NFE opportunities that might provide (Brown, 2012)
a route back into schooling or, at the very least,
provide them with the basic skills they need. UNICEF data show that women are two and a half
times more likely to be married as children if they are
Demand barriers poor than if they come from richer households: around
four in every ten women aged 20 to 49 years in the
The barriers to demand for girls’ education include poorest 20% of households were married as children,
problems related to cultural norms around gender— compared to only 16% of girls in the richest 20% of
including FGM/C and child marriage, as well as households. Child marriage is also more prevalent in
poverty, ethnicity, child labour and orphanhood—and rural areas. Regionally, 56% of women aged 20 to 49
these problems often work in combination to keep years in South Asia, 46% in West and Central Africa,
girls out of school. They are only reinforced by failures and 38% in Eastern and Southern Africa were married
to enforce not only compulsory education laws, before the age of 18, often to much older men. At the
but also laws of crucial relevance to girls, such as country level, India is home to one-third of all child
legislation on the minimum age for marriage and the brides globally (UNICEF global databases, 2014).
prohibition of FGM/C.
Some parents see child marriage as a way to protect
In patrilineal societies, where it is sons who inherit from their daughters and the family from the shame of
their fathers, girls may be seen as ‘temporary’ family premarital sex and pregnancy outside marriage,
members who will soon be given away by marriage to and encourage their daughters to marry as soon
join another household (UNICEF and UIS, 2012b). as they reach puberty, as is the case in Nigeria’s

Box 3.4 Measuring the impact of child marriage on education
Approximately one in three young women aged 20 to 24 years in 2012 were married before the age of 18
in developing countries, and one in nine were married before their fifteenth birthday (UNICEF, 2014b). Child
marriage often means the end of education, but it is not clear precisely how parental views on the value of
education for girls link to child marriage, with decisions on both issues often tied together. This makes it
difficult to assess the precise linkages between child marriage and educational attainment, but it seems clear
that education itself acts as a safeguard against child marriage. One study found that secondary education
for girls was a crucial factor in increasing the age of marriage of girls in India, Indonesia, Pakistan, Sri Lanka,
Taiwan of China and Thailand (Mathur et al., 2003).
One approach to analysing the links between education and child marriage is to analyse the decision to
marry. Field and Ambrus (2008) used the timing of puberty as the instrumental variable for the age at first
marriage in Bangladesh, finding that each additional year of delay in the age of marriage increased schooling
by 0.22 year and the likelihood of literacy by 5.6 percentage points. Nguyen and Wodon (2014b) generated
markedly similar results by examining current and past incidence of child marriage, finding that each year of
marriage before the age of 18 in Africa reduces the probability of secondary school completion by 6.5 points.
northern areas (UNICEF and UIS, 2012f). A related

practice, called ‘child betrothal,’ commonplace in In sub-Saharan Africa, the poorest
South Asia, sees a girl staying with her own family boys are expected to achieve
until a certain age, at which point she goes to her universal primary completion by
groom’s home (Edmeades and Hayes, 2014). This 2069. The poorest girls will not
practice creates uncertainties around girls’ education reach this target until 2086
and often interrupts their schooling at upper primary
or lower secondary school levels. An extreme but
rare practice is marriage by abduction, which has and a clean water supply, for example. The Global
been seen as a legitimate way to procure a bride, Initiative on Out-of-School Children study in Ghana
particularly in parts of the Horn of Africa (IRIN, 2007). found that such strenuous work cuts into their time
for school and homework and affects their ability
Other cultural norms affect the demand for girls’ to concentrate, resulting in underachievement
education in less perceptible ways. Even in and eventual drop-out (UNICEF and UIS, 2012d).
schools with functioning and separate toilets, for This only seems to confirm parental beliefs that
example, demand for girls’ education can fluctuate there are few economic benefits to be gained by
in communities with strict cultural norms around educating girls.
menstruation. In many communities in India,
menstruation is seen as ‘unclean’ and girls may be Poor, rural parents often send their children
kept at home during their periods (Thakre et al., (particularly girls) to urban families in the hope
2011). The resulting monthly absences see girls that they will be enrolled in school in exchange for
falling behind in their studies and may well lead to some domestic chores (Gustafsson-Wright and
drop-out. Pyne, 2002)—a common practice in several African
countries (through the fosterage or confiage system)
Cultural expectations can also combine with poverty as noted in the Global Initiative on Out-of-School
to fuel the child labour that keeps girls out of school. Children studies in Ghana, the Democratic Republic of
Most children out of school worldwide live in the the Congo, Haiti and some other countries. In reality,
poorest households, but poverty may merge with it often results in heavy domestic child labour, leading
strict cultural norms to deny girls, in particular, an to school absenteeism and drop-out. In Haiti, one
education. Girls are expected to fetch and carry in every ten children is subject to this practice called
fuel and water in households that lack electricity restavek, three-quarters of them girls (Smith, 2014).

Box 3.5 Out-of-school boys and boys who fall behind
Cultural norms and gender roles can work against demand for education for boys as well as girls. Entrenched
concepts of masculinity can result in more boys being out of school, particularly at the secondary level,
in certain commercial, agricultural and pastoralist contexts where their labour is seen as vital for family
livelihoods (Gustafsson-Wright and Pyne, 2002; also Global Initiative on Out-of-School Children studies for
Bolivia, 2011; Nigeria, 2012f; Pakistan, 2013b).
In Latin America and the Caribbean, in particular, boys are falling behind on key education indicators,
particularly at the secondary level. In 2011, there were 106 girls for every 100 boys enrolled in secondary
school across the region. Of the 30 countries in Latin America and the Caribbean with available data, 21
reported a gender gap in secondary enrolment that left boys lagging behind (UIS Data Centre, 2014). Boys
in school are also more likely than girls to repeat school years—a well-known trigger for school drop-out.
An analysis of repetition rates in primary and lower secondary education showed that rates are higher for
boys than girls at both levels, in every country in the region, with the exception of Saint Kitts and Nevis
(UIS, 2012b).
Boys also account for the vast majority of children living and working on the world’s streets—many of whom
are out of school. Analysis of children on the streets in 75 Brazilian cities indicates that nearly three-quarters
(around 72%) are boys (Consortium for Street Children, 2012).
In Senegal, thousands of young boys are sent to Girls who have lost one or both parents and who live
urban areas to receive education from a ‘Marabout’, a in poverty are particularly vulnerable to educational
religious teacher, but instead of going to school they exclusion. Girls orphaned as a result of AIDS,
are sent to beg for money on the street. for example, often drop out of school to care for
younger siblings and may be urged to marry early
In many countries, poverty combines with geography for their own security (UNICEF and UIS, 2012b).
and ethnicity to keep girls out of school, and Although more children have lost their fathers than
schooling disparities between girls and boys from mothers to AIDS in Zimbabwe, it is maternal death
socially excluded groups are much larger than in the that seems to have the greatest impact on their
wider population. Because remote rural populations chances of completing primary school (Nyamukapa
tend to be scattered and may not have a government and Gregson, 2005). Findings from a review of 244
school nearby, their children—particularly girls—tend studies in developing countries also indicate that
to enter the school system about two years late orphanhood is a risk factor for early sexual practices
and seldom make up enough lost ground to move (Mmari and Sabherwal, 2014). Orphaned girls are
beyond primary education. Rural areas in Bolivia, particularly vulnerable to sexual abuse and to being
Guatemala and Peru all have lower enrolment and forced into prostitution, which limits any chance of an
attendance for girls than boys in secondary school, education.
reflecting the scarcity of local schools, the heavy toll
of domestic work on rural girls, and the discrimination Breaking the barriers
they face within several ethnic groups (Glick, 2008).
The ambition must be to ensure that even the most
In India, education gender gaps are more severe vulnerable and disadvantaged girl has access to
among scheduled castes and tribes (UNICEF and UIS, a school within a reasonable distance from her
2014d). Poverty in sub-Saharan Africa threatens the home—a school that meets her most basic needs for
region’s chances of achieving universal primary school safety, privacy and cleanliness, that delivers the best
completion for decades to come, but even the poorest possible education and that places a high value on
boys are expected to achieve universal primary both her presence and her potential. The barriers to
completion by 2069—a goal that will not be reached girls’ education can be overcome if the environment
by the poorest girls until 2086 (UNESCO, 2014). around them—their homes and communities—

supports their schooling, and if schools themselves
are accessible, safe and inclusive. It will take time An initiative to establish village-
to root out the discrimination that keeps so many based schools in the north of
girls out of the classroom, but much can be done to Afghanistan closed the gender gap
ensure that schools welcome, nurture and protect all within its first year
their pupils, girls as well as boys.
First, every girl needs access to a nearby teachers could help to close the gender gap, as
school. In Afghanistan, for example, the distance to would more flexible entry qualifications. For example,
school may well determine whether a girl attends or the Girls’ Education Project in Nigeria, funded by
not, given safety concerns and traditional seclusion the UK Department for International Development
practices. The introduction of village-based schools (DFID) in partnership with UNICEF, aims to get
in Ghor Province in the north of the country, with 1 million more girls into school by 2020. The project
support from Catholic Relief Services, has resulted calls for the deployment of 10,500 female teachers
in increased enrolment and better test scores for all to rural areas where the predominance of male
students, but girls have benefited disproportionately teachers deters parents from sending their girls to
and the gender gap in enrolment was eliminated school. In return for a scholarship grant of around
completely within the first year of this initiative (Burde £200, newly-qualified female teachers commit to
and Linden, 2010). teach in rural schools for two years. In South Sudan,
where women make up about 65% of the post-
Girls need ‘girl-friendly’ schools. In Burkina war population, yet less than 10% of all teachers,
Faso, well-resourced and gender-friendly schools financial and material incentives have been given to
(BRIGHT schools) built in poor and previously under- over 4,500 girls to complete secondary school and
served rural areas have boosted the enrolment of all to women trainees to enter the teaching profession
children aged 5 to 12 years by 20%, with girls—once (UNESCO, 2014b).
again—benefiting disproportionately. These schools
are characterised by a mix of interventions spanning Child-friendly and gender-sensitive teaching
separate latrines for boys and girls, canteens and approaches should be required elements
take-home rations, textbooks and attempts to in teacher certification, as introduced by the
change institutional cultures through advocacy and Ministry of Education of Tajikistan in 2013 (UNICEF,
mobilisation, literacy training and capacity building 2013e). Tajikistan’s creation of a Center for Gender
among local partners (Kazianga et al., 2012). Pedagogies within the Ministry of Education is a
promising innovation that involves systematic gender
This links to the need for schools that have zero audits of the curriculum, textbooks and teacher
tolerance for violence of any kind, from corporal training programmes. In Ghana, in-service training
punishment to playground bullying. While the for teachers in gender and child-friendly teaching
strong enforcement of legislation to outlaw corporal approaches was found to be effective in improving
punishment is vital, ending violence in schools also girls’ enrolment and retention in schools in the most
requires the implementation of teachers’ codes of deprived areas (UNICEF and UIS, 2012d).
practice, measures to monitor and address cases
of violence, and independent and confidential Effective educational support is needed for
mechanisms that children can trust, such as hotlines, pregnant girls and young mothers, including
when reporting abuse. childcare and counselling. However, this support has
to move beyond education policies that—on paper
There is a pressing need for more female at least—allow pregnant girls or young mothers to
teachers in some contexts. Provision of continue their studies. Multisectoral approaches,
scholarships to women who want to become spanning sexual and reproductive health, child

and can support advocacy and awareness-
Research from India confirms that raising programmes for parents, youth and
good information for parents on communities to modify deeply-rooted cultural beliefs
the long-term benefits of girls’ about femininity and masculinity and make the case
education helps to keep girls in for girls’ education. A good example of the impact
school and delays their marriages of such alliances can be seen in Turkey, where the
four-year “Hey Girls, Let’s Go to School” campaign,
characterised by partnerships between public
protection and financial support are all needed institutions, civil society organizations and volunteers,
alongside education policies on this issue. resulted in the enrolment of an estimated 350,000
additional children in school (UNICEF and UIS, 2012h).
The provision of sex education courses at
primary and secondary school levels, including Few parents are averse to investing in the education
reproductive health and rights, must be intensified, of their daughters once they have good information on
given the links between early pregnancy and school its benefits. A study in India, for example, found that
drop-out. Studies across a wide range of cultures providing villagers with precise information about the
indicate that parental fears that such courses availability of jobs for girls with secondary education
promote premature sexual activity are unfounded and how to get such jobs resulted in teenage girls
(UNICEF, 1999), and greater efforts are needed to link staying longer in school, being more likely to look for
sexuality education to reproductive health services paid work and delaying marriage. Primary school-
and contraceptive provision in countries with high age girls in villages receiving such information were 5
rates of adolescent pregnancy. percentage points more likely to be in school than girls
in control villages (Jensen, 2010).
Non-formal education gives out-of-school girls
a chance to learn vital skills and, in some cases, to Altering the political economy of the household
enter or re-enter the formal education system. In in countries with high levels of poverty could ensure
Gambia, the Re-Entry Programme for Girls, initiated that parents do not have to rely on girls’ domestic
by the Ministry of Basic and Secondary Education, labour or boys’ labour outside the home. One
reaches girls who have dropped out, providing approach that is having an impact is cash transfers
extensive guidance and counselling services for granted to households on condition that children
each participant. In India, the ‘Pehchan’ project by attend school. In Bangladesh, for example, the Female
the Centre for Unfolding Learning Potentials (CULP) Secondary School Stipend Program has supported
in collaboration with the Government and UNICEF impressive progress on girls’ school enrolment—which
offers two- to three-year courses at the primary now exceeds that of boys at both primary and lower
level to rural adolescent girls who are out of school secondary levels—and on delaying marriage. Analysis
as a prelude to their reintegration into the formal of an intervention in Malawi providing cash transfers
school system, while the Hope for Teenage Mothers tied to school attendance for adolescent unmarried
organization in Kenya gives teenage mothers access girls and their families found that such conditional cash
to economic and educational opportunities through transfers are more effective in boosting girls’ enrolment
formal education, vocational training and skills than cash transfers with no conditions attached (Baird
building (Yasunaga, 2014). et al., 2011).
Alliances between governments and civil society Filling the data gap
organizations can strengthen national efforts to alter
cultural norms around gender. Such alliances can Countries have made great strides in recent years in
bolster attempts to enforce compulsory education disaggregating data by gender. As a result, we have
laws and laws on the legal minimum age for marriage a fairly clear idea of the continued gender gap among

out-of-school children and, as confirmed by the data attendance and on their energy and concentration
from countries participating in the Global Initiative in class, hampering their learning and their ability to
on Out-of-School Children, we have evidence on keep up. Those who fall further and further behind
the likelihood of girls entering school later, dropping are at even greater risk of dropping out of school
out before they complete basic education, or never altogether. Child labour is also linked to school
attending at all. However, while such data tell us what absenteeism and tardiness, only adding to the risk of
is happening, they do not necessarily tell us why. school drop-out.
As this section has highlighted, we do not yet fully
understand the interplay between decisions around Regionally, sub-Saharan Africa has the highest
child marriage, the withdrawal of girls from school prevalence rate for child labour at 27%, followed by
and perceptions of the value of girls’ education. South Asia at 12%, but South Asia accounts for the
There is now a clear need for data—both quantitative highest absolute numbers of child labourers—an
and qualitative—that drill down to look more closely estimated 77 million (UNICEF, 2014b).
at the precise impact of gender norms on the
likelihood of being out of school. Most out-of-school children are engaged in some
form of work. Equally, children who work are more
3.5 CHILD LABOURERS AND SCHOOL likely to drop out of school (Rosati and Guarcello,
PARTICIPATION18 2014). Understanding the interplay between these
two trends is, therefore, critical to achieving two
“My name is Ruslan and I am 14 years old. I do key goals for children: education for all and the
not go to school, since I work in Osh Bazaar. I elimination of child labour. All the evidence suggests
come here before 8 o’clock in the morning and that child labour and school participation are
leave for home at 8 o’clock in the evening… I incompatible, and that the more onerous the labour,
have no other options because my family has the greater the impact on schooling.
nobody else to feed us. I would like to attend
school, but I don’t want to study with children This conclusion has been reinforced by analysis
much younger than me.” carried out for this report by Understanding Children’s
Boy from Kyrgyzstan Work (UCW), which examined the way in which child
(UNICEF and UIS, 2012e) labourers intersect with out-of-school children in the
countries that participated in the Global Initiative on
UNICEF estimates that 15% of children aged 5 to Out-of-School Children.20 The study focused on child
14 years (or approximately 150 million children in all) labourers aged 7 to 14 years and their likelihood of
are engaged in some form of child labour (UNICEF, never enrolling in school or dropping out.
2014b) (see Box 3.6). Almost one-third of child
labourers are thought to be engaged in hazardous Drawn from the UCW analysis, Figure 3.5 highlights
work that threatens their health, safety or emotional child labour as a policy concern in almost all
well-being (ILO/IPEC, 2013).19 countries analysed. The sub-Saharan African
countries stand out as having especially high child
While most child labourers are in school, their labour rates, with around 45% of children aged 7 to
labour may act as a ‘push’ factor in decisions to 14 years in Ethiopia, 37% in Ghana and more than
leave school prematurely, putting pressure on their 30% in Zambia involved in child labour. While the
18 This section draws on “Child Labour and Out-of-School Children:

Evidence from 25 Developing Countries”, a background paper and 20 South Asia: Bangladesh, India, Pakistan, Sri Lanka; East Asia and the
study prepared for this report by Furio Rosati and Lorenzo Guarcello, Pacific: Cambodia, Indonesia, Philippines, Timor-Leste; Latin America
Understanding Children’s Work, Italy. and the Caribbean: Bolivia, Brazil, Colombia, Mexico; Central and Eastern
19 There are also an estimated 48 million children aged 15 to 17 years Europe and Commonwealth of Independent States: Kyrgyzstan, Romania,
engaged in child labour, which poses a threat to their continued education Tajikistan, Turkey; sub-Saharan Africa: Democratic Republic of the Congo,
(ILO-IPEC, 2013). Ethiopia, Ghana, Liberia, Mozambique, Nigeria, Sudan, Zambia.

Box 3.6 Defining child labour
The child labour indicator used in this report is adapted from UNICEF’s standard definition and refers to
the percentage of children aged 7 to 14 years who were involved in child labour according to the following
thresholds:
■■ children aged 7 to 11 years who performed at least one hour of economic activity in the week before
the survey;
■■ children aged 12 to 14 years who performed at least 14 hours of economic activity in the week before
the survey; and
■■ children aged 7 to 14 years who performed at least 28 hours of household chores in the week before
the survey.
This indicator draws on three international conventions on child labour: ILO Convention No. 138, the
United Nations Convention on the Rights of the Child, and ILO Convention No. 182, as well as the
resolution on child labour statistics adopted at the 18th International Conference of Labour Statisticians
(ICLS) in 2008 (ILO, 2009).
such as caring for siblings, cooking or cleaning,

Child labourers are at greater risk these contribute to the everyday consumption of a
of being out of school than children household and are not seen as economic activities
who are not working that contribute to household or national income.21
Current definitions stipulate a 28-hour per week
threshold beyond which household chores are
rates are generally lower in the sample countries considered child labour. By this definition, household
from other regions, there are some important chores form a much smaller component of child
exceptions. Bolivia, Kyrgyzstan and Timor-Leste labour than economic activities. In every country
all have far higher child labour rates than the other studied, with the exception of Tajikistan, it is the
countries analysed in their respective regions. economic activities around family work—such
The findings confirm that child labourers are at far as working in a store or harvesting crops—that
greater risk of being out of school than children who constitute the largest component of the child
are not working. labour performed by out-of-school children. Family
work—paid or unpaid—accounts for at least 40%
The more hours children work, the greater the of all out-of-school child labourers in the eight
likelihood that they will miss out on an education. The sample countries where this information is available
UCW analysis also confirms that out-of-school child (Democratic Republic of the Congo, Ghana, India,
labourers aged 7 to 14 years work for many more Liberia, Nigeria and Tajikistan) and rises to over 80%
hours than child labourers who attend school. The in Ethiopia and Mozambique.
most marked difference in working hours is seen in
Turkey, where out-of-school child labourers work for Boys had a higher child labour rate than girls in
an average of 45 hours each week, compared with Bangladesh, Ethiopia, Pakistan, Philippines and
an average of 15 hours for their peers who are in Sudan. In contrast, girls were more likely to be
school. This suggests that the time intensity of child caught up in child labour in the Democratic Republic
labour, as well as the type of work, matters for school of the Congo, Mozambique and Tajikistan (Rosati
attendance. and Guarcello, 2014). It is worth noting, however,
that household chores are overwhelmingly the
What type of work do child labourers do? While
household chores may take up a child’s time, 21 As defined in the System of National Accounts.

Figure 3.5 Children involved in child labour and its impact on school attendance
Percentage of children involved in child labour, 7-14 years, by country
BANGLADESH 10
14
SOUTH
ASIA INDIA
PAKISTAN* 13
SRI LANKA 10
CAMBODIA* 10
EAST ASIA
AND THE
PACIFIC
INDONESIA 5
PHILIPPINES* 7
TIMOR-LESTE* 18
BOLIVIA 29
CARIBBEAN
AMERICA
AND THE
BRAZIL* 5
LATIN
COLOMBIA 6
MEXICO 6
KYRGYZSTAN 30
CEE/CIS
ROMANIA* 1
TAJIKISTAN 12
TURKEY* 3
CONGO, DR 18
ETHIOPIA 45
SUB-SAHARAN AFRICA
GHANA 37
LIBERIA 24
MOZAMBIQUE 27
NIGERIA 27
SUDAN* 19
ZAMBIA* 30
0 10 20 30 40 50 %
Percentage of children who are out of school, 7-14 years, by child labour status and country
Child labourers Children not in child labour
BANGLADESH 48
40
SOUTH
INDIA
ASIA
PAKISTAN* 88
SRI LANKA 10
CAMBODIA*
23
EAST ASIA
AND THE
PACIFIC
INDONESIA
14
PHILIPPINES*
19
TIMOR-LESTE*
37
BOLIVIA 6
CARIBBEAN
AMERICA
AND THE
BRAZIL* 4
LATIN
COLOMBIA 23
MEXICO 21
KYRGYZSTAN 1
CEE/CIS
ROMANIA* 27
TAJIKISTAN 9
TURKEY* 56
CONGO, DR 22
SUB-SAHARAN AFRICA
ETHIOPIA 37
GHANA 29
LIBERIA 55
MOZAMBIQUE 18
NIGERIA 25
SUDAN* 75
ZAMBIA* 20
0 10 20 30 40 50 60 70 80 90 %
Notes: * Denotes child labour statistics which vary from the international definition of child labour due to limitations of the household survey or
census used. Data for Pakistan, Sudan and Timor-Leste cover children aged 10-14 years, for Turkey they cover 6-14 years. Data for Cambodia,
Pakistan, Philippines, Romania and Zambia do not include information about hours spent in household chores: the definition of child labour in
these countries is based on hours in employment only. Data for Sudan do not include information about hours spent in employment and
household chores, the definition of child labour is based on involvement in employment only. Data for Brazil use the international definition,
although Brazilian national legislation does not allow light work for children aged 12-14 years.
Sources: Understanding Children’s Work calculations based on Bangladesh Labour Force Survey, 2005-2006; Bolivia Encuesta de Trabajo Infantil (SIMPOC), 2008;
Brazil Pesquisa Nacional por Amostra de Domicílios, 2011; Cambodia Labour Force and Child Labour Survey (SIMPOC), 2012; Colombia GEIH: Módulo de Trabajo
Infantil, 2012; DR Congo MICS, 2010; Ethiopia DHS, 2011; Ghana MICS, 2006; India DHS, 2005-2006; Indonesia Child Labour Survey (SIMPOC), 2009; Kyrgyzstan
Child Labour Survey, 2007; Liberia DHS, 2007; Mexico Encuesta Nacional de Ocupación y Empleo: Módulo de Trabajo Infantil, 2011; Mozambique MICS, 2008;
Nigeria MICS, 2011; Pakistan Labour Force Survey 2007-2008; Philippines Labour Force Survey (SIMPOC), 2001; Romania Child Labour Curve (SIMPOC), 2000;
Sri Lanka Child Activity Survey (SIMPOC), 1999; Sudan Fifth Population and Housing Census, 2008; Tajikistan MICS, 2005; Timor-Leste Survey of Living Standards,
2007; Turkey Child Labor Survey (SIMPOC), 2006; Zambia Labour Force Survey, 2008 DataLink: http://dx.doi.org/10.15220/2014/ed/sd/7/f3.5

responsibility of girls. The 28-hour threshold has heads proving to be a risk factor for child labour in
important implications for estimates of girls’ child every country with available data. The lower the level
labour relative to that of boys. A lower threshold would of household education, the greater the likelihood
mean higher child labour rates, especially for girls. that children will be out of school.
The barriers Breaking the barriers
As shown in relation to other barriers to the universal The relevance and quality of education are not a
inclusion of children in education, poverty lies given: getting working children into school requires
at the very heart of the child labour barrier, with more flexible and responsive education systems
poor households less able to afford the direct and and improved learning environments. Above all,
indirect costs of education and more likely to need primary education must be free and accessible to
any additional income that can be earned by their all, including children who are, or were, working. It
children. As a result, children from such households also requires reforms that go far beyond education
are more likely to be involved in child labour. Across itself, including broader changes in public policy
all countries with available data, the wealthier the that empower families to choose education over
household, the lower the rates of child labour labour. This means addressing social and economic
(UNICEF global databases, 2014). disparities through social protection, livelihoods
assistance and access to social services, as well
This is no surprise, as better-off households do not as advocacy and awareness-raising to tackle the
need their children’s productivity or wages to make harmful social norms that perpetuate child labour
ends meet and the opportunity cost of schooling is, (UNICEF, 2014a).
therefore, lower. But the UCW analysis also found
that household poverty does more than increase Taken together, the evidence from the countries
the likelihood of child labour: it also increases the involved in the Global Initiative on Out-of-School
impact of that labour on education. Across most Children underscores the important linkages
of the countries surveyed, child labourers from the between child labour and the risks of being out
very poorest households are far more likely to be of school. These linkages reinforce the need to
out of school than child labourers from better-off invest in improved schooling, to mitigate poverty
households. and household vulnerability, and to raise household
awareness of the value of education and the damage
Seven-year-old Saritha, a girl from Colombo, caused by child labour as part of a broader strategy
begs with her mother near the Dehiwala mosque. to address both child labour and non-attendance at
This helps them to earn money. However, it is an school.
offence to go begging with children and, if the
police catch them, they are taken to court. Some There are strong links between being a child
days, they have no food in the mornings, but labourer and being out of school and the two
there is always something for lunch and dinner. challenges must be addressed together. On the
Her parents never went to school. The family lives one hand, child labour needs to be reduced in order
in a small wooden hut built on unauthorised land to increase school attendance. On the other hand,
near the Keththaramaya. The house does not increased school attendance is the most effective
have basic amenities other than a water tap. way to reduce child labour. It is essential to develop
OOSCI Country Study on Sri Lanka approaches that improve education access, quality
(UNICEF and UIS, 2013c) and relevance, so that families not only have the
opportunity to invest in their children’s education
Lack of education is, in itself, a red flag for child as an alternative to child labour but also find that
labour, with low educational levels among household the returns to schooling make their investment

worthwhile. This also links to the affordability of Social protection needs to be expanded to
schooling and the need to ensure that free education prevent the use of child labour as a household
really is just that: free, with no hidden costs that survival strategy in hard times. Social protection
might act as barriers for the poorest families. should contain basic social security guarantees to
ensure that anybody in need has access to essential
Second-chance and non-formal learning health care and income security at a nationally-
opportunities are needed to compensate for defined minimum level (at the very least) throughout
the adverse educational consequences of child their lives.
labour. It is vital to reach former working children
and other out-of-school children, including those Filling the data gap
who live and work on the streets, with educational
opportunities that are part of a broader push for their It is vital to improve the evidence base to
social reintegration. Such opportunities are critical to inform policy and ensure the effective targeting of
prevent large numbers of children carrying a burden interventions. The UCW analysis has confirmed the
of disadvantage into adulthood, permanently harmed relationship between child labour and schooling, but
by their early work experiences. there are many unanswered questions beyond this
general pattern. What impacts do different kinds
In the six largest cities of Bangladesh, for example, of work and the number of working hours have on
learning centres under the Basic Education for Hard- schooling? What is the impact of work on learning
To-Reach Urban Working Children project provide life achievement? Why are there large variations
skills-based, non-formal basic education for working across countries in the way in which work affects
children aged 10 to 14 years who have either never schooling?
been to or have dropped out of school (UNICEF and
UIS, 2014d). Joint programmes of the International Research has shown that both school access and
Labour Organization (ILO) and the International the quality of education help to keep children in
Programme on Elimination of Child Labour (IPEC) on school and out of work by influencing key household
non-formal or transitional education aim to reintegrate decisions concerning their education. However,
former child workers into formal school in countries information on school quality or its proxies (such as
such as Brazil, India, Mongolia and Nicaragua. the experience of teachers, class sizes, curricula,
Through Bridge Schools or intensive transitional school management and organization, or school
education programmes, former working children violence) is not always available and is generally not
are supported to catch up the years of study they collected through household surveys in conjunction
have ‘lost’ and reach an adequate academic level for with child labour information. Administrative data may
enrolling in formal school or vocational training (ILO, be of some help, but—again—are not always cross-
2009; Yasunaga, 2014). referenced with household surveys. This lack of data
severely limits the number of indicators that can
Mustafa, aged 11, is just starting to learn to actually be used to study the relationship between
read. While his peers are in 5th grade, Mustafa is child labour and education.
attending a 1st–3rd grade remedial programme
(Catch Up Curriculum Programme). He used to It is time to open the ‘black box’ of child labour and
collect paper with his brother during the days and look more closely at the effect of different forms of
had school attendance problems, but now both work on the chances of children enrolling in school,
of them have started attending the programme staying in school and maximising the benefits of their
regularly. education.
OOSCI Country Study on Turkey
(UNICEF and UIS, 2012h)

3.6 THE LANGUAGE BARRIER22 The vision behind the EFA goals and the education-
related Millennium Development Goals is that school
By the late 1990s, the Zambian National Reading participation allows people to lead happier, healthier
Committee could conclude: “What was for a and more productive lives. However, most education
long time seen as a reading problem in Zambian systems are designed by and for the most dominant
schools was, in fact, a language problem. Reading group in society. This group could be the ethnic
was being introduced in a language which was for majority in a particular country, but is, in some cases,
most pupils a foreign and alien language.” a dominant minority that holds the reins of economic,
OOSCI Country Study on Zambia social and political power (Kosonen and Benson, 2013).
(UNICEF and UIS, 2014g forthcoming) In either case, those who are not part of this group may
well be socially excluded in certain contexts.
An estimated 2.3 billion people lack The end result can be serious educational harm for
access to education in their own language children for whom the languages spoken at home,
as well as their own cultural values and experiences,
do not feature in their education. Instruction in a
Children from socially-excluded groups language learners do not yet understand inhibits
are significantly more likely to not their literacy and learning (see Box 3.7) and, very
attend school in Bolivia, Ecuador, India importantly, devalues their cultural identities. This
and Lao People’s Democratic Republic devaluation can be implicit through, for example, the
absence of images that reflect their lives or culture
in school materials, or explicit, with children banned
Most out-of-school girls are from from wearing their traditional clothing or speaking the
socially-excluded groups, according to language they use at home in the classroom.
one study across 16 countries
The implication is that children may not be attracted
by an educational system that seems to have little
An estimated 2.3 billion people, nearly 40% of the place for them. Some families cope by sacrificing
world’s population, lack access to education in their scarce resources so that their children can learn the
own language—a clear stumbling block to their language of instruction; others may reject schools as
learning that will not be removed by getting more socially irrelevant or pedagogically ineffective.
children into the classroom (Walter and Benson,
2012). Language overlaps with ethnicity, poverty, The impact is clear in a number of countries
rural life, religion and gender in ways that can exclude participating in the Global Initiative on Out-of-School
children from gaining access to or completing even a Children. Children from marginalised social groups are
basic education (UNESCO, 2010b). When it comes two to three times more likely than those from other
to policies on the language to be used in education, groups to be out of school in Bolivia, Ecuador, India and
policymakers face a mix of challenges and trade- Lao People’s Democratic Republic. Research indicates
offs: how can they ensure both universal access and that girls from these socially-excluded groups face the
successful learning outcomes in multilingual contexts, greatest risk of education exclusion and are about two
while at the same time balancing nation-building times more likely to be out of school than boys in the
ambitions against the need for tailored education for same countries (Lockheed and Lewis, 2012).
each individual child?
Barriers
22 This section draws on “School Access for Children from Non-Dominant The problems children face in the classroom stem
Ethnic and Linguistic Communities”, a background paper prepared for this
report by Carol Benson, Independent, United States. from problems in the wider environment, particularly

Box 3.7 Grade 4 Portuguese lesson for Changana speakers, Mozambique
An exchange between a primary school teacher and pupils in Mozambique illustrates the futility of a lesson in
a language the pupils do not understand, even when the content is simple enough for children of pre-school
age. Portuguese is taught from Grade 1 in Mozambique.
Teacher: What can you see in this picture here? [Illustration of boy with three body parts labeled]
Students: [Silence]
Teacher: What can you see here?
Carla: I can see a boy
Teacher: What?
Some students: [Echoing Carla’s answer] I can see a boy.
Teacher: There is a boy...Is it just a boy that you can see here?
Students: [Timidly] Yes.
Teacher: What?
Students: [Different answers] Yes/No
Teacher: What else can you see here?
Students: [Silence]
Source: Chimbutane, 2011
discriminatory or weak policies on language in

education. National and international languages Multilingual education may carry
dominate education policy in every country: the policy considerable costs, but these
challenge is to ensure that the language of instruction are outweighed by the long-term
reflects the way in which children learn and teachers benefits to children in terms of
teach. The evidence shows that teaching in a mother learning outcomes
tongue or in a commonly understood language in the
initial grades is far more effective as it links with the
language of home. back in the 1950s (UNESCO, 1953), there has been
relatively little coordination among the responses
There are serious issues around resources and from donors.23 Scrutiny of early reading failure has
costs. It is true that converting education to children’s led to support for initial literacy in children’s home
home languages—particularly in countries where languages, with data from early grade reading
dozens of languages are spoken, such as Niger assessments (Schroeder, 2013) supporting the
and South Africa—requires major investment in development of improved reading approaches that
linguistic development, materials production and integrate the language children speak at home, but
teacher training. However, policymakers should such programmes tend to be temporary measures
balance this against the social costs of out-of-school rather than systematic approaches.
children and high per-pupil expenditures as a result
of repetition, failure and drop-out. Cost-benefit Supply barriers
analyses in Guatemala and Senegal demonstrate that
mother tongue-based multilingual education (MLE) A number of Global Initiative on Out-of-School
programmes carry considerable initial costs, but that Children studies, including those from Pakistan and
these costs are, eventually, outweighed by the benefits the Democratic Republic of the Congo (UNICEF and
to children in terms of learning outcomes and decrease UIS, 2013b; 2013d), call for the expansion of MLE in
over time (Vawda and Patrinos, 1999; Heugh, 2011).
23 See also the 2010 International Conference on Language, Education and
the Millennium Development Goals—a milestone in the level of recognition
Donor support is not yet up to speed on this issue. granted to learners’ home languages by low-income countries and
donors. http://www.seameo.org/LanguageMDGConference2010/about.
Despite UNESCO’s call for mother tongue education html

Demand barriers
Ghana’s non-formal education
programme School for Life features There is a false perception that families fail to
classes in mother tongue languages for demand education in their home languages. The
children in disadvantaged communities. Ghana Global Initiative on Out-of-School Children
It has helped over 120,000 children study, for example, calls “lack of parental awareness
to date, with 82% of them making the of the importance of schooling” a major factor in
transition to formal education children being out of school, but it also stresses how
irrelevant and incomprehensible lessons can be for a
child who is taught in a dominant language that they
non-dominant languages to improve the quality and do not understand (UNICEF and UIS, 2012d).
provision of basic education as a means to reduce
drop-out and make education more attractive for out- There is, however, a clear need to partner with
of-school children. Despite the growing recognition of communities to address certain attitudes that can
the importance of MLE, however, these approaches hamper education. The Nigeria Global Initiative on
are often limited in depth and breadth. They are Out-of-School Children study, for example, discusses
shallow in terms of the extent to which they build on how non-literate parents believe school-educated girls
literacy and learning foundations in children’s home will reject future husbands, and how Quranic teachers
languages and narrow in terms of the number of convince Hausa families to shun Western education
such languages used for curriculum, training and (UNICEF and UIS, 2012f). This challenge could be
materials (Ouane and Glantz, 2011). They are also eased if parents were involved in developing curricular
often underfunded and lack necessary capacity approaches and were part of the negotiations between
development to implement and maintain the MLE traditional values and national educational aims.
programme.
There are also issues connected to gender, language
There has also been inadequate investment in and ethnicity. Most out-of-school girls worldwide are
home language and bilingual materials and teacher from socially-excluded groups, mainly as a result
development, an issue raised in the Viet Nam and of the mismatch between their own language and
other Global Initiative on Out-of-School Children culture and their experience at school, according
studies that have called for more and better-quality to a study of gender-within-ethnicity disparities
materials and teacher preparation (UNICEF and UIS, in school participation in 16 countries, including
2014h). Even bilingual intercultural programmes countries involved in the Global Initiative on Out-of-
in Bolivia and Ghana are hampered by a lack of School Children: Bolivia, India, Nigeria, Pakistan and
materials and training, which hurts teaching quality Romania (Lockheed and Lewis, 2012).
and learner motivation. The Bolivia Global Initiative
on Out-of-School Children study, for example, found MLE programmes in Guinea-Bissau, Mozambique
that teachers of indigenous children spoke Spanish and Niger seem to be attractive to girls in particular,
at least 75% of the time, with a negative impact on keeping them in school for longer and allowing
their pupils’ comprehension and motivation (UNICEF them to do better because they trust local bilingual
and UIS, 2011)—a problem that can increase the risk teachers and can use their home languages to
of drop-out. In Mozambique, teachers are posted demonstrate their capabilities (Benson, 2005).
outside their language communities to encourage
national unity, yet this practice limits their ability to Breaking the barriers
take a bilingual approach to teaching Portuguese,
a language few teachers have mastered. This may The first step has to be to create an environment
account, in large part, for the poor scores in the that enables education in the languages
SACMEQ assessments (UNICEF and UIS, 2014a). that children use in their everyday lives. For

that to happen, national recognition of the value their communities, providing equitable access to
of multilingual education is a must. Multilingual national curricular goals, including the learning of
education based on the languages children speak dominant languages.
at home attracts children to school and helps to
keep them there by addressing the needs of all Decentralised processes are needed to maximise
learners, including the most marginalised. It does local linguistic resources and the inclusion of linguistic
so by using the best languages of pupils and proficiency as part of teachers’ job portfolios.
teachers, while helping children to learn the dominant Creative, low-cost solutions are available to address
languages needed for their future participation in such challenges as linguistically mixed classrooms
society. Schools that work in languages children or languages that do not have written standards.
can understand allow pupils to participate and Strategies include organising multi-grade classrooms
demonstrate what they know, and encourage the by language (Kosonen, 2006) and community-based
involvement of their families. It is a cultural right. Such language development (UNESCO, 2007).
education must be reinforced by enabling policies
and adequate resourcing that permit good, quality Non-formal educational approaches have a key
implementation and lower per-pupil expenditure. The role to play, given their emphasis on instruction in
costs of multilingual education may be substantial, the language children use at home and their use of
but the pay-offs include bilingual, biliterate learners educational content that is often linked very closely to
who are likely to develop the critical thinking skills the daily lives of their pupils. Appropriate non-formal
and self-confidence needed to live better-informed, curricula, materials, pedagogies and the use of
happier, healthier and more productive lives. appropriate language of instruction can help out-of-
school children from minority groups to learn in safe
Policy alone is not enough: it must be backed and appropriate environments and prevent potential
by resources and implementation. Because discrimination (Yasunaga, 2014).
choices about the medium of instruction are central
to educational access, they should be built into MLE design needs to be consistent with
national strategic plans, resourced appropriately language, literacy and learning research.
and structured for implementation at grassroots and Educational quality is affected positively by the
technical levels to meet the specific needs in the local use of home languages even for a few years, as
context (see Table 3.1). demonstrated by the ‘early-exit’ approaches seen
in, for example, Cameroon and the Philippines
Typically, programmes based on children’s home (Walter, 2013). However, research suggests a need
languages have higher levels of participation, success for the extended use of home languages and for
and enjoyment (as well as parental involvement) and the continued development of biliteracy (reading
lower levels of repetition and drop-out, particularly and writing) across the curriculum to maximise the
among girls. This reinforces the need to collect and benefits (Cummins, 2009) (see Table 3.1). Among
disaggregate data on such programmes and their low-income countries, Eritrea and Ethiopia currently
impact (Benson, 2014). offer the strongest approaches: eight full years of
primary schooling in their most widely-spoken, non-
The ‘pull’ factors to address supply barriers include dominant languages (Walter, 2013; Skutnabb-Kangas
respect for cultural and religious traditions and local and Heugh, 2012). In Ethiopia, this translates into
calendars, along with MLE programmes based on eight years of education offered in seven languages
learners’ home languages and policies that enable (not including English) (Heugh et al., 2007).
appropriate linguistic and cultural approaches to
influence the curriculum. Demand—the ‘push’ Under Romania’s Law of Education (2011), in all
factor—can and should be created by tailoring districts where ethnic minorities account for at least
educational services to the needs of children and 10% of the total population, free public schooling is

Table 3.1 Effective implementation of multilingual education (MLE)
Educational language policy
■■ A good policy allows mother tongues to be used for schooling/literacy; a better policy specifies the approach
to mother tongue-based MLE and how it will be implemented.
■■ The ministry of education should make it clear to teachers and parents (via official channels and public media)
that previous prohibitions of the home language no longer apply.
Implementation strategies
■■ Public awareness and involvement is essential; MLE should be discussed widely at all levels.
■■ Offering MLE as an option in Mozambique has allowed implementation to follow demand rather than taking a
top-down approach that communities may resist.
■■ Autonomous Education Councils representing the major ethnolinguistic groups in Bolivia have facilitated
implementation by raising community awareness of how MLE works.
■■ Decentralised educational decision-making has allowed Ethiopian regions to implement up to eight years of
mother tongue schooling, depending on demand and resources.
Teacher recruitment, training and placement
■■ The training of teachers from the same linguistic communities as their students is only part of the picture; they
require standardised literacy, academic vocabulary and bilingual methodologies.
■■ Where there are too few teachers, it is preferable to raise the capacity of non-professionals from the same
linguistic communities as their students; in Bolivia special ‘pedagogical secondary schools’ prepare young
local women to teach in their home communities.
■■ There may be a need for specialised subject teachers in second or third languages, whose language
proficiency should be developed, assessed and accredited.
■■ Teacher educators, administrators and personnel should also receive training in MLE.
Language and materials development
■■ With support from NGOs, university linguists and educators, teachers and linguistic community members can
develop pedagogical vocabulary and materials in the language spoken at home.
■■ Using the Language Experience Approach (the promotion of reading and writing through the use of personal
experiences and oral language), learners can write their own reading materials.
■■ MLE materials should be as colourful as any other materials, but low-cost, local publishing alternatives may
be more practical to get essential print resources into MLE classrooms.
■■ Viet Nam is piloting bilingual side-by-side content learning materials for Grades 3 to 5 to support bilingual
methodologies and the continued development of vocabulary and skills in both languages.
Curriculum and assessment
■■ The MLE curriculum should be based on the national curriculum, with the exception that most language skills
will be taught initially in the home language and transferred gradually to second and third languages.
■■ Assessment of learning outcomes can often be carried out bilingually to ensure understanding.
Monitoring and evaluation
■■ Learner statistics should be maintained and disaggregated by language, sex and age.
■■ The language that children speak at home should be used to check for understanding.
Source: Benson, 2014

provided in their home language, although Romanian minority groups to learn in safe and appropriate
remains compulsory. Under the law, primary environments and prevent potential discrimination.
education in the mother tongue is guaranteed to
national minorities and, for secondary education, Ghana, for example, has rolled out one of Africa’s
classes in the languages of national minorities are most successful complementary education
organised at the request of parents and guardians. programmes, School for Life, with classes in
As a result of policies to promote the Romani mother tongue languages for children aged 8 to
language in the curriculum, the enrolment of Roma 14 years proving particularly attractive for out-of-
children has improved in the past decade but lack of school children. The aim is to ensure that children
Romanian language knowledge continues to be an achieve basic literacy within the space of nine
obstacle and a drop-out risk factor. Children aged 7 months. To date, the programme has helped over
to 16 years from Romani-speaking families are still 120,000 children in northern Ghana from the most
two and a half times more likely to be out of school disadvantaged districts and communities. An external
than those from non-Romani speaking households evaluation found impressive results, with 82% of
(UNICEF and UIS, 2012g). the children participating in School for Life making
the transition to formal education, completing their
Ethiopia’s top-down policy as implemented primary education and making the transition to lower
by regional education bureaus functions in secondary school (UNICEF and UIS, 2012d).
homogeneous regions but strains limited resources
in the linguistically-diverse south of the country Filling the data gap
(Skutnabb-Kangas and Heugh, 2012). Bolivia’s
bilingual intercultural education approach began with It is crucial to get the research methodology and
top-down policies, phased-in implementation and the data collection strategies right. A number of
education councils that respond to challenges among Global Initiative on Out-of-School Children studies,
non-dominant communities, but subsequent lack including those in Bolivia and Nigeria (UNICEF and
of resources and political distancing has weakened UIS, 2011; 2012f), use language as a proxy for
its implementation (López, 2005; UNICEF and UIS, ethnicity to illustrate disparities between groups, but
2011). In Mozambique, the offer of MLE as an option such data have little explanatory power. It seems,
has allowed time for public demand to be met by in fact, that data on linguistic and cultural issues
decentralised implementation, but the chronically are more relevant than data on ethnicity alone for
under-resourced programme depends on small NGOs educational decision-making. To analyse the effect of
and a number of linguists (Chimbutane and Benson, language on educational success or failure, repetition
2012). Under-resourcing may cause communities to and drop-out, it is vital to have data on the language
reject MLE as a result of low-quality implementation. children use at home in relation to the language of
instruction at school. This includes youth and adult
It is essential to address structural barriers. literacy targets: how, for example, is literacy measured
Given adequate financial and technical resources, and in which language(s) relative to the learner’s own
decentralised implementation allows relevant language, given that an adult head of household
responses to local linguistic and cultural needs. responding to a survey may do so in a language that
These include instruction based on the home differs from the one spoken by their children?
languages of learners and teachers, the involvement
of parents in choices around curriculum delivery, Assessment data should also be analysed in relation
and the adaptation of school calendars to local to languages: what is the learner’s home language
lifestyles, all within reasonable national guidelines. In relative to the language of instruction and to the
addition, appropriate non-formal curricula, materials, language of assessment? According to Ethiopia’s
pedagogies and the use of appropriate language national assessments, learners taught and tested in
of instruction can help out-of-school children from their own languages tend to do better in all subjects

not adapted or equipped to meet their basic needs,
In Ethiopia, learners taught and lacking everything from accessible school buildings
tested in their own languages do to teachers who have been trained to teach in
better in all subjects, including inclusive settings. These children miss out because
English, than learners with a continuing stigma around disability excludes them
language mismatch from the wider society around them, which includes
education. And they miss out because they are, very
often, ‘below the radar’ of current data collection.
(including English) than learners with a language
mismatch (Heugh et al., 2012). MLE programmes It is hard to know how many children have
should test strategically in one or more languages disabilities. In 2011, the World Report on Disability
(Mdube-Shale et al., 2004), because testing only in estimated that more than 1 billion people (or 15%
the dominant language masks whether any difficulty of the global population) live with some form of
lies in understanding the content, understanding the disability, with estimates for the number of children
test questions or expressing knowledge in the test up to the age of 14 living with disabilities ranging
language. An innovative dual-language assessment in between 93 million and 150 million (WHO, 2011).
Niger, for example, found that results were highest for However, such global estimates are speculative and
learners taught and tested in their own language, and have been in circulation since 2001, and they are
lowest for those taught and tested in the dominant derived from data of quality too varied and methods
language (Hovens, 2002). too inconsistent to provide any reliable number of
children with disabilities.
3.7 EXCLUDED AND UNCOUNTED:
CHILDREN WITH DISABILITIES24 Attempts to generate global figures have been
hampered by the lack of a common definition
Yasas, an 18-year-old from Nuwara Eliya, of disability (UNICEF, 2013a). Not surprisingly,
Sri Lanka, has never been to school as he international and national disability prevalence rates
has a serious speech impediment with acute fluctuate wildly, depending on the different surveys
stammering. He uses his hands to express used and the different questions they ask.
himself and his family understands him easily.
His parents were unaware of speech therapy, In principle, a child with a disability has the same right
although this could have helped him to a certain to an education as any other child, as set out in the
extent, and didn’t have the time or money to 2006 UN Convention on the Rights of Persons with
take Yasas to Kandy or Colombo for therapy. Disabilities (CRPD) which came into force in 2008
OOSCI Country Study on Sri Lanka (see Box 3.8).
(UNICEF and UIS, 2013c)
In practice, however, children with disabilities are
Children with disabilities are among the most denied this right disproportionately. Household survey
disadvantaged in terms of missing out on data from 13 low- and middle-income countries
education, being ‘invisible’ in the data and being show that children with disabilities aged 6 to 17 years
overlooked in responses to out-of-school children are significantly less likely to be enrolled in school
(UNESCO, 2013). than their peers without disabilities. A 2004 study in
Malawi found that a child with a disability was twice
Too often, children with disabilities are denied an as likely to have never attended school than a child
education because education systems are simply without a disability (UNICEF, 2013a).
24 This section draws on “Children with Disabilities”, a background paper India has achieved close to universal enrolment in
prepared for this report by Natasha Graham, Senior Disability Advisor,
Partnership for Child Development, Imperial College, London. primary education. However, the figures for children

with disabilities are staggering: out of 2.9 million
children with disabilities in India, 990,000 children The barrier to participation in
aged 6 to 14 years (34%) are out of school. The education is often a bigger problem
percentages are even higher among children with than the disability itself
intellectual disabilities (48%), speech impairments
(36%) and multiple disabilities (59%) (SRI-IMRB
Survey, 2009). India has made tremendous efforts many others do not, and the social impact can
to make its education system more inclusive. Under often outweigh any health or medical implications
the Right to Education Act, all children have the (WHO, 2013).
right to go to school. Additional financial resources
have been provided for children with disabilities to As a result of intense efforts by individual people
attend mainstream schools and for the adaptation with disabilities and the groups that represent
of school infrastructure. Resource centres focused them, there is growing consensus that definitions
on inclusive education have been established to of disability should include social as well as the
support clusters of schools, and large numbers of more traditional medical determinants. The World
teachers have been trained on inclusive education. Health Organization (WHO) defines ‘disabilities’
To accommodate a greater number of children with as an umbrella term that covers impairments (see
disabilities, further progress is needed (UNICEF and Box 3.9), limitations on activities and restrictions
UIS, 2014d). on participation. Similarly, the CRPD defines those
with disabilities as people who have long-term
Evolving perspectives on disability physical, mental, intellectual or sensory impairments
that interact with various barriers external to
Disability is a complex, varied and evolving themselves—whether physical or cultural, or related
phenomenon. While some health conditions to communication and attitudes—to hinder their full
associated with disability result in poor health, and effective participation in society on an equal
Box 3.8 Education in the United Nations Convention on the Rights of Persons with Disabilities
The United Nations Convention on the Rights of Persons with Disabilities (CRPD), which came into force in
May 2008, states that persons with disabilities should be guaranteed the right to inclusive education at all
levels, regardless of age, without discrimination and on the basis of equal opportunity. It has been ratified
by most of the countries that have taken part in the Global Initiative on Out-of-School Children and marks
a global shift in thinking about disability. It entails a shift in attitudes from viewing people with disabilities as
objects of charity and medical treatment to being full and equal members of society.
Article 7 is dedicated specifically to children with disabilities:
1. States Parties shall take all necessary measures to ensure the full enjoyment by children with disabilities of
all human rights and fundamental freedoms on an equal basis with other children.
2. In all actions concerning children with disabilities, the best interests of the child shall be a primary
consideration.
3. States Parties shall ensure that children with disabilities have the right to express their views freely on all
matters affecting them, their views being given due weight in accordance with their age and maturity, on an
equal basis with other children, and to be provided with disability and age-appropriate assistance to realize
that right.
The Convention also includes specific references to education, particularly in Article 24, which states
that persons with disabilities should be guaranteed the right to inclusive education at all levels, without
discrimination and on the basis of equal opportunity, and children with disabilities shall not be excluded from
free and compulsory primary education or from secondary education on the basis of disability.

regions, including CEE/CIS, the emphasis often
Box 3.9 Impairments and disabilities tends to be on the individual’s diagnosis and their
In general, an impairment is an injury, illness perceived inability to take part in society (UNICEF and
or congenital condition that causes a loss UIS, 2013a), rather than on society’s ability to include
or difference in physiological function. An
impairment is not, in itself, a disability. One the individual. Progressive shifts are being seen in
example would be children with reduced vision some CEE/CIS countries, however, with a growing
who have the eyeglasses that allow them to focus on the ability of schools to include children with
participate fully in the classroom.
disabilities, rather than on the inability of a child to
A disability is the loss or limitation of equal function in school.
opportunities to participate in society as a result
of social, institutional and environmental barriers.
One example would be children with reduced The prevailing approach to disability—be it medical
vision who do not have eyeglasses and who or social—will influence the policy response to out-
cannot, as a result, participate in school on an of-school children with disabilities. As shown in
equal basis with other children.
Table 3.2, there are marked differences between the
two approaches in terms of perceptions of disability
and measures to ensure that children with disabilities
basis with others. When viewed through the lens receive an education.
of a social approach, people with disabilities are, in
effect, no longer ‘disabled’ if they can participate The barriers
fully in society. For a child with disabilities, this
includes full participation in the classroom. The barriers to the education of children with
disabilities can range from the immediate and
The starting point for the measurement of functional tangible, such as the lack of a ramp to enter a school
capacity under this social approach was the building or the cost of transport, to the long-term
International Classification of Functioning, Disability impact of social norms and stigma that reinforce their
and Health (ICF) developed by WHO (2001). In exclusion.
2007, WHO built on the ICF classification to publish
the International Classification of Functioning, The sheer lack of reliable and comparable data on
Disability and Health for Children and Youth (ICF- children with disabilities—their numbers, the nature
CY) (WHO, 2007). The first unifying framework of their disabilities and their educational needs—
to describe the impact of context on a child’s only adds to the serious barriers they face to their
functioning encompasses several environmental education, making it difficult to develop effective
factors—including a child’s ability to participate policies and budgets for their inclusion.
in education—that should be examined when
recording a profile of whether a child can function Supply barriers
in society. In 2012, WHO and other stakeholders
agreed to merge ICF and ICF-CY into one Children with particular physical disabilities may be
classification to arrive at a comprehensive ICF confronted by school facilities that are inaccessible to
that addresses all aspects of functioning across a them, from classrooms to toilets. Children with visual
person’s lifespan. or hearing impairments struggle in environments
with inadequate light or poor acoustics, while wider
This positive global shift towards a social approach problems with transport often prevent children with
to disability represented by the CRPD is not yet fully disabilities from making the journey to school in the
embedded in practice. The ways in which disability first place. In Kyrgyzstan, for example, few (if any)
is measured and programme responses remain school buildings and classrooms are accessible to
predominantly medical, with a continued focus on children with disabilities. Transportation for children
specific physical or mental impairments. In some with disabilities has been reported as being too costly

Table 3.2 Approaches to schooling children with disabilities
Traditional/medical approach (which may include ‘integration’

into a mainstream school/classroom) Social/inclusionary approach
The focus is on the needs of ‘special’ students. The focus is on the rights of all students.
The focus is on the student. The focus is on the classroom.
The aim is to change/remedy the student. The aim is to change the school.
Programmes for students. Strategies for teachers.
The student is assessed by a specialist. Teaching/learning factors are assessed.
Programmes are diagnostic/prescriptive. The emphasis is on collaborative problem solving.
The student is placed in an appropriate programme. The regular classroom is adaptive and supportive.
The premise is that the student with special needs will benefit The premise is that all students benefit from full inclusion.
from being integrated.
The interventions are technical (special teaching). The emphasis is on good quality teaching for all.
Source: Adapted from Porter (1995); Walker (1995) in Thomas, Walker and Webb (1998)
for families and is not covered by the small monthly

allowance provided by the government (UNICEF and The emphasis is too often on the
UIS, 2012e). individual’s perceived inability to
function in society, rather than on
Children with disabilities are particularly society’s ability to accommodate
disadvantaged by inflexible curricula and examination the individual
systems, combined with non-inclusive teaching
methods. Country reports from a number of countries
that participated in the Global Initiative on Out-of- who are eligible for inclusive schooling but some
School Children—Ghana, Kyrgyzstan, Romania, parents don’t want this.”
Sri Lanka and Tajikistan—paint a picture of limited Teacher quoted in the OOSCI Country Study on
training for teachers on how to teach in inclusive Romania (UNICEF and UIS, 2012g)
settings, and little adaptation of school programmes
and teaching materials to the needs of children with Although the concept of inclusive education has
disabilities (UNICEF and UIS, 2012d; 2012e; 2012g; been promoted internationally for more than a
2013c; 2013e). The lack of appropriate training and decade, the term itself is often poorly defined and
support for teachers to teach children with disabilities government policies may be unclear and poorly
in regular schools has been cited as a factor in their implemented. A number of countries taking part
unwillingness to include these children in their classes in the Global Initiative on Out-of-School Children
(UNICEF, 2013a). reported that some schools were unaware that such
a policy even existed. Such a poor climate for the
Demand barriers inclusion of children with disabilities in education
acts as a brake on demand. Several Global Initiative
“Some parents don’t want their own child to be on Out-of-School Children country studies cited
assessed because they are ashamed of what the negative attitudes towards children with disabilities
neighbours might say … he is handicapped; so as a major factor in whether children enrol or fail to
they don’t do it although the child has a problem complete their education. In some countries there is
… Some parents have simply refused their a persistent and common belief that a child with a
children’s assessment, they can’t accept this disability is a ‘punishment’ imposed on a family for
label, that their son or daughter is this way and past mistakes, a belief that may be internalised by
so the child has to suffer. We even have children their parents. In the absence of effective inclusive

education policies, the resulting shame and stigma The data challenges
are thought to keep many children with disabilities
hidden from view at home. Global Initiative on Out-of-School Children studies
across the board concur: the barriers to the
The Ghana country study, for example, noted a education of children with disabilities are reinforced
high degree of stigmatisation, with a recent study by the invisibility of children with disabilities in the
in the northern rural area finding that children with data. Most country studies revealed a chronic lack
disabilities are viewed by parents as not having of quality administrative or household survey data
any (or only a very limited) capacity to learn. These to understand how many children with disabilities
parental perceptions combine with the failure of there are and what kinds of disabilities they face.
schools to promote inclusive education to keep This makes it difficult to gauge how many are out of
children with disabilities out of the classroom school, why they are out of school and how best to
(UNICEF and UIS, 2012d). Similarly, in Kyrgyzstan fulfil their right to an education. Most Global Initiative
the public perception of disability and the frequent on Out-of-School Children countries reported that
teasing and bullying of children with disabilities were they lack the necessary assessment tools and
cited as significant social barriers to their education methodologies to identify children with disabilities,
(UNICEF and UIS, 2012e). which makes it difficult to collect disaggregated
disability data and, by extension, to plan and
While looking at demand barriers, it is important to budget for appropriate services, infrastructure and
consider how disability, poverty, gender, ethnicity resources.
and geographic location intersect, given that the
school participation of children with disabilities is The collection of data is hampered by the
often determined by the interplay of such factors. serious and persistent differences in definitions of
It is clear that children with disabilities face disability, alongside the mass of methodologies
multiple forms of discrimination that lead to their and measurement instruments that are used to
exclusion from society and education. Girls with identify the children concerned. As a result, existing
disabilities experience double discrimination, which prevalence estimates of childhood disability vary
places them at higher risk of being out of school, to such an extent that cross-country comparisons
experiencing gender-based violence, sexual abuse, become almost meaningless—from less than 1%
neglect, maltreatment and exploitation (United in countries like China to almost 50% in the Central
Nations, 2014). African Republic (UNICEF Global databases, 2014). It
is often the case that different instruments within one
Poverty and disability are often inter-related. single country generate conflicting rates of disability.
According to the 2011 World Report on Disability,
80% of people with disabilities live in developing The Global Initiative on Out-of-School Children
countries and they are disproportionately represented Regional Report on CEE/CIS, for example, points
among the poor. Disability is both a cause and out that only 1.5 million children in the region are
consequence of poverty: with limited opportunities officially registered by their governments as having
for education and economic participation, people a disability, but this is likely to leave millions more
with disabilities often experience life-long poverty unaccounted for (UNICEF and UIS, 2013a). Many
and exclusion. Poverty itself can contribute to of these uncounted children may be in school
disability, linked as it is to poorer access to nutrition but lack the specific and appropriate support
and healthcare and a greater risk of being exposed they need to prevent poor learning outcomes and
to dangerous working conditions. Households that drop-out. Each country’s estimate comes from a
include people with disabilities carry the direct cost combination of hospital registries of children that
of their care, which often results in lower standards are identified at birth as having a disability; data
of living. on the number of children living in institutions or

attending special schools or classes; and lists of
children registered by parents or doctors as having Negative public perceptions of
disabilities. This excludes many children with disability, coupled with teasing
disabilities that develop after birth, children who and bullying, can be barriers to the
are not registered as a result of stigma, children education of children with disabilities
deemed ‘uneducable’, or children under the care
of different ministries that may not be captured
in any data-sharing. However, it is common that Under-reporting is a major issue and may well reflect
children from socially-vulnerable groups may be the reluctance of parents to report a child’s disability,
disproportionately assigned to ‘special schools’ given the continued stigmatisation of children
when they do not have any impairment but with disabilities and their families, or their lack of
rather are non-native speakers of the language of awareness of their child’s disability in the absence of
instruction or are simply the poorest of the poor. effective screening services.
As a result, the definitions used in the CEE/CIS
region are country-specific and are not necessarily Breaking the barriers
harmonised with each other or with the global
definition (UNICEF, 2011b). Ultimately, the education of children with disabilities
hinges on the removal of the many barriers they face,
The Global Initiative on Out-of-School Children from the lack of physical access to classrooms to
Regional Report on South Asia cited the invisibility the stigma that keeps them hidden away at home.
of children with disabilities as a significant factor in It also requires the provision of appropriate support
their exclusion from education. In India, for example, and an understanding of their functioning and needs,
a nationwide survey of 99,226 households in 2009 all backed by robust and comparable data. When
identified 1.5% of primary and lower secondary the school environment is welcoming and sensitive
school-age children as having a disability. As stated to the needs of children with disabilities, even simple
earlier, an estimated 34% of primary and lower adjustments can make an immense difference, as
secondary school-age children with disabilities were shown in Namibia:
out of school, compared to a national average of
4% (UNICEF and UIS, 2014d). The design of the Simenda was struggling to cope at secondary
questionnaire is thought to have had a major impact school in rural Namibia until his hearing
on the reported disability prevalence, identifying impairment was diagnosed. His teachers were
children with disabilities through two questions: briefed on supportive strategies to help him in
a filter question (“Does [name] have a disability?”) class, such as allowing him to sit wherever he
and a general question on the type of disability. could hear and checking that he was following
No information was gathered on either severity or the lesson. After two terms, his results in class
functioning. As a result, the low prevalence rate in tests had substantially improved—to the eighth
India may capture only those children with the most highest result in a class of around 30 pupils.
severe or apparent disabilities. Not surprisingly, OOSCI regional report for Eastern and Southern
similar prevalence rates are reported in surveys from Africa (UNICEF and UIS, 2014b)
other developing countries that use these same basic
questions (UNICEF and UIS, 2013a). Above all, governments need to deliver on their
repeated promises, including those made under the
An additional challenge is bias in reporting. CRPD, to ensure that all children have an equal
While children with visible impairments may right to education. This requires inclusive education
be acknowledged, those with mild or ‘hidden’ policies that benefit all children, with or without
disabilities, such as learning or psycho-social disabilities, by ensuring that teaching responds to
impairments, may well be overlooked. individual differences and diverse abilities as a matter

Table 3.3 A checklist for the effective implementation of inclusive education
The enabling legislative and policy environment

■■ Harmonise existing legislation with the relevant international conventions, such as the Convention on the Rights of
Persons with Disabilities (particularly Article 24 on education) and the Convention on the Rights of the Child, to create
education systems that are inclusive at all levels.
■■ Remove legislative barriers to the participation of children and teachers with disabilities in education. For example, abolish
laws and policies that prevent the admission of children with disabilities into school or that do not allow people with
disabilities to become teachers.
■■ A good education policy or sector plan includes the needs of children with disabilities through the adaptation of existing
measures, such as the provision of training on inclusive teaching and the provision of textbooks in accessible formats.
■■ A good education policy or sector plan also incorporates targeted measures for children with disabilities. These include
the provision of allowances/funds for transport or assistive devices and technology for children with disabilities.
Implementation strategies
■■ Strengthen data on children with disabilities, align definitions and instruments with international standards and the ICF
framework, and synchronise data collection on children with disabilities between ministries and national statistical offices.
■■ Plan and implement public awareness campaigns to change attitudes towards children with disabilities.
■■ Promote inclusive early childhood care and education programmes. Early intervention and stimulation can enhance the
development of children with disabilities and foster inclusive and non-discriminatory attitudes among children from the
start.
■■ Design social protection programmes that account for the additional costs of disability experienced by households with
an adult or child with a disability, which can pose significant barriers to school participation.
Teacher recruitment, training and placement
■■ Embed the principles and strategies of inclusive education and inclusive teaching strategies throughout the teacher
training curriculum.
■■ Design teacher training to incorporate practical experience of teaching in inclusive settings with adequate interaction with
children with different types of disabilities.
■■ Establish teacher support systems through periodic monitoring, peer support, support from resource teachers and
resource centres, and the exchange of knowledge and support.
■■ Train and enable teacher educators to adopt an inclusive orientation and to put that orientation into practice in inclusive
settings.
■■ Remove barriers to the education and recruitment of people with disabilities as teachers and take reasonable measures
to enable them to participate in the education system. This can break down stereotypes around disability and provide role
models for learners with disabilities.
■■ Promote whole school-based training on inclusive education that includes community members.
■■ Apply the expertise available in the special education system to support inclusion of children with disabilities in general
schools.
Accessible schools
■■ Promote standards, budgets and monitoring to ensure accessible school construction and the provision of water,
sanitation and hygiene facilities that are designed to be universally accessible.
■■ Provide assistive devices, alternative or augmented forms of communication such as sign language and accessible
materials such as textbooks and other learning materials in Braille or as audio books.
■■ Promote positive images of people with disabilities in textbooks and other school materials.
Curriculum and assessment
■■ The MLE curriculum should be based on the national curriculum, with the exception that most language skills will be
taught initially in the home language and transferred gradually to second and third languages.
■■ Assessment of learning outcomes can often be carried out bilingually to ensure understanding.
Monitoring and evaluation
■■ Data from education management information systems (EMIS) should record information on learners which can be
disaggregated by disability.
■■ EMIS should include data on the accessibility of schools.
Multi-sectoral approaches
■■ Adopt multi-sectoral approaches and coordination of services between ministries of education, health, social welfare,
transport, etc. to ensure the seamless provision of the support and services necessary for the education and school
participation of children with disabilities.
■■ Strengthen linkages with community-based rehabilitation services to identify, prepare, enrol and sustain the educational

of course. A checklist for the effective implementation
of inclusive education is set out in Table 3.3. Existing prevalence estimates of
childhood disability vary to such an
Several recent publications and reports have extent that cross-country comparisons
recommended key actions to be taken by become almost meaningless
governments to include children with disabilities
in mainstream policies, systems and services
(WHO, 2012). There are two main strands: broad materials, processes and assessments need to
social reforms beyond the education sector and be accessible and applicable to every child. These
reforms within the sector itself. should be reinforced by teachers who have received
practical training and ongoing support for teaching
Broader social reforms should promote inclusive in inclusive settings and who have the backing of
education for children with disabilities at all levels schools and communities that are committed to
(including early childhood education) and support the inclusive approaches. At the most practical level,
practice and culture of inclusion across education schools need to be built or retrofitted to ensure that
systems by reviewing national policies in all relevant they are accessible for children with disabilities:
sectors—health and social, as well as education. The ramps instead of stairs, for example, and doorways
aim is to ensure that policies are inclusive and aligned wide enough for wheelchair users.
with international conventions and commitments
and that multi-sectoral strategies promote the A number of countries provide solid examples of
inclusion and participation of children with disabilities what works. Kenya’s Oriang Inclusive Education
in different spheres. Sector-wide strategies, scheme, for example, focuses on and addresses
programmes and budgets should be reviewed to the constraints to the education of children with
determine whether they include concrete actions to disabilities and has increased the number of children
support children with disabilities and their families. with disabilities attending five state primary schools.
Every country needs a comprehensive multi-sectoral Its key interventions include: access to sensory-
national strategy and plan of action for children stimulation learning materials and assistive devices,
with disabilities that follow an inclusive approach financial support to adapt school environments,
to address family support, community awareness and building close links between community health
and mobilisation, human resource capacity, workers and teachers (United Nations, 2011).
coordination and service provision. Such reforms
must be bolstered by clear lines of responsibility In Bangladesh, the Centre for Disability in
and mechanisms for coordination, monitoring and Development (CDD) is working to address the
reporting across all relevant sectors. lack of teachers who have disabilities through its
inclusive education trainers, several of whom are
In particular, social reforms must aim to build positive visually impaired or have other impairments. In
attitudes towards disability. For example, the ‘It’s Mozambique, Ajuda de Desenvolvimento de Povo
About Ability’ campaign in Montenegro, launched by para Povo, a national NGO, has worked with the
the government in partnership with UNICEF, reduced national organization for people with disabilities,
the percentage of people who found it unacceptable ADEMO, to train student teachers to work with
for a child with disability to attend the same class as children with disabilities and to train student teachers
their own child from 64% in 2010 to 39% in 2012. who have disabilities (UNICEF, 2013a). Serbia’s ‘big
bang’ approach to inclusive education is outlined in
Education sector reforms need to ensure that all Box 3.10.
education strategies and action plans are inclusive,
first and foremost, and that they therefore include Non-formal education also offers a pathway for the
children with disabilities. Curricula and learning educational inclusion of children with disabilities

Box 3.10 Towards inclusive education in Serbia
In partnership with the World Bank, UNICEF and other partners and stakeholders, the government of Serbia
has adopted a ‘big bang’ approach towards realising inclusive education.
Enabling Legislative Environment: Serbia enacted the Law on the Foundations of the Education System in
September 2009, which is imbued with the principles of equal opportunity, inclusion, anti-discrimination
and the best interest of the child and demands systemic changes. For example, the Law prescribes
that school enrolment policies have to be unconditional and inclusive and abolishes the need for an
assessment of the child’s capacity and skills as a pre-condition for enrolment. Instead such assessments
are carried out during the course of the first year of schooling and serve as the basis for designing an
individual education plan as needed, aimed at facilitating each child’s learning and inclusion in the school
community.
Teacher Training: Within a short period of two years, Serbia provided in-service professional training
for some 15,000 teachers in Serbia (about 20% of the total teaching staff) to work in classrooms with
children with disabilities and change mind-sets of individuals and educational institutions in order to fully
understand, accept and ultimately embrace inclusion.
Public awareness campaigns: The Ministry of Education, Science and Technological Development, in
partnership with non-governmental and international organizations, media and local partners, organized a
campaign entitled “All to School—Future for All”. The campaign focused on changing perceptions about
inclusion and helped to build a critical mass of supporters from all parts of society—including parents,
politicians and professionals in the health, social welfare and education sectors.
Horizontal learning: In Serbia, a Network for Support of Inclusive Education was established in 2010 by
the Ministry of Education and in cooperation with the World Bank and UNICEF with the aim to provide
inclusive education models, provide capacity building for inclusive education in pre-primary and primary
education, ensure the presence of trained professionals to provide continuous development, as well
as direct coaching, consulting and supervision to schools. It supported capacity building for inclusive
education in pre-primary and primary education. Model inclusive schools were established in 14 locations.
By 2014, over 2,000 practitioners and school advisors were included in different forms of knowledge
exchange, including observance of actual classroom practice and discussions on challenges and solutions
for inclusive education. The student population covered by schools that are supported through the network
is around 150,000 students.
Strengthening inclusive learning environments: With the aim of empowering schools to implement inclusive
education, Serbia designed a programme of grants for small school projects. These grants, financed under
a World Bank loan, have been implemented in over 30% of schools in 96% of Serbian municipalities. While
varying in scope and focus, the school initiatives were primarily directed to capacity building of staff, the
elimination of physical and communication barriers for inclusion of children with disabilities, the promotion
of cooperation with parents, and local community awareness-raising on the importance of inclusive
education.
Monitoring: In December 2010, four months after the entry into force of the Law, UNICEF supported an
independent, rapid assessment in order to identify bottlenecks and constraints in the implementation of
the Law’s inclusive provisions. It provided important insights into what was and was not working, and
formed the basis for an improvement plan, which was then implemented, including the initiation of the
development of a monitoring framework. To enable the tracking of progress of Serbia’s implementation of
inclusive education laws and policies, UNICEF, together with the Government Unit on Social Inclusion and
Open Society Foundation Serbia, supported development of the Framework for Monitoring of Inclusive
Education in Serbia. The monitoring framework consists of indicators at school, municipal and national
levels that are, to a large extent, correlated with each other and enable the flow of information in both
directions (bottom-up and top-down). For each of the indicators, the framework includes input, process,
and output/outcome targets.

who might otherwise miss out. In Bangladesh, for than, students from government schools (Nath,
example, BRAC primary schools use a learner- 2002; Yasunaga, 2014).
centred approach. Teachers, all female, are recruited
from local communities and receive an initial 12-day Filling the data gap
training course from BRAC, plus in-service monthly
training. Class schedules are flexible and schools The collection, analysis and reporting of good,
include students with disabilities, with BRAC quality data on children with disabilities and the
providing corrective surgeries (when appropriate) extent of inclusiveness in the school system can
and assistive devices free of charge to students help governments to meet their commitments
who need them. BRAC builds ramps to improve the to the education of children with disabilities. It is
accessibility of its schools, and classroom policies crucial that the definitions of disability used by
for children with disabilities include sitting them at each country comply with international standards
the front, studying in pairs, the inclusion of disability- and that data collection uses measurement
related issues in textbooks and awareness-building procedures that provide internationally-comparable
among classmates and teachers. BRAC students prevalence rates.
can sit for the government examination that marks
the end of primary school, and their results show that What is needed is data collection based on a
they can often compete with, if not perform better broad and consistent definition of disability to
Box 3.11 A way forward on the data

Partnership is essential for a reliable and globally-relevant monitoring and reporting system on child disability.
No single entity can hope to capture the complex ways in which the barriers within education systems and in
the wider environment combine to keep children with disabilities out of school.
Many initiatives are being undertaken by UNICEF and its partners to address the need for comparable and
reliable data on children with disabilities.
A manual is being prepared for the production of statistics on children with disabilities to guide those
collecting data on this issue. The manual, guided by inputs from 40 international experts, will set out
conceptual and theoretical issues on the measurement of disability in children and review methods and tools
that have been used to collect data in this area.
UNICEF and the Washington Group on Disability Statistics have developed a survey module on child
functioning and disability for use in household surveys and censuses. The model reflects current thinking
around disability and can produce internationally-comparable data on children aged 2 to 17 years. The
module will explore their ability to take part in a range of activities and social interactions and look beyond
simple yes/no answers to better reflect the degree of disability and its impact on a child’s daily life.
The two organizations are also working on a related survey module to measure the school environment and
children’s participation in education, with an emphasis on measuring the barriers to the education of children
with disabilities and their solutions. The module will cover attitudes, as well as accessibility, getting to school
and affordability.
A team of international experts is working with UNICEF to create a toolkit and methodological guidelines for
in-depth assessment of the limits and restrictions children face, based on existing examples of best practice
in low-income countries. This uses the International Classification of Functioning, Disability and Health
for Child and Youth (ICF-CY) as the framework for an approach to disability focused on the barriers to the
Source: UNICEF, 2014c

capture the scale of the challenges, combined with Any reform that aims to improve the lives of children
expert assessment and follow-up to respond to the with disabilities needs to be driven by the best
educational needs of each individual child. possible evidence, with effective data collection
on disability tied directly to service provision (see
There are welcome efforts to move away from the Box 3.11). Data that identify gaps in service delivery
classification of children by their type of disability for children with disabilities should be used to
and towards assessment of the way in which they advocate for the sustainable financial and technical
are able to function in society and in school. The support that will—among many other benefits—
provision of education and individualised support bring these children into the classroom and keep
should be based on such tailored assessment of them there.
functioning of a child within a given environment.
Two children may be diagnosed with cerebral palsy, A number of countries are already changing the way
for example, but have markedly different functional in which they measure and respond to disability,
capacities and needs. Similarly two children using with Cambodia a prime example of a country that is
wheelchairs may have vastly different levels of school mobilising data collection to respond to the needs of
participation depending on how enabling or disabling individual children (see Box 3.12).
their environment is.
Box 3.12 Making the invisible child visible in Cambodia

“Knowing the situation about children with disabilities will allow Cambodia to plan and provide quality
education for ALL of Cambodia’s children.”
Nath Bunroeun, Secretary of State for Education, Youth and Sport
The 2008 census in Cambodia reported that just 1.4% of the country’s people had some form of disability
(Cambodian National Institute of Statistics, 2014), a strikingly low rate that may well have been linked to
confusion over terminology.
In 2010, the Cambodian Ministry of Education, Youth and Sport (MOEYS) mounted a national survey with
support from the Global Partnership for Education (GPE) to gather data on all out-of-school children,
including children with disabilities, to better understand the links between these two groups. A ten-question
screening instrument was used to identify children aged 2 to 9 years with potential impairments, disabilities
or any other major health issues. Children who screened positively were referred for additional health
screening by a team of doctors, psychologists, and hearing and vision specialists to confirm the presence of
an impairment or disability and to provide treatment wherever necessary.
The results confirmed that many children had easily-treatable health conditions, such as partial hearing
loss caused by untreated ear infections, and that about 5% of children with poor eyesight simply needed
eyeglasses to read properly and participate fully in school. Most of these conditions had been undiagnosed
prior to the survey.
The disaggregated disability data generated by the 2010 survey has made children with disabilities more
‘visible’ in Cambodia. As a result, the MOEYS is designing a national disability screening approach for all first
graders, including eyesight tests, to increase the enrolment of children with impairments and disabilities. An
inclusive education training module has been developed and approved by the MOEYS, and pilot projects to
mainstream education for children with disabilities in 18 provinces are beginning to inject greater equity into
Cambodia’s education system.
MeiMei, a 9-year-old third grader in Takéo Province, began to miss school because of headaches and
was no longer the good student she had once been. She struggled in class because she could not see
clearly what was written on the board. A disability screening confirmed her poor vision as the source
of her headaches and this was corrected with a simple pair of glasses. She is now back at school and
flourishing in her studies.
Source: Global Partnership for Education (n.d.)

There is no doubt that the inclusion of children in remote rural areas, and from child labourers to
with disabilities in education will carry a significant children whose home language differs from that used
financial cost related to the training and re-training in the classroom. In each case, governments need
of teachers, the re-modelling of schools to make to commit significant resources to reach the world’s
them accessible, and the provision of specialised out-of-school children, but the long-term benefits
support and equipment in the classroom. However, in terms of health, prosperity, social cohesion and
these costs are outweighed by the positive impact of national productivity are well worth the price. The
inclusion, not only on the children concerned but also next chapter looks at this issue in more detail, aiming
on their fellow pupils, their schools and communities. to close the current knowledge gap on the true costs
of universal primary education.
A similar argument applies to all children profiled in
this chapter, from those caught up in war to girls

Chapter 4
Financing needs for out-of-school

children
4.1 INTRODUCTION
The standstill in global progress on
This chapter builds on the findings of previous the number of out-of-school children
chapters in order to set out recommendations reinforces the need to reconsider
on the all-important financing of policies affecting the resources required to provide
out-of-school children. Chapter 2 uses the latest education for every child
administrative and household survey data to
reveal the magnitude of the challenge and outlines
where, and towards whom, we must target our wide and targeted interventions that are crucial for
interventions. Chapter 3 describes the barriers to decision makers at the national and sub-national
school inclusion and provides a menu of proven levels. Indeed, transferring resources toward the
policies and strategies to overcome them. The most marginalised requires a dramatic shift from
fact is, however, that the resources—financial and the existing resource allocations whereby wealthier,
human—that are available to tackle the barriers are urban areas receive disproportionately more
limited. Policymakers who are deciding where and resources than poorer, rural areas with more need.
how to spend public financing need solid information
on the cost of getting all children into school and the The chapter presents a new model focused on out-
expected impact of the interventions they select to of-school children that provides policymakers with
address this challenge. an overall picture of the costing implications for both
expansion and targeted strategies. This innovative
The standstill in global progress on reducing the approach is elaborated using available data from
number of out-of-school children reinforces the a country that still has a long way to go to achieve
need to reconsider the resources required to provide universal primary education and that faces some
education for every child. of the greatest and most pressing challenges: the
Democratic Republic of the Congo.
The enrolment of all out-of-school children and
adolescents of primary and lower secondary school If they are to enact the system-wide and targeted
age must take into account both the costs of interventions listed in Chapter 3, policymakers need
system-wide expansion of education and targeted answers to three key questions.
interventions to reach the most marginalised
children. Rather than presenting a new global mm What are the optimal levels of each of these
estimate of the cost to enrol all primary and lower programmes?
secondary school-age out-of-school children and mm How should they be distributed within the country?
adolescents, this chapter will drill down into the mm Should particular programmes be prioritised or
costs of, and challenges for, financing the system- accelerated?

dropout and the timeline for the achievement of
The selection of targeted interventions global education goals.
is hampered by a lack of information
on effectiveness and costs One major flaw of these early models identified by
Glewwe et al. (2006) was their assumption that
supply-side considerations, such as the availability
The selection of targeted interventions is, at present, of school places and the number of teachers,
hampered by a lack of information on effectiveness are—invariably—the most binding constraints on
and costs. However, an initiative by UNICEF and the participation in basic education. As discussed in
World Bank seeks to compile data on the effectiveness Chapter 3, however, the mere expansion of the
of targeted interventions and uses a new policy tool existing education offer will not ensure enrolment of
called the Simulations for Equity in Education (SEE) children who face specific types of disadvantage (for
to reproduce the impact of policy options on the example, see World Bank, 2004). Children are out of
enrolment of marginalised children. This chapter school for a variety of reasons: many of them linked to
therefore also presents results for a sub-national demand-side barriers, such as social norms around
analysis of gender-specific policies in the Balochistan gender, stigma that works against the enrolment of
province of Pakistan using the SEE approach. children with disabilities or the failure to teach children
in the language they use at home. Furthermore, some
4.2 AN EVOLVING UNDERSTANDING OF supply-side failures—and in particular poor quality
FINANCING NEEDS FOR UNIVERSAL education—cannot be remedied by simply expanding
BASIC EDUCATION the current education infrastructure but require
improved teaching, among other reforms.
Alongside the progress made toward universal basic
education over the past two decades, there has Such shortcomings in the early models were
been progress in our understanding of the financial addressed by the Education and Policy Data Center
resources required to achieve that goal. A number (EPDC) and UNESCO in a background paper for
of models that estimate the financing needs for the 2010 EFA Global Monitoring Report (EPDC and
universal enrolment (or completion) at different levels UNESCO, 2009). Using the EPDC’s High-Level
of education have emerged since 2000. Early models Interactive Projections model, the paper estimated
used a linear costing approach to assess global the financing required to achieve four of the six EFA
education financing needs, estimating the spending goals: early childhood education, primary education,
on primary education that would be needed to lower secondary education and adult literacy. The
expand existing education systems to accommodate authors augmented the average per-pupil spending
universal enrolment, based on average recurrent approach by factoring in the additional spending
spending per pupil (Devarajan et al., 2002). More required to reach marginalised children. However, as
sophisticated models incorporated capital expenses a result of data constraints, the model’s treatment
(Brossard and Gacougnolle, 2001) and accounted of marginalisation was necessarily somewhat
for improvements in the quality of education provision crude, estimating the size of just one generic and
(Delamonica et al., 2001). These generated estimates marginalised group in each country. It did not
of annual financing needs that ranged from $6.5 account for the different types of marginalisation
billion (Bruns et al., 2003) to $17 billion (Delamonica or the different costs of interventions designed
et al., 2001). The wide range reflects the variation in specifically to reach those different groups. Assuming
the sets of countries covered by the models, as well that countries were able to meet certain domestic
as their sensitivity to underlying assumptions about contribution targets between 2008 and 2015, the
unit costs of schooling, population growth, economic annual funding gap estimated by this exercise
growth, the treatment of private education, repetition, was $24.1 billion (in US constant 2007 dollars)

for primary and lower secondary school in all low- tongue instruction for non-native speakers of that
income countries, of which 13% would be required language of instruction, or schools in remote areas,
for marginalised groups of children who experience as outlined in Chapter 3. For children facing these
“extreme and persistent disadvantage in education vulnerabilities, the cost of enrolling is higher than
that sets them apart from the rest of society” the cost of enrolling the average pupil—these are
(UNESCO, 2010b). the children for whom the Glewwe et al. critique,
with its emphasis on demand-side barriers, is most
Consideration of the needs of marginalised relevant.
populations has continued to move to the forefront
of costing approaches in recent years, with improved Data on past and expected school exposure can be
data on marginalised children and adolescents and used to estimate the costs to enrol out-of-school
the interventions to reach them. The UNICEF and children who, without interventions, will not complete
World Bank Simulations for Equity in Education primary education: the children who have left school
(SEE) model, piloted in Ghana and Pakistan in 2013, early and those who are unlikely to ever set foot in a
projects the impact of targeted interventions on classroom (as discussed in Chapter 2). This assumes
participation in education. This shift in modelling that out-of-school children who are expected to enter
is driven by the consequences of global trends school in the future, in most cases one or two years
in enrolment: recent progress toward universal late, do not need the kind of interventions necessary
basic education in the majority of countries has for children who have dropped out or who will never
confirmed that the remaining out-of-school children attend.
and adolescents are likely to be the most difficult
to reach. As a result, the general expansion of
existing education systems becomes less effective There will, inevitably, be an
in increasing enrolment, and specific targeted expansion cost in enrolling every
interventions to enrol marginalised children and out-of-school child of primary and
adolescents become ever-more important. lower secondary school age in
primary education
4.3 A MODEL FOR ESTIMATING FINANCING
NEEDS FOCUSED ON OUT-OF-SCHOOL
CHILDREN AND ADOLESCENTS There will, inevitably, be an expansion cost in
enrolling every out-of-school child and adolescent of
Children with certain profiles, such as those with primary and lower secondary school age in primary
disabilities or who do not speak the language education—an expansion cost associated with
used at school, are most likely to be out of school increasing the supply of teachers (teacher training
because they face significant barriers to education cost and salary), classrooms and materials. The
that are highly context-specific. In countries where responsibility for financing that expansion cost is
universal access—let alone completion—remains split between the public sector (the Public Expansion
a distant goal, widespread and extreme poverty cost) and households (the Household Expansion
erects serious financial barriers to schooling for cost). The total of these costs represents the
large numbers of children of school age, often financing required to create enough school places to
forcing them into child labour or child marriage. accommodate out-of-school children in the public
In countries that are within the last mile of the education system. On top of this, there will also
journey towards universal access and completion, be Targeted Intervention costs to reach children
it is the most marginalised children who face with different profiles linked to marginalisation (for
specific barriers, such as the lack of accessible example, children with disabilities, working children,
schools for children with disabilities, mother girls, and children affected by conflict), allowing them

to access the school places created by spending on The annualised Household Expansion cost is simply
expansion. The cost of enrolling children currently the sum of all household spending on educational
out of school in any country is, therefore, the sum of supplies, fees and any supplement to teacher salaries
the Public Expansion cost, the Household Expansion made in a single year. Again, these costs will be
cost and the Targeted Interventions cost. directly proportional to the number of children who
have dropped out or who are never expected to enrol.
It is helpful to derive annualised costs for enrolling
out-of-school children. The annualised Public Finally, the Targeted Interventions cost must take
Expansion costs include the construction costs for account of annual spending on each child in each
temporary and permanent classrooms, average marginalised group for every different type of
teacher salaries, the cost of training teachers, and intervention.
expenditures on supplies such as textbooks. These
costs will be directly proportional to the number of These costs can be expressed in terms of the
children who have dropped out or are expected formulae in Box 4.1.
never to enrol and, with the exception of supplies,
are inversely proportional to the pupil-teacher ratio Although school fees have been abolished in many
(because larger classes reduce costs). developing countries, households often continue
Box 4.1 Formulae for estimating the cost of enrolling out-of-school children
Annual cost of enrolling out-of-school children in country X

= Public Expansion cost + Household Expansion cost + Targeted Interventions cost
Public Expansion cost
7 n
= — µ — [ annualised construction cost per temporary classroom]
8 p
1 n
+ — µ — [ annualised construction cost per permanent classroom]
8 p
n
+ — [ average teacher salary]
p
n
+ — [ training cost per teacher]
p
+ n [ per-pupil public expenditure on supplies]
Household Expansion cost

= n [ per-pupil household spending on supplies and fees]
+ n [ per-pupil household supplement to teacher salaries]
Targeted Interventions cost
∑
n m
=
i=1
∑
j=1
dij [ annual cost of intervention per child i in group j]
where dij = 1 if child i belongs to marginalised group j, dij = 0 otherwise,

p is the target pupil-teacher ratio,
n is the number of out-of-school children in country X that have dropped out of school or are
expected never to enrol,
m is the number of types of marginalisation in country X.

to supplement government expenditure with out- cost of his or her enrolment. The proposed analysis
of-pocket payments for teachers, supplies, parent- can be applied to countries to estimate the financing
teacher association dues and other fees (World needs for enrolling their out-of-school populations,
Bank, 2009). It is important, therefore, to distinguish but it can only be used effectively if it is based on a
between expansion costs that are publicly and comprehensive understanding of the barriers faced in
privately financed. In this analysis, government a given country, as well as of the interventions needed
policy on education financing can be taken as to address marginalisation. The analysis makes a
a given, so that households and government few simplifying assumptions: a one-to-one mapping
maintain the existing cost-sharing arrangement for of marginalisation types to interventions and, unlike
education spending. Alternatively, unit costs can SEE (discussed later in this chapter), 100% efficacy
be adjusted to shift the responsibility of expansion of the interventions and perfect, costless targeting.
between public and private sources. For example, Necessary interventions are scaled-up immediately
to emulate a shift of financing from households to (in one school year) in the model, but this may not
governments, the average teacher salary (part of be feasible in practice. Pre-service teacher training,
the Public Expansion cost) could be increased with for example, takes time (a problem that could be
a commensurate decrease in the current per-pupil eased in some settings by recalling retired teachers).
household supplement to teacher salaries (part of the It also assumes constant population growth and no
Household Expansion cost). diminution of the marginalisation of the child during
the time in school, so that the intervention costs are
To spread capital costs evenly over the period of incurred every year to keep the child in school through
integration of out-of-school children into schools, primary school completion.
classroom expenses are annualised over their
expected lifetimes. To prevent either a shortage Finally, and perhaps most importantly, there is no
or oversupply of classrooms as the bulk of former consideration of the quality of education and no
out-of-school children who are now enrolled make link to learning outcomes. Improving the quality
their way through primary school, temporary and of education is a critical demand-enhancing
permanent classrooms are financed in proportion to intervention in cases where enrolment is deterred
the current numbers of out-of-school children and by the poor quality of existing schools. Improving
their annual expected inflow, respectively.25 Constant quality also reduces repetition and dropout, to
population growth and a constant age structure of which marginalised children and adolescents are
the out-of-school population are assumed, so that most vulnerable. In cases where the quality of
the annual flow is one-eighth of the total current education provision is poor as a result of relatively
out-of-school population (this assumption can be low education spending, it may be appropriate to
refined based on country-specific data). Under these set the parameters of the analysis to target future
assumptions, the ratio of temporary to permanent levels (for example, a lower pupil-teacher ratio, higher
classrooms is 7:1 (giving rise to the 7/8 and 1/8 spending on materials per pupil, etc.), rather than
multipliers observed in the first two lines of the Public current levels.
Expansion cost formula in Box 4.1). A similar issue
applies to the supply of teaching staff. In some 4.4 ILLUSTRATION: AN EQUITY-BASED
countries, this may be resolved using contract APPROACH TO ASSESSING THE COST
teachers to increase the supply temporarily. OF ENROLLING OUT-OF-SCHOOL
CHILDREN IN THE DEMOCRATIC
Another key feature of this analysis is that it accounts REPUBLIC OF THE CONGO
for the possibility that a single child may have multiple
profiles linked to marginalisation, compounding the How much would it cost to enrol a country’s out-
of-school children in primary education? In this
25 Double-shifting could also be a solution to the overflow problem,
particularly where population density is high. illustration, we apply the model outlined above, with

Table 4.1 shows that two-thirds of out-of-school
A single child may have multiple children in the Democratic Republic of the Congo
profiles linked to marginalisation, are expected to enrol in the future (UNICEF and UIS,
compounding the cost of enrolment 2013d). The model focuses on the 519,000 out-of-
school children and adolescents who are expected
never to enrol and the 830,000 early school leavers,
an emphasis on equity, to estimate the financial who are certain to require additional financing to
resources required to enrol out-of-school children ensure they complete their primary education. The
and adolescents in the Democratic Republic of the analysis that follows projects the annual cost of
Congo, holding as constant the country’s quality of enrolling these more than 1.3 million out-of-school
education and its financing shares from government children and adolescents (n in the cost functions in
and households. In many countries with a high Box 4.1) through six years of primary school, using
number of out-of-school children, governments the analysis outlined in the previous section.26
provide an insufficient share of total education
Table 4.1 Out-of-school children and adolescents of
financing, leaving households to fund their children’s primary and lower secondary age in the Democratic
right to a good quality basic education. In the Republic of the Congo, 2012
case of the Democratic Republic of the Congo,
Total children aged 6-13 years 17,036,000
the existing situation is characterised by a low
Total out-of-school children 4,022,000
share of public financing (with households covering
Dropped out 830,000
80% of recurring per-pupil costs) and low per-
Will never enrol 519,000
pupil spending relative to other African countries.
Although in calculating the annual cost of enrolment Will enrol late 2,673,000
we do not specify sources of financing or model Source: Data from EADE-RDC27 2012 (UNICEF and UIS, 2013d)
improvements in the quality of education, this
analysis could indirectly model increases in both We first estimate the Public Expansion cost (see
the public share of education financing (by shifting Table 4.2) using current public expenditure data (for
costs between Public Expansion and Household teacher salaries and supplies) from the UIS (2014)28
Expansion) and the quality of primary education (by and capital expenditure data (for rural classrooms,
increasing unit costs). given that 80% of the Democratic Republic of the
Congo’s out-of-school children live in rural areas)
With over 4 million of its 17 million children aged 6 to estimated by the World Bank (2005a).
13 years out of school, according to a 2012 national
household survey (UNICEF and UIS, 2013d), the Teacher training costs are approximated using
Democratic Republic of the Congo has one of the the estimated unit cost of pedagogical secondary
highest rates of exclusion from primary and lower school completion, which is based on statistics
secondary education in the world. Indeed, it may from SECOPE (the Democratic Republic of the
account for 3% of the global total of out-of-school Congo’s Department for Monitoring the Payment
children and adolescents of primary and lower of Teacher Salaries). We assume that classrooms
secondary age (UNESCO, 2013). The 2012 survey and teachers can be re-purposed for different grade
revealed that children in the Democratic Republic of levels as children who were once out of school (a
the Congo are out of school because of a variety of
26 The example assumes that out-of-school children of lower secondary
the barriers discussed in Chapter 3—particularly high school age did not complete primary school and must, therefore, be
enrolled in primary rather than lower secondary education. This is based
rates of poverty that result in child labour and the on the high rate of overage attendance in primary school (60% of primary
school-age students are two or more years overage), as well as the short
distance between the home and the nearest school. length of lower secondary school (two years) relative to primary school (six
years) in the Democratic Republic of the Congo (UNICEF and UIS, 2013d).
Armed conflict is also a major cause of exclusion, 27 L’enquête nationale sur la situation des enfants et adolescents en dehors
de l’école en République démocratique du Congo.
as is linguistic fragmentation in a country where 242 28 The per-pupil costs presented are not unit costs. They represent total
government expenditure divided by the number of children enrolled in
languages are spoken. school (public and private).

Table 4.2 Public Expansion cost in the Democratic Republic of the Congo
Annual per-pupil Sub-total

cost (in 2010 (in millions
constant US$) US$) Notes
11.93 $2,000 per community classroom. Assumed lifespan is six
Temporary classrooms 14.1
(annualised) years, 5% cost of capital. Source: World Bank, 2005a
23.97 $9,870 per rural classroom. Assumed lifespan is 20 years,
Permanent classrooms* 4.0
(annualised) 5% cost of capital. Source: World Bank, 2005a
Teacher salaries 8.14 11.0 Based on the UIS, 2014.
Pre-service training unit costs based on 2012-2013
Teacher training 11.23 15.1
SECOPE.
Current, non-salary government expenditure, from the UIS,
Public expenditure on supplies 0.25 0.3
2014.
Public Expansion total 44.5
Note: * Assumes a 7:1 ratio of temporary-to-permanent classrooms.
Table 4.3 Household Expansion cost in the Democratic Republic of the Congo
Annual per-pupil Sub-total

cost (in 2010 (in millions
constant US$) US$) Source
Supplement to teacher salaries 10.45 14.1 Based on MEPSP (2012), deflated to 2010 constant US$
Based on MEPSP, deflated to 2010 constant US$. Includes
Household spending on supplies 17.58 23.7
school uniform
School operating fees based on MEPSP (2012) deflated to
Other fees 10.07 13.6
2010 constant US$. Includes examination fees
Household Expansion total 51.4
disproportionate number of whom would start in the spending on primary education and one-fifth of
first grade) progress through primary school. All costs the gross domestic product (GDP) per capita. This
are normalised by the Democratic Republic of the constitutes a significant burden for the poorest
Congo’s current 37:1 pupil-teacher ratio (p in Box households, which earn less than $50 per month and
4.1) (UNESCO, 2013). account for 65% of the Democratic Republic of the
Congo’s out-of-school children (UNICEF and UIS,
Household Expansion cost (see Table 4.3) is 2013d). Transferring the Household Expansion cost
estimated based on Verhaghe’s (2013) analysis of to public sources by increasing the share of education
statistics from the Ministry of Primary, Secondary and expenditure assumed by the government, so that
Professional Education (MEPSP). households spend less per pupil on supplies, fees
and teacher compensation, would go a long way
The total annual expansion cost (Public and to enrolling out-of-school children who are currently
Household) for the Democratic Republic of the Congo excluded by financial barriers, while also promoting the
is estimated at $95.9 million in 2010 US dollars. right to basic education.
Assuming that the country’s existing financing for
education is unchanged, $51.4 million of that sum Lowering household spending on education does
would be the responsibility of households. While not reduce the opportunity cost of education, for
gradual fee abolition has been underway since 2010, example through foregone earnings of a child or
households still provide 54% of total spending on adolescent (see Section 3.4 on child labour and
primary and secondary education in the Democratic school participation). For an estimated 40% of out-
Republic of the Congo (Verhaghe, 2013). Annual of-school children in the Democratic Republic of
household spending per child in primary school is the Congo, the reduction of Household Expansion
$38—nearly four times the public per-pupil recurrent costs via increased government spending would be

in the Democratic Republic of the Congo, especially
After the majority of children who for the poorest families that rely on children’s
were once out of school pass through contributions to household income. About 4%
primary education, the annual per-pupil of respondents identified the need for children’s
cost falls because capital expansion domestic work and paid labour as reasons for not
spending is no longer required enrolling, while 6% cited language difficulties in terms
of the language of instruction at school. Gender-
related barriers to enrolment were not included in
insufficient to improve enrolment (Verhaghe, 2007). the survey, but a large gender gap certainly exists
These children require a combination of targeted in a few provinces. A further 8% of out-of-school
interventions that goes far beyond expansion costs. children are excluded because they have a disability
According to the 2012 EADE-RDC survey (UNICEF or because of poor health and undernutrition,
and UIS, 2013d), the main reason for exclusion according to the survey. These are all likely to be
from education is poverty—69% of households underestimates of the true level of marginalisation,
with children out of school cited lack of money as a however, as marginalised children are less likely to be
reason for their non-enrolment, broadly consistent reached by surveys.
with Verhaghe’s (2007) assertion that 60% of out-
of-school children in the Democratic Republic of Data on the cost of interventions to overcome these
the Congo could be enrolled by removing financial barriers are scarce (UNICEF, 2014), especially for
barriers at the system level. As discussed previously, the Democratic Republic of the Congo. Table 4.4
financial barriers could be addressed by reducing shows some illustrative per-pupil costs of targeted
household responsibility for expansion spending. interventions, which are not all specific to the
Such spending could also assist the 35% of Democratic Republic of the Congo and span a wide
households that cited distance from school as the range of approaches. Specific intervention costs
reason for non-enrolment, through careful distribution for many groups (children with disabilities, child
of newly-constructed rural classrooms. labourers, orphans and children living or working on
the streets) are unavailable, preventing any precise
Other barriers identified in the 2012 EADE-RDC estimation of the Targeted Interventions cost for
survey require targeted interventions. As well the Democratic Republic of the Congo. However,
as transferring the responsibility of financing the assuming that 40% of out-of-school children require
Expansion Cost from households to the government, targeted interventions (Verhaghe, 2007), an average
cash transfers that address the opportunity costs of intervention cost is $18 (based on the subset of
education could be essential to increase enrolment interventions in Table 4.4), and 1.5 interventions
Table 4.4 Examples of Targeted Intervention costs
% of out-
of-school
children
according to Possible Cost estimate (in 2010
Profile EADE-RDC intervention constant US$) Source
Emergency 46.74 per out-of- IRC, 2011 (for the Democratic Republic of
Children in conflict areas 4.4
education school child per year the Congo)
17.46 per out-of-
Children living with School feeding Gelli, 2012 (for 32 developing countries)
school child per year
disabilities or poor health/ 6.8
nutrition* 4.04 per out-of-
Deworming Miguel and Kremer, 2004 (for Kenya)
school child per year
Mother-tongue 8% of non-capital World Bank, 2005b (for developing
Ethno-linguistic minorities 5.6
instruction per-pupil expenses countries)
Note: * Calculated by combining the percentages of children for whom disability, undernutrition and poor health were cited as reasons for being out of school.
Derived from Table 3, UNICEF and UIS, 2014c.

per year are needed for each child out of school, The exercise relies on simplifying assumptions
a preliminary estimate for the annual Targeted and is incomplete because of the scarcity of data
Interventions cost in the Democratic Republic of the on out-of-school children and interventions for
Congo is $14.7 million, or 13% of the total $111 their enrolment. As noted in Chapter 2, household
million required annually to enrol the country’s out-of- sample surveys may underestimate the number
school children (see Table 4.5). This is comparable of children who are likely to be marginalised in
to the EPDC (2009) estimate that 13% of the cost of education—though it must be noted that the EADE-
expanding access in low-income countries would be RDC survey in the Democratic Republic of the
devoted to reaching marginalised children, but more Congo (UNICEF and UIS, 2013d) was designed
data on intervention costs and target group sizes are explicitly to collect data on out-of-school children
required to refine this estimate. and included a special data collection on children
who were not living in households: those on the
Table 4.5 Total estimated annual cost of enrolling 1.3 street and in institutions.
million out-of-school children and adolescents in the
Democratic Republic of the Congo
The analysis above takes all education parameters
Source of as given. Adjusting those parameters to reflect
Cost (in financing under
million current financing improvements in quality would, of course, further
Expense type US$) % of total arrangement
44.5 40 Public
increase the expansion share of the total cost. For
Expansion example, raising total teacher compensation in the
51.4 46 Households
Targeted
Democratic Republic of the Congo from 3.3 times
Interventions 14.7 13 n/a the GDP per capita to the average for sub-Saharan
TOTAL 110.6 100 Africa (3.9 times the GDP per capita) would increase
the Expansion Cost to $101 million and the total
Note: Percentages do not sum to 100 due to rounding.
cost to $116 million. On the other hand, efficiency-
The estimated annual cost of enrolling out-of-school enhancing reforms in the Democratic Republic of
children is $111 million, or $82 per child per year the Congo, such as reducing the overhead costs
(see Table 4.5). This is higher than current spending associated with administrative bureaus and
($47 per pupil, based on Verhaghe, 2013 and UIS, regulating the growth of teacher numbers, could
2014) because of the need for capital spending lower the unit costs of providing education
(on classroom construction and teacher training) (Verhaghe, 2013) and reduce the expansion cost
and targeted interventions to reach marginalised of enrolling out-of-school children.
children. After the majority of children who were
once out of school pass through primary education,
the annual per-pupil cost would fall because capital While expanding existing education
expansion spending would no longer be required. infrastructure is necessary to
Raising $111 million would be a significant challenge increase enrolment, it is not sufficient
for the government and supporting donors: it in countries like the Democratic
is equivalent to one-quarter of the Democratic Republic of the Congo, where
Republic of the Congo’s total public spending on demand-side constraints stop children
education in 2011 and exceeds the foreign aid from accessing basic education
that the country received for basic education in
2011 by a factor of 1.4 (the Democratic Republic
of the Congo receives only 1.5% of global aid to Despite these limitations, the case study illustrates
basic education according to OECD Development the importance of expansive and targeted spending
Assistance Committee aid statistics, even though for the enrolment of out-of-school children. Roughly
household survey data indicate that it has 3% of the estimated, accounting for Targeted Interventions
world’s out-of-school children). costs increases the financing required to enrol out-

4.5 INNOVATIONS IN COSTING AND
The Simulations for Equity in SIMULATING TARGETED INTERVENTIONS
Education (SEE) model allows FOR OUT-OF-SCHOOL CHILDREN
policymakers to focus their
interventions on different groups The implication that policies based on equity may
of marginalised children, increase the cost of enrolling the remaining out-
who have different education of-school children makes it more important than
outcomes and needs ever to identify the policies that deliver the best
results. The SEE initiative was launched by UNICEF
and the World Bank in 2011 to create tools to
of-school children in primary education significantly. help policymakers select pro-equity, efficient and
While expanding existing education infrastructure is cost-effective interventions to improve education
necessary to increase enrolment, it is not sufficient outcomes, in particular for marginalised children.
in countries like the Democratic Republic of the
Congo, where demand-side constraints stop The SEE project focused on two main outputs.
children from accessing basic education. Surveys First, it aimed to shore up the existing evidence
similar to the EADE-RDC survey conducted in the on the effectiveness of targeted interventions in
Democratic Republic of the Congo would allow developing countries. Over 400 research papers
investigation into the nature of exclusion in other were reviewed and compiled as information sources
countries that are still a long way from the goal for effective education interventions. Together, this
of universal primary attendance and completion, information provides evidence-based parameters for
and into the cost of overcoming country-specific how different interventions can improve education
barriers through a combination of expansion and outcomes—crucial for countries that have only scant
targeted spending. information about their own context. Second, the
SEE simulation model provides a virtual arena where
Because they treat all students equally in their policymakers can compare the costs and outcomes
costing methodology, most of the earlier and of intervention options, focusing on the impact on
traditional models reviewed in this chapter are likely enrolment of specific marginalised groups.
to underestimate the cost of achieving universal
primary education, because they ignore the need The SEE model allows policymakers to optimise
for investments that go far beyond the expansion of the recommended programmes—including their
existing education infrastructure and programmes. scale, timing and distribution across risk groups.
The exercise starts by entering data into the
Although stylised, the model presented in this model: education outcomes for different groups of
chapter, with its focus on marginalised children and marginalised children (as identified in reports from
equity, demonstrates the magnitude of financing the Global Initiative on Out-of-School Children or
needs for the enrolment of out-of-school children. other studies), the list of proposed interventions, and
It also confirms that those needs are defined by the the parameters for their effectiveness (drawn from
situation of marginalised children who will not enrol the research on programme effectiveness). Next,
even when the number of school places increases. the policymaker sets hypothetical scenarios and the
Many of the world’s out-of-school children are the model computes estimated education improvements
hardest to reach and face the greatest hurdles in based on those inputs. In this way, it is possible to
accessing education. The equity-based example of select cost-effective interventions that target the
the Democratic Republic of the Congo, while only groups in greatest need.
partial, increases the financing needs far above the
costs of basic expansion, especially when children This model provides policymakers with a tool to
face multiple barriers to education. compare and optimise different strategies and

Table 4.6 Comparison of a system-wide and targeted intervention to improve the enrolment of girls,
Balochistan province, Pakistan
Scenario 2: Community schools

Scenario 1: Regular schools with with greater proportion of female
current teacher distribution teachers
Total cost including teachers, 2014-2023 US$524 million US$356 million
Number of new classrooms 12,000 12,000
Female teachers to be recruited 4,000 7,000
Male teachers to be recruited 8,000 5,000
Boys Girls Boys Girls

Additional children to enter school, 2014-2023
133,000 146,000 133,000 236,000
interventions. Other planning tools tend to focus and the new schools would allow 133,000 additional
on just one path, without inviting policymakers to boys to go to school and 146,000 additional girls. In
consider alternate options explicitly. It also allows contrast, the community school model in Scenario
policymakers to focus their interventions on different 2 costs only $356 million, already a considerable
groups of marginalised children, who have different financial gain. Furthermore, with its focus on hiring
education outcomes and needs. more female teachers, the community model would
result in an estimated 236,000 more girls entering
4.6 ILLUSTRATION: COMPARING THE school, over 60% more than the gains from the
IMPACT OF TARGETED INTERVENTIONS traditional model.29 The results of the SEE model
TO ENROL OUT-OF-SCHOOL GIRLS were used by policymakers in the Balochistan
IN THE BALOCHISTAN PROVINCE, Ministry of Education to plan the building of 2,000
PAKISTAN schools in remote regions of the province.
The Balochistan province in southwest Pakistan has 4.7 REACHING THE MARGINALISED MAY
among the highest rates of out-of-school children COST MORE, BUT BETTER DATA AND
and gender disparity in the country: 40% of primary INNOVATIVE TOOLS CAN HELP US
school-age boys are not in school and 57% of girls SPEND SMARTER
(UNICEF and UIS, 2013b). The lack of schools in
rural areas and parents’ reluctance to send girls to The innovative models described in this chapter,
school are among the key barriers children face to applied to the Democratic Republic of the Congo and
their schooling, as identified in the Pakistan study Pakistan, highlight the need for concerted efforts on
conducted within the Global Initiative on Out-of- three fronts in global education.
School Children (UNICEF and UIS, 2013b) and
the Balochistan Education Sector Plan (BEPS) First, the availability and quality of data on
(PPIU, 2014). Both reports propose investment in interventions for out-of-school children must
community schools and a strong focus on female continue to improve. In recent years, more
teachers for girls. The SEE model shows the benefits comprehensive data on marginalised groups have
of this targeted approach as compared to the current allowed researchers to depart from the average
approach of expanding the construction of regular unit-cost modelling approaches of the early 2000s
schools in villages (see Table 4.6).
29 Scenario 2 assumes that if a teacher in a nearby school is female, the
likelihood that a girl will not enter school is reduced by 50%. The estimated
Scenario 1 assumes that Balochistan builds regular effectiveness of the proposed intervention is based on analysis of MICS
data from Balochistan, which shows that in villages with a school, as
schools, staffed with the existing female-to-male many as one-quarter of girls do not enter, compared to only 13% of girls in
villages with community schools with female teachers. A positive effect of
teacher ratio of 1:2. The total cost would be US$524 female teachers on girls’ school attendance was also observed elsewhere
in South Asia, including rural areas of the Indian state of Rajasthan
million over the ten-year period from 2014 to 2023 (Banerjee et al., 2001) and Nepal (Bista, 2006).

and have enabled the development of improved the significant progress that has been made over
policy planning tools, such as SEE. More and better the past two decades, more resources are urgently
data are needed to improve the estimates of the needed for the most disadvantaged children.
cost and effectiveness of interventions targeted at Expanding the education system in its current
out-of-school children in developing countries. The form will not be enough to reduce out-of-school
most important limitation of the SEE model is that, prevalence. Furthermore, recent results from the SEE
where no in-country data exist, the parameters for model in Ghana and Balochistan show that using a
the effectiveness of interventions are extrapolated pro-equity approach can be more cost-effective than
from pilot studies or experience in other countries. business-as-usual approaches.
It is likely that actual effectiveness on the ground
will differ from these parameters but with careful Third, the Democratic Republic of the Congo case
consideration the differences can be minimised. study in particular underscores the importance of
Statistics and research on policies and interventions lowering unit costs of education for the poor to
to reach marginalised children contribute to make provision financially sustainable. A number
informed decisionmaking on resource allocation in of other innovative approaches to reach marginalised
the education sector and strengthen advocacy to children are being piloted worldwide. Continuing to
mobilise the resources needed to achieve universal explore, gather cost-effectiveness data and build the
basic education. evidence base on programming that focuses on all
out-of-school children and adolescents are critical to
Second, there is a clear need for equity-based reduce the financing needs for their enrolment and
approaches to financing education. Despite meet—at last—the goal of universal basic education.

Chapter 5
Conclusions and recommendations
5.1 THE BROKEN PROMISE OF While efforts to improve national administrative data
EDUCATION collection must continue to be a priority, so too
must be the continued use of household survey and
The work of the Global Initiative on Out-of-School census data to better identify the characteristics
Children, summarised in this report, has confirmed of out-of-school children. Here we can learn from
the urgent need to prioritise the world’s unfinished the experience of countries participating in the
business—out-of-school children and adolescents— out-of-school initiative: these data sources are
in the post-2015 development agenda. complementary, and harnessing the strengths of
a wide range of data sources provides a better
Despite every effort and the impressive progress understanding of school attendance and learning.
made on educational access in some countries As a result of the national studies, many countries
and regions, the world as a whole has broken a identified avenues to improve how data are collected,
fundamental promise to children: that each and every harmonised and analysed. This underscores the
one of them would be able to complete primary importance of data use as a key driver to improve
education by 2015. That promise seemed realistic quality. Using existing data can also encourage
and achievable when it formed part of the Education improved supply: countries identified an urgent need
for All goals and Millennium Development Goals. to close the data gap on the most vulnerable groups
Yet the world has failed to deliver, leaving 58 million of children, who may not be captured by existing
children and 63 million adolescents out of school data sources. Not enough is known about the extent
and unable to reach their full potential (see http:// of school exclusion among children caught up in
on.unesco.org/oosc-map). Furthermore, progress conflict, on the streets or in slums, and especially
has stalled and there are serious concerns that those with disabilities.
unless something changes—and fast—hard-won
achievements on primary enrolment could begin to Meanwhile, education systems and the societies
erode. that surround them often reinforce the barriers that
marginalise specific groups of children. This report
At the same time, post-2015 discussions are has highlighted the situation of children in conflict-
highlighting every country’s need for universal affected countries, for example, who account for just
secondary education, which is vital for national one-fifth of the world’s children of primary school age
economic prosperity and social well-being. Yet many but one-half of the world’s out-of-school children. It
governments are finding that the foundations they has explored the gender norms and discrimination
have built for universal primary education are not yet that leave more girls out of school than boys—
strong enough to enrol all children or keep them in particularly the poorest girls in rural areas, and the
the classroom, let alone lift them to the next stage of child labour that undermines learning and often
their schooling. leads to drop-out. The report has shown how the

human and financial—to fix the broken promise of
Most countries need a policy education.
framework combining three priorities:
broad investment to strengthen and This report has made a distinction between countries
expand education systems, a sharp that are nearing the goal of universal primary
focus on inclusion and the quality of education and those where large proportions—and
the education on offer, and targeted numbers—of children are still out of school. It argues
interventions for the children who are that those countries in the ‘final mile’ must strive
the very hardest to reach to break down the persistent barriers to education
faced by the most marginalised children. Countries
with the furthest to travel, on the other hand, must
lack of schooling in the languages children speak at increase their investment in approaches to expand
home can limit their participation in the classroom, and improve education systems as a whole. In
and how the right of a child with a disability to an these countries, targeted interventions to reach
education is so often denied. These barriers often marginalised children are urgently required but
work in combination to keep millions of marginalised not enough to shore up education systems that
children out of school. For those who do enrol, the are under-resourced, under-staffed and that offer
continuing barriers to their education can become education of poor quality.
too much to endure and they eventually vanish from
the classroom. The findings from the Global Initiative on Out-of-
School Children show that most countries need a
It is clear that business-as-usual approaches based policy framework consisting of three priorities: broad
on more teachers, more classrooms and more investment to strengthen and expand education
textbooks are not enough to keep the promise of a systems, a sharp focus on inclusion and the quality
quality primary education for the most disadvantaged of the education on offer, and targeted interventions
children. These children need targeted interventions for the children who are the very hardest to reach.
to sweep away the barriers to their schooling,
once and for all. At the same time, this report has This report sets out a new equity-based costing
highlighted the need for a long overdue reality check model to assess the potential scale of the investment
on the scale of the resourcing required to guarantee needed for universal primary education, recognising
a basic education for every child, including those that resources are generally limited. The costs may
children who are the very hardest to reach. be significant: the report notes that policies based
on equity will likely increase the cost of enrolling the
5.2 FIXING THE BROKEN PROMISE remaining out-of-school children, given that any
single child may face multiple barriers to education.
A mix of far-reaching policies to address
educational supply and demand While it is difficult to estimate the global amount
needed to deliver universal primary education, the
A number of policy recommendations have emerged report provides equity-focused costing analysis
from countries participating in the Global Initiative for the Democratic Republic of the Congo. It gives
on Out-of-School Children, all of which should be an idea of the scale of resources needed to get
underpinned by a revitalised political commitment to every out-of-school child in this one country into
universal primary education. While different countries the classroom: $111 million each year—equivalent
need policies that are tailored to their specific to one-quarter of the Democratic Republic of the
circumstances, every country needs to renew its Congo’s total public spending on education in
commitment, backed by the necessary resources— 2011. The report concludes that the provision of

sustainable universal education requires far greater all beneficiary children, particularly girls, in parts of
expenditure—coupled with far more effective use of Latin America and the Caribbean, South Asia and
those resources—than we see today. This makes sub-Saharan Africa. Ideally, such transfers should be
it more important than ever to sharpen the focus linked to improved learning rather than just school
on policies that deliver the best results for the most attendance in a poorly-resourced school. Moreover,
disadvantaged children. It is encouraging to note cash transfers will not work in contexts where
that the findings of the equity-based costing models schools are far and few between and not of good
in Ghana and Pakistan’s Balochistan Province show quality. The report makes the case for the scale-up of
that using pro-equity interventions can be more cost- conditional transfers within the countries where they
effective than business-as-usual approaches. are already making a difference and their expansion
to more countries and regions.
Education systems reflect their wider environments,
and what happens outside the school gates will Demand for education is also shaped by the quality
often determine whether or not a child is in the of the education on offer. The incentives for families
classroom. Measures to ensure inclusion and to send their children to school and keep them there
address the specific barriers to children’s schooling are far higher when they are confident that a school
through the reform of education systems can only has well-trained and motivated teachers, relevant
succeed when matched by measures to address learning materials and high standards, and that their
wider disadvantage and to smooth out inequities children will emerge with the skills they need for a
linked to income poverty, gender, ethnicity, language, productive adulthood.
geographic location and disability.
Lastly, even the best policies on out-of-school
Certain key measures have proven to be effective children will have little impact if delivery and
and need to be part of the policy agenda in every governance systems are weak. In some countries,
country facing the challenge of out-of-school sound policies are in place, but children and schools
children. on the ground see little of the intended effects due to
inefficiency, corruption or low capacity at the local or
On the supply side, for example, fee abolition is a district level.
crucial first step. It needs to be balanced, however,
by grants to schools and formula funding that follows Breaking the specific barriers that confront
students to ensure that schools can cope with the marginalised children
influx of new students that inevitably follows fee
abolition. It is important to note that fee abolition This report has set out measures to break five key
alone may not make education affordable for the barriers to universal basic education.
most marginalised and impoverished families. The
hidden costs of sending their children to school— Conflict. A three-pronged approach is needed to
from transportation and uniforms, to textbooks ensure that a good quality education is positioned
and informal payments to teachers, as well as the as part of wider social reforms to prevent conflict,
lost earnings from child labour—may outweigh the enable schooling to continue during conflict, and
benefits, particularly if the education on offer is of ensure that post-conflict education reforms support
poor quality. It is time to move beyond ‘fee-free’ the economic and social recovery that can prevent a
primary education to ensure that primary education is re-ignition of violence.
truly free of charge for all children and their families.
Gender discrimination. The priority is to ensure
On the demand side, cash transfers to reduce that even the most vulnerable and disadvantaged girl
poverty, particularly those that are conditional on has access to a school close to home—a school that
school attendance, have boosted enrolment for meets her most basic needs for safety, privacy and

cleanliness, that delivers the best possible education from the Global Initiative on Out-of-School Children
and that values her presence and her potential. Boys, suggest that it is time for a greater recognition of
too, need support to ensure they stay in school and the importance of non-formal or flexible learning
out of child labour. strategies for children who have been denied a
mainstream education, as well as its role as a
Child labour. Child labour must be reduced in order crucible for educational innovation. NFE should no
to increase school attendance. Given the strong longer be seen as ‘second best’ by policymakers,
links between poverty and child work, removing the practitioners and development partners.
direct and indirect costs associated with education
is the most effective way to reduce child labour. A final word on data and partnerships
More flexible and responsive education systems and
improved learning environments are needed to attract Children who are excluded from education often face
working children into the classroom and keep them multiple and overlapping disadvantages. If we are to
there. reach them, we need a much clearer picture of who
they are, where they are and exactly why they are out
Language challenges. Education programmes of school (see http://on.unesco.org/oosci-global).
based on children’s home languages have higher Without good data, governments are struggling to
levels of participation, success and enjoyment establish what and where the problems are, and
(as well as parental involvement) and lower levels therefore, an effective response is challenging.
of repetition and drop-out, especially among
girls. Schools that work in languages children This report presents the best available data on the
can understand allow pupils to participate and world’s out-of-school children. It argues that the
demonstrate what they know, and encourage the unfinished business of universal primary education
involvement of their families. and the stagnation in global trends of the number of
out-of-school children, as well as the large number of
Social, institutional and environmental barriers out-of-school adolescents, make it more important
linked to disability. The education of children than ever to invest in the improved collection and
with disabilities hinges on the removal of the many analysis of data on their needs. Given that scarce
barriers that come between them and the chance resources are a political reality, better data are
of an education, from the lack of physical access to essential to target those resources towards the
classrooms to the stigma that keeps them hidden most severe problems and towards context-relevant
away at home. It also requires the provision of interventions that have been shown to be effective.
appropriate support and an understanding of their
functioning and needs, all backed by robust and Policymakers who must decide where and how
comparable data. to spend public financing need solid information
on the cost of getting all children into school and
Thinking outside the box: The importance of the expected impact of the interventions they
non-formal education select to achieve this goal. There is also a pressing
need for better data on the specific barriers that
For many children and adolescents who are out of confront marginalised children. These include
school, the foundational skills provided through non- more rapid and flexible assessment of the needs
formal education (NFE) that takes place outside the of children caught up in fast-moving conflicts and
mainstream education system, often provided by greater disaggregation of data to see how gender
non-governmental organizations, are indispensable discrimination shapes school attendance and
in realising their right to a meaningful education. It performance. We need closer scrutiny of the ways
can provide a pathway back to regular schooling or in which child labour and non-attendance reinforce
even, in some cases, a viable alternative. Findings each other and how the languages children use at

home can become the languages they use at school. vulnerable children. Local governments, schools and
Finally, we need a concerted and global effort to communities have a crucial role to play in identifying
ensure comparable and standardised definitions and providing coordinated support to out-of-school
of disability, based on social rather than medical children and those at high risk of dropping out.
models, to end the ‘invisibility’ of children with
disabilities in the data and in the classroom. International organizations and donors must support
this work by improving international definitions of key
In short, a comprehensive evidence base is the issues (such as drop-out) as well as the standards
bedrock for effective policies to enrol more children used for data collection on out-of-school children. In
in school and to support them in the successful addition, they must also support the strengthening of
completion of a full cycle of high-quality basic national and sub-national education data systems.
education. These actors can also play a crucial role in leveraging
the data revolution that is generating technological
In every country, the appropriate mix of policies to advances in data collection, processing and analysis.
ensure that every child is in school should be guided They must ensure that governments are empowered
by the best available evidence on what is needed to make the best possible use of these advances.
and on what works. This report has stressed that Overall, the international community should draw
the barriers to universal education are complicated on the enhanced evidence base to continue to
and intertwined, and that our ability to deliver a assist governments in developing innovative and
comprehensive response—globally, regionally and inclusive policies and in implementing cost-effective
nationally—is limited by both a lack of data and a programmes.
weakness in cross-sectoral coordination among
stakeholders. Above all, this report has shown that governments,
donors and international organizations must make it
By providing evidence and recommendations on a priority to bring both financial and human resources
out-of-school data and policy, the Global Initiative on to the places and the children with the greatest
Out-of-School Children has aimed to build political needs. To make a true breakthrough, it is essential to
commitment and action to generate a real and mobilise the concerted, innovative support needed
sustained decrease in the numbers of out-of-school to reach the most vulnerable children who often go
children and adolescents worldwide. It has done so uncounted or unsupported: children with disabilities,
by linking data to policy and, very importantly, to the in conflict zones or those facing barriers to education
cross-sectoral partnerships needed to address this due to language, gender or poverty. This will also
complex issue. require strong backing from non-governmental
organizations, civil society, the private sector and
Ultimately, the responsibility for the information that communities. To put it simply: the world’s out-of-
is needed lies not only with national ministries of school children have been overlooked for too long—
education but also with any government ministry they are now everybody’s business.
that collects data or implements programmes for

References
Adukia, Anjali (2014). Sanitation and Education. Cambridge, MA: Harvard University. Available at: http://scholar.harvard.edu/
files/adukia/files/adukia_sanitation_and_education.pdf
Akresh, Richard and Damien de Walque (2008). “Armed Conflict and Schooling: Evidence from the 1994 Rwandan
Genocide”. IZA Discussion Paper No. 3516. Bonn: Forschungsinstitut zur Zukunft der Arbeit (Institute for the Study of Labor).
Alderman, Harold, John Hoddinott and Bill Kinsey (2006). “Long Term Consequences of Early Childhood Malnutrition”. Oxford
Economic Papers, 58(3): 450–474.
Back on Track and UNICEF (2011). “Education in Emergencies and Post-crisis Transition”. 2012 Consolidated Progress
Report to the Government of the Netherlands and The European Commission. New York: Back on Track (The Transition Fund)
and UNICEF. Available at: http://www.educationandtransition.org/wp-content/uploads/2011/07/2010_EEPCT_Consolidated_
Report_30June2011.pdf
Baird, Sarah, Craig McIntosh and Berk Özler (2011). “Cash or condition? Evidence from a cash transfer experiment”.
Quarterly Journal of Economics 126(4): 1709–1753.
Banerjee, Abhijit, Suraj Jacob and Michael Kremer with Jenny Lanjouw and Peter Lanjou (2001). “Promoting School
Participation in Rural Rajasthan: Results from Some Prospective Trials”. Mimeo, MIT. Cited in: Barbara Herz and Gene B.
Sperling (2004). What Works in Girls’ Education: Evidence and Policies from the Developing World. Washington, D.C.: Council
on Foreign Relations.
Benson, Carol (2005). Girls, Educational Equity and Mother Tongue-Based Teaching. Bangkok: UNESCO. Available at:
http://www.ungei.org/infobycountry/files/unesco_Girls_Edu_mother_tongue.pdf
Benson, Carol (2014). “School access for children from non-dominant ethnic and linguistic communities”. Background paper
for Fixing the Broken Promise of Education For All: Findings from the Global Initiative on Out-of-School Children. Montreal:
UNESCO Institute for Statistics.
Bista, Min Bahadur (2006). Status of Female Teachers in Nepal. Kathmandu: UNESCO.
Blattman, Christopher and Jeannie Annan, J. (2009). “The consequences of child soldiering”. Review of Economics and
Statistics. November 2010, 92(4): 882–898.
Bozzoli, Carlos, Tilman Brück and Nina Wald (2011). “Self-employment and conflict in Colombia”. DIW Berlin Discussion
Paper No. 1098. Available at: http://dx.doi.org/10.2139/ssrn.1767914
BRAC (2012). Children with Special Needs (CSN) (web-page). Available at: http://education.brac.net/children-with-special-
needs-csn
Braithwaite, Jeanine and Daniel Mont (2009). “Disability and poverty: A survey of World Bank Poverty Assessments and
implications”. ALTER – European Journal of Disability Research 3:219–232. Available at: http://siteresources.worldbank.org/
DISABILITY/Resources/280658-1239044853210/5995073-1246917324202/braithwaite_and_mont_final_pub_version_
ALTER57.pdf

Brossard, Mathieu and Luc-Charles Gacougnolle (2001). Financing Primary Education for All: Yesterday, Today and Tomorrow.
Paris: UNESCO.
Brown, Gordon. (2012). Out of Wedlock, Into School: Combating Child Marriage Through Education. London: The Office of
Gordon and Sarah Brown.
Bruns, Barbara, Alain Mingat and Ramahatra Rakotomalala (2003). Achieving Universal Primary Education by 2015: A Chance
for Every Child. Washington, D.C.: World Bank.
Bundervoet, Tom and Philip Verwimp (2005). “Civil war and economic sanctions: An analysis of anthropometric outcomes
in Burundi”. HiCN Working Paper No. 11. Brighton: Households in Conflict Network. Available at: http://www.hicn.org/
wordpress/?page_id=28
Bundervoet, Tom, Philip Verwimp and Richard Akresh (2007). “Civil war, crop failure and child stunting in Rwanda”. Policy
Research Working Paper Series 4208. Washington, D.C.: The World Bank.
Bundy, Donald (ed.) (2011). Rethinking School Health: A Key Component of Education for All. Directions in Development.
Washington, D.C.: World Bank.
Burde, Dana and Leigh L. Linden (2010). “The effect of village-based schools: Evidence from a randomized controlled trial in
Afghanistan”. Working Paper Series No. 18039. Cambridge, MA: National Bureau of Economic Research.
Burnett, Nicholas and Milan Thomas (2013). Exclusion from Education: The Economic Cost of Out-of-School Children in 20
Countries. Washington, D.C.: Results for Development Institute.
Bush, Kenneth and Diana Saltarelli (2000). “The two faces of education in ethnic conflict: Towards a peacebuilding education
for children”. Technical Report. UNICEF Innocenti Insight. Florence, Italy: UNICEF Innocenti Research Centre.
Camacho, Adriana (2008). “Stress and birth outcomes: Evidence from terrorist attacks in Colombia”. American Economic
Review 98(2): 511–515.
Cambodian National Institute of Statistics (2014). Home page. Available at: http://www.nis.gov.kh/index.php/en/ (Accessed
28 March 2014)
Carr-Hill, Roy (2013). “Missing millions and measuring development progress”. World Development 46(C): 30–44.
Chae, Sophia (2013). “Times of orphanhood, early sexual debut and early marriage in four sub-Saharan countries”. Studies in
Family Planning, 44(2): 123–146.
Chamarbagwala, Rubiana and Hilcías E. Morán (2009). “The human capital consequences of civil war: Evidence from
Guatemala”. HiCN Working Paper No. 59. Brighton: Households in Conflict Network. Available at: http://www.hicn.org/
wordpress/wp-content/uploads/2012/06/wp59.pdf
Chang, Gwang-Chol, Raphaëlle Martinez and Hilaire Mputu (2009). “Estimating the costs of education development—Case
study for Democratic Republic of Congo, Nigeria and Sudan for EFA GMR 2010 - Part I: Synthesis report”. Background paper
prepared for the EFA Global Monitoring Report 2010. Paris: UNESCO.
Chimbutane, Feliciano (2011). Rethinking Bilingual Education in Postcolonial Contexts. Bristol, UK: Multilingual Matters.
Chimbutane, Feliciano and Carol Benson (2012). “Expanded spaces for Mozambican languages in primary education: Where
bottom-up meets top-down”. In McCarty, Teresa and Nancy Hornberger (eds), “Globalization from the bottom up: Indigenous
language planning and policy in globalizing spaces and places”. Special issue of the International Multilingual Research
Journal, 6(1): 8–21.
Colclough, Christopher, Samer Al-Samarrai, Pauline Rose and Mercy Tembon (2003). Achieving Schooling for All in Africa:
Costs, Commitment and Gender. Aldershot, UK: Ashgate.
Consortium for Street Children (2012). “Street children in Brazil”. Briefing to the IPU. London: Consortium for Street Children.

Cummins, Jim (2009). “Fundamental psycholinguistic and sociological principles underlying educational success for linguistic
minority students.” In Skutnabb-Kangas, Tove; Robert Phillipson, Ajit K. Mohanty and Minati Panda (eds), Social Justice
Through Multilingual Education, pp 19–35. Bristol, UK: Multilingual Matters.
Delamonica, Enrique, Santosh Mehrotra and Jan Vandemoortele (2001). “Is EFA affordable? Estimating the global minimum
cost of education for all”. Innocenti Working Paper 87. Florence, Italy: UNICEF.
Devarajan, Shantayanan, Margaret Miller and Eric V. Swanson (2002). “Goals for development: History, prospects and costs”.
Policy Research Working Paper No. 2819. Washington, D.C.: World Bank.
DFID Nigeria (2011). Girls Education Project Phase 3 (GEP3), Business Case. Lagos: UK Department for International
Development.
Edmeades, Jeffrey and Robin Hayes, with Gillian Gaynair (2014). Improving the Lives of Married Adolescent Girls in Amhara,
Ethiopia. Washington, D.C.: ICRW.
Education Policy and Data Center and UNESCO. (2009). “Estimating the costs of achieving Education for All in low income
countries”. Background Paper prepared for the EFA Global Monitoring Report 2010. Paris: UNESCO.
Equality Now (2014). Protecting the Girl Child. Using the Law to End Child, Early and Forced Marriage and Related Human
Rights Violations. New York: Equality Now.
Field, Erica and Attila Ambrus (2008). “Early marriage, age of menarche, and female schooling attainment in Bangladesh”.
Journal of Political Economy, 116(5): 881–930.
Filmer, Deon (2005). “Disability, poverty, and schooling in developing countries: Results from 11 household surveys”. World
Bank Social Protection Discussion Paper No. 0539. Washington, D.C.: World Bank.
Foko, Borel, Beifith Kouak Tiyab and Guillaume Husson (2012). Household Education Spending: An Analytical and
Comparative Perspective for 15 African Countries. Dakar: UNESCO.
Gakidou, Emmanuela, Krycia Cowling, Rafael Lozano and Christopher J.L. Murray (2010). “Increased educational attainment
and its effect on child mortality in 175 countries between 1970 and 2009: A systematic analysis”. The Lancet, 376(9745):
959–974.
Garcia, Marito and Charity M.T. Moore (2012). The Cash Dividend: The Rise of Cash Transfer Programs in Sub-Saharan
Africa. Washington, D.C.: World Bank.
Gelli, Aulo (2012). Food for Education Works: A Review of WFP FFE Monitoring and Evaluation 2002-2006. Rome: World
Food Programme.
Ghana MoE (Ministry of Education) (2013). Education Sector Performance Report 2013. Accra: Ghana. Available at:
http://www.moe.gov.gh/docs/ESPR%202013%20Final%20August.pdf (Accessed October 2014).
Glewwe, Paul, Meng Zhao and Melissa Binder (2006). Achieving Universal Basic and Secondary Education: How Much Will It
Cost? Cambridge MA: American Academy of Arts and Sciences.
Glick, Peter (2008). “What policies will reduce gender schooling gaps in developing countries: Evidence and interpretation,”
World Development, 36(9): 1623–1646.
Global Partnership for Education (n.d.). Children with Disabilities (web page). Available at: http://www.globalpartnership.org/
focus-areas/children-with-disabilities (Accessed 23 June 2014)
Global Partnership for Education (2012). Results for Learning. Washington, D.C.: Global Partnership for Education.
Graham, Natasha (2014). “Children with disabilities”. Background paper prepared for Fixing the Broken Promise of Education
for All: Findings from the Global Initiative on Out-of-School Children. Montreal: UNESCO Institute for Statistics.

Greene, Margaret, Omar Robles, Krista Stout and Tanya Suvilaakso (2013). A Girl’s Right to Learn Without Fear: Working
to End Gender-Based Violence at School. Woking, UK: Plan International. Available at: https://www.plan.org.au/~/media/
Documents/Research%20and%20Reports/Plan%20Report%20-%20A%20girls%20right%20to%20learn%20without%20
fear.ashx
Guerrero-Serdan, Gabriela (2009). “The effects of the war in iraq on nutrition and health: An analysis using anthropometric
outcomes of children”. HiCN Working Paper No. 55. Brighton: Households in Conflict Network. Available at: http://www.hicn.
org/wordpress/wp-content/uploads/2012/06/wp55.pdf
Guichaoua, Yvan (2009). “How do ethnic militias perpetuate in Nigeria? A micro-level perspective on the Oodua People’s
Congress”. MICROCON Reasearch Working Paper No. 19. MICROCON, EU Integrated Project on Micro-Level Analysis of
Violent Conflicts. Brighton, UK: Institute of Development Studies. Available at: http://www.microconflict.eu/publications/
RWP19_YG.pdf
Gustafsson-Wright, Emily and Hnin Hnin Pyne (2002). “Gender dimensions of child labor and street children in Brazil”. World
Bank Policy Research Working Paper. Washington, D.C.: World Bank.
Hattori, Hiroyuki (2014). “Demographic and socioeconomic determinants of school attendance: An analysis of household
survey data”. Background paper prepared for Fixing the Broken Promise of Education for All: Findings from the Global
Initiative on Out-of-School Children. Montreal: UNESCO Institute for Statistics.
Herz, Barbara and Gene B. Sperling (2004). What Works in Girls’ Education: Evidence and Policies from the Developing
World. New York: Council on Foreign Relations.
Heugh, Kathleen (2011). “Cost implications of the provision of mother-tongue and strong bilingual models of education in
Africa.” In Ouane, Adama and Christine Glanz (eds), Optimising Learning, Education and Publishing in Africa: The Language
Factor, pp 255-289. Hamburg, Germany: UNESCO. Available at: http://unesdoc.unesco.org/images/0021/002126/212602e.
pdf
Heugh, Kathleen, Carol Benson, Berhanu Bogale and Mekonnen Alemu Gebre Yohannes (2007). “Final report: Study
on medium of instruction in primary schools in ethiopia”. Addis Ababa: Ministry of Education. Available at: http://www.
everythingharar.com/publication/4379_Heugh_Studyonmediumofinstruction.pdf
Heugh, Kathleen, Carol Benson, Mekonnen Alemu Gebre Yohannes and Berhanu Bogale (2012). “Implications for multilingual
education: Student achievement in different models of education in Ethiopia.” In Skutnabb-Kangas, Tove and Kathleen Heugh
(eds), Multilingual Education and Sustainable Diversity Work From Periphery to Center, pp 239-262. New York: Routledge.
Hovens, Mart (2002). “Bilingual education in West Africa: Does it work?” International Journal of Bilingual Education and
Bilingualism 5(5): 249-266.
Huebler, Friedrich (2008). “Child labour and school attendance: Evidence from MICS and DHS surveys”. Paper presented
at the Seminar on child labour, education and youth employment. Understanding Children’s Work Project, Madrid, 11-12
September.
Humphreys, M. and Weinstein, J. (2008). “Who fights? The determinants of participation in civil war”. American Journal of
Political Science, 52(2): 436–455.
ICWAC (2000). “Caught in the crossfire no more: A framework for commitment to war-affected children”. Summary by the
Chairs of the Experts’ Meeting 13-15 Sept. 2000. Winnipeg: The International Conference on War-Affected Children.
ILO (2009). “Report of the Conference” 18th International Conference of Labour Statisticians, Geneva, 24 November-5
December 2008”. ICLS/18/2008/IV/FINAL. Geneva: ILO. Available at: http://www.ilo.org/wcmsp5/groups/public/@
dgreports/@stat/documents/meetingdocument/wcms_101467.pdf.
ILO-IPEC (2013). Marking Progress Against Child Labour—Global Estimates and Trends 2000-2012. International Labour
Office, International Programme on the Elimination of Child Labour (IPEC). Geneva: ILO. Available at: http://www.ilo.org/ipec/
Informationresources/WCMS_221513/lang--en/index.htm

International Rescue Committee (2011). International Rescue Committee in the Democratic Republic of Congo: Annual Report
2011. New York: International Rescue Committee.
IRIN (2007). “Ethiopia: Surviving forced marriage”. Available at: http://www.irinnews.org/report/69993/ethiopia-surviving-

forced-marriage (Accessed 23 June 2014)
Jacoby, Hanan G. and Emmanuel Skoufias (1997). “Risk, financial markets, and human capital in a developing country”.
Review of Economic Studies, 64(3): 311-335.
Jensen, Robert, T. (2010). “Economic opportunities and gender differences in human capital: Experimental evidence for India”.
NBER Working Paper No. 16021. Cambridge, MA: National Bureau of Economic Research.
Jomaa, Lamis H., Elaine McDonnell and Claudia Probart (2011). “School feeding programs in developing countries: Impacts
on children’s health and educational outcomes”. Nutrition Reviews, 69(2): 83-98.
Justino, Patricia (2011). “Violent conflict and human capital accumulation”. CRPD Working Paper 8. Leuven, Belgium: Centre for
Research on Peace and Development, University of Leuven. Available at: https://soc.kuleuven.be/web/files/12/80/wp08.pdf
Justino, Patricia (2012). “Shared societies and armed conflict: Costs, inequality and the benefits of peace”. IDS Working
Papers, 410: 1-23.
Justino, Patricia (2014). “Barriers to education in conflict-affected countries and policy opportunities”. Background paper
prepared for Fixing the Broken Promise of Education for All: Findings from the Global Initiative on Out-of-School Children.
Montreal: UNESCO Institute for Statistics.
Justino, Patricia, Marinella Leone and Paola Salardi (2013). “Short- and long-term impact of violence on education: The case
of Timor Leste”. World Bank Economic Review, 28(2): 320-353.
Kazianga, Harounan, Dan Levy, Leigh L. Linden and Matt Sloan (2012). “The effects of girl-friendly schools: Evidence from
the BRIGHT school construction program in Burkina Faso”. Research Working Paper No. 18115. Cambridge, MA: National
Bureau of Economic Research.
Kosonen, Kimmo (2006). “Multi-grade teaching among ethnic minority children: The language issue.” In Cornish, Linley (ed.),
Reaching EFA Through Multi-Grade Teaching: Issues, Contexts and Practices, pp 239-258. Armidale: Kardoorair Press.
Kosonen, Kimmo and Carol Benson (2013). “Introduction: Inclusive teaching and learning through the use of non-dominant
languages and cultures.” In Benson, Carol and Kimmo Kosonen (eds), Language Issues in Comparative Education: Inclusive
teaching and learning in non-dominant languages and cultures, pp 1-16. Rotterdam: Sense.
Lockheed, Marlaine and Maureen Lewis (2012). “Addressing primary and secondary education for socially excluded girls.”
In Heymann, Jody and Adèle Cassola (eds), Lessons in Educational Equity: Successful Approaches to Intractable Problems
Around the World, pp 116-141. New York: Oxford University Press.
López, Luis Enrique (2005). De Resquicios a Boquerones: La Educación Intercultural Bilingüe en Bolivia. La Paz: PROEIB
Andes.
Mathur, Sanyukta, Margaret Greene and Anju Malhotra (2003). Too Young To Wed: The Lives, Rights and Health of Young
Married Girls. Washington, D.C.: International Center for Research on Women.
Mbude-Shale, Naledi, Zola Wababa and Peter Plüddemann (2004). “Developmental research: A dual-medium schools pilot
project, Cape Town, 1999-2002.” In Brock-Utne, Birgit, Zubeida Desai and Martha Qorro (eds), Researching the Language of
Instruction in Tanzania and South Africa, pp 151-168. Cape Town: African Minds.
Merrouche, Ouarda (2006). “The human capital cost of landmine contamination in Cambodia”. HiCN Working Paper 25.
Brighton: Households in Conflict Network. Available at: http://www.hicn.org/wordpress/wp-content/uploads/2012/06/wp25.pdf

Miguel, Edward and Michael Kremer (2004). “Worms: Identifying impacts on education and health”. Econometrica, 72(1):
159-217.
Mmari, Kristin and Simran Sabherwal (2014). “A review of risk and protective factors for adolescent sexual and reproductive
health in developing countries: An update”. Journal of Adolescent Health, 53(5): 562-572.
Montjourides, Patrick (2013). “Education data in conflict-affected countries: The fifth failure?” Prospects, 43(1): 85–105. Paris:
UNESCO. Available at: http://dx.doi.org/10.1007/s11125-012-9260-8
Mooijman, Annemarieke, Caroline van den Berg, Lene Odum Jensen and Donald Bundy (2005). Toolkit on Hygiene, Sanitation
and Water in Schools. Washington, D.C.: World Bank.
Nath, Samir R. (2002). “The transition from non-formal to formal education: The case of BRAC, Bangladesh”. International
Review of Education, 48(6): 517-524.
Nega, Abeyot (2012). “Improving education quality, equity and access in Ethiopia: Findings from the young lives school
component”. Ethiopia Policy Brief 1. Addis Ababa: Young Lives in Ethiopia. Addis Ababa: UK Aid and the Ministry of Foreign
Affairs, Netherlands. Available at: http://www.younglives.org.uk/files/policy-papers/yl-ethiopia-pb1-education
Nguyen, Minh Cong and Quentin Wodon (2014a). “Analysing the gender gap in education attainment: A simple framework
with application to Ghana”. Journal of International Development, 26(1): 59-76.
Nguyen, Minh Cong and Quentin Wodon (2014b). “Estimating the impact of child marriage on literacy and education
attainment in Africa.” In Wodon, Quentin (ed.), Child Marriage and Education. Washington, D.C.: The World Bank.
Nordstrum, Lee E. (2013). “Teacher supply, training and cost in the context of rapidly expanding enrolment: Ethiopia, Pakistan
and Tanzania”. Background paper prepared for the EFA Global Monitoring Report 2013/2014. Paris: UNESCO.
Norwegian Refugee Council (1999). Global IDP Project of the Norwegian Refugee Council. Available at:
http://www.internal-displacement.org
Nyamukapa, Constance and Simon Gregson (2005). “Extended family’s and women’s roles in safeguarding orphans’
education in AIDS-afflicted rural Zimbabwe”. Social Science and Medicine, 60(10): 2155-2167.
OCHA (2013). Financial Tracking Service (FTS). OCHA, Cited in UNICEF and UIS (2014), Regional Report on Middle East and
North Africa. Global Initiative on Out-of School Children. Amman: UNICEF and UNESCO Institute for Statistics.
OCHA (2014). Occupied Palestinian Territory: Gaza Emergency. Humanitarian Snapshot. Available at: http://www.ochaopt.
org/documents/humanitarian_snapshot_13july2014_opt_v1.pdf (Accessed 13 July 2014)
OECD Development Assistance Committee Statistics (2014). Available at: http://stats.oecd.org (Accessed July 2014)
OHCHR (2014). “Nigeria: UN and African experts call for immediate release of abducted girls and accountability of
perpetrators”. Press release 8 May 2014. Geneva: Office of the High Commissioner for Human Rights. Available at: http://
www.ohchr.org/EN/NewsEvents/Pages/DisplayNews.aspx?NewsID=14577&LangID=E
O’Malley, Brendan (2007). Education Under Attack: A Global Study on Targeted Political and Military Violence Against
Education Staff, Students, Teachers, Union and Government Officials, and Institutions. Paris: UNESCO.
O’Malley, Brendan (2010). Education Under Attack: A Global Study on Targeted Political and Military Violence Against
Education Staff, Students, Teachers, Union and Government Officials, and Institutions. Paris: UNESCO.
Ouane, Adama and Christine Glanz (eds) (2011). Optimising Learning, Education and Publishing in Africa: The Language
Factor, pp 255–289. Hamburg and Tunis Belvédère: UNESCO and the Association for the Development of Education in Africa
(ADEA) / African Development Bank. Available at: http://unesdoc.unesco.org/images/0021/002126/212602e.pdf
Parlow, Anton (2012). “Armed conflict and children’s health. Exploring new directions: The case of Kashmir”. MPRA Paper No.
38033. Munich: Munich Personal RePEc Archive.

Plan (2008). Paying the Price: The Economic Cost of Failing to Educate Girls. Woking: Plan International.
Porter, Gordon (1995). “Organisation of schooling: Achieving access and equality through inclusion”. Prospects, 25(2):
299-309.
PPIU, Education Department, Government of Balochistan (2014). Balochistan Education Sector Plan 2013-2018. Quetta,
Pakistan. Available at: http://www.globalpartnership.org/content/education-sector-plan-2013-2018-balochistan-province-
pakistan
Psacharopoulos, George and Harry Anthony Patrinos (2002). “Returns to investment in education: A further update”. World
Bank Policy Research Working Paper 2881. Washington, D.C.: World Bank.
Rodriguez, Catherine and Fabio Sanchez (2009). “Armed conflict exposure, human capital investments and child
labor: Evidence from Colombia”. HiCN Working Paper No. 68. Brighton: Households in Conflict Network. Available at:
http://www.hicn.org/wordpress/wp-content/uploads/2012/06/wp68.pdf
Rosati, Furio and Lorenzo Guarcello (2014). “Child labour and out of school children: Evidence from 25 developing countries”.
Background paper and study prepared for Fixing the Broken Promise of Education for All: Findings from the Global Initiative
on Out-of-School Children. Montreal: UNESCO Institute for Statistics.
Salmon, Lizette and Quentin Wodon (2014). “Children living in the streets.” In Wodon, Quentin (ed.), Education and Highly
Vulnerable Groups. Washington, D.C.: The World Bank.
Santos, Ricardo (2014). “Post-conflict returns to education: The case of Timor Leste”. Unpublished PhD thesis. Brighton:
Institute of Development Studies.
Schroeder, Leila (2013). “Teaching and assessing independent reading skills in multilingual African countries: Not as simple
as ABC.” In Benson, Carol and Kimmo Kosonen (eds), Language Issues in Comparative Education: Inclusive Teaching and
Learning in Non-Dominant Languages and Cultures, pp 245–264. Rotterdam: Sense.
Schultz, T. Paul (2002). “Why governments should invest more to educate girls”. World Development, 30(2): 207–225.
Shemyakina, Olga (2011). “The effect of armed conflict on accumulation of schooling: Results from Tajikistan”. Journal of
Development Economics 95(2): 186–200.
Sigdel, Shailendra (2014). “Process behind selecting and calculating out-of-school children estimates in India”. Background
paper prepared for Global Report on Out-of-School Children. Montreal: UNESCO Institute for Statistics.
Skutnabb-Kangas, Tove (2000). Linguistic Genocide in Education—Or Worldwide Diversity and Human Rights? Mahwah
N.J.: Lawrence Erlbaum.
Skutnabb-Kangas, Tove and Kathleen Heugh (eds) (2012). Multilingual Education and Sustainable Diversity Work From
Periphery to Center. New York: Routledge.
Smith, Amy Erica (2014). “What perpetuates child servitude? Public opinion on children’s domestic labor in Haiti”.
AmericasBarometer Insights, 103. Available at: http://www.vanderbilt.edu/lapop/insights/IO903en.pdf
Smith, Lisa C. and Lawrence Haddad (2000). Explaining Child Malnutrition in Developing Countries: A Cross-Country Analysis.
Washington, D.C.: International Food Policy Research Institute. Available at: http://www.ifpri.org/sites/default/files/pubs/pubs/
abstract/111/rr111.pdf
Social & Rural Research Institute of IMRB International (SRI-IMRB) (2009). All India Survey of Out-Of-School Children of Age 5
and in 6-13 Years Age Group. Noida, Uttar Pradesh, India: Educational Consultants India Ltd.
Soe-Lin, Shan, Tae Min Kim, Milan Thomas, Julian Schweitzer, Robert Hecht (2014). How to Close the Gap on MDGs 4 & 5 in
Africa. Washington, D.C.: Results for Development Institute.

Sommer, Marni (2010). “Where the education system and women’s bodies collide: The social and health impact of girls’
experiences of menstruation and schooling in Tanzania”. Journal of Adolescence, 33(4): 521–529.
Stromquist, Nelly (2014). “Out-of-school children: Why gender matters”. Background paper prepared for Fixing the Broken
Promise of Education for All: Findings from the Global Initiative on Out-of-School Children. Montreal: UNESCO Institute for
Statistics.
Thakre, Subhash B., Sushama S. Thakre, Monica Reddy, Nidhi Rathi, Ketaki Pathak and Suresh Ughade (2011). “Knowledge
and practice among adolescent girls in Saoner, Nagpur District”. Journal of Clinical and Diagnostic Research, 5(5): 1027-
1033.
Thomas, Gary, David Walker and Julie Webb (eds) (1998). The Making of the Inclusive School. London: Routledge.
Understanding Children’s Work (2013). Child Labour and Out-of-School Children: Evidence from 25 Developing Countries.
Rome: Understanding Children’s Work, an Interagency Research Cooperation Project.
UNESCO (1953). “The use of vernacular languages in education. Report of the UNESCO meeting of specialists (1951).
Monographs on Fundamental Education VIII.” Paris: UNESCO.
UNESCO (2003). Gender and Education for All, The Leap to Equality: EFA Global Monitoring Report. Paris: UNESCO.
UNESCO (2007). Advocacy Kit for Promoting Multilingual Education: Including the Excluded. Bangkok: UNESCO. Available at:
http://www.unescobkk.org/education/multilingual-education/resources/mle-advocacy-kit/
UNESCO (2010a). “The quantitative impact of conflict on education”. Background paper for EFA Global Monitoring Report
2011. Paris: UNESCO.
UNESCO (2010b). Reaching the Marginalized. EFA Global Monitoring Report. Paris: UNESCO. Available at: http://unesdoc.
unesco.org/images/0018/001866/186606E.pdf
UNESCO (2011). The Hidden Crisis: Armed Conflict and Education. EFA Global Monitoring Report. Paris: UNESCO.
UNESCO (2012). From Access to Equality: Empowering Girls and Women through Literacy and Secondary Education. Paris:
UNESCO.
UNESCO (2013). “Children still battling to go to school”. EFA GMR policy paper 10. Paris: UNESCO. Available at:
http://unesdoc.unesco.org/images/0022/002216/221668E.pdf
UNESCO (2014a). “Aid reductions threaten education goals”. EFA GMR policy paper 13. Paris: UNESCO. Available at:
http://unesdoc.unesco.org/images/0022/002280/228057E.pdf
UNESCO (2014b). Teaching and Learning: Achieving Quality for All. EFA Global Monitoring Report 2013/4. Paris: UNESCO.
UNESCO Institute for Statistics (UIS) (2005). Children Out of School: Measuring Exclusion from Primary Education. Montreal:
UNESCO Institute for Statistics.
UNESCO Institute for Statistics (UIS) (2006). International Standard Classification of Education (ISCED 1997). Montreal: UIS.
Available at: http://www.uis.unesco.org/Library/Documents/isced97-en.pdf
UNESCO Institute for Statistics (UIS) (2010). Global Education Digest 2010: Comparing Education Statistics Across the
World—Special Focus on Gender. Montreal: UIS.
UNESCO Institute for Statistics (UIS) (2011a). Global Education Digest 2011: Comparing Education Statistics Across the
World—Focus on Secondary Education. Montreal: UIS.
UNESCO Institute for Statistics (UIS) (2011b). Financing Education in Sub-Saharan Africa: Meeting the Challenges of
Expansion, Equity and Quality. Montreal: UIS. Available at: http://dx.doi.org/10.15220/978-92-9189-079-2-en

UNESCO Institute for Statistics (UIS) (2012a). International Standard Classification of Education (ISCED) 2011. Montreal: UIS.
Available at: http://dx.doi.org/10.15220/978-92-9189-123-8-en
UNESCO Institute for Statistics (UIS) (2012b). Opportunities Lost: The Impact of Grade Repetition and Early School Leaving—
Global Education Digest 2012. Montreal: UIS. Available at: http://dx.doi.org/10.15220/978-92-9189-120-7-en
UNESCO Institute for Statistics (UIS) (2014). UIS Data Centre (website). Available at: http://data.uis.unesco.org. (Accessed
August 2014)
UNESCO Institute for Statistics (UIS) and Education for All Global Monitoring Report (EFA GMR) (2014a). “Progress in getting
all children to school stalls, but some countries show the way forward”. GMR Policy Paper 14, UIS Fact Sheet 28. Paris:
UNESCO. Available at: http://unesdoc.unesco.org/images/0022/002281/228184e.pdf
UNESCO Institute for Statistics (UIS) and Education for All Global Monitoring Report (EFA GMR) (2014b). “Wanted: Trained
teachers to ensure every child’s right to primary education”. GMR Policy Paper 15, UIS Fact Sheet 30. Paris: UNESCO.
Available at: http://unesdoc.unesco.org/images/0022/002299/229913E.pdf
UNFPA (2006). Sexual Violence Against Women and Girls in War and Its Aftermath: Realities, Responses, and Required
Resources. New York: UNFPA.
UNHCR (2014). War’s Human Cost. Global Trends 2013. Geneva: UNHCR. Available at: http://www.unhcr.org/5399a14f9.html
(Accessed 24 June 2014)
UNHCR and OSCE (2002). Ninth Assessment of the Situation of Ethnic Minorities in Kosovo (31 May 2002). New York and
Vienna: United Nations High Commissioner for Refugees and Organization for Security and Cooperation in Europe (UNHCR/
OSCE).
UNICEF (1999). Children Orphaned by AIDS: Front-Line Responses from Eastern and Southern Africa. New York: UNICEF.
UNICEF (2007a). The State of the World’s Children: Women and Children, the Double Dividend of Gender Equality. New York:
UNICEF.
UNICEF (2007b). UNICEF Education Strategy. New York: United Nations Economic and Social Council.
UNICEF (2009). All Children, Everywhere: A Strategy for Basic Education and Gender Equity. New York: UNICEF.
UNICEF (2011a). “Situation analysis of children and women in Nigeria”. 2011 Update. Lagos: UNICEF Nigeria. Available at:
http://www.unicef.org/nigeria/SITAN_UNICEF_Nigeria_2011_FINAL_2012_Sept.pdf
UNICEF (2011b). “The right of children with disabilities to education: A rights-based approach to inclusive education in the
CEE/CIS region”. Background Note. Geneva: UNICEF Regional Office for Central and Eastern Europe and the Commonwealth
of Independent States (CEE/CIS). Available at: http://www.unicef.org/ceecis/Background_NoteFINAL(1).pdf
UNICEF (2013a). Children with Disabilities. The State of the World’s Children 2013. New York: UNICEF.
UNICEF (2013b). Equitable Access (web-page). Available at: http://www.unicef.org/education/bege_61657.html
UNICEF (2013c). Girls’ Education and Gender Equality (web-page). Available at: http://www.unicef.org/education/
bege_70640.html
UNICEF (2013d). Summary of Midterm Review of Country Programs: Central and Eastern Europe and the Commonwealth of
Independent States region. New York: UNICEF Executive Board.
UNICEF (2014a). Child Labour and UNICEF in Action: Children at the Centre. New York: UNICEF.
UNICEF (2014b). Every Child Counts: The State of the World’s Children 2014 in Numbers. New York: UNICEF.

UNICEF (2014c). Strengthening Statistics on Children with Disabilities: UNICEF’s Work and Planned Activities. New York:
UNICEF.
UNICEF (2014d). Thematic Report 2013: Basic Education and Gender Equality. New York: UNICEF.
UNICEF global databases (2014). Available at: http://www.data.unicef.org
UNICEF and Asociación Civil Educación para Todos (2011). Caracterización del perfil de la exclusión al año 2011—República
de Honduras: Breve compendio de indicadores para la identificación de los cuellos de botella en la cobertura, el acceso y la
progresión de la escolaridad. Tegucigalpa: UNICEF/Asociación Civil Educación para Todos.
UNICEF and UIS (2011). Bolivia Country Study. Global Initiative on Out-of-School Children. La Paz: UNICEF and the UNESCO
Institute for Statistics (UIS). Available at: http://dx.doi.org/10.15220/uis-unicef-cntry-bolivia-2011-sp
UNICEF and UIS (2012a). Brasil: Acesso, permanência, aprendizagem e conclusão da Educação Básica na idade certa—
Direito de todas e de cada uma das crianças e dos adolescentes. Global Initiative on Out-of-School Children. Brasilia:
UNICEF and the UNESCO Institute for Statistics (UIS). Available at: http://dx.doi.org/10.15220/uis-unicef-cntry-brasil-2012-po
UNICEF and UIS (2012b). Completar la Escuela: Un Derecho para Crecer, un Deber para Compartir. América Latina y el
Caribe. Global Initiative on Out-of-School Children. Panamá: UNICEF and the UNESCO Institute for Statistics (UIS). Available
at: http://dx.doi.org/10.15220/978-92-806-4653-5-sp
UNICEF and UIS (2012c). Estudio de País: Colombia: Construcción de una Política Nacional para Mejorar la Retención y el
Acceso de Estudiantes al Sistema Educativo. Global Initiative on Out-of-School Children. Bogota: UNICEF and the UNESCO
Institute for Statistics (UIS). Available at: http://dx.doi.org/10.15220/978-958-8514-41-3-sp
UNICEF and UIS (2012d). Ghana Country Study. Global Initiative on Out-of-School Children. Accra: UNICEF and the
UNESCO Institute for Statistics (UIS). Available at: http://dx.doi.org/10.15220/uis-unicef-cntry-ghana-2012-en
UNICEF and UIS (2012e). Kyrgyzstan Country Study. Global Initiative on Out-of-School Children. Bishkek: UNICEF and the
UNESCO Institute for Statistics (UIS). Available at: http://dx.doi.org/10.15220/978-9967-26-912-5-en
UNICEF and UIS (2012f). Nigeria Country Study. Conducted within the Conceptual and Methodology Framework (CMF).
Global Initiative on Out-of-School Children. Lagos: UNICEF and the UNESCO Institute for Statistics (UIS). Available at: http://
dx.doi.org/10.15220/uis-unicef-cntry-nigeria-2012-en
UNICEF and UIS (2012g). Romania Country Study: Analysing the Situation of Out-of-School Children in Romania. Global
Initiative on Out-of-School Children. Islamabad: UNICEF and the UNESCO Institute for Statistics (UIS). Available at: http://
dx.doi.org/10.15220/uis-unicef-cntry-romania-2012-en
UNICEF and UIS (2012h). Turkey Country Study. Global Initiative on Out-of-School Children. Ankara: UNICEF and the
UNESCO Institute for Statistics (UIS). Available at: http://dx.doi.org/10.15220/978-92-806-4725-9-en
UNICEF and UIS (2013a). Education Equity Now! A Regional Analysis of the Situation of Out-of-School Children in Central and
Eastern Europe and the Commonwealth of Independent States. Global Initiative on Out-of-School Children. Geneva: UNICEF
and the UNESCO Institute for Statistics (UIS). Available at: http://dx.doi.org/10.15220/uis-unicef-reg-cee/cis-2013-en
UNICEF and UIS (2013b). Out-of-School Children in the Balochistan, Khyber Pakhtunkhwa, Punjab and Sindh Provinces
of Pakistan. Global Initiative on Out-of-School Children. Islamabad: UNICEF and the UNESCO Institute for Statistics (UIS).
Available at: http://dx.doi.org/10.15220/uis-unicef-cntry-balochistan-2013-en
UNICEF and UIS (2013c). Out-of-School Children in Sri Lanka. Global Initiative on Out-of-School Children. Colombo: UNICEF
and the UNESCO Institute for Statistics (UIS). Available at: http://dx.doi.org/10.15220/uis-unicef-cntry-sri-lanka-2013-en
UNICEF and UIS (2013d). Rapport de l’enquête nationale sur les enfants et adolescents en dehors de l’école [République
Démocratique du Congo]. Global Initiative on Out-of-School Children. Kinshasa: UNICEF and the UNESCO Institute for
Statistics (UIS). Available at: http://dx.doi.org/10.15220/uis-unicef-cntry-congo-2013-fr

UNICEF and UIS (2013e). Tajikistan Country Study. Global Initiative on Out-of-School Children. Dushanbe: UNICEF and the
UNESCO Institute for Statistics (UIS). Available at: http://dx.doi.org/10.15220/uis-unicef-cntry-tajikistan-2013-en
UNICEF and UIS (2014a). Eastern and Southern Africa Regional Report. Global Initiative on Out-of-School Children. Nairobi:
UNICEF and the UNESCO Institute for Statistics (UIS). Available at: http://dx.doi.org/10.15220/uis-unicef-reg-esar-2014-en
UNICEF and UIS (2014b). Regional Report on Middle East and North Africa. Global Initiative on Out-of-School Children.
Amman: UNICEF and the UNESCO Institute for Statistics (UIS).
UNICEF and UIS (2014c). Regional Report: West and Central Africa. Global Initiative on Out-of-School Children. Dakar:
UNICEF and the UNESCO Institute for Statistics (UIS). Available at: http://dx.doi.org/10.15220/uis-unicef-reg-wca-2014-en
UNICEF and UIS (2014d) South Asia Regional Study: Covering Bangladesh, India, Pakistan and Sri Lanka. Global Initiative
on Out-of-School Children. Kathmandu: UNICEF and the UNESCO Institute for Statistics (UIS). Available at: http://dx.doi.
org/10.15220/978-9937-8215-6-8-en
UNICEF and UIS (2014e forthcoming). Ethiopia Country Study. Global Initiative on Out-of-School Children: Ethiopia country
study. Addis Ababa: UNICEF and the UNESCO Institute for Statistics (UIS).
UNICEF and UIS (2014f forthcoming). Framework for Monitoring Out-of-School Children and Adolescents in the ECA Region,
Geneva: UNICEF Regional Office for Central and Eastern Europe and the Commonwealth of Independent States.
UNICEF and UIS (2014g forthcoming). OOSCI Zambia Country Report. Lusaka: UNICEF and the UNESCO Institute for
Statistics (UIS).
UNICEF and UIS (2014h). OOSCI Viet Nam Country Report. Hanoi: UNICEF and the UNESCO Institute for Statistics.
UNICEF and University of Wisconsin (2008). Monitoring Child Disability in Developing Countries. Results from the Multiple
Indicator Cluster Surveys. New York: UNICEF.
UNICEF and World Bank (2013). Simulations for Equity in Education (SEE): Background, Methodology, and Pilot Results. New
York: UNICEF.
UNICEF and World Bank (2013b). Simulations for Equity in Education (SEE): Model Description and User’s Guide. New York:
UNICEF.
United Nations (2005). Security Council Resolution 1612 (2005) on Children and Armed Conflict. Available at:
http://www.un.org/ga/search/view_doc.asp?symbol=S/RES/1612(2005)
United Nations (2006). United Nations Convention on the Rights of Persons with Disabilities. New York: United Nations.
United Nations (2010). “Report of the special representative of the Secretary-General for children and armed conflict
(A/65/219)”. New York: United Nations General Assembly. Available at: http://www.refworld.org/docid/4c8f2d922.html
United Nations (2011). “Selected examples: Best practices at the international, regional, subregional and national levels for
including persons with disabilities in development efforts.” New York: United Nations.
United Nations (2013). Education First: The United Nations Secretary-General’s initiative to ensure quality, relevant and
transformative education for everyone. New York: United Nations. Available at: http://www.globaleducationfirst.org/files/
EdFirst_G29383UNOPS_lr.pdf
United Nations (2014). Disabilities web page. Available at: http://www.un.org/disabilities/default.asp?navid=13&pid=1514

(accessed 5 August 2014).
UNODC (2014). Global Study on Homicide 2013, pp 22– 24. Vienna: UNODC. Available at: http://www.unodc.org/
documents/gsh/pdfs/2014_GLOBAL_HOMICIDE_BOOK_web.pdf

Unterhalter, Elaine, Amy North, Madeleine Arnot, Cynthia Lloyd, Lebo Moletsane, Erin Murphy-Graham, Jenny Parkes and
Mioko Saito (2014). “Interventions to enhance girls’ education and gender equality”. Education Rigorous Literature Review.
London: Department for International Development. Available at: http://r4d.dfid.gov.uk/pdf/outputs/HumanDev_evidence/
Girls_Education_Literature_Review_2014_Unterhalter.pdf
USAID (2007). “Role of education and the demobilization of child soldiers”. Issue Paper No. 1. Washington, D.C.: USAID.
Valente, Christine (2011). “Children of the revolution: Fetal and child health amidst violent civil conflict”. Health, Econometrics
and Data Group (HEDG) Working Paper 11/12. York: University of York.
Vawda, Ayesha Yaqub, and Harry Anthony Patrinos (1999). “Producing educational materials in local languages: Costs from
Guatemala and Senegal”. International Journal of Educational Development 19: 287-299.
Verhaghe, Johan (2007). “School fee practice and policy in the DRC: Frais de fonctionnement ou fonds de famille?” Kinshasa.
Unpublished manuscript.
Verhaghe, Johan (2013). “La gratuité de l’enseignement primaire en République démocratique du Congo face aux défis
structurels”. Congo-Afrique LIIème 478.
Walter, Stephen (2013). “Exploring the development of reading in multilingual education programs”. In Benson, Carol
and Kimmo Kosonen (eds), Language Issues in Comparative Education: Inclusive teaching and learning in non-dominant
languages and cultures, pp 265-281. Rotterdam: Sense.
Walter, Stephen and Carol Benson (2012). “Language policy and medium of instruction in formal education”. In Spolsky,
Bernard (ed.) The Cambridge Handbook of Language Policy, pp 278-300. Cambridge: Cambridge University Press.
Watkins, Kevin (2013). Education without Borders: A Report from Lebanon on Syria’s Out-of-School Children. A World at
School.
Wodon, Quentin (ed.) (2014). Child Marriage and Education. Washington, D.C.: The World Bank.
World Bank (2002). Achieving Education for All by 2015: Simulation Results for 47 Low-Income Countries. Washington,
D.C.: Human Development Network, Africa Region and Education Department, World Bank.
World Bank (2004). Books, Buildings and Learning Outcomes: An Impact Evaluation of World Bank Support to Basic
Education in Ghana. Washington, D.C.: World Bank.
World Bank (2005a). Education in the Democratic Republic of Congo: Priorities and Options for Regeneration. World Bank
Country Study. Washington, D.C.: World Bank.
World Bank (2005b). “Education for All: In their own language”. World Bank Education Note. Washington, D.C.: World Bank.
World Bank (2007). “Measuring disability prevalence”. SP Discussion Paper No. 0706. Washington, D.C.: World Bank p.1.
World Bank (2009). Abolishing School Fees in Africa: Lessons from Ethiopia, Ghana, Kenya, Malawi, and Mozambique.
Washington, D.C.: World Bank.
World Bank (2011). World Development Report: Conflict, Security and Development. Washington, D.C.: International Bank for
Reconstruction and Development and The World Bank.
World Bank (2013). Rwanda Education Resilience Case Report. Washington, D.C.: World Bank. Available at: http://www-
wds.worldbank.org/external/default/WDSContentServer/WDSP/IB/2013/05/14/000445729_20130514145339/Rendered/PD
F/776820WP0CR0Rwanda0Box0342041B00PUBLIC0.pdf
WHO (2001). International Classification of Functioning, Disability and Health (ICF). Geneva: WHO.
WHO (2007). International Classification of Functioning, Disability and Health for Children and Youth (ICF-CY). Geneva: WHO.

WHO (2011). “Early marriages, adolescent and young pregnancies”. Report by the Secretariat. Geneva: Executive Board,
WHO.
WHO (2012). “Early childhood development and disability”. Discussion paper. Geneva: WHO.
WHO (2013). “Disability and health”. Factsheet, No. 352. Geneva: WHO. Available at: http://www.who.int/mediacentre/
factsheets/fs352/en/
WHO and World Bank. (2011). World Report on Disability. Geneva: World Health Organization. Available at: http://www.who.
int/disabilities/world_report/2011/en/ (accessed 25 July 2014).
Wils, Anababette (2014). “Simulations for Equity in Education”. Background paper prepared for Fixing the Broken Promise of
Education for All: Findings from the Global Initiative on Out-of-School Children. Montreal: UNESCO Institute for Statistics.
Yasunaga, Mari (2014). “Non-formal education as a means to meet learning needs of out-of-school children and adolescents”.
Background Paper prepared for Fixing the Broken Promise of Education for All: Findings from the Global Initiative on Out-of-
School Children. Montreal: UNESCO Institute for Statistics.

Appendix I
UNICEF regional classification
SUB-SAHARAN AFRICA
Eastern and Southern Africa, West and Central Africa, Djibouti, Sudan
Eastern and Southern Africa

Angola, Botswana, Burundi, Comoros, Eritrea, Ethiopia, Kenya, Lesotho, Madagascar, Malawi, Mauritius,
Mozambique, Namibia, Rwanda, Seychelles, Somalia, South Africa, South Sudan, Swaziland, Uganda, United
Republic of Tanzania, Zambia, Zimbabwe
West and Central Africa

Benin, Burkina Faso, Cabo Verde, Cameroon, Central African Republic, Chad, Congo, Côte d’Ivoire, Democratic
Republic of the Congo, Equatorial Guinea, Gabon, Gambia, Ghana, Guinea, Guinea-Bissau, Liberia, Mali,
Mauritania, Niger, Nigeria, Sao Tome and Principe, Senegal, Sierra Leone, Togo
MIDDLE EAST AND NORTH AFRICA

Algeria, Bahrain, Djibouti, Egypt, Iran (Islamic Republic of), Iraq, Jordan, Kuwait, Lebanon, Libya, Morocco,
Oman, Palestine, Qatar, Saudi Arabia, Sudan, Syrian Arab Republic, Tunisia, United Arab Emirates, Yemen
SOUTH ASIA
Afghanistan, Bangladesh, Bhutan, India, Maldives, Nepal, Pakistan, Sri Lanka
EAST ASIA AND THE PACIFIC

Brunei Darussalam, Cambodia, China, Cook Islands, Democratic People’s Republic of Korea, Fiji, Indonesia,
Kiribati, Lao People’s Democratic Republic, Malaysia, Marshall Islands, Micronesia (Federated States of),
Mongolia, Myanmar, Nauru, Niue, Palau, Papua New Guinea, Philippines, Republic of Korea, Samoa, Singapore,
Solomon Islands, Thailand, Timor-Leste, Tonga, Tuvalu, Vanuatu, Viet Nam
LATIN AMERICA AND THE CARIBBEAN

Anguilla, Antigua and Barbuda, Argentina, Barbados, Belize, Bolivia (Plurinational State of), Brazil, British Virgin
Islands, Chile, Colombia, Costa Rica, Cuba, Dominica, Dominican Republic, Ecuador, El Salvador, Grenada,
Guatemala, Guyana, Haiti, Honduras, Jamaica, Mexico, Montserrat, Nicaragua, Panama, Paraguay, Peru, Saint
Kitts and Nevis, Saint Lucia, Saint Vincent and the Grenadines, Suriname, Trinidad and Tobago, Turks and
Caicos Islands, Uruguay, Venezuela (Bolivarian Republic of)

CENTRAL AND EASTERN EUROPE AND THE COMMONWEALTH OF INDEPENDENT
STATES (CEE/CIS)
Albania, Armenia, Azerbaijan, Belarus, Bosnia and Herzegovina, Bulgaria, Croatia, Georgia, Kazakhstan,
Kyrgyzstan, Montenegro, Republic of Moldova, Romania, Russian Federation, Serbia, Tajikistan, The former
Yugoslav Republic of Macedonia, Turkey, Turkmenistan, Ukraine, Uzbekistan
WESTERN EUROPE, NORTH AMERICA AND AUSTRALASIA*

Andorra, Australia, Austria, Belgium, Canada, Cyprus, Czech Republic, Denmark, Estonia, Finland, France,
Germany, Greece, Holy See, Hungary, Iceland, Ireland, Israel, Italy, Japan, Latvia, Liechtenstein, Lithuania,
Luxembourg, Malta, Monaco, Netherlands, New Zealand, Norway, Poland, Portugal, San Marino, Slovakia,
Slovenia, Spain, Sweden, Switzerland, United Kingdom, United States
* The category ‘Western Europe, North America and Australasia’ is not an official UNICEF region, but it is used in this report to group all countries not belonging
to other UNICEF regions. It includes countries in which UNICEF does not operate, primarily high- and upper-middle-income countries in Australasia, Europe and
North America.

Appendix II
Definitions
Attendance data. Data on school participation collected with household surveys or population censuses.
According to the most common measure, pupils or students who have attended a given grade or level of
education at least one day during the academic reference year are counted as attending school.
Educational attainment. The highest ISCED level of education an individual has successfully completed.
This is usually measured with respect to the highest educational programme successfully completed, which is
typically certified by a recognised qualification.
Education finance
All staff compensation as % of total expenditure in public institutions. Compensation for all
staff (teachers and non-teachers) expressed as a percentage of direct expenditure in public educational
institutions (instructional and non-instructional) of the specified level of education. Financial aid to students
and other transfers are excluded from direct expenditure. Staff compensation includes salaries, contributions
by employers for staff retirement programmes, and other allowances and benefits.
Current expenditure other than staff compensation as % of total expenditure in public

institutions. Current expenditure other than for staff compensation expressed as a percentage of direct
expenditure in public educational institutions (instructional and non-instructional) of the specified level of
education. Financial aid to students and other transfers are excluded from direct expenditure. Current
expenditure other than for staff compensation includes expenditure on school books and teaching materials,
ancillary services (e.g. food and transport), and administration and other support activities.
Expenditure by level of education as % of total government expenditure on education. Expenditure

on education by ISCED level, expressed as a percentage of total general government expenditure on
education.
Expenditure on education as % of GDP (from government sources). Total general (local, regional and
central) government expenditure on education (current, capital and transfers), expressed as a percentage of
GDP.
Expenditure on education as % of total government expenditure (all sectors). Total general (local,
regional and central) government expenditure on education (current, capital and transfers), expressed as
a percentage of total general government expenditure on all sectors (including health, education, social
services, etc.). It includes expenditure funded by transfers from international sources to the government.

Government expenditure per student as % of GDP per capita. Average total general government
expenditure (current, capital and transfers) per student in the given level of education, expressed as a
percentage of GDP per capita.
Enrolment data. Data on school participation from administrative records on pupils or students officially
registered in a given grade or level of education, regardless of age.
Gross domestic product (GDP). The sum of gross value added by all resident producers in the economy,
including distributive trades and transport, plus any product taxes, minus any subsidies not included in the value
of the products.
Gross domestic product (GDP) per capita. The gross domestic product divided by mid-year population.
Gross intake ratio to the last grade of primary education. Total number of new entrants in the last
grade of primary education, regardless of age, expressed as a percentage of the population at the theoretical
entrance age to the last grade of primary education. The ratio can exceed 100% if many over- or under-aged
children enter the last grade of primary education as a result of early or late entry into primary school and grade
repetition.
International Standard Classification of Education (ISCED). A classification system that provides a

framework for the comprehensive statistical description of national educational systems and a methodology
that translates national educational programmes into internationally comparable levels of education (levels 0 to 6
according to ISCED 1997). The basic unit of classification in ISCED is the educational programme. ISCED also
classifies programmes by field of study, programme orientation and destination.
ISCED level 0: Pre-primary education. Programmes at the initial stage of organized instruction, primarily
designed to introduce very young children, aged at least 3 years, to a school-type environment and provide
a bridge between home and school. Programmes classified at this level are variously referred to as infant
education, nursery education, pre-school education or kindergarten.
ISCED level 1: Primary education. Normally designed to give pupils a sound basic education in reading,
writing and mathematics. Main criteria include beginning of systematic studies characteristic of primary
education, e.g. reading, writing and mathematics; entry into the nationally designated primary institutions or
programmes. The commencement of reading activities alone is not a sufficient criterion for classification of
an educational programme at ISCED level 1.
ISCED level 2: Lower secondary education. The lower secondary level of education generally continues
the basic programmes of the primary level, although teaching is typically more subject-focused, often
employing more specialised teachers who conduct classes in their field of specialisation. Main criteria
include: programmes at the start of level 2 correspond to the point where programmes are beginning to be
organised in a more subject-oriented pattern, more specialised teachers are conducting classes in their field
of specialisation.
ISCED level 3: Upper secondary education. Programmes at ISCED level 3 are typically designed to
complete secondary education in preparation for tertiary education or provide skills relevant to employment,
or both. Programmes at this level offer studies more varied, specialised and with in-depth instruction than

programmes at lower secondary education (ISCED level 2). They are more differentiated, with an increased
range of options and streams available.
Non-formal education. Education that is institutionalised, intentional and planned by an education provider.
The defining characteristic of non-formal education is that it is an addition, alternative and/or complement to
formal education within the process of the lifelong learning of individuals. It is often provided to guarantee the
right of access to education for all. It caters to people of all ages but does not necessarily apply a continuous
pathway-structure; it may be short in duration and/or low-intensity, and it is typically provided in the form of short
courses, workshops or seminars. Non-formal education mostly leads to qualifications that are not recognised
as formal or equivalent to formal qualifications by the relevant national or sub-national education authorities or
to no qualifications at all. Non-formal education can cover programmes contributing to adult and youth literacy
and education for out-of-school children, as well as programmes on life skills, work skills, and social or cultural
development.
Number of out-of-school adolescents of lower secondary school age. Number of adolescents of official
lower secondary school age who are not enrolled in primary or secondary education.
Number of out-of-school children of primary school age. Number of children of official primary school age
who are not enrolled in primary or secondary education.
Out-of-school adolescents. Adolescents of official lower secondary school age who are not in primary or
secondary education. Adolescents in pre-primary or non-formal education are considered out of school.
Out-of-school children. Children of official primary school age who are not in primary or secondary education.
Children in pre-primary education or non-formal education are considered out of school.
Percentage of out-of-school adolescents of lower secondary school age. Number of adolescents of

official lower secondary school age who are not enrolled in primary or secondary education, expressed as a
percentage of the population of official lower secondary school age.
Percentage of out-of-school children of primary school age. Number of children of official primary school
age who are not enrolled in primary or secondary education, expressed as a percentage of the population of
official primary school age. Children enrolled in pre-primary education are considered out of school.
Percentage of out-of-school children of primary school age enrolled in pre-primary education.

Number of children of official primary school age who are enrolled in pre-primary education, expressed as a
percentage of the population of official primary school age.
Pupil-teacher ratio. Average number of pupils per teacher at a given level of education, based on headcounts
of both pupils and teachers.
Second-chance education. Education specifically targeted at individuals who never attended school, left
school either before completion of the level of education in which they were enrolled, or completed the level
but wish to enter an education programme or occupation for which they are not yet qualified. Participants are
often older than the typical target age group for the given ISCED level programme (but not necessarily adults).
Sometimes also referred to as ‘bridging programmes’ or ‘re-integration programmes’.

Special needs education. Education designed to facilitate learning by individuals who require additional
support and adaptive pedagogical methods in order to participate and meet learning objectives in an education
programme. Reasons may include (but are not limited to) disadvantages in physical, behavioural, intellectual,
emotional and social capacities. Programmes in special needs education may follow a similar curriculum as that
offered in the parallel regular education system, but they take individual needs into account by providing specific
resources (e.g. specially-trained personnel, equipment or space) and, if appropriate, modified educational
content or learning objectives. These programmes can be offered to individual students within already-existing
education programmes or as a separate class in the same or separate educational institutions.
For more definitions, refer to the multilingual UIS online glossary at http://www.uis.unesco.org/Pages/Glossary.aspx

Appendix III
Reader’s guide
METHODOLOGY FOR ESTIMATING OUT-OF-SCHOOL CHILDREN
The UIS generates estimates of the number of out-of-school children for two age groups: children of primary
school age and adolescents of lower secondary school age. The age ranges used for primary and lower
secondary education in each country are based on the International Standard Classification of Education
(ISCED). Within each age group, only children in formal primary or secondary education are counted as in
school. Children in pre-primary education or in non-formal education are considered out of school.
The out-of-school rate is calculated as the proportion of children not enrolled in primary or secondary education.
Example: the official primary school age range in a country is 6 to 11 years. Of 100 children aged 6 to 11 years,
80 are enrolled in primary education and 5 are enrolled in secondary education. 85 children of primary school
age are in school and 15 are out of school. The primary out-of-school rate is then 15/100=15%.
The absolute number of out-of-school children and adolescents at the national, regional and global levels is
calculated by subtracting the number of primary and lower secondary school-age children and adolescents
enrolled in primary and secondary education at the national, regional and global levels from estimates of the
population of primary and lower secondary school age by the United Nations Population Division.
This methodology was also used for the national and regional studies in the Global Initiative on Out-of-School
Children.
GLOBAL AND REGIONAL AVERAGE METHODOLOGY
Although the UIS and UNICEF use different sources of data on school participation, the basic methodology used
for the calculation of regional averages of the out-of-school rate is similar. Regional averages of the out-of-school
rate are calculated as weighted averages of national out-of-school rates. National populations of primary school
age (lower secondary school age) are used as weights during the calculation of the regional percentage of
children of primary school age (adolescents of lower secondary school age) out of school.
Both the UIS and UNICEF have developed regional average methodology to account for cases of missing data.
The UIS, which uses administrative data, provides an explanation of the methodology to calculate regional
averages on its website: http://www.uis.unesco.org/Education/Pages/FAQ.aspx. UNICEF, which uses household
survey data, publishes regional estimates only if the countries with available data in that region cover at least
50% of the corresponding regional population. More information is provided in Appendix IV.

REFERENCE PERIOD
The reference year for the administrative data used for out-of-school children estimates is the academic year
ending in 2012 or the most recent year available within the period 2010 to 2013. Where a given reference
period is spread across two calendar years, the later year is cited. For example, the school year 2011/2012 is
presented as 2012.
The reference period for the household survey data used for out-of-school children estimates is within the period
2008 to 2013.
DATA SOURCES
Administrative data
The UIS compiles education statistics in aggregate form from official administrative sources at the national
level. These include data on educational programmes, access, participation, progression, completion, internal
efficiency, and human and financial resources. They cover:
• education in pre-primary, primary and secondary schools, and in colleges, universities and other tertiary
education institutions;
• education in public (or state) and private sectors; and
• special needs education (both in regular and special schools).
These data are collected annually by the UIS and its partner agencies through the following three major surveys:
the UIS education survey; the UNESCO, Organisation for Economic Co-operation and Development (OECD),
Eurostat (Statistical Office of the European Union) (UOE) Education Data Collection; and the World Education
Indicators (WEI) programme. The questionnaires for the UIS, UOE and WEI surveys can be downloaded from the
UIS website: http://www.uis.unesco.org/UISQuestionnaires
Household survey data
Multiple Indicator Cluster Surveys (MICS) are initiated by UNICEF to assist countries in collecting and analysing
data for monitoring the situation of children and women. More detailed information on MICS is available at
http://www.data.unicef.org
The Demographic and Health Surveys (DHS) are designed to collect, analyse and disseminate nationally-
representative data on population, health, HIV and nutrition in less-developed countries. More detailed
information on DHS is available at http://www.dhsprogram.com
For other data sources of national household surveys used by UNICEF in the statistical annex and analytical
chapters, please visit http://www.data.unicef.org
For other data sources used by the country and regional reports on the Global Initiative on Out-of-School
Children, please visit http://www.allinschool.org

Population data
UIS and UNICEF estimates use population data from the 2012 revision of the World Population Prospects by the
United Nations Population Division. For more information on UN Population Division estimates, please visit
http://www.un.org/esa/population
Some national OOSCI studies used national population estimates. For more information see
http://www.allinschool.org
Economic data
Data on economic indicators, such as gross domestic product (GDP) and purchasing power parity (PPP), are
based on the World Bank’s economic data release of September 2013. Data for total government expenditure
are based on the International Monetary Fund World Economic Outlook, with some additional data sourced from
national ministries of education.
Conflict data
Countries are classified as conflict-affected based on the EFA Global Monitoring Report (UNESCO, 2014b).
Other data
Other data from national and regional Global Initiative on Out-of-School Children studies can be referenced from
the OOSCI website: http://www.allinschool.org, which features an up-to-date list of all published studies and
analysis on out-of-school children and children at risk of dropping out.
Data presented in the analytical chapters may not always be included in the statistical tables but can be
referenced at the UIS Data Centre (http://www.uis.unesco.org/datacentre) and the UNICEF global statistical
databases (http://www.data.unicef.org), which include data on child health, survival, development, education
and protection.
TECHNICAL NOTE
This report features out-of-school children estimates calculated from both administrative and household survey
sources. As discussed in Chapter 2, administrative records and household surveys are two data sources which
differ in fundamental ways: who collects the data, as well as how, when and for what purpose. As a result, the
out-of-school children estimates calculated from one data source may not match those based on other data
sources.

Appendix IV
Statistical tables
LIST OF STATISTICAL TABLES
A.1 OUT-OF-SCHOOL CHILDREN / Primary and lower secondary school age / Administrative data / 2012
A.2 OUT-OF-SCHOOL CHILDREN / Primary school age / Household survey data / 2008-2013
THE FOLLOWING SYMBOLS AND FOOTNOTES ARE USED IN THE STATISTICAL TABLES
Symbol Interpretation
… No data available
* National estimation
** For country data: UIS estimation
For regional data: Partial imputation due to incomplete country coverage (between 33% and 60% of
population)
– Magnitude nil or negligible
a Data refer to the most recent year available during the period specified in the column heading
+n Data refer to the school or financial year n years after the reference year
-n Data refer to the school or financial year n years prior to the reference year
x Data refer to years or periods other than those specified in the column heading. Such data are not
included in the calculation of regional and global averages, with the exception of 2005-2006 data from
India and 2006 data from Brazil. Estimates from years prior to 2000 are not displayed
y Data differ from the standard definition or refer to only part of a country. If they fall within the noted
reference period, such data are included in the calculation of regional and global averages

TABLE A.1. OUT-OF-SCHOOL CHILDREN / Primary and lower secondary school age / Administrative data / 2012
Out-of-school adolescents of
Age group Out-of-school children of primary school age lower secondary school age
Share of children of primary
Secondary
(ISCED 1)
(ISCED 2)
Number out school age enrolled in pre- Number out of
Primary
Lower
Out-of-school rate (%) of school primary education (%) Out-of-school rate (%) school
Country MF M F MF (000) %F MF M F MF M F MF (000) %F

or territory (1) (2 ) (3) (4) (5) (6) (7) (8) (9) (10) (11) (12) (13) (14) (15)
Afghanistan 7-12 13-15 ... ... ... ... ... ... ... ... ... ... ... ... ...
Albania 6-10 11-14 ... ... ... ... ... ... ... ... ... ... ... ... ...
Algeria 6-10 11-14 1 ... ... 25 ... ... ... ... ... ... ... ... ...
Andorra 6-11 12-15 ... ... ... ... ... ... ... ... ... ... ... ... ...
Angola 6-11 12-14 14 -1
3 -1
26 -1
513 -1
89 -1
4 -1
3 -1
5 -1
12 ** , -2
... ... 166 ** , -2
...
Anguilla 5-11 12-14 ... ... ... ... ... — -1 — -1 — -1 ... ... ... ... ...
Antigua and Barbuda 5-11 12-14 15 13 16 2 55 1 1 1 25 35 15 1 30
Argentina 6-11 12-14 ... ... ... ... ... ... ... ... 1 -1 ... ... 20 -1 ...
Armenia 6-9 10-14 ... ... ... ... ... ... ... ... ... ... ... ... ...
Aruba 6-11 12-13 1 -2 ... ... 0.1 -2 ... — -2 — -2 — -2 ... ... ... ... ...
Australia 5-11 12-15 3 3 3 61 44 2 ... ... 2 2 3 28 51
Austria 6-9 10-13 ... ... ... ... ... ... ... ... ... ... ... ... ...
Azerbaijan 6-9 10-14 11 * 10 * 12 * 54 * 52 * 3* 3* 3* 13 * 12 * 13 * 85 * 49 *
Bahamas 5-10 11-13 2 -2 ... ... 1 -2 ... ... ... ... 9 -2 11 -2 7 -2 2 -2 38 -2
Bahrain 6-11 12-14 ... ... ... ... ... ... ... ... 10 10 9 5 46
Bangladesh 6-10 11-13 4 *, -2 6 *, -2 2 *, -2 621 *, -2 20 *, -2 — *, -2 — *, -2 — *, -2 22 *, -2 30 *, -2 15 *, -2 2,206 *, -2 32 *, -2
Barbados 5-10 11-13 3* , -1
3* , -1
3* , -1
1* , -1
54 * , -1
2* , -1
3* , -1
2* , -1
7* , -1
... ... 1* , -1
...
Belarus 6-9 10-14 6 6 ** 6 ** 20 48 ** 5 6 5 2 ... ... 9 ...
Belgium 6-11 12-13 1 1 1 7 47 1 1 1 ... ... ... ... ...
Belize 5-10 11-14 1 2 — 0.4 11 — — — 3 2 4 1 66
Benin 6-11 12-15 5 ... ... 83 ... — ... ... ... ... ... ... ...
Bermuda 5-10 11-13 12 -1 13 -1 12 -1 1 -1 47 -1 — -1 — -1 — -1 19 -1 23 -1 14 -1 0.4 -1 36 -1
Bhutan 6-12 13-16 8 10 7 8 40 ... ... ... 14 18 10 8 35
Bolivia 6-11 12-13 13 -1 13 -1 13 -1 194 -1 48 -1 — -1 — -1 — -1 10 -1 10 -1 10 -1 47 -1 50 -1
Bosnia and Herzegovina 6-10 11-14 ... ... ... ... ... ... ... ... ... ... ... ... ...
Botswana 6-12 13-15 ... ... ... ... ... ... ... ... ... ... ... ... ...
Brazil 7-10 11-14 ... ... ... ... ... ... ... ... ... ... ... ... ...
British Virgin Islands 5-11 12-14 15 *, -1 13 *, -1 16 *, -1 0.5 *, -1 56 *, -1 4 *, -1 4 *, -1 4 *, -1 6 *, -1 4 *, -1 9 *, -1 0.1 *, -1 68 *, -1
Brunei Darussalam 6-11 12-13 4 4 5 2 55 3 3 3 — ... ... — ...
Bulgaria 7-10 11-14 4 4 3 9 47 — — — 10 10 9 24 46
Burkina Faso 6-11 12-15 33 32 35 917 52 … … … 50 47 53 784 52
Burundi 7-12 13-16 6 -2 6 -2 6 -2 81 -2 51 -2 — -2 — -2 — -2 31 -2 28 -2 35 -2 264 -2 57 -2
Cambodia 6-11 12-14 2 — 3 29 90 — — — ... ... ... ... ...
Cameroon 6-11 12-15 8 3 14 295 83 — — — ... ... ... ... ...
Canada 6-11 12-13 ... ... ... ... ... … … … ... ... ... ... ...
Cabo Verde 6-11 12-14 3 1 4 2 78 — — — 8 7 8 2 52
Cayman Islands 5-10 11-13 ... ... ... ... ... ... ... ... ... ... ... ... ...
Central African Rep. 6-11 12-15 28 19 36 194 66 . . . 54 43 66 229 61
Chad 6-11 12-15 36 -1 28 -1 44 -1 770 -1 61 -1 — -1 — -1 — -1 ... ... ... ... ...
Chile 6-11 12-13 7 7 7 109 49 2 2 2 3 3 4 18 52

Secondary
(ISCED 1)
(ISCED 2)
Primary
Lower Out-of-school rate (%) of school primary education (%) Out-of-school rate (%) school

or territory (1) (2 ) (3) (4) (5) (6) (7) (8) (9) (10) (11) (12) (13) (14) (15)
China 7-11 12-14 ... ... ... ... ... ... ... ... ... ... ... ... ...
China, Hong Kong SAR 6-11 12-14 1* 1* 2* 5* 67 * — — — 8* 7* 9* 16 * 55 *
China, Macao SAR 6-11 12-14 ... ... ... ... ... ... ... ... 13 14 13 2 47
Colombia 6-10 11-14 14 13 14 599 49 … … … 7 8 7 263 46
Comoros 6-11 12-15 ... ... ... ... ... ... ... ... ... ... ... ... ...
Congo 6-11 12-15 8 12 4 56 26 — — — ... ... ... ... ...
Congo, DR 6-11 12-13 ... ... ... ... ... … … … ... ... ... ... ...
Cook Islands 5-10 11-14 3* ... ... —* ... ... ... ... 13 * 11 * 14 * 0.2 * 55 *
Costa Rica 6-11 12-14 7 8 7 33 45 — 1 — 12 13 12 30 48
Côte d'Ivoire 6-11 12-15 ... ... ... ... ... ... ... ... ... ... ... ... ...
Croatia 7-10 11-14 1 2 - 2 11 — — — 1 ... ... 2 ...
Cuba 6-11 12-14 3 4 3 28 46 — — — — — 1 2 95
Curaçao 6-11 12-13 … … … … … … … … … … … … …
Cyprus 6-11 12-14 2* 2* 2* 1* 44 * 1* 1* —* 1* 2* 1* 0.4 * 23 *
Czech Republic 6-10 11-14 ... ... ... ... ... ... ... ... ... ... ... ... ...
Denmark 6-12 13-15 2 2 1 9 37 1 2 1 2 2 2 4 50
Djibouti 6-10 11-14 42 +1 38 +1 45 +1 39 +1 53 +1 … … … ... ... ... ... ...
Dominica 5-11 12-14 4 -2 5 -2 3 -2 0.3 -2 33 -2 3 -2 4 -2 3 -2 5 **, -1 9 **, -1 1 **, -1 0.2 **, -1 13 **, -1
Dominican Republic 6-11 12-13 11 10 12 137 53 — — — 9 7 10 34 56
Ecuador 6-11 12-14 3 4 2 59 36 2 3 2 6 6 7 57 54
Egypt 6-11 12-14 3 ** , -1
... ... 258 ** , -2
... 1 ** , -1
1 ** , -1
— ** , -1
1 ** ... ... 64 ** ...
El Salvador 7-12 13-15 5 5 5 41 47 4 4 4 9 8 9 39 52
Equatorial Guinea 7-12 13-16 38 38 38 38 50 — — — ... ... ... ... ...
Eritrea 7-11 12-14 66 64 68 518 51 — — — 65 61 69 255 52
Estonia 7-12 13-15 3 4 3 2 39 — — — 5 5 5 2 47
Ethiopia 7-12 13-16 ... ... ... ... ... ... ... ... ... ... ... ... ...
Fiji 6-11 12-15 1 ... ... 1 ... ... ... ... 4 ... ... 2 ...
Finland 7-12 13-15 1 1 1 4 43 — — — 3 3 3 5 50
France 6-10 11-14 1 2 1 43 23 1 1 1 — ... ... 9 ...
Gabon 6-10 11-14 ... ... ... ... ... ... ... ... ... ... ... ... ...
Gambia 7-12 13-15 26 29 24 75 45 … … … 22 ** , -2
23 ** , -2
21 ** , -2
25 ** , -2
48 **, -2
Georgia 6-11 12-14 1 2 1 4 24 ... ... ... ... ... ... ... ...
Germany 6-9 10-15 — ** 1 ** - ** 13 ** 15 ** ... ... ... ... ... ... ... ...
Ghana 6-11 12-14 12 +1 13 +1 12 +1 467 +1 48 +1 11 +1 11 +1 12 +1 8 +1 5 +1 11 +1 130 +1 69 +1
Gibraltar 5-10 11-12 ... ... ... ... ... ... ... ... ... ... ... ... ...
Greece 6-11 12-14 — -1 1 -1 - -1 3 -1 19 -1 — -1 — -1 — -1 — -2 ... ... 1 -2 ...
Grenada 5-11 12-14 ... ... ... ... ... ... ... ... ... ... ... ... ...
Guatemala 7-12 13-15 5 -1 4 -1 5 -1 110 -1 53 -1 — -1 — -1 — -1 20 -1 16 -1 25 -1 213 -1 61 -1
Guinea 7-12 13-16 24 19 30 431 61 … … … 52 **, -1 43 **, -1 60 **, -1 520 **, -1 58 **, -1
Guinea-Bissau 6-11 12-14 29 -2 27 -2 31 -2 70 -2 53 -2 — -2 — -2 — -2 ... ... ... ... ...

Secondary
(ISCED 1)
(ISCED 2)
Primary
Lower

or territory (1) (2 ) (3) (4) (5) (6) (7) (8) (9) (10) (11) (12) (13) (14) (15)
Guyana 6-11 12-14 25 30 20 32 37 1 1 1 6 **, -1 ... ... 3 **, -1 ...

Haiti 6-11 12-14 ... ... ... ... ... ... ... ... ... ... ... ... ...
Holy See . . . +1 . +1 . +1 . +1 . +1 . +1 . +1 . +1 . +1 . +1 . +1 . +1 . +1
Honduras 6-11 12-14 6 7 5 67 42 2 2 2 ... ... ... ... ...
Hungary 7-10 11-14 3 4 3 13 46 1 1 1 1 1 2 5 57
Iceland 6-12 13-15 1 -1 2 -1 1 -1 0.4 -1 39 -1 — -1 — -1 — -1 3 -1 2 -1 4 -1 0.4 -1 68 -1
India 6-10 11-13 1 -1 ... ... 1,387 -1 ... ... ... ... 23 -1 23 -1 23 -1 16,396 -1 48 -1
Indonesia 7-12 13-15 5 5 4 1,336 42 — — — 13 14 11 1,674 43
Iran, Islamic Rep. 6-10 11-13 —* ... ... 3* ... ... ... ... 5* 3* 6* 146 * 61 *
Iraq 6-11 12-14 ... ... ... ... ... … … … ... ... ... ... ...
Ireland 5-12 13-15 — — - 1 34 — — — — ... ... 0.1 ...
Israel 6-11 12-14 3 -1
3 -1
3 -1
23 -1
43 -1
3 -1
3 -1
3 -1
— -1
... ... — -1
...
Italy 6-10 11-13 1 -1 1 -1 1 -1 26 -1 70 -1 — -1 1 -1 — -1 — -1 ... ... 5 -1 ...
Jamaica 6-11 12-14 ... ... ... ... ... ... ... ... ... ... ... ... ...
Japan 6-11 12-14 — ... ... 6 ... . ... ... — ... ... 4 ...
Jordan 6-11 12-15 3 2 4 25 62 — — — 6 -1
6 -1
6 -1
31 -1
50 -1
Kazakhstan 7-10 11-15 1 **, +1 2 **, +1 - **, +1 13 **, +1 14 **, +1 — +1 — +1 — +1 4 **, +1 3 **, +1 4 **, +1 42 **, +1 55 **, +1
Kenya 6-11 12-13 ... ... ... ... ... ... ... ... ... ... ... ... ...
Kiribati 6-11 12-14 ... ... ... ... ... ... ... ... ... ... ... ... ...
Korea, DPR 7-10 11-13 ... ... ... ... ... ... ... ... ... ... ... ... ...
Korea, Republic of 6-11 12-14 1 — 1 22 69 … … … — ... ... 4 ...
Kuwait 6-10 11-14 ... ... ... ... ... ... ... ... ... ... ... ... ...
Kyrgyzstan 7-10 11-15 2 1 2 6 69 1 1 1 9 *, -1 9 *, -1 9 *, -1 47 *, -1 49 *, -1
Lao PDR 6-10 11-14 4 3 5 30 60 — — — 29 26 31 178 54
Latvia 7-12 13-15 2 2 1 2 33 1 2 1 8 7 8 4 50
Lebanon 6-11 12-14 4 ** 1 ** 7 ** 18 ** 89 ** 1 1 1 20 ** 17 ** 22 ** 50 ** 58 **
Lesotho 6-12 13-15 18 20 16 62 45 ... ... ... 23 27 18 35 40
Liberia 6-11 12-14 59 -1
58 -1
60 -1
389 -1
50 -1
... ... ... ... ... ... ... ...
Libya 6-11 12-14 ... ... ... ... ... ... ... ... ... ... ... ... ...
Liechtenstein 7-11 12-15 1* , -1
... ... —* , -1
... —* , -1
—* , -1
—* , -1
5* , -1
2* , -1
9* , -1
0.1 * , -1
82 *, -1
Lithuania 7-10 11-16 2 2 2 2 47 1 1 — — ... ... 1 ...
Luxembourg 6-11 12-14 5 -1
6 -1
4 -1
2 -1
39 -1
1 -1
1 -1
1 -1
3 -1
3 -1
2 -1
0.5 -1
36 -1
Madagascar 6-10 11-14 ... ... ... ... ... … … … ... ... ... ... ...
Malawi 6-11 12-15 ... ... ... ... ... ... ... ... ... ... ... ... ...
Malaysia 6-11 12-14 ... ... ... ... ... ... ... ... 9 -1 6 -1 13 -1 154 -1 71 -1
Maldives 6-12 13-15 ... ... ... ... ... ... ... ... ... ... ... ... ...
Mali 7-12 13-15 27 22 32 637 58 — — — 45 -1 37 -1 53 -1 440 -1 58 -1
Malta 5-10 11-13 5 5 5 1 48 — — — 10 13 7 1 33
Marshall Islands 6-11 12-13 — -1 ... ... — -1 ... ... ... ... ... ... ... ... ...
Mauritania 6-11 12-15 30 32 27 169 45 ... ... ... ... ... ... ... ...
Mauritius 5-10 11-13 2 2 2 2 50 2 2 2 ... ... ... ... ...
Mexico 6-11 12-14 2 3 1 291 27 — — — 14 15 12 977 45

Secondary
(ISCED 1)
(ISCED 2)
Primary

or territory (1) (2 ) (3) (4) (5) (6) (7) (8) (9) (10) (11) (12) (13) (14) (15)
Micronesia 6-11 12-13 ... ... ... ... ... ... ... ... ... ... ... ... ...
Monaco 6-10 11-14 ... ... ... ... ... — — — ... ... ... ... ...
Mongolia 6-10 11-14 2 2 3 5 64 — — — — ... ... 0.4 ...
Montenegro 6-10 11-14 2 2 1 1 28 1 1 1 ... ... ... ... ...
Montserrat 5-11 12-14 ... ... ... ... ... ... ... ... ... ... ... ... ...
Morocco 6-11 12-14 1 +1
1 +1
1 +1
43 +1
57 +1
1 +1
1 +1
— +1
... ... ... ... ...
Mozambique 6-12 13-15 14 11 16 692 59 ... ... ... 38 33 43 665 57
Myanmar 5-9 10-13 ... ... ... ... ... … … … ... ... ... ... ...
Namibia 7-13 14-16 11 13 10 43 43 ... ... ... ... ... ... ... ...
Nauru 6-11 12-15 24 * 25 * 23 * 0.3 * 48 * 2* 2* 1* 3* ... ... —* ...
Nepal 5-9 10-12 1 **, +1 ... ... 45 **, +1 ... ... ... ... 5 **, +1 ... ... 98 **, +1 ...
Netherlands 6-11 12-14 1 1 1 16 46 — — — — ... ... 2 ...
New Zealand 5-10 11-14 1 2 1 5 40 1 1 1 — ... ... 1 ...
Nicaragua 6-11 12-14 7 -2
7 -2
6 -2
54 -2
44 -2
. -2
. -2
. -2
18 -2
18 -2
17 -2
72 -2
48 -2
Niger 7-12 13-16 36 31 42 1,049 57 — — — 78 -1 75 -1 82 -1 1,133 -1 52 -1
Nigeria 6-11 12-14 34 ** , -2
29 ** , -2
40 ** , -2
8,709 ** , -2
57 ** , -2
... ... ... ... ... ... ... ...
Niue 5-10 11-14 ... ... ... ... ... ... ... ... ... ... ... ... ...
Norway 6-12 13-15 1 1 — 2 29 — — — 1 2 — 2 5
Oman 6-11 12-14 3 3 2 7 42 2 2 2 9 12 6 13 33
Pakistan 5-9 10-12 28 * 23 * 33 * 5,370 * 57 * ... ... ... 54 49 58 6,461 52
Palau 6-10 11-13 ... ... ... ... ... ... ... ... ... ... ... ... ...
Palestine 6-9 10-15 7 7 8 33 51 1 — 1 15 17 13 98 42
Panama 6-11 12-14 8 8 8 35 51 — — — 13 14 13 28 47
Papua New Guinea 6-12 13-14 13 10 17 165 61 ... ... ... ... ... ... ... ...
Paraguay 6-11 12-14 17 -1 17 -1 18 -1 150 -1 50 -1 1 -1 1 -1 1 -1 15 -1 15 -1 15 -1 62 -1 50 -1
Peru 6-11 12-14 4 -1 4 -1 4 -1 129 -1 48 -1 — — — 6 -1 6 -1 6 -1 104 -1 49 -1
Philippines 6-11 12-14 ... ... ... ... ... ... ... ... ... ... ... ... ...
Poland 7-12 13-15 3 3 3 70 47 — — — 5 5 5 65 50
Portugal 6-11 12-14 1 2 1 8 24 1 1 1 ... ... ... ... ...
Puerto Rico 6-11 12-14 15 -1
18 -1
13 -1
48 -1
39 -1
— -1
— -1
— -1
... ... ... ... ...
Qatar 6-11 12-14 ... ... ... ... ... ... ... ... 2 -1 2 -1 1 -1 1 -1 29 -1
Republic of Moldova 7-10 11-15 9* 9* 10 * 14 * 49 * 1* 1* 1* 13 * 13 * 14 * 28 * 51 *
Romania 7-10 11-14 ... ... ... ... ... 2 2 1 ... ... ... ... ...
Russian Federation 7-10 11-15 3 3 2 151 36 2 2 2 ... ... ... ... ...
Rwanda 7-12 13-15 1 ... ... 23 ... — ... ... ... ... ... ... ...
Saint Kitts and Nevis 5-11 12-14 16 ** 18 ** 15 ** 1 ** 45 ** … ... ... 8 -1
8 -1
7 -1
0.2 -1
44 -1
Saint Lucia 5-11 12-14 17 17 17 4 51 1 1 1 12 12 12 1 51
Saint Vincent/
Grenadines
5-11 12-14 1 ... ... 0.1 ... ... ... ... 6 -2 4 -2 7 -2 0.4 -2 61 -2
Samoa 5-10 11-12 4 5 3 1 32 3 2 3 1 ... ... 0.1 ...

San Marino 6-10 11-13 7 7 7 0.1 47 — — — 7* , -1
7* , -1
7* , -1
0.1 * , -1
46 *, -1

Secondary
(ISCED 1)
(ISCED 2)
Primary
Lower

or territory (1) (2 ) (3) (4) (5) (6) (7) (8) (9) (10) (11) (12) (13) (14) (15)
Sao Tome and Principe 6-11 12-14 3 +1 3 +1 3 +1 1 +1 43 +1 — +1 — +1 — +1 12 -2 11 -2 12 -2 1 -2 53 -2

Saudi Arabia 6-11 12-14 3 **, +1 5 **, +1 2 **, +1 115 **, +1 30 **, +1 — +1 — +1 — +1 5 **, +1 ... ... 77 **, +1 ...
Senegal 7-12 13-16 21 23 18 439 43 — — — ... ... ... ... ...
Serbia 7-10 11-14 7* 7* 7* 22 * 48 * —* —* —* 4* 3* 4* 10 * 53 *
Seychelles 6-11 12-14 6 -1 ... ... 1 -1 ... 4 -1 ... ... 2 -1 ... ... 0.1 -1 ...
Sierra Leone 6-11 12-14 ... ... ... ... ... ... ... ... ... ... ... ... ...
Singapore 6-11 12-13 ... ... ... ... ... ... ... ... ... ... ... ... ...
Sint Maarten 6-11 12-13 … … … … … … … … … … … … …
Slovakia 6-9 10-14 ... ... ... ... ... ... ... ... ... ... ... ... ...
Slovenia 6-11 12-14 2 3 2 2 40 1 1 1 4 4 3 2 44
Solomon Islands 6-11 12-14 ... ... ... ... ... 7 7 7 ... ... ... ... ...
Somalia 6-11 12-13 ... ... ... ... ... ... ... ... ... ... ... ... ...
South Africa 7-13 14-15 10 ** 10 ** 9 ** 656 ** 49 ** ... ... ... — ** ... ... 5 ** ...
South Sudan 6-11 12-13 59 ** , -1
52 ** , -1
66 ** , -1
992 ** , -1
55 ** , -1
1 -1
1 -1
1 -1
... ... ... ... ...
Spain 6-11 12-15 — — - 7 26 — — — — ** ... ... 0.4 ** ...
Sri Lanka 5-9 10-13 6 6 6 108 50 ... ... ... 7 -1
8 -1
6 -1
94 -1
43 -1
Sudan 6-11 12-13 48 -1 46 -1 51 -1 2,811 -1 52 -1 ... ... ... 35 -1 32 -1 39 -1 610 -1 55 -1
Suriname 6-11 12-15 8 -1
8 -1
7 -1
5 -1
46 -1
— -1
— -1
— -1
16 -1
16 -1
15 -1
6 -1
47 -1
Swaziland 6-12 13-15 ... ... ... ... ... … … … ... ... ... ... ...
Sweden 7-12 13-15 1 — 1 3 69 — — — 4 4 4 13 47
Switzerland 7-12 13-15 1 1 - 4 20 — — — 4 4 4 10 47
Syrian Arab Republic 6-9 10-14 1 -2
... ... 19 -2
... — -2
— -2
— -2
10 9 11 255 53
Tajikistan 7-10 11-15 1 ... ... 7 ... … … … 6 -1 2 -1 9 -1 50 -1 78 -1
Tanzania 7-13 14-17 ... ... ... ... ... — — — ... ... ... ... ...
Thailand 6-11 12-14 ... ... ... ... ... … … … ... ... ... ... ...
TFYR Macedonia 6-10 11-14 8 ** 8 ** 8 ** 10 ** 48 ** 2 2 2 ... ... ... ... ...
Timor-Leste 6-11 12-14 8 -1 8 -1 9 -1 16 -1 54 -1 ... ... ... 34 -1 33 -1 34 -1 34 -1 50 -1
Togo 6-11 12-15 ... ... ... ... ... — -1 ... ... ... ... ... ... ...
Tokelau 5-10 11-13 ... ... ... ... ... ... ... ... ... ... ... ... ...
Tonga 5-10 11-14 10 ** 11 ** 9 ** 2 ** 41 ** ... ... ... ... ... ... ... ...
Trinidad and Tobago 5-11 12-14 1 -2 1 -2 2 -2 2 -2 62 -2 ... ... ... ... ... ... ... ...
Tunisia 6-11 12-14 — ... ... 1 ... ... ... ... ... ... ... ... ...
Turkey 6-10 11-13 5 4 5 313 55 — — — 1 ... ... 38 ...
Turkmenistan 7-9 10-14 ... ... ... ... ... ... ... ... ... ... ... ... ...
Turks and Caicos
Islands
6-11 12-14 ... ... ... ... ... ... ... ... ... ... ... ... ...
Tuvalu 6-11 12-15 ... ... ... ... ... ... ... ... ... ... ... ... ...
Uganda 6-12 13-16 9 -1 10 -1 8 -1 663 -1 43 -1 … … … ... ... ... ... ...
Ukraine 6-9 10-14 2 2* 1* 24 22 * ... ... ... 4 4* 4* 80 47 *
United Arab Emirates 6-10 11-14 2* 1* 3* 6* 75 * 1* 1* 1* ... ... ... ... ...
United Kingdom 5-10 11-13 — — - 7 62 — — — 2 2 3 51 53

Secondary
(ISCED 1)
(ISCED 2)
Primary

or territory (1) (2 ) (3) (4) (5) (6) (7) (8) (9) (10) (11) (12) (13) (14) (15)
United States 6-11 12-14 7 7 7 1,800 49 3 3 3 3 3 3 354 44

Uruguay 6-11 12-14 — -2 ... ... 0.5 -2 ... — -2 — -2 — -2 23 -2 21 -2 24 -2 35 -2 53 -2
Uzbekistan 7-10 11-15 9 -1 7 -1 10 -1 178 -1 57 -1 2 -1 2 -1 2 -1 6 -1 6 -1 7 -1 181 -1 55 -1
Vanuatu 6-11 12-15 ... ... ... ... ... … … … ... ... ... ... ...
Venezuela 6-11 12-14 6 4 7 191 59 1 1 2 8 10 6 134 37
Viet Nam 6-10 11-14 2 ... ... 122 ... 2 ... ... ... ... ... ... ...
Yemen 6-11 12-14 13 5 21 490 79 … … … 37 ** 26 ** 49 ** 667 ** 64 **
Zambia 7-13 14-15 2 ** 2 ** 2 ** 59 ** 45 ** ... ... ... ... ... ... ... ...
Zimbabwe 6-12 13-14 ... ... ... ... ... ... ... ... ... ... ... ... ...
REGIONAL AVERAGES
Sub-Saharan Africa 22 ** 19 ** 25 ** 32,711 ** 56 ** ... ... ... 33 ** 31 ** 36 ** 21,941 ** 54 **
Eastern and Southern Africa 15 ** 14 ** 17 ** 10,980 ** 55 ** ... ... ... 27 **, -1 24 **, -1 30 **, -1 8,474 **, -1 55 **, -1
West and Central Africa 27 ** 23 ** 31 ** 18,828 ** 57 ** ... ... ... 40 ** 37 ** 42 ** 12,535 ** 53 **
Middle East and North Africa 9 ** 8 ** 11 ** 4,301 ** 58 ** ... ... ... 12 ** 9 ** 14 ** 2,911 ** 59 **
South Asia 6 ** 6 ** 6 ** 9,810 ** 48 ** ... ... ... 26 ** 26 ** 26 ** 26,328 ** 48 **
East Asia and the Pacific 5 ** 5 ** 5 ** 6,853 ** 47 ** ... ... ... 8 ** 9 ** 8 ** 7,375 ** 46 **
Latin America and the Caribbean 6 ** 7 ** 6 ** 3,759 ** 47 ** ... ... ... 8 ** 8 ** 7 ** 2,819 ** 48 **
CEE/CIS 5 5 5 1,008 49 ... ... ... 5 ** 5 ** 5 ** 1,158 ** 51 **
W. Europe, N. America and Australasia 4 4 3 2,240 47 ... ... ... 3 3 3 1,014 50
WORLD 9 ** 8 ** 10 ** 57,781 ** 53 ** ... ... ... 17 ** 16 ** 17 ** 62,889 ** 50 **
DataLink: http://dx.doi.org/10.15220/2014/ed/sd/7/tii.1

TABLE A.2. OUT-OF-SCHOOL CHILDREN / Primary school age / Household survey data / 2008-2013a
Age group Out-of-school rate for children of primary school age (%)
Sex Location Household wealth quintile
Second Second
Primary Poorest poorest Middle richest Richest
(ISCED 1) MF M F Urban Rural quintile quintile quintile quintile quintile Source
Country or territory (1) (2 ) (3) (4) (5) (6) (7) (8) (9) (10) (11) (12)
Afghanistan 7-12 43 y 36 y 52 y 22 y 46 y ... ... ... ... ... Living Condition Survey 2011-2012
Albania 6-10 10 10 9 10 9 11 7 10 11 9 DHS 2008-2009
Algeria 6-10 4x 3x 4x 3x 5x 7x 3x 2x 2x 3x MICS 2006
Andorra 6-11 ... ... ... ... ... ... ... ... ... ...
Inquérito Integrado sobre o Bem-Estar
Angola 6-11 21 23 25 15 33 37 26 23 14 10
da População 2008-2009
Anguilla 5-11 ... ... ... ... ... ... ... ... ... ...
Antigua and Barbuda 5-11 ... ... ... ... ... ... ... ... ... ...
Argentina 6-11 1 2 1 ... ... 2 1 2 2 1 MICS 2011-2012
Armenia 6-9 3y 3y 3y 3y 2y 2y 4y 3y 3y 3y DHS 2010
Aruba 6-11 ... ... ... ... ... ... ... ... ... ...
Australia 5-11 ... ... ... ... ... ... ... ... ... ...
Austria 6-9 ... ... ... ... ... ... ... ... ... ...
Azerbaijan 6-9 27 x 26 x 28 x 26 x 28 x 28 x 29 x 25 x 30 x 22 x DHS 2006
Bahamas 5-10 ... ... ... ... ... ... ... ... ... ...
Bahrain 6-11 14 x 14 x 13 x ... ... ... ... ... ... ... MICS 2000
Bangladesh 6-10 21 23 19 23 20 28 20 17 17 19 DHS 2011
Barbados 5-10 ... ... ... ... ... ... ... ... ... ...
Belarus 6-9 8 7 10 9 7 7 7 11 10 7 MICS 2012
Belgium 6-11 ... ... ... ... ... ... ... ... ... ...
Belize 5-10 6 6 6 2 8 12 6 3 1 2 MICS 2011
Benin 6-11 24 21 27 17 28 43 29 18 12 10 DHS 2011-2012
Bermuda 5-10 ... ... ... ... ... ... ... ... ... ...
Bhutan 6-12 5 5 5 2 6 ... ... ... ... ... Living Standard Survey 2012
Bolivia 6-11 3 3 3 2 4 5 3 3 2 1 DHS 2008
Bosnia and Herzegovina 6-10 2 2 3 3 2 5 1 3 2 3 MICS 2012
Botswana 6-12 13 x 15 x 12 x 11 x 15 x ... ... ... ... ... Family Health Survey 2007
Pesquisa Nacional por Amostra de
Brazil 7-10 5x 6x 5x ... ... ... ... ... ... ...
Domicilios 2006
British Virgin Islands 5-11 ... ... ... ... ... ... ... ... ... ...
Brunei Darussalam 6-11 ... ... ... ... ... ... ... ... ... ...
Bulgaria 7-10 ... ... ... ... ... ... ... ... ... ...
Burkina Faso 6-11 48 46 50 17 55 69 58 49 39 15 DHS 2010
Burundi 7-12 15 15 16 9 16 24 18 14 13 8 DHS 2010
Cambodia 6-11 14 y 14 y 14 y ... ... ... ... ... ... ... Socio-Economic Survey 2012
Cameroon 6-11 15 13 18 6 22 40 15 8 4 1 DHS 2011
Canada 6-11 ... ... ... ... ... ... ... ... ... ...
Cabo Verde 6-11 ... ... ... ... ... ... ... ... ... ...
Cayman Islands 5-10 ... ... ... ... ... ... ... ... ... ...
Central African Republic 6-11 27 22 32 14 34 43 37 27 18 10 MICS 2010
Chad 6-11 48 45 52 29 53 60 58 51 43 26 MICS 2010

Encuesta Caracterización
Chile 6-11 9y 10 y 8y 9y 8y ... ... ... ... ...
Socioeconómica Nacional 2011

Second Second
China 7-11 3y 3y 3y 3y 4y ... ... ... ... ... Population Census 2010
China, Hong Kong SAR 6-11 ... ... ... ... ... ... ... ... ... ...
China, Macao SAR 6-11 ... ... ... ... ... ... ... ... ... ...
Colombia 6-10 9 10 8 9 9 10 8 10 9 7 DHS 2010
Comoros 6-11 69 x 69 x 69 x 59 x 71 x 76 x 74 x 68 x 64 x 61 x MICS 2000
Congo 6-11 8y 8y 8y 8y 8y 10 y 7y 8y 8y 8y DHS 2011-2012
Congo, DR 6-11 25 23 28 14 30 35 32 27 24 8 MICS 2010
Cook Islands 5-10 ... ... ... ... ... ... ... ... ... ...
Costa Rica 6-11 4 4 4 3 5 6 3 4 3 2 MICS 2011
Côte d'Ivoire 6-11 32 y 28 y 36 y 27 y 35 y 43 y 30 y 37 y 28 y 20 y DHS 2012
Croatia 7-10 ... ... ... ... ... ... ... ... ... ...
Cuba 6-11 ... ... ... ... ... ... ... ... ... ...
Curaçao 6-11 ... ... ... ... ... ... ... ... ... ...
Cyprus 6-11 ... ... ... ... ... ... ... ... ... ...
Czech Republic 6-10 ... ... ... ... ... ... ... ... ... ...
Denmark 6-12 ... ... ... ... ... ... ... ... ... ...
Enquête Djiboutienne auprès des Ménages
Djibouti 6-10 31 y 29 y 32 y ... ... ... ... ... ... ...
pour les Indicateurs Sociaux 2012
Dominica 5-11 ... ... ... ... ... ... ... ... ... ...
Encuesta Nacional de Hogares de
Dominican Republic 6-11 ... 9y 8y ... ... ... ... ... ... ...
Propósitos Múltiples 2012
Encuesta Nacional de Empleo,
Ecuador 6-11 4y 4y 3y 3y 4y ... ... ... ... ...
Desempleo y Subempleo 2013
Egypt 6-11 12 y 11 y 13 y 10 y 13 y 18 y 12 y 12 y 11 y 7y Family Condition Survey 2009
El Salvador 7-12 ... ... ... ... ... ... ... ... ... ...
Equatorial Guinea 7-12 40 x 39 x 40 x ... ... ... ... ... ... ... MICS 2000
Eritrea 7-11 43 y 43 y 44 y 21 y 53 y 69 y 59 y 39 y 16 y 16 y Population and Health Survey 2010
Estonia 7-12 ... ... ... ... ... ... ... ... ... ...
Ethiopia 7-12 35 36 35 14 39 48 42 38 30 14 DHS 2011
Fiji 6-11 ... ... ... ... ... ... ... ... ... ...
Finland 7-12 ... ... ... ... ... ... ... ... ... ...
France 6-10 ... ... ... ... ... ... ... ... ... ...
Gabon 6-10 13 y 13 y 13 y 13 y 11 y 16 y 13 y 11 y 10 y 14 y DHS 2012
Gambia 7-12 37 39 36 25 46 53 41 38 34 18 MICS 2010
Georgia 6-11 4 5 4 3 5 8 5 3 3 4 Reproductive Health Survey 2010
Germany 6-9 ... ... ... ... ... ... ... ... ... ...
Ghana 6-11 27 28 26 20 33 39 33 23 18 14 MICS 2011
Gibraltar 5-10 ... ... ... ... ... ... ... ... ... ...
Greece 6-11 ... ... ... ... ... ... ... ... ... ...
Grenada 5-11 ... ... ... ... ... ... ... ... ... ...
Guatemala 7-12 ... ... ... ... ... ... ... ... ... ...
Enquête Démographique et de Santé et
Guinea 7-12 42 37 47 16 53 68 55 46 25 11
à Indicateurs Multiples 2012
Guinea-Bissau 6-11 33 31 35 17 44 48 44 35 16 13 MICS 2010

Second Second
Guyana 6-11 5 6 4 4 6 9 3 5 4 3 DHS 2009
Haiti 6-11 23 y 23 y 22 y 14 y 27 y 35 y 28 y 21 y 13 y 8y DHS 2012
Holy See . ... ... ... ... ... ... ... ... ... ...
Honduras 6-11 7y 8y 7y 6y 8y 11 y 7y 6y 4y 7y DHS 2011-2012
Hungary 7-10 ... ... ... ... ... ... ... ... ... ...
Iceland 6-12 ... ... ... ... ... ... ... ... ... ...
National Family Health Survey
India 6-10 17 x 15 x 18 x 12 x 18 x 30 x 19 x 12 x 8x 4x
2005-2006
Indonesia 7-12 6 6 5 5 6 9 6 5 4 3 DHS 2012
Multiple Indicator Demographic and
Iran, Islamic Rep. 6-10 3 4 3 3 5 ... ... ... ... ...
Health Survey 2010-2011
Iraq 6-11 10 7 13 6 16 21 10 6 4 2 MICS 2011
Ireland 5-12 ... ... ... ... ... ... ... ... ... ...
Israel 6-11 ... ... ... ... ... ... ... ... ... ...
Italy 6-10 ... ... ... ... ... ... ... ... ... ...
Jamaica 6-11 2 3 1 2 2 3 2 3 2 1 MICS 2011
Japan 6-11 ... ... ... ... ... ... ... ... ... ...
Jordan 6-11 2 2 2 2 2 3 1 3 1 1 DHS 2012
Kazakhstan 7-10 1 1 1 1 1 1 1 1 — — MICS 2010-2011
Kenya 6-11 13 15 12 6 15 28 11 10 8 4 DHS 2008-2009
Kiribati 6-11 15 17 13 ... ... ... ... ... ... ... Census of Population and Housing 2010
Korea, DPR 7-10 1 1 1 0.4 2 ... ... ... ... ... MICS 2009
Korea, Republic of 6-11 ... ... ... ... ... ... ... ... ... ...
Kuwait 6-10 ... ... ... ... ... ... ... ... ... ...
Kyrgyzstan 7-10 2 2 1 3 1 1 1 2 1 3 DHS 2012
Lao PDR 6-10 15 15 15 5 17 29 18 9 5 3 Social Indicator Survey 2011-2012
Latvia 7-12 ... ... ... ... ... ... ... ... ... ...
Lebanon 6-11 2 2 2 ... ... ... ... ... ... ... MICS Style 2009
Lesotho 6-12 11 13 9 7 12 17 12 8 8 6 DHS 2009
Liberia 6-11 66 x 64 x 67 x 44 x 79 x 84 x 81 x 72 x 62 x 34 x DHS 2007
Libya 6-11 ... ... ... ... ... ... ... ... ... ...
Liechtenstein 7-11 ... ... ... ... ... ... ... ... ... ...
Lithuania 7-10 ... ... ... ... ... ... ... ... ... ...
Luxembourg 6-11 ... ... ... ... ... ... ... ... ... ...
Enquête Nationale du Suivi des OMD
Madagascar 6-10 31 y 32 y 29 y 14 y 34 y 46 y 32 y 27 y 20 y 18 y
2012-2013
Malawi 6-11 15 16 14 7 16 25 19 14 11 4 DHS 2010
Malaysia 6-11 ... ... ... ... ... ... ... ... ... ...
Maldives 6-12 6 7 5 6 6 6 6 5 5 7 DHS 2009
Mali 7-12 43 40 45 20 50 64 53 45 30 15 MICS 2010
Malta 5-10 ... ... ... ... ... ... ... ... ... ...
Marshall Islands 6-11 ... ... ... ... ... ... ... ... ... ...
Mauritania 6-11 39 41 38 28 45 55 46 34 32 17 MICS 2011

Second Second
Mauritius 5-10 ... ... ... ... ... ... ... ... ... ...
Mexico 6-11 3x 3x 3x ... ... ... ... ... ... ... Census of Population and Housing 2005
Micronesia 6-11 ... ... ... ... ... ... ... ... ... ...
Monaco 6-10 ... ... ... ... ... ... ... ... ... ...
Mongolia 6-10 4 5 4 3 6 7 7 3 2 2 MICS 2010
Montenegro 6-10 3x 3x 2x 3x 2x 8x 1x 1x —x —x MICS 2005-2006
Montserrat 5-11 ... ... ... ... ... ... ... ... ... ...
Morocco 6-11 11 x 9x 12 x 4x 17 x 23 x 14 x 5x 3x 3x MICS 2006-2007
Mozambique 6-12 23 y 23 y 23 y 14 y 26 y 33 y 31 y 24 y 16 y 9y DHS 2011
Myanmar 5-9 10 10 9 7 11 19 9 6 5 5 MICS 2009-2010
Namibia 7-13 13 y 14 y 12 y 10 y 15 y ... ... ... ... ... Population and Housing Census 2011
Nauru 6-11 3y 3y 2y ... ... ... ... ... ... ... Population and Housing Census 2011
Nepal 5-9 6 4 9 3 7 9 9 6 3 1 DHS 2011
Netherlands 6-11 ... ... ... ... ... ... ... ... ... ...
New Zealand 5-10 ... ... ... ... ... ... ... ... ... ...
Living Standards Measurement
Nicaragua 6-11 30 29 31 24 36 ... ... ... ... ...
Survey 2009
Enquête Démographique et de Santé et
Niger 7-12 50 45 54 17 55 66 60 55 43 19
à Indicateurs Multiples 2012
Nigeria 6-11 41 y 38 y 43 y 29 y 48 y 73 y 44 y 27 y 25 y 30 y DHS 2013
Census of Population and Housing
Niue 5-10 — — — ... ... ... ... ... ... ...
2011
Norway 6-12 ... ... ... ... ... ... ... ... ... ...
Oman 6-11 ... ... ... ... ... ... ... ... ... ...
Pakistan 5-9 36 33 40 25 41 61 41 29 23 13 DHS 2012-2013
Palau 6-10 ... ... ... ... ... ... ... ... ... ...
Palestine 6-9 7 7 7 8 3 10 8 7 6 4 MICS 2010
Panama 6-11 ... ... ... ... ... ... ... ... ... ...
Papua New Guinea 6-12 ... ... ... ... ... ... ... ... ... ...
Paraguay 6-11 12 13 11 11 13 ... ... ... ... ... Encuesta Permanente de Hogares 2008
Peru 6-11 3 3 3 3 3 4 3 2 2 2 DHS 2012
Philippines 6-11 12 x 12 x 11 x 10 x 13 x 21 x 12 x 8x 6x 8x DHS 2003
Poland 7-12 ... ... ... ... ... ... ... ... ... ...
Portugal 6-11 ... ... ... ... ... ... ... ... ... ...
Puerto Rico 6-11 ... ... ... ... ... ... ... ... ... ...
Qatar 6-11 ... ... ... ... ... ... ... ... ... ...
Republic of Moldova 7-10 1 1 2 2 1 2 1 2 1 1 MICS Summary Report 2012
Romania 7-10 ... ... ... ... ... ... ... ... ... ...
Russian Federation 7-10 ... ... ... ... ... ... ... ... ... ...
Integrated Household Living Conditions
Rwanda 7-12 8y 9y 7y 7y 9y 13 y 9y 7y 8y 4y
Survey 2010-2011
Saint Kitts and Nevis 5-11 ... ... ... ... ... ... ... ... ... ...
Saint Lucia 5-11 1 — 1 1 — ... ... ... ... ... MICS 2012
Saint Vincent/Grenadines 5-11 ... ... ... ... ... ... ... ... ... ...

Second Second
Samoa 5-10 12 y 12 y 11 y 11 y 12 y 15 y 12 y 10 y 12 y 9y DHS 2009
San Marino 6-10 ... ... ... ... ... ... ... ... ... ...
Sao Tome and Principe 6-11 6 6 6 6 7 12 9 6 2 3 DHS 2008-2009
Saudi Arabia 6-11 ... ... ... ... ... ... ... ... ... ...
Senegal 7-12 38 40 37 19 50 53 43 36 31 22 DHS 2010-2011
Serbia 7-10 1 2 1 1 2 4 — 1 — 2 MICS 2010
Seychelles 6-11 ... ... ... ... ... ... ... ... ... ...
Sierra Leone 6-11 26 27 24 21 28 42 32 24 17 12 MICS 2010
Singapore 6-11 ... ... ... ... ... ... ... ... ... ...
Sint Maarten 6-11 ... ... ... ... ... ... ... ... ... ...
Slovakia 6-9 ... ... ... ... ... ... ... ... ... ...
Slovenia 6-11 ... ... ... ... ... ... ... ... ... ...
Solomon Islands 6-11 35 x,y 38 x,y 31 x,y 28 x,y 36 x,y 42 x,y 36 x,y 39 x,y 33 x,y 22 x,y DHS 2007
Somalia 6-11 77 x 75 x 79 x 59 x 88 x 96 x 93 x 81 x 66 x 47 x MICS 2006
South Africa 7-13 ... ... ... ... ... ... ... ... ... ...
South Sudan 6-11 74 71 77 57 79 89 85 78 66 46 MICS 2010
Spain 6-11 ... ... ... ... ... ... ... ... ... ...
Sri Lanka 5-9 ... ... ... ... ... ... ... ... ... ...
Sudan 6-11 25 22 28 11 31 45 38 23 8 3 MICS 2010
Suriname 6-11 5 5 4 4 6 9 4 2 1 4 MICS 2010
Swaziland 6-12 4 4 3 3 4 5 3 5 2 1 MICS 2010
Sweden 7-12 ... ... ... ... ... ... ... ... ... ...
Switzerland 7-12 ... ... ... ... ... ... ... ... ... ...
Syrian Arab Republic 6-9 13 x 13 x 14 x 11 x 15 x ... ... ... ... ... MICS 2006
Tajikistan 7-10 3 2 4 2 3 5 4 2 2 2 DHS 2012
Tanzania 7-13 20 21 18 9 23 32 26 17 12 7 DHS 2010
Thailand 6-11 4 4 4 5 4 6 5 4 4 3 MICS 2012
TFYR Macedonia 6-10 2 2 2 1 2 4 2 1 1 — MICS 2011
Timor-Leste 6-11 28 29 27 20 30 40 31 28 22 16 DHS 2009-2010
Togo 6-11 11 9 14 6 14 20 15 8 6 4 MICS 2010
Tokelau 5-10 ... ... ... ... ... ... ... ... ... ...
Tonga 5-10 ... ... ... ... ... ... ... ... ... ...
Trinidad and Tobago 5-11 2x 2x 2x ... ... 6x 2x 1x 1x 2x MICS 2006
Tunisia 6-11 2 2 2 1 3 4 2 2 1 2 MICS 2011-2012
Turkey 6-10 7y 7y 8y 7y 9y 13 y 7y 5y 4y 4y DHS Style 2008
Turkmenistan 7-9 ... ... ... ... ... ... ... ... ... ...
Turks and Caicos Islands 6-11 ... ... ... ... ... ... ... ... ... ...
Tuvalu 6-11 2 x,y 3 x,y 1 x,y 3 x,y 1 x,y 1 x,y 2 x,y 1 x,y 5 x,y — x,y DHS 2007
Uganda 6-12 19 19 19 15 19 27 21 16 15 13 DHS 2010
Ukraine 6-9 — — — — — — 1 — — — MICS 2012
United Arab Emirates 6-10 ... ... ... ... ... ... ... ... ... ...
United Kingdom 5-10 ... ... ... ... ... ... ... ... ... ...

Second Second
United States 6-11 ... ... ... ... ... ... ... ... ... ...
Uruguay 6-11 ... ... ... ... ... ... ... ... ... ...
Uzbekistan 7-10 4x 4x 4x 3x 5x 6x 4x 4x 4x 3x MICS 2006
Vanuatu 6-11 23 y 24 y 22 y 23 y 23 y 25 y 28 y 21 y 20 y 20 y DHS (Preliminary) 2013
Venezuela 6-11 8x 9x 7x ... ... 14 x 7x 4x 2x 2x MICS 2000
Viet Nam 6-10 2 2 2 2 2 5 2 2 — 2 MICS 2010-2011
Yemen 6-11 30 x 25 x 36 x 17 x 36 x 56 x 32 x 27 x 19 x 14 x MICS 2006
Zambia 7-13 28 y 29 y 28 y 20 y 33 y ... ... ... ... ... Census of Population and Housing 2010
Zimbabwe 6-12 13 y 13 y 12 y ... ... ... ... ... ... ... Population Census 2012
REGIONAL AVERAGES
Sub-Saharan Africa 30 29 31 19 34 48 35 27 21 15
Eastern and Southern Africa 24 25 24 13 28 36 29 24 18 11
West and Central Africa 35 33 37 22 42 57 40 30 25 19
Middle East and North Africa 12 10 13 7 17 ... ... ... ... ...
South Asia 20 18 22 15 22 33 21 15 10 7
East Asia and the Pacific 4 4 4 4 5 ... ... ... ... ...
Latin America and the Caribbean 6 7 6 ... ... ... ... ... ... ...
CEE/CIS 5 4 5 5 5 9 5 4 3 3
Western Europe, North America and Australasia ... ... ... ... ... ... ... ... ... ...
WORLD 17 16 18 11 21 33 23 17 13 9
DataLink: http://dx.doi.org/10.15220/2014/ed/sd/7/tii.2

ALL IN SCHOOL
Out-of-School
Children Initiative
UNICEF AND UNESCO INSTITUTE FOR STATISTICS
Fixing the Broken Promise of Education for All,

published by the UNESCO Institute for Statistics
and UNICEF, presents the latest statistical evidence
from administrative records and household surveys
to better identify children who are out of school
and the reasons for their exclusion from education.
It aims to inform the policies needed to reach
these children and finally deliver the promise of
Education for All.
Based on a series of national and regional studies

and policy analysis by leading experts, the
report explains why better data and cross-sector
collaboration are fundamental to the design of
effective interventions to overcome the barriers
facing out-of-school children and adolescents.
While highlighting the way forward for system-
wide policies to improve educational quality
and affordability, the report also presents the
information needed for targeted approaches to
address the compounding effects of disadvantage
faced by children caught up in armed conflict,
girls, working children, children with disabilities,
or members of ethnic or linguistic minorities.
This report presents a roadmap to improve the

data, research and policies needed to catalyse
action for out-of-school children as the world
embarks on a new development agenda for
education.
ISBN 978-92-9189-162-7
9 789291 891627
See the data!

Global OOSCreport Full Web - 217 PDF

Uploaded by

Copyright:

Available Formats

Global OOSCreport Full Web - 217 PDF

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Global OOSCreport Full Web - 217 PDF

Uploaded by

Copyright:

Available Formats

FIXING THE BROKEN PROMISE

OF EDUCATION FOR ALL

United Nations Children’s Fund (UNICEF)

Published in 2015 by:

UNESCO Institute for Statistics

Tel: +1 514 343 6880

Typesetting: Em Dash Design / www.emdashdesign.ca

Download the data in this report

Chapter 2. Data on out-of-school children and adolescents . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17

Chapter 3. Barriers and policy solutions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39

Chapter 4. Financing needs for out-of-school children. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 89

Findings from the Global Initiative on Out-of-School Children 3

Appendix I. UNICEF regional classification. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 121

Appendix II. Definitions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 123

Appendix III. Reader’s guide. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 127

Appendix IV. Statistical tables. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 131

4 Fixing the Broken Promise of Education for All

Findings from the Global Initiative on Out-of-School Children 5

Findings from the Global Initiative on Out-of-School Children 7

Irina Bokova Anthony Lake

8 Fixing the Broken Promise of Education for All

Findings from the Global Initiative on Out-of-School Children 9

AIDS Acquired immunodeficiency syndrome

10 Fixing the Broken Promise of Education for All

Findings from the Global Initiative on Out-of-School Children 11

Findings from the Global Initiative on Out-of-School Children 13

14 Fixing the Broken Promise of Education for All

Chapter 2 of this report presents the latest global

Findings from the Global Initiative on Out-of-School Children 15

Data on out-of-school children and

Findings from the Global Initiative on Out-of-School Children 17

Lower secondary out-of-school rate

Source: UNESCO Institute for Statistics, August 2014 DataLink: http://dx.doi.org/10.15220/2014/ed/sd/7/f2.1

18 Fixing the Broken Promise of Education for All

Sub−Saharan Africa South Asia Rest of world

Male 7.6 million

Note: The data refer to the regional classification used by UNICEF.

Findings from the Global Initiative on Out-of-School Children 19

Sub−Saharan Africa South Asia Rest of world

16.1 million Male 13.7 million

9.5 million Male 10.2 million

Note: The data refer to the regional classification used by UNICEF.

% Number (in millions) % Number (in millions)

Source: UNESCO Institute for Statistics, August 2014

20 Fixing the Broken Promise of Education for All

In relative terms, 42 countries with data were able

Box 2.1 When is a child considered to be ‘out of school’?

Findings from the Global Initiative on Out-of-School Children 21

MIDDLE EAST AND Yemen 0.5

22 Fixing the Broken Promise of Education for All

Likely to enter Unlikely to ever Million children

WEST AND 14 26 60 18.8

SOUTH ASIA 28 15 57 9.8

Findings from the Global Initiative on Out-of-School Children 23

24 Fixing the Broken Promise of Education for All

Primary school age 1-2 years older 3 or more years older

% Number (million) % Number (million)

Findings from the Global Initiative on Out-of-School Children 25

26 Fixing the Broken Promise of Education for All