89989 Preventing Child Marriage: Lessons from World Bank Group Gender Impact Evaluations Jennifer Parsons and Jennifer McCleary-Sills Gender Group, World Bank Group Synopsis Child marriage is closely linked to poverty and girls’ educational opportunities. Across 18 of the 20 countries with the highest prevalence of child marriage, girls with no education are up to six times more likely to marry than girls with a secondary education. Girls living in poorer households are almost twice as likely to marry before the age of 18, compared with girls in higher-income households. Drawing on results from seven impact evaluations (IEs) supported by the World Bank Group (WBG), this brief highlights lessons learned about what works to delay age of marriage and reduce child marriage. Effective programs will ultimately increase girls’ educational attainment, increase girls’ value and opportunities in society, and reduce poverty. Promising interventions are beginning to address the drivers of child marriage. However, more rigorous and long-term impact evaluations are needed to better understand intervention components and delivery mechanisms that are effective in preventing child marriage. Background Child marriage is a fundamental barrier to the achievement of international commitments for development and gender equality. In developing countries, one-third of girls are married before age 18, and one in nine is married before age 15 (UNICEF, 2013). These figures include girls in formal marriages and living in informal unions, which are referred to together as “child marriage” (UNICEF, 2005). If present trends continue, more than 142 million girls will be married before the age of 18 Source: World Bank Group 2014. in the next decade, or 39,000 girls each day (UNFPA, 2012). Such early marriages remain and 75 percent in Niger and Bangladesh, pervasive despite legal prohibition in many respectively. While child marriage prevalence countries. As shown in this map, across the in India is not among the highest, the sheer 111 countries for which data are available, the size of its population means that India has the prevalence of child marriage ranges from 2 highest number, accounting for one-third of the percent in Algeria and Libya to as high as 74 world’s child brides (World Bank Group, 2014). Reducing Child Marriage: Lessons from World Bank Group Gender Impact Evaluations 1 Average regional prevalence ranges from 21 or to save on the bride price—the younger the percent in Africa to 17 percent in Latin America bride, the lower the dowry her family has to and the Caribbean, to 8 percent in Eastern and pay. Families often perceive marriage as a way Southern Europe (Raj and Boehmer, 2013). to provide for their daughter’s future, believing However, these averages can mask it will improve her economic and social considerable variation across and within circumstances. The reality, however, is that countries. girls who marry young are more likely to remain poor even after marriage (WBG, 2014; In addition to violating the rights of millions of Mathur et al., 2003). The girls most likely to girls every year, child marriage brings marry early are those with the least education significant development impacts at the and lowest economic status (Raj, 2010). Given individual, community, and society levels. the connection between poverty and access to Girls who marry before the age of 18 are at education, girls in poor households face greater risk of poor health outcomes, including overlapping disadvantages and greater risk of HIV and other sexually transmitted infections, being married before age 18. early pregnancy, and early childbearing (Hindin and Fatusi, 2009). In developing countries, pregnancy-related causes are the largest contributor to mortality among girls ages 15 to 19, killing nearly 70,000 girls each year (WBG, 2014; UNFPA, 2013). Early childbearing also increases the risk of complications such as obstetric fistula, which commonly occurs among girls who give birth before their bodies are physically mature. This can cause chronic incontinence, often resulting in stigmatization and social exclusion (WBG, 2014). In addition to the effect on a girl’s sexual and reproductive health, marrying before age 18 increases the odds of experiencing intimate partner violence by 22 percent (WBG, 2014). The lack of power Where girls have greater educational and that young brides experience in their economic opportunities, they are more likely to relationships is often exacerbated by large age take up those opportunities than to have differences between them and their husbands, children in their teenage years (WBG, 2014). which further constrain their voice and agency Lack of education can be both a risk factor for within marriage (UNICEF, 2005; WBG, 2014). and an outcome of child marriage. Evidence from Bangladesh and Sub-Saharan Africa Risk Factors for Early Marriage suggests that women who married early are High prevalence of child marriage is both a over five percentage points less likely to be symptom and a consequence of poverty. The literate and over eight percentage points less practice is driven by social norms and likely to have any secondary education expectations and by gendered discrimination (Nguyen and Wodon, 2013; Field and Ambrus, that devalues women and girls and their right 2008). A study across 18 of the 20 countries to make choices for themselves. It is also with the highest prevalence of child marriage driven by limited choices for poor families. Girls found that girls with no education were up to living in poor households are almost twice as six times more likely to marry young compared likely to be married before the age of 18 with girls who had received secondary compared with girls in higher-income education (ICRW, 2006). Across Africa, each households, as are rural girls compared with additional year a girl is married before age 18 those from urban areas (WBG, 2014). Girls are reduces her probability of literacy by about six often married off by their families to pay debts percentage points, probability of having at least Reducing Child Marriage: Lessons from World Bank Group Gender Impact Evaluations 2 some secondary schooling by eight percentage attendance rates. In the Zomba program in points, and probability of secondary school Malawi, girls received US$1-5 per month, in completion by almost seven percentage points addition to school fees, if their monthly school (Wodon, 2014). We see from this evidence that attendance rate was above 80%. The child marriage is closely associated with unconditional cash transfer (UCT) arm in the illiteracy and lower levels of education, and Zomba program offered identical cash that more schooling seems to raise the age of transfers but without a school attendance marriage. requirement. Evidence from the World Bank Group Additional Educational Components Two of the six IEs looked at CCT programs In recent years, the World Bank Group has with additional education components that increased investments in gender-informed supplemented the cash incentive for sending programs, research, and impact evaluation. girls to school. In addition to cash based on At the same time, there has been greater school attendance, one program (Bandiera et programmatic effort to reduce child marriage al., 2012) included girls clubs, vocational and across the globe, with programs increasingly life skills courses, mentors, and recreation addressing poverty, education, and lack of activities as a part of the intervention, and a economic opportunities as drivers of child second (Khandker et al., 2003) included a marriage. From the 161 IE papers within the subsidy for school uniforms, school curriculum enGENDER IMPACT (eGI) database, we reform and infrastructure improvements. identified those measuring programmatic impact on child marriage. Age at marriage is an outcome measured across seven of these eGI evaluations. The programs evaluated were overwhelmingly designed with the aim of increasing educational attainment for girls, with increasing age at marriage as a secondary goal. This brief explores strategies and program components to identify intervention strategies that are effective in delaying marriage. (See Table 1 for a summary of the included IEs.) Summary of Programs and Evaluations The third IE in this category evaluated a Conditional Cash Transfer Programs school-based reproductive health curriculum Three of the seven IEs evaluated conditional intervention in Kenya, specifically focusing on cash transfer (CCT) programs that were “cash the arm of the intervention that reduced the for attendance” interventions (Baez et al., cost of education with the aim of keeping 2011; Baird et al., 2011; Baird et al., 2010), children in school longer (Duflo et al., 2006). with one of those IEs comparing the conditional versus unconditional arms of an Structural Interventions intervention (Baird et al., 2011). The final IE in this analysis examined structural factors in labor market supply and demand and Programs offered financial incentives, their impact on child marriage in Bangladesh conditional on girls’ attendance in school. In a (Heath and Mobarak, 2012). Analysis program for girls in Pakistan, girls received a compared girls in villages with garment stipend of approximately US$3.50 per month if factories versus those in villages without they were enrolled in grades six through eight factories. of government schools and had at least 80% Reducing Child Marriage: Lessons from World Bank Group Gender Impact Evaluations 3 So What Works? Implications for Future incorporated the following components into Programming programming in addition to the cash transfer: • assigning mentors for girls Multiple strategies were used across the • providing recreational activities and evaluated programs to improve girls’ education vocational skills for girls and delay marriage. Nearly all interventions • improving education facilities evaluated included a cash transfer, and • providing subsidies for uniforms. additional educational and structural components were used in half of the IEs. Here we examine which programmatic components appear to have the greatest effect on reducing child marriage. Conditional Cash Transfer Programs In all the CCT programs, significant decreases occurred in early marriage, measured by the proportion of girls in the treatment group married by age 18 at the end of the intervention compared with girls in the control group. In the IE of the Zomba program in Malawi (Baird et al., 2011), there were substantial and statistically significant delays in Structural Interventions marriage and childbearing among girls in the In one IE that examined the impact of available UCT compared to those in the control group. garment factory jobs and the connection to The observed impact among girls in the CCT education level and age at marriage among arm was smaller and not significant. The girls (Heath and Mobarak, 2012), household largest impact among girls in the UCT arm was surveys assessed whether a girl was married among those who had dropped out of school before age 16 or 18. In villages where garment after the program began, which suggests that factories were present, girls were statistically the decision to delay marriage is influenced not significantly less likely to be married by age 16 only by increased educational attainment, but and age 18 compared to girls in villages also by the economic needs and stability of a without factories, suggesting opportunities for girl and her family. employment provide alternatives to early marriage for girls in these areas. Additional Educational Components All these IEs reported significant decreases in child marriage among those who were a part of Components of Effective Programming the interventions. Among those who received The findings from these seven IEs resonate education at a reduced cost, a delay was seen with evaluations of other child marriage in the age at marriage, compared with those prevention efforts. Here we look at two for whom the cost of education was not offset examples of programs in India and Ethiopia. In through program participation. 1994, the Haryana State Government in northern India launched Apni Beti Apna Dhan IEs of programs that were found to delay (ABAD, or “Our Daughter, Our Wealth”), a marriage address multiple drivers of early program that gave poor families an incentive to marriage, including poverty, access to keep their daughters in school and unmarried education, and limited livelihood opportunities. until age 18 and sought to improve family and Cash transfer programs were shown to have community perceptions about the value of girls. positive impacts across all IEs. Also effective An impact evaluation showed that girls in the were extended CCT programs that treatment group were significantly more likely to be in school and unmarried at age 18 than Reducing Child Marriage: Lessons from World Bank Group Gender Impact Evaluations 4 girls in the control group, although the impact Limitations on attitudinal changes about the value of girls While our findings are consistent with other are less clear (Nanda et al., 2014; WBG, impact evaluations of what works to prevent 2014). child marriage, there are some limitations to In the Awi zone of Ethiopia, in the rural Amhara this analysis. The universe of IEs included in Region, one in five girls is married before her this analysis is limited to only those carried out 15th birthday and nearly half (44 percent) are or supported by the World Bank Group and married by age 17. Virtually all (92 percent) of subsequently included in the enGENDER these marriages are arranged by their families. IMPACT database. The IEs selected for this In 2004, the regional government partnered database apply a gender lens to either with the Population Council on a two-year programming content or the data analysis and project to delay marriage and keep girls in evaluation process. We therefore acknowledge school. This program, Berhane Hewan (“Light that this database is not inclusive of all existing evaluations of strategies to prevent child for Eve”), provided families with cash conditioned on their daughters’ remaining marriage. In addition, much of the evidence unmarried and in school for the duration of the linking education to child marriage comes from program. Other elements included social cross-sectional survey data, so attributing mobilization of girls led by female mentors, causality is difficult. provision of school supplies, livelihood training for out-of-school girls, and “community conversations” on early marriage and reproductive health. Two years later, girls (ages 10 to 14) enrolled in the program were only one-tenth as likely to be married compared with girls in the control group. Participants were also three times more likely to be in school compared with girls in the control group (Erulkar and Muthengi, 2009; WBG, 2014). Components of these two programs are consistent with the key intervention Conclusion components seen in the seven IEs included here: CCTs, teaching life and vocational skills, Evidence points to a number of drivers of early providing mentors, increasing access to free marriage, including poverty, social norms, and education, engaging in health awareness gender-based discrimination. More longitudinal conversations with girls and their communities, data analysis and continued impact and addressing the underlying economic evaluations of programs are necessary to drivers of child marriage. From the initial determine the direct and indirect long-term success of the programs in Ethiopia and India effects that programs have on delaying as well as results from the seven eGI marriage. Based on these eGI evaluations, and evaluations, it is evident that the drivers of supported by evidence from other evaluations, child marriage—poverty, limited opportunities it is clear that programs and strategies must for educational attainment and vocational address multiple social and economic drivers training, and the low perceived value of girls in to delay marriage. It is also critical to provide society—must be addressed simultaneously to individual and family-level resources within reduce the number and proportion of girls interventions. Families make decisions about marrying before age 18. their daughters’ marriages within the context of social norms, financial constraints, and labor market factors. The practice of child marriage Reducing Child Marriage: Lessons from World Bank Group Gender Impact Evaluations 5 also continues in many countries despite laws education and financial opportunity can either regulate age at marriage. directly or indirectly delay marriage. In addition to promoting programmatic strategies that increase educational attainment and economic opportunities for girls, preventing child marriage requires strengthened legal and policy frameworks, to ensure increased awareness and greater enforcement of existing laws. Early marriage is a complex public health issue that impacts girls’ and women’s health, educational attainment, and economic opportunities as well as broader development objectives for communities and countries. Programs and strategies that increase girls’ access to enGENDER IMPACT enGENDER IMPACT is an online gateway for Gender-Related Impact Evaluations. At www.worldbank.org/engenderimpact you will find profiles summarizing key information about World Bank Group funded gender-related impact evaluations. These profiles are organized around priority areas for policy action, including: reducing health disparities, shrinking education and skills gaps, increasing economic opportunities, boosting voice and agency, and addressing gender-based violence. enGENDER IMPACT aims to share knowledge from previous evaluations and encourage more and better evaluations in key gender topics. Reducing Child Marriage: Lessons from World Bank Group Gender Impact Evaluations 6 References: Baez, J. E., Alam, A., and Del Carpio, X. V. (2011). Does Cash for School Influence Young Women’s Behavior in the Longer Term? Evidence from Pakistan. Baird, S., McIntosh, C., and Özler, B. (2011). Cash or condition? Evidence from a cash transfer experiment. The Quarterly Journal of Economics, 126(4), 1709-1753. Baird, S., Chirwa, E., McIntosh, C., and Özler, B. (2010). The short-term impacts of a schooling conditional cash transfer program on the sexual behavior of young women. Health Economics, 19(S1), 55-68. Bandiera, O., Buehren, N., Burgess, R., Goldstein, M., Gulesci, S., Rasul, I., and Sulaiman, M. (2012) Empowering Adolescent Girls: Evidence from a Randomized Control Trial in Uganda. Duflo, E., Dupas, P., Kremer, M., and Sinei, S. (2006). Education and HIV/AIDS prevention: evidence from a randomized evaluation in Western Kenya. World Bank Policy Research Working Paper. Erulkar, A. and Muthengi, E. (2009). Evaluation of Berhane Hewan: A Program to Delay Marriage In Rural Ethiopia, International Perspectives on Sexual and Reproductive Health. 35(1): 6–14. Field, E. and Ambrus, A. (2008). Early Marriage, Age of Menarche, and Female Schooling Attainment in Bangladesh. Journal of Political Economy 116(5): 881–930. Heath, R., and Mobarak, A., (2012) "Does Demand or Supply Constrain Investment in Education? Evidence from Garment Sector Jobs in Bangladesh”. Hindin, M. and Fatusi, A. (2009). Adolescent Sexual and Reproductive Health in Developing Countries, International Perspective on Sexual and Reproductive Health. 35(2): 58–62. International Center for Research on Women (ICRW). (2006), Too Young to Wed: Education & Action Toward Ending Child Marriage, Brief on Child Marriage and Domestic Violence. Khandker, S., Pitt, M., and Fuwa, N. (2003). Subsidy to promote girls' secondary education: the female stipend program in Bangladesh. Mathur, S., Greene, M., and Malhotra, A. (2003). Too Young to Wed: Education and Action Toward Ending Child Marriage. Washington, DC: International Center for Research on Women. Nanda, P., Datta, N., and Das, P. (2014). Impact on Marriage: Program Assessment of Conditional Cash Transfers. Nguyen, MC and Wodon, Q. (2013). Estimating the Impact of Child Marriage on Literacy and Education Attainment in Africa. Washington, DC: World Bank. Raj, A. (2010). When the Mother Is a Child: The Impact of Child Marriage on Health and Human Rights of Girls, Archives of Diseases in Childhood 95: 931–35. Raj, A. and Boehmer, U. (2013). Girl Child Marriage and Its Association with National Rates of HIV, Maternal Health, and Infant Mortality Across 97 Countries. Violence Against Women 19: 536– 51. World Bank Group (2014). Voice and Agency: Empowering Women and Girls for Shared Prosperity. UNFPA, State of the World’s Population 2013. UNICEF, (2005). Early Marriage: A Harmful Traditional Practice. UNICEF, State of the World’s Children (New York: UNICEF, 2013). http://www.unicef.org/sowc2013/ UNFPA, Marrying Too Young: End Child Marriage (New York: UNFPA, 2012). Wodon, Q. Child Marriage and Education. Washington, DC: World Bank, forthcoming. Reducing Child Marriage: Lessons from World Bank Group Gender Impact Evaluations 7 Photo Credits: 1. Map: The World Bank Group (2014). Voice and Agency: Empowering Women and Girls for Shared Prosperity, page 103. The report can be downloaded here: http://www.worldbank.org/en/topic/gender/publication/voice-and-agency-empowering-women-and-girls- for-shared-prosperity 2. Students at Al Shami Girls Secondary School. The school, established in 2004, is very important for the education of the girls in the Beit Our Titha community near Ramallah. Photo: Arne Hoel 3. A computer class at the Female Experimental High School in Herat. The school is benefitting from the Education Quality Improvement Program (EQUIP)whose objective is to increase access to quality basic education, especially for girls. Photo: Graham Crouch 4. Life for students at Tailulu in the Tongan capital Nukua'lofa, is set to change with the arrival of high- speed broadband internet in Tonga. Photo: Tom Perry 5. History teacher Nicoleta Caraiman (center) helps her students during class exercise at the Scoala Frumusani School in the village of Frumusani, Romania. Photo: Dominic Chavez 6. High school students in La Ceja, Department of Antioquía, Colombia. Photo: Charlotte Kesl Reducing Child Marriage: Lessons from World Bank Group Gender Impact Evaluations 8 Table 1: Summary of Education outcomes measured by WBG Gender IEs Programmatic Approach Conditional Unconditional Additional School Labor Cash Cash Educational Based Supply Transfer Transfer Components SRH and Demand Impact Impact Evaluation Citation Does Cash for School Influence Young Women's Baez et  Behavior in the Longer Term? Evidence from Pakistan al., (2011) The Short-Term Impacts of a Schooling Conditional Baird et  Cash Transfer Program on the Sexual Behavior of al., (2010) Young Women in Malawi Cash or Condition Evidence from a Cash Transfer Baird et   Experiment al., (2011) Subsidy to Promote Girls' Secondary Education: The Khander   Female Stipend Program in Bangladesh et al., (2003) Empowering Adolescent Girls: Evidence from a Bandiera,   Randomized Control Trial in Uganda et al., 2012 Education and HIV/AIDS Prevention: Evidence from a Duflo et   randomized evaluation in Western Kenya al., (2006) Does Demand or Supply Constrain Investments in Heath and  Education? Evidence from Garment Sector Jobs in Mobarak, Bangladesh (2012) Reducing Child Marriage: Lessons from World Bank Group Gender Impact Evaluations 9