87854 Gender-Based Violence Prevention: Lessons from World Bank Impact Evaluations April 30, 2014 Sveinung Kiplesund and Matthew Morton Gender and Development, World Bank Gender-based violence (GBV) is widespread, with adverse repercussions for individual wellbeing and development more broadly. Yet, knowledge on what works to prevent GBV is limited. Drawing on lessons from three impact evaluations supported by the World Bank, this brief highlights opportunities for addressing GBV in economic empowerment interventions. Interventions that address gender norms and engage men have shown some promise from a small number of World Bank evaluations. It is, however, clear that more rigorous impact evaluations are needed to better understand how to effectively prevent GBV, especially in developing countries. Gender-based violence (GBV) is widespread and and further the World Bank’s twin goals of eliminating 1 devastating. It takes many different forms, including extreme poverty and boosting shared prosperity. rape, physical abuse, workplace harassment, honor killings, child marriage, economic violence and female genital mutilation/cutting, among others. A recent study by the World Health Organization found that over 35 In recent years, the Bank’s portfolio of work on GBV percent of women worldwide have experienced physical has expanded. Yet, we only identified three impact or sexual intimate partner violence (IPV) or non-partner evaluations measuring GBV as an outcome, and, sexual violence (Garcia-Moreno et al., 2013). With among these, only one (Gupta et al., 2013) addressed nearly one-third of (ever-partnered) women GBV as a primary programmatic focus. Two additional experiencing violence by a partner, IPV is the most impact evaluations have looked at related areas such common form of violence women experience. The as the impact of psychotherapy on survivors’ wellbeing impacts of GBV extend beyond the individual survivor, in the Democratic Republic of Congo (Bass et al., 2013) affecting children, households, workplaces, and and the role of messages and information about HIV communities. Studies have documented large risk in changing sexual behaviors of youth in Kenya development impacts. Conservative estimates from a (Dupas, 2011). However, these are not discussed here number of countries put the economic costs of lost as they do not measure GBV exposure or perpetration. productivity due to intimate partner violence between 1 and 2 percent of GDP—more than what most governments spend on primary education (1.5 percent) (Duvvury et al., 2013). Designing and implementing The table in Annex 1 gives a brief overview of key projects that help prevent GBV are critical. They can features of the three interventions. The Women’s improve the lives of millions of women and their families Income Generating Support (WINGS) Program (Blattman et al., 2013) involved a cash transfer and business skills intervention aimed primarily at vulnerable women in Northern Uganda. The core 1 While this brief largely focuses on GBV against women, it intervention included regular supervision and advice on should be noted that many survivors of GBV are male. Gay, business activities. The program had no impact on bisexual and transgendered individuals are also often targeted. Gender-based Violence Prevention: Lessons from World Bank Impact Evaluations 1 intimate partner violence and no statistically significant Dialogue Groups regularly, physical IPV was results for other empowerment measures such as significantly reduced. women’s independence and status in the community. The intervention did expand women’s economic The study in Cote d’Ivoire was not designed to measure opportunities, with monthly cash earnings doubling and the economic benefits of the loans scheme, although savings tripling, but this economic empowerment did pre- and post-tests suggest that there were a number of not appear to translate to reductions in violence positive economic impacts. This is in line with other exposure during the evaluation period. In one of the trials, which have found VSLA to have an overall treatment arms, male partners also received business positive impact on household savings, assets and skills training in order to facilitate more cooperation consumption expenditure. (Annan et al., 2013 and Ksoll between partners. This component had a small and et al., 2013). significant positive effect on couples’ communication and relationships, but no observed impact on partner Emerging Policy Lessons violence. The Empowerment and Livelihood for Adolescents (ELA) program was also implemented in Uganda Programs to increase women’s economic (Bandiera et al. 2012). The program delivered empowerment may reduce GBV by increasing women’s "Adolescent Development Clubs," which provided safe bargaining power and ability to leave abusive places for positive social interactions along with relationships, though there are also risks that increased vocational and life skills to girls aged 14-20. The clubs empowerment could threaten household patriarchies were led by female mentors who taught courses on and exacerbate violence. The three interventions each income generation with a focus on micro-enterprises. had components aimed at increasing women’s Life skills courses covered topics such as sexual and economic opportunities. Bandiera et al. (2012) found reproductive health, family planning, rape, conflict that the ELA program positively impacted young resolution and leadership. The study found that the women’s likelihood of engaging in income-generating share of girls reporting sex against their will dropped activities as well as significantly reducing the share of from 21 percent to almost zero. Furthermore, the adolescent girls reporting having sex unwillingly. program significantly increased entrepreneurial skills Blattman et al. (2013), on the other hand, found no and participation in income-generating activities, in reduction in intimate partner violence, but large addition to a number of positive outcomes related to increases in women’s income and wealth. These mixed condom use, teen pregnancies rates and knowledge findings are consistent with the results of a recent about risky behaviors (HIV and pregnancy knowledge). systematic review: five studies found protective associations and six documented a risk association The third intervention took place in Cote d’Ivoire between women’s involvement in income generation through women-only Village Savings and Loans and experience of past year violence (Vyas & Watts, Associations (VSLA) and Gender Dialogue Groups 2009). Protective elements can be effectively built into (GDG) (Gupta et al., 2013). The intervention was women’s empowerment interventions, as was done in aimed at female VSLA participants and their male Cote d’Ivoire where gender dialogue groups partners. VSLA is a methodology designed by CARE supplemented savings groups to address norms, International in the 1990s to increase saving attitudes, and relationship dynamics. opportunities and capital acquisition through the creation of groups. In this intervention, the VSLA provided simple savings and loan facilities. In addition to the standard VSLA model, half of the groups also Gender norms and attitudes are shaped by history and included a “Gender Dialogue Group.� The curriculum of context, and are often resistant to change. Social norms these Gender Dialogue Groups was designed to help about gender roles and violence can influence the risk participants (both male and female) discuss norms and of experiencing GBV. Since gendered expectations and attitudes regarding financial decisions, the value of limitations on agency can establish a pattern of women in the household, gender equality and the use subordination and voicelessness, changing norms can of violence. Adding the Gender Dialogue Groups to the be an important part of a comprehensive approach to VSLA program showed statistically significant increases violence prevention. The ELA program in Uganda had in control over household economic resources. positive effects on attitudes related to adolescent girls’ Physical, sexual, and emotional IPV also decreased, desired age of marriage and child bearing. The WINGS although findings were not statistically significant. program did not find any significant changes in gender Among women and men who attended the Gender attitudes. In Cote d’Ivoire, where gender norms were tackled head on through Gender Dialogue Groups, Gender-based Violence Prevention: Lessons from World Bank Impact Evaluations 2 significant results were found. The Gender Dialogue highlighted the importance of exposure, documenting Group magnified the programs effect on reducing that when partners (both male and female) regularly gender inequitable attitudes. Other evidence reviews attended gender dialogue groups, there were larger and suggest that addressing underlying gender and significant reductions in frequency of past year physical relationship norms can be a critical element of effective partner violence. GBV prevention (Heise 2011 and Ricardo et al., 2011). Implications for Program Design There are too few World Bank impact evaluations on Creating groups or a safe space where participants can this topic to establish firm policy recommendations from meet and have positive interactions is an important the work. However, the results are consistent with component of many programmatic approaches to broader evidence which points to a few key issues for changing behaviors, attitudes and norms related to program design that should be taken into account and violence (Paluck & Ball, 2010). Although the Blattman evaluated more extensively. study found that the introduction of a group dynamics element aimed at enhancing the quality of group  Consider adding strategic design features interactions increased income; it had little impact on to economic empowerment interventions that psychosocial measures, social support and symptoms address gender-based violence and/or of depression. Gender-based violence outcomes were underlying norms. These features can and should not reported for this part of the analysis. address underlying social and cultural norms enabling GBV, such as decision making dynamics, household power relations, and social sanctions/stigma related to specific behaviors. Involving men in interventions that aim to empower  Engaging men and boys can increase the women and reduce violence can be critical. In one likelihood of success. Rather than engaging men treatment arm of the WINGS intervention in Uganda, only as potential perpetrators of violence, programs male partners were given an opportunity to take part in should leverage men’s influence as critical decision the program from the start. It incorporated discussions makers and potential agents of change, as well as about gender relations and exercises to improve recognize men’s susceptibility to violence communication and problem solving skills into the themselves. Getting their buy-in is important. In business skills training. While the component did not some contexts it makes sense to engage men and significantly reduce violence, the involvement of men boys separately, while in other settings this can be had positive impacts on the couples ’ interactions and done through mixed sex groups (Ricardo et al., women’s health. Other impact evaluations have found 2011). promising evidence of the impacts of interventions  Consider the duration and intensity of engaging men and boys on attitudes related to components aimed at reducing GBV. It takes violence, but more research is needed to understand time to change deep-rooted norms and behaviors, what kinds of program models work best and for whom and few short-term or one-off interventions have (Ricardo et al., 2011). proven effective in reducing violence. Implications for Research Design For any intervention it is important to consider the Research in the field of GBV is clearly needed and duration (length of intervention) and frequency (of important. While great strides have been made over the activities) needed to achieve the desired outcome. The past decades documenting and raising awareness evidence base on this is still very small; few about GBV, there is still limited evidence around what interventions have compared multiple frequencies and works, and what doesn’t, in prevention. As the Bank durations (Ricardo et al., 2011). Reviews do, however, expands its work in the area of GBV, it is imperative suggest that short-term and one-off interventions are that the knowledge base about effective approaches to less likely to show positive effects on attitudes, preventing violence and responding to the needs of knowledge and behaviors related to GBV perpetration those who experience GBV also evolves. Some key or victimization than longer term ones. It is difficult to recommendations: draw any clear conclusions from the three interventions evaluated here. But it is notable that, for the two studies  Explore active ingredients. The design of the finding reductions in levels of GBV, the components three impact evaluations reviewed here made it addressing GBV lasted for an extended period of time difficult to determine the effect of different (over three months). The study in Cote D’Ivoire components. In the ELA intervention from Uganda, Gender-based Violence Prevention: Lessons from World Bank Impact Evaluations 3 for example, it is not possible to determine if it was Statistics Manual from the United Nations Statistics the life skills training, the economic empowerment Division and the Researching Violence Against component or a safe space and social interactions Women guide (see Annex 1 for these and other that resulted in lower rates of unwanted sex. This resources). In addition to using appropriate makes replication of positive results more measures, it is important to include researchers challenging, as it is unclear which programmatic with experience working on GBV. They can provide elements, and in what sequence and combination, technical expertise, and further ensure the safety of are essential for reducing GBV. Going forward participants and research staff. impact evaluation designs can expand understanding of key components. Complementary Finally, additional rigorous impact evaluations are process studies should be considered, as they can needed. Although it can be challenging and costly to help assess process and implementation factors. implement randomized control trials in low-income  Include valid and reliable GBV-related contexts, there is an urgent need to invest in a stronger measures following ethical guidelines. As was evidence base. Currently, the vast majority of evidence shown by both Blattman et al. (2013) and Bandiera from rigorously evaluated interventions comes from a et al. (2012) it is useful to include GBV measures in few high income countries. While some lessons can be interventions that do not focus solely on GBV. drawn from these, the generalizability is questionable Gender-based violence measures need to follow and needs to be tested in other contexts. Therefore, existing guidelines and validated examples to there is a need to test programs designed for ensure that data on GBV are both valid and preventing GBV and other interventions that the ethically collected. Collecting data on GBV often literature suggests may have positive or negative involves different requirements to ensure study spillover effects on GBV. For example, a better participants’ safety than for most other outcomes understanding of impacts of cash transfers, measured in economic research. A number of microfinance, education, leadership training and public resources are available for practitioners wishing to works on GBV is needed. include GBV measures, including the Gender 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 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 the knowledge from previous evaluations and encourage more and better evaluations in key gender topics. Gender-based Violence Prevention: Lessons from World Bank Impact Evaluations 4 References Annan, J., Bundervoet, T., Seban, J., and Costigan, J. (2013) A Randomized Impact Evaluation of Village Savings and Loans Associations and Family-Based Interventions in Burundi. New York: International Rescue Committee. Bandiera, O., Buehren, N., Burgess, R., Goldstein, M., Gulesci, S., Rasul, I., & Sulaiman, M. (2012). Empowering Adolescent Girls: Evidence from a Randomized Control Trial in Uganda. Bass, J. K., Annan, J., McIvor Murray, S., Kaysen, D., Griffiths, S., Cetinoglu, T., . . . Bolton, P. A. (2013). Controlled Trial of Psychotherapy for Congolese Survivors of Sexual Violence. New England Journal of Medicine, 368(23), 2182-2191. Blattman, C., Green, E., Annan, J., Jamison, J., Bureau, C. F. P., Aryemo, F., & Segura, A. (2013). Building Women’s Economic and Social Empowerment Through Enterprise An Experimental Assessment of the Women’s Income Generating Support (WINGS) Program. Dupas, P. (2011). Do Teenagers Respond to HIV Risk Information? Evidence from a Field Experiment. American Economic Journal: Applied Economics, 3(1), 1-34. Duvvury, N., Callan, A., Carney, P. & Raghavendra, S. (2013). Intimate Partner Violence: Economic Costs and Implications for Growth and Development. Women's Voice, Agency & Participation Research Series. Washington, D.C., World Bank Garcia-Moreno, C., Pallitto, C., Devries, K., Stockl, H., Watts, C., & Abrahams, N. (2013). Global and regional estimates of violence against women: prevalence and health effects of intimate partner violence and non-partner sexual violence. Geneva: World Health Organization, Gupta, J., Falb, K., Kpebo, D., Seban, J., & Annan, J. (2013). Village Savings and Loans Associations, Gender Dialogue Groups, and Gender-Based Violence against Women in Cote d'Ivoire: Findings from a randomized community trial. Gupta, J., Falb, K., Lehmann, H., Kpebo, D., Xuan, Z., Hossain, M. and Annan, J. (2013). Gender norms and economic empowerment intervention to reduce intimate partner violence against women in rural Cote d'Ivoire: a randomized controlled pilot study. BMC International Health and Human Rights, 13(1), 46. Heise, L. (2011). What works to prevent partner violence? An Evidence Overview: London: London School of Hygiene and Tropical Medicine. Ksoll, C., Lilleør, H. B., Lønborg, J.H., and Rasmussen O. D. (2013) Impact of Village Savings and Loans Associations: Evidence from a Cluster Randomized Trial, The Rockwool Foundation Research Unit, Study Paper No. 56 Paluck, E. L., & Ball, L. (2010). Social norms marketing aimed at gender-based violence: A literature review and critical assessment. New York: International Rescue Committee. Ricardo, C., Eads, M., & Barker, G. (2011). Engaging boys and young men in the prevention of sexual violence: A systematic and global review of evaluated interventions. Pretoria: South Africa: Oak Foundation. Vyas, S., & Watts, C. (2009). How does economic empowerment affect women's risk of intimate partner violence in low and middle income countries? A systematic review of published evidence. Journal of International Development, 21(5), 577-602. Gender-based Violence Prevention: Lessons from World Bank Impact Evaluations 5 Annex 1 Overview of Interventions Study Country Programmatic/ Intervention Frequency Length Target Sample GBV indicators and Main results implementing and duration between population size established measures agency/partner of gender baseline used related and component follow up Blattman Uganda AVSI Uganda  Cash Business 18 months Vulnerable 1800  Intimate partner  No impact on intimate et al., and Innovations transfer training lasted people victimization (emotional partner violence 2013 for Poverty  Business five between and physical)  Significant impact on Action training consecutive 14-30 (86%  Items from the DHS women’s economic days (not female) questionnaire were used empowerment focused on GBV specifically) Bandiera Uganda BRAC Uganda,  Vocational Clubs are open 24 months Adolescent 4800  Forced sex victimization  Significant decrease in girls et al., University training five days a girls 14-20  Included a question on who reported they had sex 2012 College London  Life skills week. Life skills (average “unwilling sex in past year� unwillingly in the past year and World Bank training courses take age 16) in the questionnaire  Significant increase in place regularly entrepreneurial skill and and girls decide participation in income- when and how generating activities many to take Gupta et Cote International  Village VSLA meetings Approx. Women 934  Domestic violence The GDG addition did help al., 2013 d’Ivoire Rescue Savings took place 24 months over 18. victimization (physical, reduce overall levels of IPV. Committee, Yale and Loan every week. The GDG emotional, sexual and Sig. reductions in economic School of Public Association included economical) and violence. Non-sig. Health and (VSLA) The treatment male acceptability of being reductions in physical and Innovations for arm also partners beaten by husband.  Gender emotional abuse and a non- Poverty Action Dialogue received eight  Items from the WHO study sign. increase in sexual bi-weekly GDG on domestic violence and violence. Physical IPV was Group (GDG) sessions women’s health in past 12 significantly reduced among months. women and men most adherent to the curriculum Gender-based Violence Prevention: Lessons from World Bank Impact Evaluations 6 Annex 2 Resources for researching GBV Ellsberg, M., & Heise, L. (2005), Researching Violence Against Women: A Practical Guide for Researchers and Activists. Washington DC, United States: World Health Organization, PATH. Available at http://www.path.org/publications/files/GBV_rvaw_complete.pdf Measure DHS. (2013). Domestic Violence Module. Washington DC: USAID. Available at: http://www.measuredhs.com/pubs/pdf/DHSQM/DHS6_Module_Domestic_Violence_28March2013_DHS QM.pdf Partners for Prevention. (2010). United Nations Multi-Country Study on Men and Violence: Post-Conflict Men’s Questionnaire. Bangkok: United Nations. Available at: http://www.partners4prevention.org/sites/default/files/post-conflict_mens_questionnaire_final.pdf. Sexual Violence Research Initiative. (2013) Research Tools. (The page brings together methodologies based on best practice in gender-based violence research). Available at http://www.svri.org/tools.htm United Nations Statistics Division. (2013), Gender Statistic Manual, Available at: http://unstats.un.org/unsd/genderstatmanual/ United Nations Statistics Division. (2013), Guidelines for Producing Statistics on Violence against Women: Statistical Surveys, Available at: http://unstats.un.org/unsd/gender/docs/guidelines_VAW.pdf World Health Organization. (2005). WHO multi-country study on women’s health and domestic violence against women: summary report of initial results on prevalence, health outcomes and women’s responses. Geneva: WHO. Available at: http://www.who.int/gender/violence/who_multicountry_study/en/ World Health Organization. (2007). WHO ethical and safety recommendations for researching, documenting and monitoring sexual violence in emergencies. Geneva: WHO. Available at: http://www.who.int/gender/documents/OMS_Ethics&Safety10Aug07.pdf World Health Organization. (2001). Putting women first: ethical and safety recommendations for research on domestic violence against women. Geneva: WHO. Available at: http://www.who.int/gender/violence/womenfirtseng.pdf Gender-based Violence Prevention: Lessons from World Bank Impact Evaluations 7