CONFERENCE EDITION THE COST OF NOT EDUCATING GIRLS EDUCATING GIRLS AND ENDING CHILD MARRIAGE: A PRIORITY FOR AFRICA (SYNTHESIS BRIEF) QUENTIN WODON, CHATA MALE, CLAUDIO MONTENEGRO, HOA NGUYEN, AND ADENIKE ONAGORUWA NOVEMBER 2018 THE COST OF NOT EDUCATING GIRLS EDUCATING GIRLS AND ENDING CHILD MARRIAGE: A PRIORITY FOR AFRICA (SYNTHESIS BRIEF) QUENTIN WODON, CHATA MALE, CLAUDIO MONTENEGRO, HOA NGUYEN, AND ADENIKE ONAGORUWA BACKGROUND TO THE SERIES This study is part of a series of notes at the World Bank Through lower expected earnings in adulthood and higher on the potential cost of not educating girls globally. fertility over their lifetime, a lack of education for girls Despite substantial progress over the last two decades, leads to higher rates of poverty for households. This is due girls still have on average lower levels of educational to both losses in incomes and higher basic needs from attainment than boys in many countries, especially larger household sizes. Data on subjective perceptions also at the secondary and tertiary levels. As documented suggest that higher educational attainment is associated with by the World Development Report 2018, when it perceptions of higher well-being among women. comes to actual learning, while girls tend to outperform boys in reading, they score lower in math and science Low educational attainment for girls may also weaken tests in many countries. Together with occupational solidarity in communities and reduce women’s participation segregation and social norms that discourage women in society. Lack of education is associated with a lower to take full advantage of labor market opportunities, proclivity to altruistic behaviors, and it curtails women’s voice this leads to large gaps in earnings between men and and agency in the household, at work and in institutions. women. In addition, low educational attainment for Fundamentally, a lack of education disempowers women and girls has potential negative impacts on a wide range girls in ways that deprive them of their basic rights. of other development outcomes not only for the girls themselves, but also for their children, families, At the level of countries, a lack of education for girls can lead communities, and societies. The objective of the series to substantial losses in national wealth. Human capital wealth of notes is to document these potential impacts and is the largest component of the changing wealth of nations, their economic costs. ahead of natural capital (such as oil, minerals, and land) and produced capital (such as factories or infrastructure). By Low educational attainment affects girls’ life trajectories reducing earnings, low educational attainment for girls leads in many ways. Girls dropping out of school early are to losses in human capital wealth and thereby in the assets more likely to marry or have children early, before they base that enables countries to generate future income. Low may be physically and emotionally ready to become educational attainment for girls is also associated with higher wives and mothers. This may affect their own health. population growth given its potential impact on fertility It may also affect that of their children. For example, rates. This may prevent some countries from ushering the children of mothers younger than 18 face higher risks transition that could generate the demographic dividend. of dying by age five and being malnourished. They Finally, low educational attainment for girls may lead to may also do poorly in school. Other risks for girls and less inclusive policy-making and a lower emphasis on public women associated with a lack of education include investments in the social sectors. Overall, the message is intimate partner violence and a lack of decision-making clear: educating girls is not only the right thing to do. It also ability in the household. makes economic and strategic sense for countries to fulfill their development potential. 1 | T HE COST OF NOT EDUCATING GIRLS EDUCATING GIRLS AND ENDING CHILD MARRIAGE: A PRIORITY FOR AFRICA SYNTHESIS BRIEF | NOV 2018 BACKGROUND focus of this study is on the potential impacts and cost of low educational attainment for girls and child marriage This study was prepared ahead of the second African girls’ as opposed to lack of learning while in school (given data summit on ending child marriage organized by the African constraints). Specifically, the study documents associations Union in Ghana. It is adapted from a recent global study on of low educational attainment for girls with six domains the cost of not educating girls, and also relies on previous of interest: (1) earnings and standards of living; (2) child work on the economic impacts of child marriage. While nine marriage and early childbearing; (3) fertility and population in ten girls complete their primary education and three in growth; (4) health, nutrition, and well-being; (5) agency and four complete their lower secondary education globally, the decision-making; and (6) social capital and institutions. proportions are much lower in Africa. In the case of sub- Saharan Africa especially, despite progress over the last two Within the six domains, more than 50 different outcomes decades, just over two thirds of girls complete their primary are considered. For most outcomes, estimates of correlations education, and four in ten complete lower secondary school. are obtained using household survey data for, depending This is in part because the prevalence of child marriage on the indicator, up to 40 countries in sub-Saharan Africa and early childbearing remains very high in the region. In and North Africa. For outcomes measured through addition, even when they remain in school, girls just like boys Demographic and Health Surveys, results are based on suffer the consequences of a learning crisis by which too analysis for a core set of 13 countries. Finally, based on the many children do not acquire the foundational skills that a analysis of the impacts of low educational attainment, child functional education system ought to ensure, especially again marriage, and early childbearing for girls on a wide range of in Africa. other domains, selected costs associated with these impacts are measured. For the measurement of such costs, because More needs to be done to improve educational opportunities this study was prepared for the second summit of the for girls. This requires among others putting an end to African Union on ending child marriage organized in Accra in the practice of child marriage and reducing the risk of November 2018, the focus is on costs associated specifically early childbearing. To make the case for investments in with child marriage. The framework guiding the work is given girls’ education and towards ending child marriage, the in Figure 1. Figure 1: Conceptual Framework Economic Benefits Six Domains of Potential Impacts Development Outcomes Higher EDUCATION FOR GIRLS Earnings Earnings and Standards of Living World Bank twin goals: Child Marriage and Early Childbearing Ending extreme Gains in poverty and Complex Direct Welfare boosting shared Fertility and Population Growth and Indirect prosperity (growth for Health, Nutrition, and Well-being Potential Impacts the bottom 40 percent) Budget Agency and Decision-making Savings Social Capital and Institutions Other Source: Authors Benefits NOV 2018 | THE COST OF NOT EDUCATING GIRLS EDUCATING GIRLS AND ENDING CHILD MARRIAGE: A PRIORITY FOR AFRICA SYNTHESIS BRIEF | 2 KEY RESULTS the early grades, thus hindering the progression of girls to higher educational attainment. But the implication is that FINDINGS ACROSS DOMAINS while primary schooling lays the foundation for future learning, it is essential to enable girls to pursue their • Education matters for all children, but even more education through the secondary level and to ensure that so for girls than boys because of the link for girls learning occurs in order to reap the benefits from more between dropping out of school and marrying or education. This in turn also underscores the importance having children early: Many of the potential impacts of ending child marriage and early childbearing since of education on development outcomes apply to they lead girls to drop out of secondary school. both boys and girls. When a child does not finish secondary school, or does not learn what is needed to FINDINGS BY DOMAIN OF IMPACTS function productively as an adult, potential costs are high for boys and girls alike in terms of lost earnings. • Earnings and standards of living: Women with primary But not educating girls is especially costly in part education (partial or completed) earn 19 to 30 percent because of the close relationships between educational more than those with no education at all depending attainment, child marriage, and early childbearing, on the model used for the estimations (Figure 2). By and the risks that they entail for young mothers and contrast, women with secondary education may expect their children. In addition, occupational segregation to make more than twice as much, and women with by gender between paid and unpaid (housework and tertiary education almost five times as much as those care) work, and between types of employment and with no education. Secondary and tertiary education are sectors, which is in part due to social norms, also leads also associated with higher labor force participation, and to especially high potential costs for girls. Finally, especially full-time work. Finally, women with secondary although this is not discussed in this study, it is also and tertiary education report higher standards of living worth mentioning that girls and women in contexts compared to those with primary education or lower. of fragility and violence are especially vulnerable to For example, women with a secondary education are the consequences of low educational attainment. more likely to state that they have enough money to buy food versus women with primary education or • While primary schooling is necessary, it is not sufficient less. When considering the impact of child marriage because for many indicators, having a primary on earnings for women, estimates suggest that women education does not make a large difference versus who married early may have earnings on average eight having no education at all. The gains associated with percent lower across 12 countries than if they had educational attainment tend to be substantial only with a married after 18. This leads to a loss in national earnings secondary education. This is likely in part a reflection of (including all men and women) of about one percent the failure of schools to deliver learning of basic skills in for those countries, which is far from negligible. Figure 2: Gains in Earnings by Education Level (Versus No Education) 600% 500% 400% 300% 200% 100% 0% Primary Secondary Tertiary No additional controls Without location and sectoral controls Source: Authors. The Figure displays average marginal potential impacts. 3 | T HE COST OF NOT EDUCATING GIRLS EDUCATING GIRLS AND ENDING CHILD MARRIAGE: A PRIORITY FOR AFRICA SYNTHESIS BRIEF | NOV 2018 • Child marriage and early childbearing: Each additional up to three fourths since early childbearing goes hand year of secondary education is associated with lower in hand with child marriage. This also means that when risks of marrying as a child and having a child before age assessing benefits from educating girls at the secondary 18 by about seven percentage points on average (see level, we should include benefits from reducing child Table 1). If universal secondary education were achieved, marriage and early childbearing. By contrast, primary child marriage could be virtually eliminated, and the education is not associated with lower risks of child prevalence of early childbearing could be reduced by marriage and early childbearing in most countries. Table 1: Potential Impact of Educational Attainment on Child Marriage and Early Childbearing Reduction in risk per additional year of secondary education Reduction in risk of child marriage -7.5 Reduction in risk of early childbearing -6.6 Source: Authors based on Demographic and Health Surveys. Note: Estimates based on country-level analysis for 13 African countries. All estimated potential impacts are statistically significant except for one country for early childbearing. • Fertility and population growth: A higher level of modern contraceptive use of a third from the base. education is associated with a substantial reduction If girls were better educated, and if child marriage in total fertility (see Figure 3). As a result, universal were to be drastically reduced thanks to universal secondary education could reduce total fertility by secondary education, annual rates of population a third in 18 developing countries considered for the growth could be reduced substantially, with larger analysis. About two thirds of this potential impact impacts in countries that have not yet achieved the could come from educational attainment itself, and demographic transition. This could generate a large one third from ending child marriage. Universal demographic dividend. Again, the potential impact of secondary education could also lead to an increase in primary education in all these areas is much smaller. Figure 3: Reduction in Total Fertility (Versus No Education) Primary Completed Some Secondary Secondary Completed Higher Education 0% 5% 10% 15% 20% 25% 30% 35% 40% Source: Authors. The Figure displays average marginal potential impacts. NOV 2018 | THE COST OF NOT EDUCATING GIRLS EDUCATING GIRLS AND ENDING CHILD MARRIAGE: A PRIORITY FOR AFRICA SYNTHESIS BRIEF | 4 • Health, nutrition and well-being: Universal secondary half. Reductions in under-five mortality of about one education could increase women’s knowledge of HIV/ fifth versus baseline rates could also be achieved in AIDS and their ability to make decisions about their those countries. With the important exception of own healthcare by one tenth nationally. Women’s under-five mortality for which in a few countries a psychological well-being could also improve and the substantial decrease is observed when mothers have risk of intimate partner violence could decrease. In primary education, the gains from universal primary many countries, a higher level of education for mother education in the area of child health appear once again is associated with a reduction in the risk of stunting to be limited. In the case of child marriage and early for their young children (see Figure 4). As a result, childbearing, gains also tend to be limited nationally in countries where potential impacts are statistically given that few children are born of mothers younger significant, universal secondary education for mothers than 18, but marginal impacts tend to be large. could reduce stunting rates for their children by almost Figure 4: Potential Reduction in Under-five Stunting (Versus Less than Primary Completed) Primary Completed Some Secondary Secondary Completed Higher Education 0% 5% 10% 15% 20% 25% 30% Source: Authors. The Figure displays average marginal potential impacts. • Agency and decision-making: A higher level of Finally, having a secondary education is associated educational attainment is associated with more decision- with higher birth registration in some countries. While making ability for women in their household (see results are not robust across countries, effects are Figure 5). As a result, achieving universal secondary substantial when statistically significant. As with the education could increase by more than one tenth other indicators, while some benefits could result from women’s reported ability to make decisions within universal primary education, they would be smaller. the household, from baseline values. Women with Benefits from ending child marriage in those areas, secondary education report lower satisfaction rates with while occasionally statistically significant, tend not to be basic services than women with no education, which large, and are smaller than for educational attainment. may reflect a more realistic assessment of their quality. Figure 5: Gain in Decision-making Ability (Versus No Education) Primary Completed Some Secondary Secondary Completed Higher Education 0% 2% 4% 6% 8% 10% 12% 14% Source: Authors. The Figure displays average marginal potential impacts. 5 | T HE COST OF NOT EDUCATING GIRLS EDUCATING GIRLS AND ENDING CHILD MARRIAGE: A PRIORITY FOR AFRICA SYNTHESIS BRIEF | NOV 2018 • Social capital and institutions: Achieving universal at US$41 billion for Africa, so the loss in wealth may secondary education could enable more women be equivalent to a few years of international aid. to display altruistic behaviors such as volunteering, donating to charity, and helping strangers, with a • Lost human capital wealth due to lower earnings for change of up to 17 percent from baseline values stunted children: Stunting in early childhood leads (this could be because better educated women have to losses in earnings in adult life. Stunting rates could more opportunities to display altruistic behaviors). A be reduced if child marriage and early childbearing secondary education is also associated with a higher were eliminated, which could generate gains in likelihood for women of reporting being able to rely human capital wealth. The magnitude of those gains is on friends when in need and it could affect how however likely to be smaller than the direct effect on women perceive their countries’ institutions, although women’s earnings, but it is still likely to be substantial. in this specific area more work would be needed to confirm the robustness of those relationships. • Loss in total wealth per capita from population For this set of indicators, the potential gains from growth: Child marriage is associated with higher primary versus no education at all cannot be measured rates of fertility and population growth. This in given data limitations. The effects of child marriage turn reduces levels of total wealth per person, cannot be estimated with the available data. especially in countries that have high population growth. The gains in wealth per capita that could result from lower population growth by ending ECONOMIC COSTS: child marriage could be initially smaller than those estimated for women’s earnings. They are THE CASE OF CHILD estimated at US$ 26 billion in the first year of the elimination of child marriage for 13 countries. Yet MARRIAGE these gains would however be cumulative over time, rivalling just within a few years the losses in wealth • Given that this study was prepared ahead of the from women’s lower earnings due to the impact African Union’s second summit on child marriage, of child marriage on educational attainment. the focus for the analysis of economic costs is on impacts associated with child marriage, as opposed to the (larger) costs from low educational attainment for girls. While only a small set of economic costs are estimated, the costs are clearly large. Estimates are in terms of losses in human capital wealth defined as the value of the future earnings of the labor force. • Lost human capital wealth due to lower earnings for women: Lower earnings for women in adulthood due to child marriage lead to losses in human capital wealth defined as the present value of the future earnings of the labor force. The loss in human capital wealth incurred today because women were married early in their youth is estimated at US$63 billion for 12 countries that account for half of the continent’s population. For the continent overall, simple extrapolation suggest that losses could be twice as large. For perspective, annual total net Official Development Assistance (ODA), which consists of disbursements of loans made on concessional terms net of repayments of principal, was estimated in 2016 NOV 2018 | THE COST OF NOT EDUCATING GIRLS EDUCATING GIRLS AND ENDING CHILD MARRIAGE: A PRIORITY FOR AFRICA SYNTHESIS BRIEF | 6 SOLUTIONS TO END • Targeted interventions to reach especially vulnerable girls: The literature suggests that targeted interventions CHILD MARRIAGE especially in the forms of incentives to keep girls in school may have large benefits. Three types of • Again, given the aim of this study to inform the African interventions were recently reviewed on the basis Union’s second summit on child marriage, the focus of the available evidence from experimental or for discussing policies is on interventions needed to quasi-experimental studies: (1) There is a need for end child marriage. Multiple interventions are likely to interventions to expand economic opportunities be needed to end child marriage, but since keeping girls for adolescent girls who dropped out of school and in school is key, improving education opportunities for who are unlikely to be able to return; (2) Imparting girls should be a priority. Typically, various interventions adolescent girls with life skills and reproductive health aiming to end child marriage and early childbearing knowledge is also important, whether girls are in school and improve opportunities for girls tend to be managed or out of school. Evidence suggests that safe space by different Ministries, including ntot only Ministries clubs where girls may discuss issues of sexual and of Education, but also Ministries of Health, Ministries reproductive health as well as other topics with female of Population, Ministries of Labor, and Ministries of mentors may be an effective means of achieving this. Gender or Women and Children’s Affairs. Strategies (3) However, according to the literature, the most may be defined at the Ministry level, or through effective targeted interventions to delay marriage and inter-ministerial committees. Yet since keeping girls childbearing are those that enable girls to remain in in school is essential to end child marriage and early school, especially through incentives offsetting the childbearing, providing education opportunities for out-of-pocket and opportunity costs of schooling. girls is especially important. In practice, three-pronged strategies are likely needed: (1) General basic conditions • Efforts to change gender-based social norms: Child must be met for access to education and learning; marriage, early childbearing, and low educational (2) Targeted interventions must be implemented attainment for girls are rooted in social norms to reach vulnerable girls; and (3) Efforts must be that perpetuate gender inequality. To tackle this undertaken to change gender-based social norms. challenge, beyond general conditions that education systems should meet and targeted interventions • General conditions for access to education and learning. to reach vulnerable girls, additional community- In many countries, there is a need to build secondary based interventions that involve all members of the schools closer to where girls live or provide modes of community may be an effective means of changing transportation and in some cases boarding to enable these norms. Such interventions should target men them to attend schools, especially at the secondary and community leaders apart from women. Finally, level. Providing adequate water, sanitation and hygiene adequate laws – for example on the minimum facilities for girls is also important, as is the need to age for marriage without exceptions for parental address the risk of violence and sexual harassment and judicial consent, are also essential, but often either at or en route to school. It is also essential to not sufficient on their own to achieve change. ensure that schools improve learning outcomes and provide girls with appropriate skills. Among various entry points that can be used to that end, the following SUMMARY OF can be mentioned (1) reducing disadvantages that girls face in remote communities due in part to poor KEY FINDINGS targeting of Government resources; (2) creating a more inclusive school culture for girls; (3) providing girls Table 1 below provides the main estimated potential with role models—including through female teachers; impacts by domain. Potential impacts are summarized by and (4) raising the returns to secondary education for showing gains from a secondary education in comparison women at the local level through better employment to no education at all, factoring in the benefits from the opportunities. This list is by no means exhaustive and virtual elimination of child marriage that would follow if the appropriate entry points vary between countries. all girls completed their secondary schooling. In most 7 | T HE COST OF NOT EDUCATING GIRLS EDUCATING GIRLS AND ENDING CHILD MARRIAGE: A PRIORITY FOR AFRICA SYNTHESIS BRIEF | NOV 2018 cases, potential impacts are estimated for the completion education. This is also the case for the gains associated with of secondary school, but in some cases the potential the elimination of child marriage since they are assumed to impacts are for both partial and completed secondary follow from universal secondary education. It should again school combined. In virtually all cases, estimates of the be emphasized that what is measured is associations, not potential impacts of low educational attainment for girls – necessarily causal impacts. In addition, for some indicators, or equivalently of gains associated with higher educational especially in the case of agency and decision-making, attainment as captured by secondary education, are large. and social capital and institutions, the data often pertain As documented in more detailed in the study, most gains to reported behaviors and perceptions, thereby making are associated with secondary as opposed to primary interpretation more tentative. Table 1: Selected Potential Benefits from Universal Secondary Education for Girls Domain Estimated Potential impacts Earnings and standards of living Expected earnings in adulthood more than doubled Increase in labor force participation or working full time by up to one tenth Gain in perceptions of standards of living of up to one tenth Child marriage and early childbearing Virtual elimination of child marriage Reduction in early childbearing by up to three fourths Fertility and population growth Reduction in total fertility by about one third Increase in contraceptive use by a third from base Reduction in population growth by 0.6 percentage point Health, nutrition and well-being Increase in women’s knowledge of HIV/AIDS by one tenth Increase in women’s decision-making ability for health by more than a fourth Increase in women’s psychological well-being Reduction in under-five mortality rate by up a fifth Reduction in under-five stunting rate by almost half Agency and decision-making Women more likely to be able to make decisions in the household Women possibly more likely to better assess the quality of basic services Increase in likelihood of birth registration by one third Social capital and institutions Women more likely to report altruistic behaviors Women more likely to report ability to rely on friends when in need Women possibly more likely to be more able to assess institutions/leaders Source: Authors. Table 2 provides the valuation of the two potential impacts in the tens of billions of trillions of dollars. These estimates for which a monetary cost is computed assuming the are only orders of magnitude since they depend on models elimination of child marriage. The analysis focuses on losses in and assumptions, but they demonstrate that the potential human capital wealth related to (i) lost earnings for women; cost of child marriage is high for the girls and societies overall, and (ii) higher fertility rates and thereby annual population and the potential cost of not educating girls would be even growth rates due to child marriage. The estimated costs run larger. Table 2: Selected Economic Costs of Child Marriage, Selected Countries Estimates costs Value Loss in human capital wealth due to US$ 63 billion lower earnings (12 countries) Loss in welfare (total wealth per capita) due to high population US$ 26 billion loss in first year, cumulative and rising over time growth (13 countries) Source: Authors. NOV 2018 | THE COST OF NOT EDUCATING GIRLS EDUCATING GIRLS AND ENDING CHILD MARRIAGE: A PRIORITY FOR AFRICA SYNTHESIS BRIEF | 8 To conclude, low educational attainment for girls and child As with any empirical work of this nature, estimates of marriage can have pervasive potential impacts ranging from potential impacts and costs are subject to two important lower earnings and standards of living to lower psychological caveats. First, estimates from available observational well-being and agency for girls and women. Possibly in part data do not permit establishing causal relationships. Thus, because educational investments at the secondary level when referring to potential ‘impacts’, the analysis should provide an option value to continue investing to acquire be taken as only suggestive of what might be achieved further skills later in life, the benefits from education are with higher educational attainment for girls and women, much larger at the secondary than at the primary level. a reduction in child marriage and early childbearing, and This is even more the case when considering that achieving related policy changes. What is measured are associations universal secondary school completion for girls could virtually between educational attainment and child marriage and eliminate the risk of child marriage. Countries need to other development outcomes. For several of the outcomes ensure that child marriage is eliminated and that all girls can considered, whether these associations reflect casual go to school and acquire foundational cognitive and socio- relationships can be corroborated by evidence from existing emotional skills while in school. While the public and private empirical studies that are able to more credibly establish cost of ending child marriage and providing universal quality causality. But for other outcomes, fewer such studies are primary and secondary education for all girls could be far available. Second, simulations of the benefits of increasing from negligible, the potential returns to this investment could girls’ education or ending child marriage obtained from the be much larger. Increasing investments in girls’ education estimates of potential impacts do not account for broader clearly makes economic sense. It is also the right thing to do. effects in the economy arising from an expansion in the number of better educated girls or women, or a sharp reduction in child marriage. The economics literature ANNEX: CONTRIBUTION suggests that these effects can be sizable, particularly lowering the overall returns to schooling in the labor market AND LIMITS OF THE STUDY in the case of higher educational attainment. Thus, estimates only provide orders of magnitude of potential impacts and This note summarizes findings from a research program at costs, not precise values of ultimate potential impacts taking the World Bank to document the potential negative impacts into account general equilibrium effects. of low educational attainment for girls and child marriage, and some of the related economic costs. The fact that investing in girls is essential for development is not news. The contribution of this study is to illustrate the potential negative effects of not investing in girls in a slightly more comprehensive way, with more recent survey data, and for a larger set of countries than done so far. By pulling together evidence on the associations between low educational attainment and child marriage for girls and multiple socio- economic domains in many countries, the analysis can help foster greater mobilization towards girls’ education and ending child marriage. 9 | T HE COST OF NOT EDUCATING GIRLS EDUCATING GIRLS AND ENDING CHILD MARRIAGE: A PRIORITY FOR AFRICA SYNTHESIS BRIEF | NOV 2018 Recommended citation for this study: Wodon, Q., C. Montenegro, H. Nguyen, and A. Onagoruwa. 2018, Educating Girls and Ending Child Marriage: A Priority for Africa – Synthesis Brief. The Cost of Not Educating Girls Notes Series. Washington, DC: The World Bank. The full study is available on the World Bank website. The study was prepared by a team at the World Bank with support from the Children’s Investment Fund Foundation and the Global Partnership for Education ahead of the African Union’s Second African Girls’ Summit on Ending Child Marriage in Africa. The work builds on previous studies at the World Bank on the economic impacts of child marriage (jointly with ICRW), the changing wealth of nations, and the cost of gender inequality in earnings. The authors are grateful to Rafael Cortez, Bénédicte de la Brière, Lucia Hanmer, Qaiser Khan, Oni Lusk-Stover, Sameera Maziad Al Tuwaijri, Harry Patrinos, and Jeffrey Waite for valuable peer review comments on this study for Africa and the previous global study on the cost of not educating girls from which part of this study is adapted, and to Omar Arias and Meskerem Mulatu for additional comments and advice. Luis Benveniste provided additional guidance. Weight Creative formatted the study for dissemination. The team is also grateful to Erin McCarthy and Linda Weisert at the Children’s Investment Fund Foundation and Louise Banham and Jane Elizabeth Davies at the Global Partnership for Education for support. The findings, interpretations, and conclusions expressed in this study are entirely those of the authors and should not be attributed in any manner to the World Bank, its affiliated organizations or members of its Board of Executive Directors or the countries they represent. Citation and the use of material presented in this study should take into account its provisional character. The World Bank does not guarantee the accuracy of the data included in this work. Information contained in this study may be freely reproduced, published or otherwise used for noncommercial purposes without permission from the World Bank. However, the World Bank requests that the original study be cited as the source. © 2018 The World Bank, Washington, DC 20433. NOV 2018 | THE COST OF NOT EDUCATING GIRLS EDUCATING GIRLS AND ENDING CHILD MARRIAGE: A PRIORITY FOR AFRICA SYNTHESIS BRIEF | 10 THE COST OF NOT EDUCATING GIRLS EDUCATING GIRLS AND ENDING CHILD MARRIAGE: A PRIORITY FOR AFRICA SYNTHESIS BRIEF