88179 Voice Agency and Empowering women and girls for shared prosperity Cover: A woman raises her hand to speak at a community meeting in Aurangabad, India. Photo: Simone D. McCourtie / World Bank Group. Above: This is a woman activist from the Rwanda Women Network taken at a meeting of the Village of Hope project. Photo: Mary Ellsberg. Voice Agency and Empowering women and girls for shared prosperity This work is a product of the staff of The World Bank Group with external contributions. The findings, interpre- tations, and conclusions expressed in this work do not necessarily reflect the views of The World Bank Group, its Board of Executive Directors, or the governments they represent. The World Bank Group does not guarantee the accuracy of the data included in this work. The boundaries, colors, denominations, and other information shown on any map in this work do not imply any judgment on the part of The World Bank Group concerning the legal status of any territory or the endorsement or acceptance of such boundaries. Nothing herein shall constitute or be considered to be a limitation upon or waiver of the privileges and immuni- ties of The World Bank, all of which are specifically reserved. This work is subject to copyright. Because the World Bank Group encourages dissemination of its knowledge products, this work may be reproduced in whole or in part for non-commercial purposes so long as full attribution to this work is given. Copyright 2014 by the World Bank Group Design: Miki Fernández / ULTRAdesigns, Inc. Contents Foreword ............................................................................................................................ ix Acknowledgements ........................................................................................................... xi Abbreviations ...................................................................................................................... xii Executive Summary.............................................................................................................. xv CHAPTER 1 FRAMING THE CHALLENGE: NORMS, CONSTRAINTS AND DEPRIVATIONS Introduction .......................................................................................................................................... 2 Why agency? ........................................................................................................................................ 2 Why does agency matter for development? ................................................................................. 6 Our conceptual framework ............................................................................................................... 7 Overlapping disadvantage and deprivations ............................................................................... 9 Overlapping disadvantages ......................................................................................................... 9 Multiple deprivations .................................................................................................................... 12 Focus on key drivers: social norms and the law ........................................................................... 15 Norms shape agency .................................................................................................................... 15 National laws .................................................................................................................................. 19 Customary laws and plural legal systems ................................................................................... 20 Endnotes ............................................................................................................................................... 22 CHAPTER 2 ENHANCING WOMEN’S AGENCY: A CROSS-CUTTING AGENDA The role of cross-cutting public actions ......................................................................................... 28 Changing social norms ....................................................................................................................... 30 Changing norms by working with men and boys, households and communities ................. 30 Broadcast media ........................................................................................................................... 31 A progressive legal framework ........................................................................................................ 34 Constitutions and principles of equality ..................................................................................... 35 Supporting effective implementation and enforcement ......................................................... 36 Expanding access to justice through customary processes .................................................... 37 Increasing women’s agency through sectoral policies and programs..................................... 38 Expanding economic opportunities and training ..................................................................... 38 Designing gender responsive social protection ........................................................................ 41 Increasing gender equality in education .................................................................................... 43 Endnotes................................................................................................................................................ 47 Voice and Agency empowering women and girls for shared prosperity CHAPTER 3 FREEDOM FROM VIOLENCE Gender-based violence as a development challenge.................................................................. 55 How large is the challenge? .............................................................................................................. 57 Prevalence of intimate partner violence...................................................................................... 58 Reporting and responses.............................................................................................................. 63 Costs of violence.................................................................................................................................. 64 Individual-level effects................................................................................................................... 65 Family-level effects........................................................................................................................ 66 Economy-wide effects................................................................................................................... 66 Who is worst affected?.................................................................................................................. 68 The state of the evidence: What works?......................................................................................... 71 Boosting positive gender norms.................................................................................................. 71 Legal reform and responses......................................................................................................... 74 Social support and services.......................................................................................................... 77 Economic empowerment (plus).................................................................................................... 79 Integrating violence prevention into other sectoral interventions........................................... 80 Endnotes................................................................................................................................................ 82 CHAPTER 4 CONTROL OVER SEXUAL AND REPRODUCTIVE HEALTH AND RIGHTS The nature of the challenge............................................................................................................... 95 How great is the challenge?.............................................................................................................. 97 Fertility Choices.............................................................................................................................. 97 Early sexual activity and pregnancy ............................................................................................ 99 Child marriage ............................................................................................................................... 102 Program and policy evidence: What works?.................................................................................. 105 Engaging men and other gatekeepers........................................................................................ 105 Information and services: Improving access and quality........................................................... 106 Alternatives to early marriage....................................................................................................... 109 Implementation of legal responses............................................................................................. 111 Endnotes................................................................................................................................................ 113 Contents | v CHAPTER 5 CONTROL OVER LAND AND HOUSING Women’s control over land and housing as a development challenge.................................... 125 Potentially transformative effects ................................................................................................ 125 What do ownership and control mean? ...................................................................................... 128 How large is the challenge?............................................................................................................... 129 Existing disparities.............................................................................................................................. 129 Marital status and work.................................................................................................................. 131 Laws and norms.............................................................................................................................. 132 Default marital property regimes................................................................................................. 134 Weak implementation of laws....................................................................................................... 136 The state of the evidence: What works?......................................................................................... 137 Ensuring gender equality under the law..................................................................................... 137 Improving implementation............................................................................................................ 140 Endnotes................................................................................................................................................ 143 CHAPTER 6 AMPLIFYING VOICES Why women’s voice and participation matters............................................................................. 151 Driving social change for women’s agency: the role of ICTs...................................................... 152 Using ICTs to increase voice and participation........................................................................... 153 Addressing the digital divide........................................................................................................ 154 Collective action as a catalyst for change...................................................................................... 154 Women’s political voice and participation can be transformative............................................ 156 Underrepresentation and biased gender norms........................................................................ 159 Attitudes toward female leadership affect women’s political participation........................... 162 Quotas can reduce barriers to women’s participation.............................................................. 163 Implications for policy......................................................................................................................... 163 Increasing access to ICTs and making content relevant............................................................ 163 Supporting collective action......................................................................................................... 164 Increasing accountability............................................................................................................... 166 Opportunities to expand women’s political participation........................................................ 167 Endnotes................................................................................................................................................ 170 vi | Voice and Agency empowering women and girls for shared prosperity CHAPTER 7 CLOSING GAPS IN DATA AND EVIDENCE Monitoring change and assessing progress .................................................................................. 179 Country-level data.......................................................................................................................... 179 Available data and indicators....................................................................................................... 182 Closing data and evidence gaps.................................................................................................. 182 Composite indices......................................................................................................................... 188 Program and project-level indicators and emerging good practice........................................ 188 Experience at the World Bank Group.......................................................................................... 190 Endnotes................................................................................................................................................ 192 Annex I: Background Papers ............................................................................................................. 195 BOXES Box 1.1 What is agency?.................................................................................................................. 3 Box 1.2 Women’s mobility: Evidence on freedom of movement................................................ 5 Box 1.3 Multiple disadvantages among adolescent Hmong girls.............................................. Box 1.4 Norms about masculinity shape men’s behavior............................................................ 16 Box 2.1 In Tostan’s footsteps........................................................................................................... 32 Box 2.2 New African constitutions reshaping gender norms: South Africa and Kenya........... 34 Box 2.3 How the Adolescent Girls Initiative is helping to overcome agency constraints........ 40 Box 2.4 Promoting gender equality in school and improving the learning environment........ 44 Box 2.5 Curriculum reform to promote gender equality through education............................ 45 Box 3.1 Key messages ..................................................................................................................... 54 Box 3.2 What is GBV? Stories from survivors in Tonga................................................................. 56 Box 3.3 The Domestic Violence Module of the Demographic and Health Surveys................. 60 Box 3.4 Serving survivors and empowering women in the Great Lakes.................................... 65 Box 3.5 Mobilizing communities against violence: Lessons from SASA!................................... 74 Box 3.6 Increasing women’s access to justice in Papua New Guinea......................................... 77 Box 3.7 Coordinated support for survivors of violence in the Solomon Islands....................... 78 Box 3.8 Integrating violence: Innovations in a transport project in Brazil................................. 81 Box 4.1 Key messages...................................................................................................................... 94 Box 4.2 Engaging traditional “gatekeepers” in adolescent sexual and reproductive health in Malawi.................................................................................................................. 106 Box 4.3 Social accountability and service delivery....................................................................... 107 Box 4.4 Promising approach to delaying marriage in Ethiopia................................................... 110 Box 4.5 Egypt: Closing the gap between laws and practice....................................................... 112 Box 5.1 Key messages...................................................................................................................... 124 Box 5.2 Can control over land reduce gender-based violence?................................................. 127 Box 5.3 What do we mean by ownership and control?................................................................ 128 Contents | vii Box 5.4 Main types of marital property regimes.......................................................................... 135 Box 5.5 Expanding women’s access to land rights in Honduras................................................. 138 Box 5.6 Tools to guide gender land assessments........................................................................ 140 Box 5.7 Lessons from Vietnam’s land reform process.................................................................. 141 Box 6.1 Key messages ..................................................................................................................... 150 Box 6.2 I believe in the power of the voice of women.................................................................. 153 Box 6.3 Women’s movements drive reforms................................................................................. 157 Box 6.4 “If anyone listens, I have a lot of plans”: Women’s participation in Afghanistan......... 158 Box 6.5 Women are underrepresented at senior levels in the justice sector............................ 160 Box 6.6 Differing levels of political engagement.......................................................................... 161 Box 6.7 Mapping and creating safe spaces for girls in South Africa.......................................... 165 Box 6.8 Working with civil society to prevent gender-based violence in Haiti......................... 166 Box 7.1 Key messages ..................................................................................................................... 178 Box 7.2 Selected internationally-agreed indicators on agency.................................................. 180 Box 7.3 Internationally-agreed indicators for measuring violence against women.................. 181 Box 7.4 Selected data sources and examples of measures of women’s agency...................... 183 Box 7.5 Analysis of current data gaps and possible ways forward............................................. 184 Box 7.6 Measuring and expanding agency in Latin America and the Caribbean..................... 187 Box 7.7 Using new measures of women’s autonomy in Chad: The Relative Autonomy Index ................................................................................................................. 189 Box 7.8 Selected World Bank Group Gender Core Sector Indicators related to agency........ 190 Box 7.9 Examples of World Bank Group project indicators related to women’s agency........ 191 FIGURES Figure 1.1 Agency, endowments, economic opportunities and drivers of change...................... 8 Figure 1.2 Share of 15–19 year olds completing school grades (current cohort) by wealth quintile and gender.......................................................................................... 10 Figure 1.3 Share of women experiencing agency deprivations in three domains ....................... 13 Figure 1.4 Share of women experiencing agency deprivations in three domains in Niger......... 13 Figure 1.5 Correlation between education levels and deprivations in control over resources, child marriage and condoning wife beating........................................ 13 Figure 1.6 Correlation between women’s work and lack of control over household resources.......................................................................................................... 14 Figure 1.7 Correlation between wome’s level of education and lack of sexual autonomy.......... 14 Figure 3.1 Change in percentage of women who believe wife beating can be justified ............. 57 Figure 3.2 Share of ever partnered women who have experienced physical or sexual intimate partner violence or both, in their lifetime......................................................... 58 Figure 3.3 Share of women experiencing physical or sexual intimate partner violence or both, by ethnic group.................................................................................................... 59 Figure 3.4 Share of women experiencing physical or sexual intimate partner violence or both and at least one control on their behavior ........................................................ 61 Figure 3.5 Reasons for not seeking help............................................................................................ 64 viii | Voice and Agency empowering women and girls for shared prosperity Figure 3.6 Economy-wide costs of intimate partner violence, selected countries........................ 67 Figure 3.7 Share of women who condone wife beating, by age of respondent............................ 72 Figure 3.8 Cumulative number of countries with legislation against domestic violence............. 75 Figure 3.9 Countries with specific legislation criminalizing marital rape........................................ 76 Figure 4.1 Education as a key driver of women’s ability to exercise choice in relationships........ 96 Figure 4.2 Pregnancy desires of married adolescent girls............................................................... 101 Figure 4.3 Child marriage prevalence in 111 countries..................................................................... 103 Figure 5.1 Higher shares of women’s housing ownership are associated with more gender equality......................................................................................................... 127 Figure 5.2 Share of women and men who report owning housing or land.................................... 130 Figure 5.3 Joint and sole reported ownership of land by gender................................................... 130 Figure 5.4 Probability of rural land ownership, by marital and working status.............................. 132 Figure 5.5 Women’s reported ownership of housing or land, by type of marital property regime..................................................................................................... 135 Figure 5.6 Probability that a woman will report owning housing or land in rural areas, conditional on type of property regime and marital status........................................... 136 Figure 6.1 Share of women in parliament by region and world, 1997–2013................................... 159 Figure 6.2 Percentage of female justices in constitutional courts................................................... 161 Figure 6.3 Rate of participation in political discussion with friends................................................ 161 viii Figure 6.4 Attitudes toward women leaders reflected in the share of women in parliament...... 162 Figure 6.5 Urban spatial area considered safe by sex and grade ................................................... 165 TABLES Table 1.1 Legal restrictions on married women’s agency............................................................... 20 Table 2.1 Selected illustrations of programs to enhance agency.................................................. 29 | ix Foreword O ur flagship World Development better suited. Women then enter a smaller Report 2012 demonstrated that gen- range of jobs with lower barriers to entry, der equality and economic devel- less stability, and lower wages, continuing a opment are inextricably linked. It showed vicious circle of inequality. Overwhelmingly, that equality not only guarantees basic girls and women also perform the unpaid rights but also plays a vital role in promot- work of caregiving, for which they are often ing the robust, shared growth needed to end penalized with poverty in old age. extreme poverty in our increasingly compet- Norms over time may become legalized itive, globalized world. The persistent con- discrimination, which imposes its own steep straints and deprivations that prevent many economic cost. As the 19th-century philos- of the world’s women from achieving their opher John Stuart Mill wrote, laws start “by potential have huge consequences for indi- recognizing the relations they find already viduals, families, communities, and nations. existing.… Those who had already been com- The 2012 report recognized that expanding pelled to obedience became in this manner women’s agency—their ability to make deci- legally bound to it.” Rightly, he added, what sions and take advantage of opportunities— “color, race, religion, or in the case of a con- is key to improving their lives as well as the quered country, nationality, are to some men, world we all share. sex is to all women,” their subordinate status Voice and Agency: Empowering Women often codified by law. Today, in 128 countries, and Girls for Shared Prosperity represents a laws in fact treat men and women differ- major advance in global knowledge on this ently—making it impossible, for example, for critical front. The vast data and thousands a woman to obtain independently an ID card, of surveys distilled here cast important light own or use property, access credit, or get on the nature of constraints women and a job. These constraints are fundamentally girls continue to face globally. unjust. They are also economically unwise. As an anthropologist, I especially wel- The good news is that social norms can come the report’s focus on social norms, and do change. This report identifies prom- which act as powerful prescriptions for how ising opportunities and entry points for men and women should behave. Even where lasting transformation, such as interven- women can legally own property, they may tions that reach across sectors and include not, because those who do become outcasts. life-skills training, sexual and reproductive Even where girls go to school and take an health education, conditional cash trans- interest in math, teachers and parents may fers, and mentoring. It finds that address- direct them away from certain studies and ing what the World Health Organization has jobs for which social norms say boys are identified as an epidemic of violence against x | Voice and Agency empowering women and girls for shared prosperity women means sharply scaling up engage- backed by evidence and data, is long over- ment with men and boys. due. The World Bank Group’s twin goals of ending extreme poverty and boosting The report also underlines the vital role shared prosperity demand no less than the information and communication technolo- full and equal participation of women and gies can play in amplifying women’s voices, men, girls and boys, around the world. expanding their economic and learning oppor- tunities, and broadening their views and aspi- The World Bank Group is committed rations. As Pakistan’s young activist Malala to accelerating and enhancing equality in Yousafzai said of herself and her peers during everything we do and to shining a spot- our conversation at the World Bank Group light on inequality wherever we find it. This in 2013, “We spoke, we wrote, we raised our report does both. It should inform the global voices” through the media. “We spoke and development agenda going forward and we achieved our goal. Girls are going back to advance momentum toward a better future school and are allowed to go to the market.” for all. A bold new path toward equality, Jim Yong Kim grounded in fundamental human rights and President, World Bank Group | xi Acknowledgements Voice and Agency: Empowering Women and Advisory Group: Gary Barker, Promundo; Girls for Shared Prosperity is a product of Lourdes Beneria, Cornell University; Eppu the World Bank Group’s Gender and Devel- Mikkonen-Jeanneret, HelpAge International; opment team, under the leadership of Jeni Cheryl Doss, Yale University; Mary Ells- Klugman, Director. The task was managed berg, George Washington University; Naila by Lucia Hanmer, Lead Economist. The main Kabeer, London School of Economics; Sunita contributors to analysis and writing were Kishor, Demographic and Health Surveys; Sarah Twigg, Jennifer McCleary-Sills, Tazeen Stephan Klassen, University of Goettingen; Hasan and Julieth Andrea Santamaria Kathleen Kuehnast, US Institute of Peace; Bonilla, with special thanks to Sarah Twigg Susan Markham, National Democratic Insti- for her coordination role. tute; Lori Michau, Raising Voices; Andrew Background analysis was undertaken by Morrison, IDB; Kathleen Newland, Migration Emma Samman, Matthew Morton, Sarah Policy Institute; Agnes Quisumbing, Interna- Haddock, Josefina Posada, and Sofia Trom- tional Food Policy Research Institute; Char- mlerova, with thanks to Zuzana Boehmova, lotte Watts, London School of Hygiene and Anjali Fleury, Alicia Hammond, Sveinung Tropical Medicine; Alicia Yamin, Harvard Kiplesund, Nazia Moqueet, Sarah Nedolast, University; and Lawrence Yanovitch, GSMA Marie-Anne Nsengiyumva, Milad Pournik, Foundation, as well as Sabina Alkire of the and for various inputs. Oxford Poverty and Human Development Initiative, and Alison Evans. We are grateful to Caroline Anstey for her support in initiating the work, and to The team would also like to acknowledge the Government of Sweden and the Nordic the advice and support of World Bank Group Trust Fund for their financial support. The colleagues, in particular: Ana Revenga, Acting Umbrella Facility for Gender Equality will PREM Vice President, and Luis Benveniste, support dissemination efforts. TrustLaw Louise Cord, Luis-Felipe Lopez-Calva, and Connect of the Thomson Reuters Foundation, Vijayendra Rao, who acted as peer reviewers; and Serena Grant, in particular, are thanked Sarah Iqbal and the Women, Business and the for supporting several country studies. Law team for collaboration; and the World Bank Group’s Gender Board members and The report draws on 15 thematic and others for valuable comments and inputs. country papers, listed in Annex I. It is informed by more than a dozen consulta- Led by Sarah Jackson-Han, Amy Adkins tions since December 2012 in venues rang- Harris, Malcolm Ehrenpreis and Maura Leary ing from Managua, Nicaragua to Kathmandu, provided communications support. Admin- Nepal, and benefited greatly from the collec- istrative support was provided by Maureen tive wisdom and research of our Technical Itepu, Ngozi Kalu-Mba, and Mame Niasse. xii | Voice and Agency empowering women and girls for shared prosperity Abbreviations ABAD Apni Beti Apna Dhan (Our Daughter Our Wealth) AGALI Adolescent Girls’ Advocacy and Leadership Initiative AGI Adolescent Girls Initiative CCT conditional cash transfer CEDAW Convention on the Elimination of All Forms of Discrimination against Women CFR Council on Foreign Relations CSI Core Sector Indicator DHS Demographic and Health Survey EDGE Evidence on Data and Gender Equality (initiative) ELA Empowerment and Livelihoods for Adolescents EU European Union FAO Food and Agriculture Organization FEMME Foundations to Enhance Management of Maternal Emergencies FGM/C female genital mutilation/cutting FPAM Family Planning Association of Malawi GDP gross domestic product GBV gender-based violence HIV/AIDS human immunodeficiency virus/acquired immune deficiency syndrome IAEG-GS Inter-agency and Expert Group on Gender Statistics IBRD International Bank of Reconstruction and Development ICAN International Civil Society Action Network ICPD International Conference on Population and Development ICRW International Center for Research on Women ICT information and communication technology IDA International Development Association IFC International Finance Corporation IMAGES International Men and Gender Equality Survey IPV intimate partner violence IRIS Integrated Risk Information System IUD Intrauterine Device KNOMAD Global Knowledge Partnership on Migration and Development KOFAVIV Commission of Women Victims for Victims LSMS–ISA Living Standards Measurement Study–Integrated Surveys on Agriculture Abbreviations | xiii MCH maternal and child health NSP National Solidarity Programme (Afghanistan) OECD Organisation for Economic Co-operation and Development OPHI Oxford Poverty and Human Development Initiative PATH Honduras Land Administration Program PRSC Poverty Reduction and Support Operation Credit RAI Relative Autonomy Index RALAS Reconstruction of Aceh Land Administration System RBF results-based financing RCT randomized control trial RMS Results Measurement System SEWA Self Employed Women’s Association SIGI Social Institutions and Gender Index SMS Short Message Service STI sexually transmitted infection TUP Targeting the Ultra Poor (program) UFGE Umbrella Facility for Gender Equality UN United Nations UNFPA United Nations Population Fund UNICEF United Nations Children’s Fund UNSD United Nations Statistics Division UN Women United Nations Entity for Gender Equality and the Empowerment of Women USAID United States Agency for International Development VAW Violence Against Women WEAI Women’s Empowerment in Agriculture Index WHO World Health Organization Young women attend a workshop on gender in Nepal. Photo: Mary Ellsberg/GWU | xv Executive Summary Why voice and agency? on others. For me, this is a source of pride, B my husband asking [my advice]. Now there y ratifying the Convention on the isn’t this machismo. There is mutual respect. Elimination of All Forms of Discrimi- Together we decide.” Similarly, one man in nation against Women (CEDAW), 188 Vietnam commented that “happiness and states have committed to advancing gender equality are related. If the husband under- equality by confronting “any distinction, stands that and is supporting and helping exclusion, or restriction made on the basis his wife … the happiness of the family will be of sex which [impairs] the enjoyment or reinforced.” Full and equal participation also exercise by women … of human rights and requires that all people have voice—mean- fundamental freedoms.” Alongside CEDAW, ing the capacity to speak up and be heard, which came into force in 1979, the 1995 Bei- from homes to houses of parliament, and to jing Platform of Action and various United shape and share in discussions, discourse, Nations Security Council resolutions provide and decisions that affect them. universally accepted benchmarks. These benchmarks include recognition of wom- Increasing women’s voice and agency are en’s right to sexual and reproductive health, valuable ends in themselves. And both voice the right to be free from gender-based vio- and agency have instrumental, practical lence, and equal rights for women and men value too. Amplifying the voices of women to access and control land—rights that and increasing their agency can yield broad establish a clear framework for our global development dividends for them and for report on voice and agency. At the same their families, communities, and societies. time, accumulating evidence and experience Conversely, constraining women’s agency have made clear that tackling poverty and by limiting what jobs women can perform boosting shared prosperity demand that all or subjecting them to violence, for example, people have the opportunity to realize their can create huge losses to productivity and potential and participate fully in all aspects income with broader adverse repercussions of life. for development. We argue that overcoming these deprivations and constraints is central At the individual level, this requires to efforts to end extreme poverty and boost agency—meaning the capacity to make deci- shared prosperity. sions about one’s own life and act on them to achieve a desired outcome, free of vio- Nor is this a zero-sum game. Increasing lence, retribution, or fear. Agency is some- women’s agency need not curtail men’s times defined as “empowerment.” As an agency, and men and boys stand to gain Ecuadorean woman said, “I have free space, from gender equality that improves the to decide for myself, no longer dependent economic and psychological well-being of xvi | Voice and Agency empowering women and girls for shared prosperity all household members, as many men have societies more resilient and more prosper- come to recognize. “The woman helps the ous. For example: man manage the household,” one urban man ■■ Delays in marriage are associated with in Niger said during discussions undertaken greater educational achievement and to inform this report. “It’s a partnership. We lower fertility. And lower fertility can want it that way. Here, in town, a man does increase women’s life expectancy and better when his wife contributes.” has benefits for children’s health and education. Context ■■ When more women are elected to office, There has been unprecedented progress policy-making increasingly reflects the in important aspects of the lives of girls priorities of families and women. and women over recent decades. Yet even where gender gaps are narrowed, system- ■■ Property ownership can enhance wom- atic differences in outcomes often persist, en’s agency by increasing the social including widespread gender-based-vio- status of women, amplifying their voice, lence and lack of voice. These deprivations and increasing their bargaining power and constraints sometimes reflect per- within the household. sistent violations of the most basic human Recognizing agency constraints in devel- rights. And in many instances, constraints opment project design can also improve are magnified and multiplied by poverty effectiveness. Use of reproductive health and lack of education. services by adolescents, for example, is bet- ter where projects address mobility con- Voice and Agency: Empowering women straints and train providers to address pos- and girls for shared prosperity is a major new report by the World Bank that shines a spot- sible issues of stigma. This fact underlines light on the value of voice and agency, the the broader significance of understanding patterns of constraints that limit their reali- how agency constraints operate and how zation, and the associated costs, not only to policies and public action can lift those con- individual women but to their families, com- straints and enhance agency. munities, and societies. It highlights prom- The good news is that promising direc- ising policies and interventions, and it iden- tions for enhancing agency are emerging. tifies priority areas where further research Moreover, the global momentum to tackle and more and better data and evidence are this agenda is growing. This trend is perhaps needed. Underlining that agency has both most vivid in the case of ending gender-based intrinsic and instrumental, concrete value, violence, a major focus of this report. The this report puts advancing women’s voice number of countries recognizing domes- and agency squarely on the international tic violence as a crime has risen from close development agenda. to zero to 76 in just 37 years. In countries Removing constraints and unleashing with legislation against domestic violence, women’s full productive potential can yield women’s acceptance of wife beating is lower. enormous dividends that help make whole This finding suggests the value of enacting Executive Summary | xvii legislation that criminalizes violence. At world, no place is less safe for a woman the same time, laws are not a panacea, and than her own home, with more than 700 awareness of the law and effective implemen- million women globally subject to phys- tation and enforcement are critical. ical or sexual violence at the hands of their husbands, boyfriends, or partners. Where do we stand? As shown in the map in figure 1, regional rates of such violence range from 21 Expanding agency is a universal challenge. percent in North America to 43 percent Agency constraints and deprivations affect in South Asia. Across 33 low- and mid- women and girls in all countries, what- dle-income countries, almost one-third ever their income level. The basic facts are of women say that they cannot refuse sex sobering: with their partners. ■■ Gender-based violence is a global ■■ Many girls have limited control over epidemic, affecting women across their sexual and reproductive rights. all regions of the world. In most of the On present trends, more than 142 million FIGURE 1 Share of women who have experienced physical or sexual intimate partner violence during their lifetime Europe and Central Asia 29% North America 21% Middle East and North Africa 40% East Asia and the Pacific 30% South Asia 43% Sub-Saharan Africa 40% Latin America and the Caribbean Australia and 33% New Zealand 28% Source: Preliminary analysis of WHO (World Health Organization), global prevalence database (2013) using World Bank regions. Note: Areas shaded in grey are not calculated or do not have relevant data. xviii | Voice and Agency empowering women and girls for shared prosperity girls will be married before the age of 18 women’s lives that are important for promot- in the next decade. And each year, almost ing women’s agency and gender equality. one in five girls in developing countries becomes pregnant before her 18th birth- Determinants and drivers day. The lifetime opportunity costs of This report focuses on key drivers and deter- teen pregnancy have been estimated to minants of voice and agency. What we see range from 1 percent of annual gross in practice is a series of compounding con- domestic product in China to as much as straints. Some arise from women’s and girl’s 30 percent in Uganda, measured solely limited endowments (health, education, and by lost income. In developing countries, assets) and economic opportunities. Even pregnancy-related causes are the largest where endowments and economic opportu- contributor to the mortality of girls ages nities are better, social norms about gender 15–19—nearly 70,000 deaths annually. roles are limiting. This problem is evident, ■■ Fewer women than men own land and for example, in gender roles surround- housing. In some cases, this differential ing child care and housework. Even when is wide. In Burkina Faso, for example, women are taking on more work outside more than twice as many men as women of the home, they typically remain largely (65 percent and 31 percent, respec- responsible for housework and child care. tively) report owning a house. In many Social norms can limit women’s mobil- countries, women can access land only ity and ability to network, restrict women’s through male relatives. representation in politics and government, ■■ In too much of the world, women are and be enshrined in discriminatory laws grossly underrepresented in formal and practices. Unequal power relationships politics and positions of power. World- within households and in society as a whole wide, women account for less than 22 have broad-based effects. Gender-based percent of parliamentarians and fewer violence, for example, is associated with than 5 percent of mayors. Rates vary social norms and expectations that reinforce across countries and regions. In Nordic inequality and place the choices of women countries, for example, women hold 42 and girls outside their realm of control. percent of parliamentary seats, and in Legal discrimination is pervasive. In Rwanda, the share is close to two-thirds. 2013, 128 countries had at least one legal Agency has multiple dimensions and is difference between men and women, rang- inevitably context specific. To enable global ing from barriers to women obtaining official coverage and add value, this report limits identification cards to restrictions on owning its focus to four central domains of women’s or using property, establishing their credit- agency: freedom from violence, control over worthiness, and getting a job. Twenty-eight sexual and reproductive health, ownership countries—mainly in the Middle East and and control of land and housing, and voice North Africa and South Asia—had 10 or more and collective action. At the same time, it differences. In 26 countries, statutory inheri- recognizes that these are just a few areas of tance laws differentiate between women and Executive Summary | xix men. In 15 countries, women still require From other studies, we know that inti- their husbands’ consent to work. Other laws mate partner violence is more frequent limit women’s agency in marriages and fam- and severe in poorer groups across such ily life. Laws and legal institutions also play a diverse settings as India, Nicaragua, and central role in prohibiting gender-based vio- the United States. lence and in enabling women to realize their ■■ Ethnic minority status can further mag- reproductive health rights. nify disadvantage. Nearly three-quarters Laws and social norms interact. Women’s of girls out of school globally belong to land ownership, for example, is determined ethnic minorities in their countries. by a complex interaction between some- New analysis of Demographic and Health times contradictory sets of statutory laws, Survey data from 54 countries reveals that customs, and norms. Social norms, custom- women often experience deprivations ary practices, the inaccessibility and weak and constraints across multiple domains capacity of institutions, and, in many cases, of agency at the same time. We find most women’s lack of awareness pose important women (four in five) lack control over barriers to the full realization of women’s household resources, believe gender-based land rights. violence is justified under certain circum- stances, or were married before they turned Overlapping disadvantage 18. Just as striking, almost half of all women and agency deprivations report being deprived in more than one of Constraints on agency do not occur in these areas, and almost one in eight expe- a vacuum and differ in nature across and riences all three (figure 2). However these within countries. A banker in Beijing may be averages mask vast differences across struggling to balance elder care with work countries. In Niger, for example, almost all while also facing glass ceilings in her career. women experience at least one constraint These challenges clearly differ in nature and (figure 3). scope from those faced by adolescent girls Agency deprivations and constraints are hoping to attend school in a low-income linked to other disadvantages—particularly developing country. Here we explore over- access to education. Figure 4 shows that lapping disadvantages—that is, the system- about 90 percent of women with a primary atic exclusion that many people experience education or less experience at least one as a result of multiple inequalities that limit of the deprivations shown in figure 2, and their life chances. For example: almost two-thirds experience all three. This ■■ Poverty increases the likelihood of agency finding contrasts with about 1 in 5 and 1 in deprivations. Girls living in poor house- 20, respectively, of women with a second- holds are almost twice as likely to marry ary education or higher. Almost 1 in 5 rural before the age of 18 as girls in higher-in- women with a primary education experience come households, as are girls from rural all three deprivations compared with 1 in areas versus their urban counterparts. 100 urban women with a higher education. FIGURE 2 Share of women experiencing overlapping agency deprivations in three domains Lack of control over household resources Absence of deprivations 42 21 10 8 11 13 Condones Child marriage wife beating 12 43 10 15 51 Source: Voice and Agency 2014 team estimates based on Demographic and Health Survey data for 54 countries using latest data available from 2001-2012. FIGURE 3 Share of women experiencing overlapping agency deprivations in three domains in Niger Lack of control over household resources Absence of deprivations 85 1 6 15 19 45 Condones wife beating Child marriage 8 71 3 3 75 Source: Voice and Agency 2014 team estimates based on Demographic and Health Surveys for Niger using data from 2006. FIGURE 4 Correlation between education levels and deprivations in control over resources, child marriage and condoning wife beating 90% 65% 18% 5% Suffer at least Suffer three Suffer at least Suffer three one deprivation deprivations one deprivation deprivations Women with primary Women with secondary education or less education and higher Source: Voice and Agency 2014 team estimates based on Demographic and Health Survey data for 54 countries using latest data available from 2001-2012. Executive Summary | xxi Overarching policies and more likely to prevent a stranger’s act of measuring progress violence. The report identifies promising entry points Progressive constitutions and legal for public actions to promote women’s reforms can support the transformation of agency. These entry points include poli- social norms surrounding agency. We focus cies that change social norms and the law, on three core areas: ensuring that all sources alongside programs to promote economic of law adhere to principles of gender equal- opportunities, social protection, and edu- ity; supporting effective implementation cation, where well-designed interventions and enforcement of laws; and expanding and new approaches to implementation are access to justice for all women, including demonstrating significant benefits for wom- through customary processes. en’s agency. The evidence on effectiveness of Expanding women’s economic oppor- interventions designed to combat violence tunities can have wide-ranging benefits, against women and other agency depriva- including benefits for women’s agency. tions is also investigated in depth. Research on norms and agency drawing on Addressing social norms is critical data from women and men in 20 countries because adverse norms underpin and rein- in all regions, for example, concluded that “women’s ability to work for pay… may be force the multiple deprivations that many one of the most visible and game-changing women and girls experience. Although events in the life of modern households and there is no silver bullet for promoting all communities.” But not all work is equally changes in norms, evidence suggests a empowering—working conditions matter, need for public actions that both enhance as does the type of work that women do. women’s and girls’ aspirations and change Among the promising new approaches are behaviors of women and men, boys and programs that tackle norms and provide girls, so that social norms become gen- young women and girls with new infor- der-equal. Promising ways to promote such mation and opportunities, as in Uganda’s changes include working with men, boys, Empowerment and Livelihoods for Adoles- households, and communities, as in Austra- cent program, which provides girls with life- lia, where the Male Champions of Change skills training and local market-informed initiative works with male chief executive vocational training. In addition to the eco- officers and leaders throughout business nomic benefits, participating girls have and the federal government to push for demonstrated much greater control over significant and sustainable increases in the sexual and reproductive health. representation of women in leadership. Similarly, promoting awareness of progres- Social protection can be transformative. sive laws can help stimulate changes in Programs that go beyond protection per se norms and behavior. Evidence across eight and include elements to tackle regressive countries, for example, found that men who gender norms have had promising results. were aware of laws addressing violence Such elements have included addressing against women were nearly 50 percent child care responsibilities; increasing access xxii | Voice and Agency empowering women and girls for shared prosperity to finance and assets; increasing skills, of gender norms. Effectively engaging men, self-confidence, and aspirations; incentiviz- boys, communities, and traditional authori- ing girls’ schooling; and providing informa- ties to change norms around violence, mar- tion and building awareness about gender riage, reproduction, household gender roles, issues and rights. and the roles of women and men in public life have helped to promote women’s agency Education has major significance in this in countries as diverse as Australia and story, with a focus beyond achieving basic Senegal. levels to quality and content. Around the world, we see that better educated women Several types of interventions have been are often better able to make and imple- shown to expand women’s and girls’ sex- ment decisions and choices, even where ual autonomy and control over reproduc- gender norms are restrictive. In South tive decisions. They include programs that Asia and the Middle East and North Africa, promote more gender-equitable communi- women with more education are less likely cation and decision-making and improve- to have to ask their husband’s or family’s ments in access to and quality of informa- permission to seek medical care. In all tion and health services. Interventions to regions, women with more education also expand life opportunities for women and tend to marry later and have fewer chil- girls offer promise when they include provi- dren. Enhanced agency is a key reason why sion for safe spaces, life skills, and job skills. children of better educated women are Women’s sexual and reproductive agency less likely to be stunted: educated mothers can also be supported through more equi- have greater autonomy in making decisions table laws related to marriage and property, and more power to act for their children’s among others, provided such legislation is benefit. coupled with strong implementation and enforcement. Promising interventions Reform of discriminatory laws, particu- targeting agency deprivations larly in the realm of family, inheritance, and property law, is an important first step for Promising interventions to tackle violence, advancing women’s access to land and hous- enhance sexual and reproductive health, ing. But legal reform must be coupled with increase access to assets, and enable voice actions to improve implementation and typically have multiple components and enforcement, gender-sensitive land admin- engage at different levels. They address istration, collection of richer sex-disaggre- norms and involve the wider community— gated data, and monitoring of results. engaging men, boys, women, and girls. And they reflect commitment over time—one-off Women’s collective action and autono- or short-term interventions are less likely mous women’s movements play a pivotal to be effective. As already indicated, a com- role in building the momentum for progres- mon factor of successful approaches across sive policy and legal reform. Development all of the domains explored in the report is agencies and partners can help to enable an acknowledgment of the powerful role change, including through knowledge Executive Summary | xxiii exchange, support for innovative and locally Data gaps and the way ahead driven pilot programs to shift behaviors, and help in capturing and sharing good prac- The data challenges are large. We can estab- tices. Such support should embody large ele- lish profiles of women’s voice and agency by ments of local problem solving and learning using proxies to measures specific aspects, by doing. such as exposure to violence, levels of unmet need for contraception, prevalence of At the same time, new information and female land ownership, and representation communication technologies (ICTs) are of women in politics. But to better capture opening up new spaces for collective action progress toward gender equality, greater and women’s participation in public life. investments are needed. We need to develop Along with the media, ICTs are shaping the new measures and invest in higher-quality aspirations and hopes of the next generation data that more accurately reflect constraints of women and girls. These wider horizons on and expressions of agency, hold gov- can be especially valuable for women and ernments and development agencies such girls whose mobility and opportunities are as the World Bank Group to account, and most restricted. incorporate these findings into our everyday work and decisions. *** This report distills an array of data, studies and evidence to shine a spotlight on the pervasive deprivations and constraints that face women and girls worldwide—from epidemic gender-based violence to laws and norms that prevent women from own- ing property, working, making decisions about their own lives and having influence in society. It identifie some promising programs and interventions to address these deprivations and constraints. Policymakers and stakeholders need to tackle this agenda, drawing on evidence about what works and systematically tracking progress on the ground. This must start with reforming discriminatory laws and follow through with concerted policies and public actions, including multi-sectoral approaches that engage with men and boys and challenge adverse social norms. There is much to gain. Increasing wom- en’s voice and agency is a valuable end in its own right. And it underpins achieve- ment of the World Bank Group’s twin goals of eliminating extreme poverty and boosting shared prosperity for girls and boys, women and men, around the world. Shanghai, China—Mother and daughter riding bicycle by night in the street in Shanghai. Photo: © Thinkstock.com/jacus Chapter 1 Framing the Challenge: Norms, Constraints and Deprivations 2 | Voice and Agency empowering women and girls for shared prosperity Introduction This chapter starts by elaborating the case for strengthening agency and closing Women’s voice and agency matter. They gaps. Next, we present a framework for matter for the shared prosperity of people understanding how agency is affected by alive today and for the prospects of a better drivers of other gender equality outcomes. life for future generations. We go on to present evidence on agency, Enormous development progress has looking at how deprivations in women’s occurred in the past two decades—many agency in certain domains overlap with people are living longer and healthier lives, other disadvantages that women experi- becoming educated, and having more access ence, and we present new findings on the to goods and services.1 Yet in the world today, different ways in which women experience 1.2 billion people live in extreme poverty; such deprivations and how they overlap. 774 million adults are illiterate; 783 million Finally, we examine two drivers—social people have no access to clean water; and norms and laws and legal institutions—in 2.5 billion lack adequate sanitation.2 greater depth. Often, the same people face all of those Why agency? deprivations simultaneously, and common markers of disadvantage include being a The 2012 World Development Report: Gen- woman, being poor, and lacking education. der Equality and Development4 highlighted Less often discussed is that far too many wide-ranging and unprecedented progress women and girls also are unable to own in important aspects of the lives of girls and basic assets such as land, have no voice in women over recent decades. More coun- their communities and governments, have tries than ever guarantee women and men limited control over their sexual and repro- equal rights under the law in such areas as ductive rights, and lack freedom from vio- property ownership, inheritance, and mar- lence, even in their own homes. One of the riage.5 Gender gaps in primary schooling most alarming facts is that more than 800 have narrowed in many countries. Globally, million women alive today have experi- more women than men attend university, enced either physical or sexual partner vio- and women are now living longer than men lence or non-partner sexual violence during in every region of the world.6 In all but a their lifetimes.3 Freedom from these kinds handful of countries, women have the right of deprivations is a fundamental aspect of to vote and stand for election. well-being that is too often denied. At the same time, the World Development Tackling these gaps matters—and Report 2012 and the broader literature— expanding voice and agency is central—for including Esther Duflo’s article Women reducing extreme poverty and boosting Empowerment and Economic Development shared prosperity. Moreover, the expansion and a body of work by Naila Kabeer includ- of agency is a universal challenge. Gaps exist ing, Paid Work, Women’s Empowerment in all countries, regardless of their level of and Inclusive Growth—establish that even income. where such gaps are narrowed, systematic Chapter 1 Framing the Challenge: Norms, Constraints and Deprivations | 3 differences in outcomes often persist.7 We are often at a systematic disadvantage in see that situation in economic opportuni- their ability to make effective choices in a ties, where there has been too little prog- range of spheres, from making decisions at ress in closing key gaps, as our recent global home, deciding what kind of work to do, to report Gender at Work8 documented. choosing whether or when to get married and how many children to have to becoming This report focuses on what we see as a politically active. Agency is an outcome that key driver of persistent gaps: limited agency. matters in its own right; it is an important Agency is about the ability to make effec- driver of other aspects of gender equality; tive choices and to transform those choices and it also has value as a process, as we will into desired outcomes (box 1.1). Women show. Box 1.1 What is agency? “Agency” is the ability to make decisions about one’s own life and act on them to achieve a desired outcome, free of violence, retribution, or fear. The ability to make those choices is often called empowerment. Agency is critical at the individual level, as demonstrated by women worldwide who have spoken of having agency and of being empowered in many different ways: My opportunity is that I have free space, to decide for myself, no longer dependent on others. For me, this is a source of pride, my husband asking me [my advice] now there isn’t this machismo … there is mutual respect … together we decide.—Leticia from Ecuador What is an empowered woman? A woman who is able to work and able to fulfil the needs that she has identified for herself. As long as she has the strength to work, she can solve all her problems. Empowered women can manage their own lives, no matter what the circumstances. —Egypt Pathways fieldwork An empowered woman is one who can help herself and others, who has a job, knows about herself and her environment and her community. You cannot stay in the house and be empowered. If you join societies, organizations, communities, and other social things, even spiritually, you will be empowered. If you are enlightened, empowerment will follow.—Tsikata and Darkwah from Ghana I am free and I have some power; my partner has the same: sovereign decisions are freedom and power.—Urban man, neighborhood of Olzstyn, Poland Agency is also about group and collective action, as we examine in chapter 6: Men used to shut us up and say we shouldn’t speak. Women learned to speak up in a sangathan [group]. Earlier, we couldn’t speak up even at home. Now, we can be more assertive and also go out. I am able to help other women gain confidence as well.—Woman leader of savings group, Vejpur Village, Sabarkantha, Gujarat Source: Narayan et al., 1999; Kabeer, 2013; Boudet et al., 2012; Agarwal, 2010.9 4 | Voice and Agency empowering women and girls for shared prosperity The reality of making choices and exer- participation in politics and by freedom of cising agency is complex and varies enor- movement. Given its complexity, agency is mously, even within the same country. Valu- inherently difficult to measure. Chapter 7 able insights can be drawn from qualitative discusses some of the commonly used tech- approaches to capture that diversity. One niques and the agenda for closing data gaps. valuable and recent source is On Norms and Agency,10 which draws on discussions with One way to measure agency is to use over 4,000 women and men in 20 countries data on what people say they do in differ- and explores the nature and effects of gen- ent domains of life. Such actions are called der differences and inequalities in their lives expressions of agency. This approach has the to uncover the role of gender norms and advantage that it is more comparable and roles. We commissioned additional work in more objective than data on personal per- Chad11 and Niger,12 which provides further ceptions of agency (how much freedom and insights. choice people say they have) in the abstract— which can be distorted by social norms and The gender equality agenda requires may not be comparable across individuals, let much greater progress in expanding agency. alone countries. This report advances the agenda by exam- ining and helping to understand the facts, Building on World Development Report while systematically documenting what we 2012, we investigate several specific expres- don’t know, and by presenting new evidence sions of agency: that allows us to compare across communi- ties, countries, and regions of the world. We ■■ Freedom from the risk of violence deepen the evidence base in ways that foster ■■ Control over sexual and reproductive a better understanding of constraints and health and rights possible ways forward. Just as importantly, ■■ Ability to own and control land and we examine how change can come about housing and identify policies and programs that have worked to increase agency and we identify ■■ Voice and influence in society promising future options. At the same time, Each of these topics is explored in the we underline that this is a sphere where chapters that follow. This mirrors the generalizations are especially risky, and approach laid out in World Development that country-specific diagnostic work and Report 2012, although we do not devote a dialogue are essential to move ahead at the separate chapter to freedom of mobility. national level. Mobility is critical and is a recurring theme Agency is difficult to capture in its in different parts of this report, with high- entirety. Agency can be expressed in many lights presented in box 1.2. Mobility is being ways: through personal relationships, pursued separately under the Global Knowl- autonomy in decision-making, and ability to edge Partnership on Migration and Develop- amass endowments—such as land or prop- ment (KNOMAD) initiative13 and is beyond erty, education, or good health—through the scope of this report. Chapter 1 Framing the Challenge: Norms, Constraints and Deprivations | 5 Box 1.2 Women’s mobility: Evidence on freedom of movement Mobility includes an individual’s physical capacity to move freely beyond the household, as well as the ability to move across social and economic spheres. It helps women and men build and maintain social and professional networks and enables participation in the economy and civic life. Physical mobility influences social and economic mobility, and significantly affects women’s and girls’ opportunities and choices. In Bangladesh, for example, women working outside the home, in both formal and informal employment, are more likely to vote and to do so according to their own decisions compared with women who work from home or do not work for pay. Yet, widespread restrictions on women’s mobility arise from social norms. Women who leave the house too often may be branded “bad girls” or “bad wives.” Only a good-for-nothing, a ‘girl of the wind,’ can ride a taxi motorbike.—Women’s focus group, Garin Malam, Communauté Urbaine de Zinder, Niger Our analysis from 52 developing countries found that one in three women agree that wife-beating is justified for going out without permission; the same number of women report having no say over visits to family or friends. Women’s mobility can change with age; for example, only 20 percent of respondents ages 45 to 49 have no say over visits to family and friends, compared with 60 percent of those ages 15 to 19. On average, women report greater freedom of movement if they are educated, identify themselves as head of the household, or belong to a richer household. For example, 43 percent of women without an education have no say in decisions about visits to friends and family, compared with 17 percent of those with a higher education. Women’s mobility can be restricted by laws that dictate the need for her husband’s consent to work outside the home or that restrict the types of jobs or industries in which she may work. A number of countries in Europe and Central Asia, for example, have lengthy lists of jobs that are prohibited for women; in Russia, the list includes 456 jobs, ranging from woodworker to truck driver in agriculture. Such restrictions originated from a desire to protect women; but in practice, they limit women’s earnings potential and opportunities outside the home. Lack of appropriate transport is often a constraint, and unsafe public spaces where women risk sexual harassment and other forms of gender-based violence make it difficult or impossible for women to move about freely. In Yemen, for example, lack of mobility is cited as a major restriction on girls’ ability to go to school, and one-third of women deliver babies without medical care because appropriate transport is lacking. In central Afghanistan, mobility limits women’s ability to find work, socialize, and build networks. Lack of social mobility can limit women’s success as entrepreneurs, as in Sierra Leone, where limited access to networks is a key factor in restricting women’s ability to scale-up microenterprises. Source: Kabeer, 2013; Boudet et al., 2012; World Bank, forthcoming; World Bank and IFC, 2013; World Bank, 2011; Echavez, 2012; Cherie Blair Foundation for Women, 2014.14 6 | Voice and Agency empowering women and girls for shared prosperity Why does agency matter improves child nutrition, even after taking for development? into account other factors linked to better nutrition, such as household wealth, moth- Agency has intrinsic value: the ability to er’s height, breast-feeding practices, water, exercise choice and to take action is import- and sanitation: ant in its own right. Amartya Sen power- fully argued this value in Development as ■■ In Ethiopia, one year olds whose moth- Freedom, published in 1999, and the intrin- ers had a primary school education sic value of process freedoms is now an along with access to antenatal care were accepted element of mainstream develop- 39 percent less likely to have stunted ment thinking.15 More than a decade ago, growth. World Development Report 2000/2001: ■■ In Vietnam, infants whose mothers Attacking Poverty included voicelessness as had attained a lower-secondary educa- a social dimension of poverty. That report tion were 67 percent less likely to have found that poverty meant lack of freedom of stunted growth.18 choice and action and lack of power to con- Lack of agency can prevent women and trol one’s life: Poverty is like living in jail, liv- girls from accessing a range of services— ing in bondage, waiting to be free—quoting a from health services to secondary education young Jamaican woman.16 (see chapters 2, 3, and 4) to new information Agency also has instrumental value— technologies and services (see chapter 6). that is, expanding agency is likely to bring Women’s lack of agency can compromise broader gains for development and to the effectiveness of development projects, advance the agenda of eliminating poverty effectively lowering the return on invest- and sharing prosperity. We show in chapter ments. Explicit attention to project design 3 that the estimated costs of gender-based in a whole range of sectors—from service violence for a range of countries runs from provision to to the expansion of economic 1.2 percent to 3.7 percent of gross domes- opportunities can address this risk. For tic product, equivalent to what many gov- example, in Peru, incorporating compo- ernments spend on primary education (see nents to increase the aspirations of female figure 3.5). And, if girls are educated but beneficiaries of a social protection program their work choices are restricted, then the doubled the effect of the intervention.19 forgone costs in productivity and income Similarly, a recent study of Pakistan shows can be huge. The lifetime opportunity costs that—despite the efforts to improve mater- of adolescent pregnancy range as high as 12 nal health—the use of health services has percent of annual GDP in India and 30 per- been far from universal; the average uptake cent in Uganda.17 of prenatal care and postnatal care is 50 Agency has intergenerational benefits percent and 21 percent, respectively. But the too. A large body of evidence shows that uptake of those services increases by up to women have greater agency when they are 10 percent when women’s decision-making educated and that, in turn, benefits their power increases by 1 percent, after taking children. For example, a mother’s education into account the effect of other factors, such Chapter 1 Framing the Challenge: Norms, Constraints and Deprivations | 7 as poverty and education.20 In other words, multidimensional and are considered a range of public programs can produce sub- in relation to: economic opportunities, stantially stronger results when the women for example, jobs and entrepreneurship; those programs are designed to serve under- endowments, for example, investments in stand their choices and are empowered to education and health and asset ownership; exercise them. and agency - the focus of this report. Women’s limited voice in society and In figure 1.1, agency, economic opportu- the economy is reflected in their underrep- nities, and endowments are connected and resentation in politics and in government, often (but not always) mutually reinforcing as well as in the corporate world. Women gender equality outcomes. The left side of the make up fewer than 22 percent of parlia- chart shows sets of determinants, the impor- mentarians and, increasing by only half a tance of which varies by context. A particu- percent per year since 19906 progress has lar focus of this report is on laws (statutory been glacial.21 Important exceptions exist: as well as customary and religious), as well in Rwanda almost two-thirds of the parlia- as on social norms. The latter are also being mentarians are women.22 Yet despite such explored in the upcoming World Development high representation, the country still has the Report 2015: Mind and Culture. second-largest share of women experienc- But the causality is not only one-way— ing intimate partner violence (56 percent), gender equality outcomes also affect how reminding us that representation on its own markets and institutions work and how is unlikely to be sufficient.23 In the corporate households allocate resources and make world, women hold just under one-quarter decisions. For example, more educated of senior management positions and fewer women may go to court to use the law to than one-fifth (19 percent) of board seats.24 establish their right to inherit property, as These constraints are not confined to the in Ethiopia,26 demonstrating practices that public arena. In most of the world, no place influence norms and change how formal is less safe for a woman than her own home, institutions function. Similarly, greater eco- a shocking fact that we investigate in chapter nomic opportunities for women can change 3. The gravity of this challenge is increasingly social norms about girls’ education and the well-recognized, as evidenced by propos- status of female children. Such a change is als to include commitments to end violence evident in Bangladesh, where a 2012 sur- against women and girls in the global devel- vey found that, compared with data from opment framework beyond 2015.25 1979, most women interviewed wanted fewer children and no longer cared much Our conceptual framework whether they had boys or girls. Indeed, some expressed a preference for daughters.27 This report adopts the framework of World Development Report 2012, in which the In practice, women’s agency is often lim- functioning of households, markets, and ited by a series of compounding constraints. institutions and their interactions shape These constraints can arise from limited gender outcomes. Gender outcomes are endowments and economic opportunities. 8 | Voice and Agency empowering women and girls for shared prosperity FIGURE 1.1 Agency, endowments, economic opportunities, and drivers of change JOBS AND SOCIAL PROTECTION SOCIAL NORMS Economic opportunities Informal institutions Gender equality MARKETS HOUSEHOLDS EDUCATION LAWS AND AGENCY LEGAL Endowments INSTITUTIONS Formal institutions Source: Adapted from World Development Report 2012: Gender Equality and Development. For example, across 18 of the 20 countries home, they also continue to bear most of the with the highest prevalence of child mar- responsibility for housework and child care. riage, girls with no education were up to six Differences in the amount of time spent on times more likely to marry than girls who housework range from 50 percent more time had received a secondary education.28 And, by women than men in Cambodia and Swe- as we show in the next section, constraints den to about three times more in Italy and six on agency are typically worse for women times more in Iraq.29 Social norms can limit and girls who also experience other sources women’s mobility and their ability to net- of disadvantage, in particular poverty, work, restrict their political representation, minority ethnicity, and lack of education, and result in discriminatory legal rules and among others. For example, girls living in practices. They shape power relationships poorer households are almost twice as likely within households and in the society more to marry before the age of 18, compared broadly. Evidence from the United States, with girls in higher-income households. for example, shows that despite comparable backgrounds, accomplished women are less Even where endowments and economic likely than men to believe they meet the cri- opportunities are closer to gender parity, teria to run for public office.30 social norms about gender roles impose limitations. Limitations can be seen in gen- Equally important to gender equality and der roles surrounding child care and house- poverty reduction outcomes is how those out- work. Even when women work outside the comes are achieved—agency has normative Chapter 1 Framing the Challenge: Norms, Constraints and Deprivations | 9 value as a process. For policy, valuing agency expressions of agency that follow: (a) free- as a process to help achieve development dom from violence, (b) control over sexual outcomes means paying attention to how and reproductive health and rights, (c) con- women and men can engage with policy for- trol over land and housing, and (d) voice and mulation and implementation—whether they collective action. are included in consultative processes and whether their views and needs are taken into Overlapping disadvantages account in project design. For example, roads In almost every region of the world, certain may be planned and built by including consul- groups of people face systematic social exclu- tations with communities that give women a sion as the result of multiple inequalities that voice and take into account gender-specific constrict their life chances.—Naila Kabeer31 security concerns. Taking steps to ensure that those who are frequently excluded have access A sustainable path toward ending to information and services that increase their extreme poverty and promoting shared agency—for example, text messages can be a prosperity involves creating an inclusive feasible method for sending information and society not only in terms of economic welfare service referrals, especially for adolescents— but also in relation to the voice and agency illustrates how valuing agency as a process has of all people and groups.32 Worldwide, some benefits for development outcomes. people are more likely to be disadvantaged than others, and common markers of such Overlapping disadvantage disadvantage include being a woman, being and deprivations disabled, being young or old, and being a member of a minority ethnic group.33 It is obvious that constraints on agency do not come about in a vacuum and differ in Those inequalities can intersect to produce a nature across and within countries. A banker “multiplication of disadvantage” for instance, in Beijing may be struggling to balance elder the intersection of gender, age, ethnicity, care with work and facing glass ceilings in and place of residence can have significantly her career. Those challenges clearly differ in more deleterious effects than the effects of nature and scope from the struggles of ado- gender alone.”34 Somali women refugees lescent Pakistani girls hoping to go to school, in East Africa, for example, face multiple which we illustrate in chapter 6. exclusions that stem from their ethnicity, religion, and refugee status, which are com- It is important to set the stage for the pounded by being female. Those are among socioeconomic correlates of agency depri- the themes explored in a recent World Bank vation, as well as about how deprivations Group report, Inclusion Matters.35 Increasing can accumulate. This section presents new women’s agency requires addressing other results that draw on Demographic and markers of disadvantage as well—for exam- Health Surveys (DHS) for 54 countries, ple, poverty, ethnicity, or location. which importantly inform the cross-cutting agenda for policies and public actions in the Poverty often increases gender gaps, next chapter and the analysis of the specific as illustrated in figure 1.2 for the case of 10 | Voice and Agency empowering women and girls for shared prosperity FIGURE 1.2 Share of 15–19 year olds completing school grades (current cohort) by wealth quintile and gender 100% 100% 80 80 60 60 40 40 20 20 0 0 1 2 3 4 5 6 7 8 9 1 2 3 4 5 6 7 8 9 Grade Grade INDIA 2005-2006 NIGER 2012 100% 100% 80 80 60 60 40 40 20 20 0 0 1 2 3 4 5 6 7 8 9 1 2 3 4 5 6 7 8 9 Grade Grade CAMBODIA 2010 ZAMBIA 2010 100% 100% 80 80 60 60 40 40 20 20 0 0 1 2 3 4 5 6 7 8 9 1 2 3 4 5 6 7 8 9 Grade Grade TANZANIA 2010 INDONESIA 2012 Richest 20% – Male Richest 20% – Female Poorest 40% – Male Poorest 40% – Female Source: Voice and Agency 2014 team estimates based on World Bank., 2009. Educational Attainment and Enroll- ment around the World. http://econ.worldbank.org/projects/edattain. Chapter 1 Framing the Challenge: Norms, Constraints and Deprivations | 11 schooling.36 In a recent study across 27 school globally belong to ethnic minorities countries, boys from the poorest households in their countries. Here are some country were nine percentage points more likely to examples from the Latin American region, complete their primary education than girls, where this challenge is being increasingly whereas boys and girls from rich households well-recognized: were equally likely to complete primary school. In India, the median boy and girl ■■ In Guatemala, 60 percent of indigenous aged 15–19 in the richest quintile reaches women are illiterate, compared with 40 grade 10 (with no gender gap), whereas the percent of indigenous men and 30 per- median boy in the poorest quintile reaches cent of nonindigenous women. only grade six, and the median girl only ■■ In Bolivia, the probability that a Quechua- grade one.37 In contrast, in Cambodia, Tan- speaking woman will complete secondary zania and Zambia, girls and boys from poor school is 28 percentage points lower than families are equally less likely than their bet- that for a Spanish-speaking man. ter-off peers to complete higher grades. In sum, opportunities for girls compared with ■■ In Brazil, white men generally earn the boys vary by country—but they are always highest wages for any level of education, worse for poor children. whereas black women earn the least. Ethnic minority status can further mag- Gender, youth, and ethnicity may work nify disadvantage. It has been estimated together to limit agency, as among adoles- that nearly three-quarters of the girls out of cent Hmong girls in Vietnam (box 1.3). Box 1.3 Multiple disadvantages among adolescent Hmong girls Vietnam has achieved tremendous progress in development, but the gains have been unevenly enjoyed. In Ha Giang Province in northern Vietnam, Hmong children and adolescents, especially girls, face multiple disadvantages. Among the Hmong ethnic minority, poverty rates exceed 80 percent, compared with 20 percent for Vietnam as a whole. Traditional preference for sons and filial piety mean that girls spend long hours on domestic chores— only 4 percent of Hmong girls are enrolled in secondary school and, as adults, their opportunities for paid work are negligible. Rates of child marriage and total fertility for Hmong are double the rate for the Kinh majority ethnic group. Hmong girls also have inadequate access to sexual and reproductive health information and services. In recent fieldwork in Ta Lung commune, girls often struggled to imagine future lives that differed from those of their mothers. Most Hmong girls reported very little say in family or community decisions, where they faced discrimination on the grounds of both their gender and their age. That was reflected in one mother’s plans for her children: “My daughter will finish grade nine only, and then she will get married. My son won’t go anywhere; he will live with me, so I let him reach the high grade.” Sources: Nicola Jones et al. 2013; http://data.worldbank.org/country/vietnam 12 | Voice and Agency empowering women and girls for shared prosperity Intimate partner violence (IPV) is striking points (figure 1.3). First, most (four more frequent and severe among poorer in five) women experience at least one of groups across such diverse countries as these three deprivations of agency. Second, India, Nicaragua, and the United States— almost half of all women report agency-re- though not in others, such as South Africa. lated deprivations in more than one area of Being a member of an ethnic minority or their lives. Third, nearly one in eight women lower caste can also worsen the threat of gen- experience agency-related deprivations in der-based violence. Gender-based violence is all three areas. Finally, and not least, the like- explored further in chapter 3, but some strik- lihood differs sharply for different groups. ing examples are highlighted here: However these averages make vast dif- ■■ In India, Muslim women and women ferences across countries. In Niger for exam- from scheduled castes are the most often ple almost all women experience at least exposed to intimate partner violence. A one constraint, and almost half (45 percent) recent study found that 35 percent of experience all three (figure 14). Muslim women and 41 percent of women Further analysis shows that agency from scheduled castes were exposed to deprivations are linked to other sources of physical violence; of those women, 15 disadvantage, especially lack of education. percent and 18 percent, respectively, Figure 1.5 shows that about 90 percent of were exposed to emotional violence; and women with no more than a primary edu- 11 percent of both groups were exposed cation experience at least one of the con- to sexual violence. straints shown in figure 1.3, and almost ■■ In Australia, indigenous women are five 65 percent experience all three, compared times more likely to be subject to domes- with about one in five and one in 20, respec- tic violence, 38 times more likely to be tively, of women with a secondary education hospitalized for assault, and 10 times or higher. Almost one in five (18 percent) more likely to die from assault than non- of rural women with a primary education experienced all three deprivations com- indigenous women. pared with one in 100 urban women with a higher education. Multiple deprivations Figure 1.6 shows how agency depri- We turn now to examine how women can vations in one area—lack of control over experience multiple deprivations of agency household resources—vary according to at the same time. We look at three key areas whether women work or not and the type where deprivations are widespread: control of work women do. On average, women over household resources; attitudes that who worked in waged employment have expose women to increased risk of gen- more control over household resources than der-based violence and whether they were those who are paid in-kind and those who married as children. do not work outside the home. For exam- The evidence from 54 countries covered ple, in Mozambique, fewer than 20 percent by the DHS immediately highlights several of women wage-earners lack control over FIGURE 1.3 Share of women experiencing overlapping agency deprivations in three domains Lack of control over household resources Absence of deprivations 42 21 10 8 11 13 Condones Child marriage wife beating 12 43 10 15 51 Source: Voice and Agency 2014 team estimates based on Demographic and Health Survey data for 54 countries using latest data available from 2001-2012. FIGURE 1.4 Share of women experiencing overlapping agency deprivations in three domains in Niger Lack of control over household resources Absence of deprivations 85 1 6 15 19 45 Condones wife beating Child marriage 8 71 3 3 75 Source: Voice and Agency 2014 team estimates based on Demographic and Health Surveys for Niger using data from 2006. FIGURE 1.5 Correlation between education levels and deprivations in control over resources, child marriage and condoning wife beating 90% 65% 18% 5% Suffer at least Suffer three Suffer at least Suffer three one deprivation deprivations one deprivation deprivations Women with primary Women with secondary education or less education and higher Source: Voice and Agency 2014 team estimates based on Demographic and Health Survey data for 54 countries using latest data available from 2001-2012. 14 | Voice and Agency empowering women and girls for shared prosperity FIGURE 1.6 Correlation between women’s work and lack of control over household resources Work paid Work paid Did not work in cash in kind last year Cameroon Côte d’Ivoire Dominican Republic Share of women Ghana in lacking control over household resources Haiti Honduras 0% to 20% Mali 21% to 40% Mozambique 41% to 60% Nepal 61% to 80% Uganda 81% to 99% Zambia Source: Voice and Agency 2014 team estimates based on Demographic and Health Surveys using latest available data from 2006-2012. FIGURE 1.7 Correlation between women’s level of education and lack of sexual autonomy No education Primary Secondary Higher Armenia Cambodia Share of women Cameroon who lack sexual autonomy Côte d’Ivoire Haiti 0% to 20% Honduras 21% to 40% Mozambique 41% to 60% Nepal 61% to 80% Uganda 81% to 99% Zimbabwe Source: Voice and Agency 2014 team estimates based on Demographic and Health Surveys using latest available data from 2010-2012. Chapter 1 Framing the Challenge: Norms, Constraints and Deprivations | 15 household resources, compared with 41 to households are also critical, but they are not 60 percent of women who did not work in our focus here.47 the past year. Norms shape agency Education levels are also highly cor- related with a woman’s degree of sexual Different types of norms shape agency. Here, autonomy—measured here by whether a we focus on social norms and international woman says that she is able to refuse sex, legal norms that can be implemented at the to ask her partner to use a condom, or both national level through legislation. (figure 1.7). Again, women with higher edu- cation have greater agency in all 10 coun- Social norms tries shown here, and secondary education The idea that social norms affect develop- often has major benefits too. In Cameroon, ment outcomes is not new; it was estab- Côte d’Ivoire, and Mozambique, for example, lished in the early development literature. 61 percent to 80 percent of women with no The Journal of Economic Development and education lack sexual autonomy, compared Cultural Change, founded in 1952 focused with fewer than 20 percent of women with on the micro level and drew on sociology, higher education. economics, and analysis of tradition to This analysis underlines that agency advance the understanding of development deprivations are not experienced in isola- and growth. Today, the importance of under- tion; many women face multiple constraints standing micro level behaviors for policy is and deprivations in different areas of their reemerging with the convergence of ideas lives simultaneously. At the same time, across economics, politics, philosophy, and women who experience other forms of dis- behavioral psychology. The upcoming World advantage, especially poverty, and lack of Development Report 2015 will examine how education, are more likely to face depriva- policy design which recognizes psycholog- tions in their agency as well. This point has ical and cultural factors, can better achieve important implications for policy, which are development goals. explored further in chapter 2. Social norms are powerful prescriptions reflected in formal structures of society and Focus on key drivers: social in its informal rules, beliefs, and attitudes.48 norms and the law They are reinforced by sanctions, which can The expansion of agency encompasses an be positive or negative, imposed by people enormous agenda. To make a valuable con- belonging to the same reference group or tribution to development thinking and prac- by the state.49 Social norms define what is tice, we narrow our detailed investigation to deemed appropriate behavior and desirable selected drivers and determinants that are attributes for women, men, boys and girls, particularly relevant to policy and program creating gender roles.50 Across the 20 coun- decisions, and where recent evidence sheds tries studied in On Norms and Agency, for light on promising directions—namely, example, men and women reported similar social norms and the law. Markets and ideals for men’s and women’s roles: men are 16 | Voice and Agency empowering women and girls for shared prosperity Box 1.4 Norms about masculinity shape men’s behavior Men’s actions and behaviors are subject to norms of masculinity just as women’s are subject to norms of femininity. Norms about masculinity are often characterized as being aggressive, risk-taking, virile, unemotional, and dominant over women. Even though men often may have more decision- making power and resulting agency than women, they also may feel pressured to conform to societal expectations. For example, men may not agree to use contraception if it will reflect negatively on their virility. In the words of one Tanzanian woman: You cannot tell men to use birth control; they want children. The more they have, the more manly they appear to be.—Young woman, Zabibu village, Tanzania Men are also expected to be providers and protectors of their families. When they are unable to meet those roles, they may compensate by finding alternative ways to prove their masculinity. They may engage in risky behavior: becoming more sexually active, drinking excessively, or becoming more aggressive to demonstrate their masculinity to peers. To maintain an appearance of invulnerability, men may be less inclined to use health services that they need. Norms about masculinity can also result in aggression and violence toward women. Harmful norms can be transmitted across generations; a boy who has witnessed his father beat his mother is much more likely to become a perpetrator of violence as an adult. As a result of seen as providers, heads of households, and constraints operate for men as well as for benevolent decision-makers, whereas ideal women (box 1.4). Transgression of social women are depicted as obedient, caring, and norms may reflect not just on the individual good mates for their husbands and as being but also on family members, and in some responsible for all of the housework and cases also on the community.53 care of all members of the household.51 Social norms affect both day-to-day and If an individual’s behavior is seen as con- major decisions. The workings and inter- forming to a social norm, status and commu- actions of formal and informal institutions, nity acceptance can be secured. Conversely, markets, and households are all affected by behaviors that stray from or conflict with social norms in ways that are often inexplicit. prevailing norms may be subject to negative Social norms affect decisions about school- social sanctions. As in Niger, where, notwith- ing, who in the household gets a job and who standing a woman’s right to inherit land does the majority of unpaid care and house- under Qur’anic law few women claim their work. Across six developed and developing right to inheritance: “Women cannot inherit countries, for example, World Development land. They live with their husbands.”52 These Report 2012 found that women devote one Chapter 1 Framing the Challenge: Norms, Constraints and Deprivations | 17 patriarchal power structures within society that give men power and status over women, men have little incentive to challenge those norms. The good news, however, is that norms about masculinity can and do shift; a recent analysis of the International Men and Gender Equality Survey has shown that younger generations of men are more supportive of gender equality and more likely to engage in household tasks. Similarly, evidence from urban Niger suggests that both women and men are becoming more open to sharing economic responsibility for providing for the household: The woman helps the man manage the household. It’s a partnership. We want it that way. Here, in town, a man does better when his wife contributes.—Men’s focus group, Lazaret, Communauté Urbaine de Niamey And men have much to gain from gender equality. Data show that men in more gender-equal relationships are more likely to be happy; men who participate actively as fathers are more likely to have better physical and mental health; and men who take on greater care-giving roles report benefits to their friendships and relationships with their children, as well as improved relationships with their spouses. Studies from developed countries have found that men who are more actively involved in care giving live longer and report lower rates of mental and physical health problems, including high blood pressure and cardiovascular disease. Source: Fleming et al., 2013; Boudet et al., 2012; World Bank, forthcoming; Brown et al., 2003; Holt-Lunstad et al., 2009.54 to three more hours a day to housework than society and politics in the rest of the world.56 men, two to 10 times the amount of time to Norms affect people’s daily actions and act as child or elder care, and one to four hours less an underlying and sometimes unconscious a day to market activities.55 factor that affects processing information, making decisions, and taking actions. One Social norms influence which occupa- obvious implication is that interventions that tions men and women work in and how fail to address the underlying norms may markets work—who is trusted, who is hired, have limited effects. For example, success in and how contracts are entered into. They eliminating female genital mutilation started affect women’s ability to participate in for- when interventions tackled community-level mal institutions, such as parliaments, and norms about the eligibility of young women to hold leadership positions in businesses, for marriage, rather than focusing on the legal institutions, and other associations. For health risks alone.57 example, in the countries of the Middle East and North Africa, women hold only 7 per- The good news is social norms about gen- cent of the seats in parliament, despite the der roles do change and evolve over time. expansion of women’s participation in civil Girls’ education is a recent example; many 18 | Voice and Agency empowering women and girls for shared prosperity people all over the world are now as likely over the past six years in Canada and South to send their daughters to school and on to Korea, acceptance rose by at least 10 per- higher education as their sons: over the last centage points.61 decade, the gender gap in school enrollment However, gender norms may even be per- has closed in many countries; and globally, petuated by those who are most adversely the gap between the number of boys and affected. As we explore in the next chapter, girls enrolled in secondary school is now only large numbers of women who are subjected 3 percent, with 69 percent of girls and 72 to violence say they think husbands are jus- percent of boys enrolled.58 And norms about tified in beating their wives. men’s and women’s roles are also changing in many places, as discussed in box 1.4. The role Norms can also persist even when cir- of men in fostering progressive norm change cumstances change.62 Such persistence may is explored further in chapters 2 and 3. reflect various reasons: Social norms can shift in response to ■■ Cultural beliefs distort views: Evidence market incentives. Better earnings pros- from psychology and other sciences pects for women may encourage parents shows that the beliefs of individuals to invest in their girls’ schooling. In Ban- shape what they pay attention to and how gladesh, for example, the increasing value they interpret it. People have cognitive placed on women as economic earners bias that can lead them to misinterpret explains why parents now educate their new information so that it reinforces daughters as well as their sons. As Shanu, their initial beliefs, and they continue to a 36-year-old married woman said: “Ear- act in the same way.63 lier, people used to educate their sons but ■■ Widespread practices reinforce views: not their daughters. Now girls get the same When nearly all people adhere to a social education, they have the same value, they norm and when the consequences of get the same respect.”59 departing from the norm are significant, Even norms deeply embedded in cul- voluntary compliance with the practice can tural and religious beliefs can change. A be universal. For example, one reason that notable recent example in some countries female genital cutting continues to be prac- has been changing norms about homosex- ticed is that parents believe that they are uality. Coupled with legal reform, those acting in the best interests of their daugh- changes together now mean partnerships ters, as young women who are not cut will other than marriages are recognized under not be able to marry in their communities.64 the law, and same-sex couples can marry. In ■■ Elite groups suppress dissent: When the United States, public acceptance of gay social norms benefit an elite group, that and lesbian relationships has increased by group is better able to suppress dissent 19 percentage points in the past 12 years, and maintain the status quo. Laws, for with 59 percent of Americans reporting example, can be deliberately shaped to that they believe such relationships are ensure that men have greater control morally acceptable in 2013.60 Similarly, and economic advantage. Chapter 1 Framing the Challenge: Norms, Constraints and Deprivations | 19 The law acquire, own, manage, and dispose of property. CEDAW has been used effectively by women’s More generally, norms can be enshrined groups to mobilize for reform of discrimina- in law. Different sources of law may reflect tory laws.66 However, as documented by the norms that may be at odds with each other. CEDAW Committee, which undertakes regular International legal rights—such as those reviews at the national level, implementation reflected in the International Convention continues to be hampered by state reserva- on Human Rights—reflect moral and social tions about family and citizenship laws, by fail- norms that are broadly held and prescribe ure to fully integrate their treaty obligations international standards. At the national into national laws, and by weak implementa- level, constitutions embody norms that are tion of the laws that exist.67 Yemen, for exam- supposed to be universal for the nation. Sev- ple, ratified CEDAW in 1984, but then in 1999, enty-five countries have constitutions that introduced a number of regressive laws, such include the principle of nondiscrimination as requiring spousal permission for women to by gender, for example. work outside the home.68 Here, we briefly review the legal con- text, showing how international norms are National Laws generally supportive of equality, but that Regressive laws and legal institutions can national laws often limit women’s agency, impede women’s agency and perpetuate dis- including in their marriages and family life criminatory norms in many ways. What do and in their ability to access and control we mean by regressive? We mean laws that, property. We return to the role of the law in for example, restrict women from working later chapters, including its role in prohib- in certain jobs or restrict married women iting gender-based violence and in enabling from being the head of household. Those are women to realize their reproductive health extensively documented in the World Bank rights and to own land. Group’s Women, Business and the Law data- base and report.69 For example, in the Dem- International norms ocratic Republic of Congo, married women need their husband’s permission to work The 1979 Convention on the Elimination of and to register land or a business. In Egypt All Forms of Discrimination against Women and Jordan, if a woman opts for a no-fault (CEDAW), ratified by 188 states, provides a divorce (khula) to avoid having to establish key foundation for fostering progressive gen- grounds for the divorce, she must forgo all der norms to expand agency.65 It covers such financial claims.70 Custody laws that favor key areas as equality in marriage and family fathers, as in Iraq and Jordon, can act as an life, mobility, citizenship rights, and family for- additional disincentive to divorce. In Nepal, mation, and is a key reference document for married daughters are excluded from inher- this report. Articles 15 and 16 of CEDAW call iting property from their parents. for equal rights for spouses in marriage (for example, to choose where to live and work), Table 1.1 illustrates the ways in which as well as equal rights of women and men to laws can restrict married women’s agency, 20 | Voice and Agency empowering women and girls for shared prosperity TABLE 1.1 Legal restrictions on married women’s agency Restriction (total number of states States in which married women are restricted in brackets) compared to married men Benin, Burkina Faso, Cameroon, Chad, Democratic Republic of Congo, Republic of Congo, Gabon, Guinea, Haiti, Islamic Republic of Iran, Choose where Jordan, Kuwait, Malaysia, Mali, Nicaragua, Niger, Oman, Rwanda, to live (25) Saudi Arabia, Senegal, Sudan, Syrian Arab Republic, United Arab Emirates, West Bank and Gaza, and Republic of Yemen Guinea, Islamic Republic of Iran, Jordan, Kuwait, Lebanon, Madagascar, Confer citizenship on Malaysia, Mali, Mauritania, Nepal, Oman, Saudi Arabia, Sudan, Syrian her children (16) Arab Republic, United Arab Emirates, and West Bank and Gaza Bolivia, Cameroon, Chad, Democratic Republic of Congo, Gabon, Get a job without Guinea, Islamic Republic of Iran, Jordan, Kuwait, Mauritania, Niger, permission (15) Sudan, Syrian Arab Republic, United Arab Emirates, and West Bank and Gaza Travel outside Islamic Republic of Iran, Jordan, Kuwait, Malaysia, Oman, Sudan, the home (9) Syrian Arab Republic, West Bank and Gaza, and Republic of Yemen Travel outside Oman, Saudi Arabia, Sudan, and Syrian Arab Republic the country (4) Source: World Bank and International Finance Corporation, Women, Business and the Law 2014. including their mobility. Women’s choices national statutes.71 The effect of custom- about where to live can be constrained by ary laws on women’s agency varies across family laws and citizenship laws, for exam- countries and depends on the area of ple, such as laws that restrict awarding cit- law—for example, in some matrilineal and izenship to a non-national husband or the matrilocal traditions, women have clear children of their marriage. and independent land rights, although such rights tend to be the exception rather than Customary laws and plural legal the rule. In some cases, customary laws systems may be grounded in discriminatory tradi- Plural legal systems are common; 58 of tional practices,72 as in Kenya, where cus- the 143 countries for which information toms dictate that a married women’s con- is available formally recognize either cus- trol over property can be exercised only tomary or religious law, or both, alongside with her husband’s consent.73 Chapter 1 Framing the Challenge: Norms, Constraints and Deprivations | 21 Out of the 31 countries that recognize 2004, the Supreme Court overruled a case customary law, 11 exempt such laws from in which a lower court had found the cus- the constitution.74 And even where custom- tom that only male children could inherit ary and religious laws are subject to con- their father’s property to be unconstitu- stitutional principles, in practice, constitu- tional, on the basis that customary law tional oversight is often lacking. In Nigeria, could not be undermined merely because it customary law is supposed to be subject to did not recognize a role for women.75 constitutional principles of equality. Yet in *** This chapter has set the stage for the more detailed investigations that follow. We have seen that expanding women’s agency is critical for individual and collective well-being, and for the promotion of shared prosperity and eliminating poverty. Yet, many women—especially those who are poor and lacking education—experi- ence agency deprivations across a range of domains, and they often face multiple constraints simultaneously. These empirical patterns illuminate the need to consider policy and programming design options as part of a comprehensive approach to advancing agency and gender equality. We have explored the importance of social norms and laws and legal institutions as drivers of gender equality outcomes, and made the case for increased policy atten- tion in those areas, which is the focus of the next chapter. 22 | Voice and Agency empowering women and girls for shared prosperity Endnotes 9 Deepa Narayan, Robert Chambers, Meera Kaul Shah, and Patti Petesch, Global Synthesis: United Nations Development Programme 1 Consultations with the Poor (Washington, (UNDP), Human Development Report 2010: DC: World Bank, 1999); Kabeer, “Paid The Real Wealth of Nations—Pathways to Work, Women’s Empowerment”; Ana Maria Human Development (New York: United Munoz Boudet, Patti Petesch, and Carolyn Nations); and Charles Kenny, Getting Better: Turk, On Norms and Agency: Conversations Why Global Development Is Succeeding— about Gender Equality with Women and Men And How We Can Improve the World Even in 20 Countries (Washington, DC: World More (New York: Basic Books, 2011). Bank, 2012); Bina Agarwal, “Gender and World Bank, Annual Report 2013 (Wash- 2 Green Governance: The Political Economy ington, DC: World Bank, 2013); UNESCO, of Women’s Presence within and beyond Teaching and Learning: Achieving Quality for Community Forestry,” (New York: Oxford All, EFA Global Monitoring Report 2013/4 University Press, 2010). (Paris: UNESCO Publishing, 2014); and UN 10 Countries in the study included Afghanistan, Water, “Facts and Figures,” http://www. Bhutan, Burkina Faso, Dominican Repub- unwater.org/water-cooperation-2013/ lic, Fiji, India, Indonesia, Liberia, Moldova, water-cooperation/facts-and-figures/en/. Papua New Guinea, Peru, Poland, Serbia, World Health Organization, Global and 3 South Africa, Sudan, Tanzania, Togo, Viet- Regional Estimates of Violence against nam, West Bank and Gaza, and the Republic Women: Prevalence and Health Effects of of Yemen. Intimate Partner Violence and Non-Partner 11 Sabina Alkire, Pierre Pratley and Ana Vaz, Sexual Violence (Geneva: World Health “Women’s Autonomy in Chad: Measurement Organization, 2013). and Distinctiveness,” Women’s Voice, World Bank, World Development Report 4 Agency and Participation Research Series: 2012: Gender Equality and Development World Bank, Washington, DC, forthcoming. (Washington, DC: World Bank, 2011). 12 World Bank, Republic of Niger: Gender, World Bank and International Finance 5 Agency and Economic Development in Niger Corporation, Women, Business and the Law (Washington, DC: World Bank, forthcoming). 2014: Removing Restrictions to Enhance 13 For more information on KNOMAD, see Gender Equality (London: Bloomsbury its website at http://www.knomad.org/. Publishing, 2013). See also Miriam Temin, Mark Montgomery, Since 1980, women have been living longer 6 Sarah Engebretsen, and Kathyrn M. Barker, than men in all world regions; before 1980, “Girls on the Move: Adolescent Girls men lived longer than women in South Asia. and Migration in the Developing World.” World Development Report 2012. Population Council, Washington, DC, 2013. Esther Duflo, “Women Empowerment 7 ht t p://w w w.popcouncil.org/uploads/ and Economic Development,” Journal of pdfs/2013PGY_GirlsOnTheMove.pdf. Economic Literature, 50 no. 4(2012): 1051- 14 Kabeer, “Paid Work, Women’s Empower- 79; Naila Kabeer, “Paid Work, Women’s ment”; Boudet et al., On Norms and Agency; Empowerment and Inclusive Growth: World Bank, Republic of Niger; World Bank Transforming the Structures of Constraint,” and International Finance Corporation, (New York: UN Women, 2013). Women, Business and the Law 2014; Middle World Bank, Gender at Work: A Companion 8 East and North Africa Region Transport and to the World Development Report on Jobs Energy Unit, Gender and Transport in the (Washington, DC: World Bank, 2014). Middle East and North Africa Region: Case Chapter 1 Framing the Challenge: Norms, Constraints and Deprivations | 23 Studies from West Bank and Yemen, Report MEASURE DHS and ICF International, No. 54788-MNA (Washington, DC: World 2012). Bank, 2011); Chona R. Echavez, “Gender and 24 “Women in Senior Management: Setting Economic Choice: What’s Old and What’s the Stage for Growth,” Grant Thornton New for Women in Afghanistan,” Afghani- International, Chicago, 2013, http://www. stan Research and Evaluation Unit, Kabul, gti.org/files/ibr2013_wib_report_final.pdf. 2012; “Sierra Leone Women Entrepre- neurs,” Cherie Blair Foundation for Women, 25 The recent report of the High-Level Panel 2014, http://www.cherieblairfoundation. on post-2015 stated, ”Barriers [to women’s org/sierraleone-women-entrepreneurs/. opportunity] can only be removed when there is zero tolerance of violence towards Amartya Sen, Development as Freedom, 15 women and girls and when they have full (Oxford: Oxford University Press, 1999). and equal rights in political economics and World Bank, World Development Report 16 public spheres.” High-Level Panel of Eminent 2000/2001: Attacking Poverty (Oxford: Persons on the Post-2015 Development Oxford University Press, 2001), 16. Agenda, A New Global Partnership: Eradicate Measured by the young mother’s forgone 17 Poverty and Transform Economies through annual income over her lifetime. Jad Sustainable Development (New York: United Chaaban and Wendy Cunningham. 2011. Nations, 2013), 17. “Measuring the Economic Gain of Investing 26 Meaza Ashenafi and Zenebeworke Tadesse, in Girls: The Girl Effect Dividend,” Policy Women, HIV/AIDS, Property and Inheritance Research Working Paper 5753, World Bank, Rights: The Case of Ethiopia (New York: United Washington, DC, 2011. Nations Development Programme, 2005). UNESCO, Teaching and Learning, 167. 18 27 Naila Kabeer, “The Rise of the Daughter- Elizaveta 19 Perova and Renos in-Law: The Decline of Missing Women Vakis, “Improving Gender and Development in Bangladesh.” Anokhi Magazine, Octo- Outcomes through Agency: Policy Lessons ber 2012, http://www.anokhimagazine. from Three Peruvian Experiences,” World com/percolator-talk/rise-of-daughter-in- Bank, Lima, 2013. law#sthash.dxZ8zAtj.dpuf. Xiaohui Hou and Ning Ma, “The Effect 20 28 “Child Marriage and Education,” Inter- of Women’s Decision-Making Power on national Center for Research on Women, Maternal Health Services Uptake: Evidence Washington, DC, 2006, http://www.icrw. from Pakistan,” Health Policy and Planning org/files/images/Child-Marriage-Fact- 28, no. 2 (2013): 176–84. Sheet-Education.pdf. Computed from IPU data available in World 21 29 María Inés Berniell and Carolina Sán- Development Indicators for 1990 and yearly chez-Páramo, “Overview of Time Use Data from 1997 through 2012, for representation Used for the Analysis of Gender Differences of women in single or lower house of in Time Use Patterns,” Background paper parliament. for the WDR 2012. Inter-Parliamentary Union: “Women in 22 30 Richard L. Fox and Jennifer L. Lawless, “Gen- National Parliaments,” http://www.ipu. dered Perceptions and Political Candida- org/wmn-e/world.htm. cies: A Central Barrier to Women’s Equality in Electoral Politics,” American Journal of National Institute of Statistics of Rwanda, 23 Rwanda Ministry of Health, and ICF Political Science 55, no. 1 (2011): 55–73. International, Rwanda Demographic 31 Naila Kabeer, “MDGs, Social Justice and and Health Survey 2010 (Calverton, MD: the Challenge of Intersecting Inequalities,” 24 | Voice and Agency empowering women and girls for shared prosperity Policy Brief No. 3, Centre for Development 46 East Asia and Pacific (Cambodia, Indonesia, Policy and Research, London, 2011. the Philippines, Timor-Leste), Europe and 32 World Bank, The World Bank Group Goals: Central Asia (Albania, Armenia, Azerbaijan, End Extreme Poverty and Promote Shared Moldova, Ukraine), Latin America and the Prosperity (Washington, DC: World Bank, Caribbean (Bolivia, Colombia, Dominican 2013). Republic, Guyana, Haiti, Honduras, Nica- ragua, Peru), Middle East and North Africa 33 Kabeer, “MDGs, Social Justice”. (Arab Republic of Egypt, Jordan, Morocco), 34 World Bank, Inclusion Matters: The Founda- South Asia (Bangladesh, India, Maldives, tion for Shared Prosperity (Washington, DC: Nepal), Sub-Saharan Africa (Benin, Burkina World Bank, 2013), 74–75. Faso, Burundi, Cameroon, Democratic Republic of Congo, Republic of Congo, Côte 35 World Bank, Inclusion Matters. d’Ivoire, Ethiopia, Gabon, Ghana, Guinea, 36 United Nations Girls’ Education Initiative, Kenya, Lesotho, Liberia, Madagascar, 2012. Malawi, Mali, Mozambique, Namibia, Niger, 37 World Bank, World Development Report Nigeria, Rwanda, São Tomé and Príncipe, 2012, 74. Senegal, Sierra Leone, Swaziland, Tanzania, Uganda, Zambia, Zimbabwe). 38 World Bank, Maureen A. Lewis and Mar- laine E. Lockheed, Inexcusable Absence: Why 47 For more on economic opportunities, see 60 Million Girls Aren’t in School and What World Bank, Gender at Work. to Do about It (Washington, DC: Center for 48 Boudet et al., On Norms and Agency, 24. Global Development, 2006). 49 John LeJeune and Gerry Mackie, “Social 39 Garcia-Verdú Chioda and Muñoz Boudet, Dynamics of Abandonment of Harmful cited in World Development Report 2012, 75. Practices.” Innocenti Research Centre, 40 World Bank, Inclusion Matters, xvi. United Nations Children’s Fund, Florence, Italy, 2008. Kabeer, “Paid Work, Women’s Empowerment.” 41 50 Boudet et al., On Norms and Agency, 24. 42 Nicola Jones, Elizabeth Presler-Marshall, and Tran Thi Van Anh, Gender Justice: Lis- 51 World Bank, (EN51) World Development tening to the Aspirations and Priorities of Report 2012, 80. Hmong Girls in Vietnam, Country briefing, 52 Men’s focus group, Morey, Département of Overseas Development Institute, London, Keita, Région of Tahoua, reported in Repub- 2013. lic of Niger. 43 Rachel Jewkes, “Intimate Partner Violence: 53 Michel Dorais and Simon L. Lajeunesse, Causes and Prevention,” Lancet  359, no. Dead Boys Can’t Dance: Sexual Orientation, 9315 (2002): 1423–29. Masculinity, and Suicide (Montreal: McGill- 44 Koustuv Dalal and Kent Lindqvist, “A Queen’s University Press, 2004). National Study of the Prevalence and Cor- 54 Paul J. Fleming, Gary Barker, Jennifer relates of Domestic Violence among Women McCleary-Sills, Matthew Morton, “Engaging in India,”  Asia-Pacific Journal of Public Men and Boys in Advancing Women’s Agency: Health 24, no. 2 (2012): 265–77. Where We Stand and New Directions,” 45 Shelley Burchfield and Antoinette Bray- Women’s Voice, Agency, and Participation brook, “Improving Law and Justice Out- Research Series 2013, No.1, World Bank, comes for Indigenous Women and Chil- Washington, DC 2013; Boudet et al., On dren,” Indigenous law Bulletin  7, no. 12 Norms and Agency, Republic of Niger: Gender, (2009): 6–9. Agency and Economic Development in Niger; Chapter 1 Framing the Challenge: Norms, Constraints and Deprivations | 25 Stephanie L. Brown, Randolph M. Nesse, 64 Mackie, Female Genital Cutting. Amiram D. Vinokur, and Dylan M. Smith, 65 There are 99 signatories and 188 parties to “Providing Social Support May Be More CEDAW. See United Nations Treaty Collection Beneficial than Receiving it: Results from a website, https://treaties.un.org/Pages/ Prospective Study of Mortality,” Psychological ViewDetails.aspx?src=TREATY&mtdsg_ Science  14, No. 4 (2003): 320–27; Julianne Holt-Lunstad, Wendy Birmingham, Adam no=IV-8&chapter=4&lang=en. M. Howard, and Dustin Thoman, “Married 66 Mala Htun and S. Laurel Weldon, “The with Children: The Influence of Parental Civic Origins of Progressive Policy Change: Status and Gender on Ambulatory Blood Combating Violence against Women in Pressure.”  Annals of Behavioral Medicine 38, Global Perspective, 1975–2005,” American No. 3 (2009) : 170–79. Political Science Review 106, no. 3 (2012): World Bank, World Development Report 55 548–69. 2012, 80. 67 See “Overview of Convention,” http://www. World Bank, Opening Doors: Gender Equality 56 un.org/womenwatch/daw/cedaw/. and Development in the Middle East and 68 World Bank: Opening Doors. North Africa (Washington DC: World Bank, 2013). 69 World Bank and International Finance Corporation, Women, Business and the Law Gerry Mackie, “Female Genital Cutting: 57 2014. The Beginning of the End,” in Female Circumcision: Multidisciplinary Perspectives, 70 World Bank: Opening Doors. ed. Bettina Shell-Duncan and Ylva Herlund, 71 World Bank and International Finance (Boulder, CO: Lynne Reinner Publishers, Corporation, Women, Business and the Law 2009), 245–82. 2014. UNESCO, Teaching and Learning. 58 72 Tanja Chopra and Deborah Isser, “Access to Naila Kabeer, Simeen Mahmud, and Sakiba 59 Justice and Legal Pluralism in Fragile States: Tasneem, “Does Paid Work Provide a The Case of Women’s Rights,” Hague Journal Pathway to Women’s Empowerment? on the Rule of Law 4, no.(2012): 337–58. Empirical Findings from Bangladesh,” IDS Working Paper 375, Institute of 73 Patricia Kameri-Mbote, “The Land Has Its Development Studies, London, 2011, 36. Owners! Gender Issues in Land Tenure under Customary Law.” paper presented at the “U.S., Record-High Say Gay, Lesbian 60 UNDP–International Land Coalition Work- Relations Morally OK: Americans’ Tolerance shop titled “Land Rights for African Develop- of a Number of Moral Issues Up since 2001,” ment: From Knowledge to Action,” Nairobi, Gallup Politics, May 20, 2013. Kenya, October 31–November 3, 2005. “The Global Divide on Homosexuality: 61 Greater Acceptance in More Secular 74 Those countries are Botswana, Chad, Fiji, and Affluent Countries,” Pew Research Ghana, India, Lesotho, Malaysia, Mauritius, Global Attitudes Project, June 2013, Sierra Leone, Sri Lanka, and Zambia. http://www.pewglobal.org/2013/06/04/ 75 Mojekwu v. Mojekwu, cited by Mary Hallward- the-global-divide-on-homosexuality/. Driemeier and Tazeen Hasan, Empowering World Bank, See World Development Report 62 Women: Legal Rights and Economic 2012, 174, box 4.7. Opportunities in Africa (Washington DC: World Bank, 2012) World Bank, World Development Report 63 2012, 174, box 4.7. Shanghai, China—Mother and daughter riding bicycle by night in the street in Shanghai. The girl is playing video game. Photo: © Thinkstock.com/jacus Chapter 2 Enhancing Women’s Agency: A Cross-Cutting Agenda 28 | Voice and Agency empowering women and girls for shared prosperity The role of cross-cutting goal is to help enable progress by outlining public actions promising directions that can help bring about change. Our framework in figure 1.1 This chapter lays out some key policy and suggests several entry points for policy. Here, programming responses to widespread we focus on changing social norms and the deprivation of voice and agency. We are law, which can have potentially cross-cutting motivated by the fact that many women and significance, alongside programs around eco- girls face multiple and overlapping depri- nomic opportunities, social protection, and vations on their agency, which points to the education, in which well-designed interven- importance of broad-based and systemic tions and new approaches to implementation change. Of particular importance are chang- are proving to have significant benefits for ing regressive social norms, moving beyond women’s agency. There is more general evi- legal reforms and amendments on paper dence which suggests that women’s agency to changes in practice, and addressing the can be positively affected by interventions in multiple forms of disadvantage that women other sectors—transport and infrastructure, and girls often face at the same time. If these for example—which is potentially important fronts can be successfully tackled, real and but not explored here. transformational change could ensue. To help set the scene table 2.1 illustrates a We have seen encouraging elements of selection of programs that have successfully change in practice. In Turkey, for example, enhanced agency at the same time as advanc- extending the compulsory education age ing other development outcomes. We have by three years changed parents’ and girls’ already seen in chapter 1 that education, aspirations for the future—in just five years, especially at the secondary level and above, the share of 15-year-old girls that were mar- reduces the likelihood that women will expe- ried fell by 50 percent and the probability of rience agency deprivations. For example, a giving birth by age 17 fell by 43 percent.1 In social protection program in Malawi pro- Rwanda, the implementation of changes to vided a cash transfer that was conditional land titling rules designed to help women on keeping girls in school. The program had register land has improved access to land for a number of agency-enhancing results: girls married women.2 received more schooling, and early marriage But clearly, results such as these are not yet was reduced along with teenage pregnancy a generalized nor a global phenomenon. Our and self-reported sexual activity of teenage girls. We discuss all the programs illustrated in table 2.1 in greater depth in this chapter The quality of education is likely to be and the chapters that follow. central to the post-2015 global frame- work —it is estimated that 250 million It is important to underline that although children are currently unable to read there is promising evidence, the results are write or do basic arithmetic, 130 million context-specific. Clearly there is no one-size- of whom are in school. fits-all solution, but rather a need to distill promising directions. Chapter 2 Enhancing Women’s Agency: A Cross-Cutting Agend | 29 TABLE 2.1 Selected illustrations of programs to enhance agency Selected outcomes Sexual and Freedom from reproductive health Voice and collective Entry point violence and rights Control over assets action Work with Promundo’s Program In Uzbekistan, men, In Australia, the men, boys, P promotes men’s religious leaders, Male Champions households, and roles as gender- and mothers-in law of Change initiative communities equitable caregivers. are trained on sexual promotes women’s and reproductive leadership. health. Legal reform In Papua New Compulsory In Rwanda, In Burkina Faso, a Guinea, new education laws in Mandatory joint quota law is tied to legislation created Turkey delayed age titling increased federal campaign specialized Family of marriage and married women’s funding and offers and Sexual Violence early childbirth. land ownership to extra funding Units and survivor- over 80 percent. incentives to parties centered training with at least 30 and strengthened percent women. prosecution. Expansion In El Salvador, In Uganda, The Haiti of economic one-stop centers the BRAC Adolescent Girls opportunities provide vocational Empowerment Initiative provides training, access and Livelihoods mentorship services to microfinance, for Adolescents to girls through childcare services, program provided local internships and and crisis support girls with life skills non-governmental for victims of and vocational organizations. violence. training. Social protection In Peru, the Juntos In Malawi, a In West Bengal, In Nicaragua, an Conditional Cash conditional cash the Targeting the Asset Transfer Transfer program transfer program Ultra Poor Program Program that is associated kept girls in reduced food included beneficiary with a 9 percent school longer and insecurity and contact with decrease in physical significantly reduced increased access to community leaders violence and 11 early marriage, assets. changed aspirations percent decrease in teenage pregnancy, and goals. emotional violence. and self-reported sexual activity. Education In India, a Gender The Forum of In Yemen, youth Equality Movement African Women councils taught in Schools Educationalists young women program included Centre of Excellence and men conflict discussions of in 14 countries prevention and violence, emotions provides gender- mitigation skills, management, and responsive training allowing them to relationships. for teachers and work together to sexual education for advocate and to girls. resolve community disputes. 30 | Voice and Agency empowering women and girls for shared prosperity Changing social norms A number of programs show promis- ing new ways to challenge common norms Adverse social norms underpin and rein- about gender roles by engaging men pro- force the multiple deprivations that many actively in change.4 Two current examples women and girls experience. So, what include the following: sorts of policies and public actions can affect these norms? No silver bullet exists, ■■ In Brazil, Promundo supports the imple- mentation of alcohol control policies but evidence suggests that public actions to help limit health risks and self-de- are needed on two broad fronts. First, to structive behavior and to reduce gender enhance women’s and girls’ own sense of violence, alongside programs that help capacity and their aspirations to depart men to access associated health services.5 from existing limiting gender norms and their associated behaviors.3 Second, to ■■ Australia’s Male Champions of Change change behaviors of women and men, boys initiative works with 21 male chief exec- and girls so that social norms become gen- utive officers, department heads, and nonexecutive directors from across Aus- der-equal. As we will discuss in chapter 6, tralian business and the government politics and collective action are important to push for significant and sustain- on both these fronts. able increases in the representation of women in leadership roles as a business Changing norms by working priority.6 with men and boys, households and communities Engaging men in challenging the accept- ability of violence is essential, as the next Norms often have unconscious effects on the chapter shows. Although gender-based vio- decisions people make and the actions they lence is often framed as a “women’s issue,” take, or norms can operate partly outside many groups worldwide are working to of peoples awareness. Norms often persist extend this challenge beyond a single-sex even when circumstances change. movement. For example, the United States– Women’s agency cannot increase in iso- based organization Men Stopping Violence lation from the wider community. Men, boys, includes community activities and training community leaders, and family elders who to engage men, with specifically targeted support gender equality are key allies and programs for fathers as community allies, stakeholders in changing gender norms. in the fight to end violence against women and girls.7 Promundo’s Program P, devel- Increasing women’s agency is not a zero- oped as part of the global MenEngage cam- sum game. Increasing women’s agency need paign, aims to promote men’s practices as not curtail men’s agency. And, as box 1.4 gender-equitable caregivers, in the process showed, men have plenty to gain from gen- preventing violence against women and der equality, which contributes to increased children and transforming the social institu- economic and psychological well-being of all tions that influence men’s caregiving prac- household members. tices.8 In Burundi, men started an awareness Chapter 2 Enhancing Women’s Agency: A Cross-Cutting Agend | 31 campaign by traveling to nearby villages to health issues, existing gender roles, and gen- share personal experiences around their der stereotypes. The program led to a 74 per- transition from violence to nonviolence.9 cent increase in health promotion activities in the pilot regions and a 72 percent increase Interventions that engage with men in knowledge and awareness of reproductive and boys to change violent behaviors need health issues among men and mothers-in-law to start early. We will see in the next chap- who participated in community sessions.15 ter that men who witnessed their mothers being beaten by a partner were more than Where social norms are deeply 2.5 times more likely to have ever perpe- entrenched, as is often the case, it is critical trated violence against their own partners. to design interventions that include com- Recent research from East Asia and the munity leaders and those with the power Pacific shows that 58 percent of men who to endorse change. Successful programs to have perpetrated rape, report having done reduce gender-based violence have enlisted so for the first time as adolescents.10 community leaders and have sought endorsement from the broader community. Family planning programs that include Some examples include the following: men with their spouses can spur greater buy-in and have a strong effect on the uptake ■■ In Senegal, community-based awareness and continuing use of contraception.11 This campaigns enlisted the support of reli- effect is in part because opposition from hus- gious leaders to successfully reduce the bands constrains the use of maternal health accepted practice of female genital cut- care and contraceptives.12 Understanding ting (see box 2.1). such opposition is important. Communica- ■■ In Afghanistan and Ethiopia, community tion is often an influential predictor of family awareness projects have successfully planning, with women who characterize com- reduced the incidence of early marriage.16 munication with their husbands as difficult ■■ In Morocco, the endorsement by reli- or infrequent reporting limited use of family gious authorities of the draft family code planning.13 Conversely, open communica- helped facilitate its passage into law.17 tion and joint decision-making about fam- ily planning can promote contraceptive use ■■ In Jordan, the Religious Leaders Program and the achievement of fertility intentions.14 built the capacities of male and female New approaches to reproductive and sexual religious leaders to advocate for family health have engaged men, senior household health. Six months after the training, reli- members, and community leaders, as well as gious leaders identified several modern women of reproductive age. In Uzbekistan, methods as acceptable under Islamic law for example, a community-based approach and expressed more positive attitudes trained men, religious leaders, and moth- toward family planning.18 ers-in-law in reproductive and sexual health Broadcast media education. These trainees then facilitated dialogue with other men and mothers-in-law Low barriers to access mean that broad- in their communities to discuss reproductive cast media—radio and television—can be 32 | Voice and Agency empowering women and girls for shared prosperity Box 2.1. In Tostan’s footsteps Tostan, which means “breakthrough” in the Wolof language, is an international nonprofit organization working in eight African countries. It was founded in 1991 in Senegal and now operates in more than 450 communities in Djibouti, Guinea, Guinea-Bissau, The Gambia, Mali, Mauritania, Senegal, and Somalia. Tostan spreads awareness of human rights issues using social networks. Known primarily for its success in accelerating the abandonment of female genital cutting, Tostan has also achieved results in governance, health, economic growth, education, and environment, as well as child protection, empowerment of women and girls, and early childhood development. Tostan works through a Community Empowerment Program. It assigns trained facilitators to live in participating villages for three years. The facilitator is fluent in the local language and is from the same ethnic group as the community. In each village, two classes of 25 to 30 participants are held, one for adults and one for adolescents. The classes draw on customary African oral traditions, such as theater, storytelling, dance, artwork, song, and debate in their teaching. Participants are taught basic management and literacy skills to help them implement small projects. In another program, villagers learn about human rights, including the right to health and the right to be free from all forms of violence, and discuss how new practices can help build a healthier community. Participants follow up by speaking with friends and family and by traveling to other villages to raise awareness—an outreach system first used by Demba Diawara, a Senegalese Imam. The Imam recognized that just as tradition requires female genital cutting for a girl to marry, ending the practice requires an agreement between the groups whose children marry one another. The Imam walked from village to village to raise awareness, and consequently, 13 neighboring villages pledged to abandon the practice of female genital cutting. Tostan’s work has led to changes in social norms. One reason is that other communities that are considering abandoning the custom can send delegates to the successful villages to see that an alternative is possible, collective abandonment works, daughters’ reputations remain undamaged, the effort does not bring shame to the community, and no supernatural punishments occur. Another reason is that incorporating participatory human rights teaching into its approach ennobles the process of norm revision—individuals are not rejecting the bad but embracing the good. Tostan’s approach has encouraged 5,423 villages in Senegal to move away from female genital cutting, as well as child and forced marriage. The movement has spread to communities in other African countries, with hundreds of other villages joining the pledge. The government of Senegal has now adopted Tostan’s development model and is working with the organization to end female genital cutting by 2015. Sources: Tostan; Dugger, 2011; and Le Jeune and Mackie, 2008.19 Chapter 2 Enhancing Women’s Agency: A Cross-Cutting Agend | 33 a powerful tool for changing social norms they listen to provide information on a and enhancing women’s agency. In India, broad range of topics such as HIV/AIDS for example, the arrival of cable television and children’s health.23 The programs was associated with significant increases in reduce the feeling of isolation for women women’s reported autonomy, decreases in in an often physically sparse environ- the reported acceptability of wife-beating, ment and provide valuable information and decreases in reported son preference. to them in their own language, empow- Female school enrolment also increased, ering them to deal with issues within along with increased birth spacing. The their family and community. impacts were stronger where women held ■■ In Nepal, the “Samajhdari” weekly more traditional attitudes—in places where radio show instigates productive dia- women had formerly held high preferences logue about sex, teaching women how for sons, the share preferring sons fell 20 per- to speak unhesitatingly; how to think centage points with the arrival of cable tele- critically about their rights and choices vision, compared with a 12 percent decline and about the causes, consequences, and overall.20 Television was also identified as an interconnectedness of HIV/AIDS and important medium through which Afghan gender-based violence; and how to orga- women learned about new laws on gender nize collective action to minimize them.24 equality as well as about male and female ■■ In post-genocide Rwanda, radio pro- relations in other societies.21 In South Africa gramming designed to challenge social and Nicaragua, television shows encourage norms of deference to authority had sub- dialogue about gender issues and challenge stantial impacts in terms of increasing traditional social norms. In South Africa, willingness to express dissent and reduc- exposure to a television series focused on ing the likelihood of listeners deferring domestic violence was linked to an increase to local officials when solving local prob- in help-seeking and support-giving behav- lems. 25 iors. A similar series in Nicaragua was asso- ciated with a 62 percent greater probability Through transmission of messages of having talked to someone about domestic relating to culturally appropriate behavior, violence, HIV, homosexuality or the rights of media play a strong influencing role over youth.22 social and political norms, as individuals and collectives begin to discuss and adopt In rural areas, where television may the messages across all facets of society. be less accessible, radio programming can support social norm change, empowering women, increasing their access to informa- tion and their connection to social support In South Africa, exposure to a television networks. Some examples include: series focused on domestic violence was linked to an increase in help-seeking ■■ In Mozambique, research suggests that and support-giving behaviors. around 95 percent of women listen to community radio and that the programs 34 | Voice and Agency empowering women and girls for shared prosperity And yet, the media can be a double- content and format. In Bangladesh, watch- edged sword. The media can reflect and ing television was associated far more pos- influence social norms and promote posi- itively with women’s empowerment than in tive change, but it can also reinforce gender Egypt. Qualitative analysis in this case sug- stereotypes and perpetuate gender biases, gests that the positive impacts were a result especially among children.26 Recent anal- of Bangladeshi women watching talk shows, ysis of popular family films and television which cover many contemporary issues.28 in the United States, for example, found that males outnumber females by a ratio of A progressive legal 3 to 1 in G-rated family films. This has not framework changed since 1946. More than 80 percent National laws play an important role in sanc- of all working characters in family films pro- tioning violence and discrimination, and duced from 2006-2009 were male and not reforms can help transform social norms one female character was depicted in the surrounding women’s agency. Here, the field of medical science, law, or politics, or focus is on how progressive constitutions as a business leader.27 Evidence from other and law reforms can support gender norm countries similarly suggests a varied impact changes and how to promote their effective of the media on women, depending on the implementation. Box 2.2 New African constitutions reshaping gender norms: South Africa and Kenya The purpose of the South African Constitution is the “creation of a non-racial and non-sexist egalitarian society underpinned by human dignity, the rule of law, a democratic ethos, and human rights.” Women’s movements—including the African National Congress Women’s League, the Women National Coalition of South Africa, and the lesbian, gay, bisexual, and transgender movement—played a critical role in the struggle against apartheid and the framing of the post-apartheid constitution in South Africa. The 1996 constitution enshrines concepts of dignity, equality, and protection of vulnerable groups. Its bill of rights explicitly prohibits discrimination “directly or indirectly against anyone on one or more grounds including … gender, sex, pregnancy, marital status….” South Africa’s post-apartheid constitution is the first in the world to outlaw discrimination based on sexual orientation. Chapter 2 Enhancing Women’s Agency: A Cross-Cutting Agend | 35 Constitutions and principles Constitutions can provide the frame- of equality work to ensure that laws are nondiscrimina- tory and comply with international treaties, There are many examples of constitutions including CEDAW. In doing so, constitutions evolving to accommodate new norms about can promote affirmative protections for gender equality—from women’s suffrage women and girls. For these protections to be in the late 19th and early 20th centuries to effective, all laws, whether statutory, custom- South Africa’s post-apartheid constitution, ary, or religious, must be subject to constitu- the first in the world to outlaw discrimina- tional protections on non-discrimination. tion based on sexual orientation. Several Constitutional oversight can be an new constitutions during the last decade, effective tool to address discrimination. including the Kosovo Constitution (2008), Non-governmental organizations, women’s the Tunisia Constitution (2014), and the networks, and other groups have used the South Africa and Kenya Constitutions (box constitution to challenge discriminatory 2.2), reflect widespread popular consulta- statutory or customary laws in court, as in tions and lobbying by women’s networks South Africa, where customary inheritance and embody principles of nondiscrimina- laws that favor male heirs were declared tion and gender equality. unconstitutional and invalid in 2005.30 In The constitution affirms the role of customary law within the country’s legal framework and makes all laws subject to the principles of non-discrimination. The new constitution catalyzed new legislation and case law, which reshaped the definition of family, relations between family members, and the meaning of marriage. The 1998 Domestic Violence Act, for example, covers “persons who are or were married to each other according to any law, custom, or religion and persons of the same or opposite sex who live or have lived together in a relationship in the nature of a marriage.” In Kenya, Article 60 of the 2010 constitution specifically calls for “the elimination of gender discrimination in law, customs, and practices related to land and property.” It also established constitutional oversight of customary law, and international treaties automatically become part of the national legal framework upon ratification. The new constitution requires the government to take legislative and other measures to redress any disadvantage suffered by individuals or groups because of past discrimination. These reforms have already influenced how disputes are decided in courts. In 2011, a judge declared that a divorced daughter was allowed to inherit her parents’ estate because the constitution eliminated all gender discrimination related to land. But in 2014, a new Matrimonial Property Act stripped women of their rights to family property unless they can prove financial contribution. Sources: Andrews, 2005; Sloth-Nielsen and Van Heerden, 2010; Rukunga v. Rukunga, 2011.29 36 | Voice and Agency empowering women and girls for shared prosperity Kuwait, the constitutional protection of Women often face multiple barriers to the equality under the law was used success- full enjoyment of their legal rights. And the fully in 2009 to challenge discriminatory individuals and institutions responsible for laws that required a husband’s signature on implementing the law too often do not have a wife’s passport application.31 full knowledge or understanding of what the law provides or do not have the resources It is, however, very difficult to address and capacity to implement and enforce the structural inequalities and discriminatory law. Public officials posted to remote rural social norms using formal laws alone. In areas, for example, may be unaware of new Rwanda, women can claim land equally laws. On both sides, norms that reinforce the under statutory law but the discriminatory status quo are a constraint. social norms, practices, and power relations that underlie the customary application of At best, statutory courts will almost always land rights continue.32 Reforms designed to be the last resort, especially in developing overcome discriminatory customary prac- countries and for the poor. Scarcity and geo- tices need to be accompanied by processes graphical remoteness of courts, weak capacity that address underlying norms, for example, of legal personnel and legal services, and a lack by working with traditional leaders to incor- of legitimacy and accountability of legal insti- porate human rights values into tribunals tutions—particularly in fragile and conflict-af- or other customary decision-making bod- fected states—may make the formal justice ies and by including women as judges and system inaccessible for most. Measures such decision makers. For example, in Uganda, 30 as lowering court fees, providing legal aid, percent of judges in local council customary paralegal services, mobile court services, and courts must legally be women, and opera- providing training and institutional capacity tional guidelines have been developed to building for police, lawyers, paralegals, court incorporate human rights and gender sensi- staff, and judges can all help expand the access tization training of court staff.33 and reach of the formal justice system. In Jor- dan, the World Bank is helping to provide low- Supporting effective cost legal services for women,34 and in Indo- implementation and enforcement nesia, paralegal services are helping women bring claims to court and to access basic iden- For progressive laws to make a difference in tification documents, such as birth and mar- practice people must be aware of their rights riage certificates.35 The ability to access this and all citizens must have equal access to type of documentation can improve women’s justice. Yet, as we saw in the last chapter and ability to access other services, including land will see in our discussions of violence against titles, as reviewed in chapter 5. women and land and property ownership in the chapters that follow, this equality is often Partnerships can help overcome barri- not the case: rural women and men often ers to access to justice and help implemen- have limited awareness and limited access tation. For example, service providers and to legal aid and appeal mechanisms, which advocates working on women’s economic prevents them from exercising their rights. empowerment or protection from violence Chapter 2 Enhancing Women’s Agency: A Cross-Cutting Agend | 37 can forge effective partnerships with legal services. Such partnerships can encompass People’s awareness of laws makes a legal aid providers teaming up with domes- difference. In eight countries men who tic violence counselors in women’s shelters reported awareness of laws addressing or bundling the delivery of legal aid with violence against women were nearly existing services women already access, 50 percent more likely to prevent a such as midwifery services or microcredit stranger’s act of violence. schemes.36 The Philippine Judicial Academy and the Ateneo Human Rights Centre worked together to create an interactive guide on new ways to share information quickly and gender-based violence, which was used to providing effective ways to communicate track legal cases, provide technical guidance with peer groups and to reach out to groups for judges, and train court personnel.37 who may otherwise be excluded (these issues Awareness of laws makes a difference. For are discussed in greater depth in chapter 6). example, in eight countries men who reported awareness of laws addressing violence against Expanding access to justice women were nearly 50 percent more likely to through customary processes prevent a stranger’s act of violence.38 Dissem- At least 58 countries formally recognize ination of information about laws and legal either customary or personal law,42 and rights can be effective to raise awareness. The Huairou Commission’s Tool for Change, a man- customary, religious, or community-based ual on human rights law, seeks to raise aware- redress mechanisms may be the most read- ness of land rights and HIV/AIDS in Cameroon, ily available dispute resolution instruments Uganda, and Zambia.39 Legal empowerment for most people, especially the poor or those strategies, through legal literacy programs, living in rural areas. The United Kingdom legal aid, or alternative dispute resolution Department for International Development mechanisms, can also help ensure that prin- has estimated that in some countries, as ciples of equality and nondiscrimination are much as 80 percent of the population turn to translated into practice.40 Kosovo’s Law on informal justice systems to resolve disputes.43 Reproductive Health, for example, guarantees For example, in Tanzania, where 44 percent the right of information and education on sex- of women reported spousal abuse in 201044 ual and reproductive health for every person and where reporting of such incidents to the and charges the Ministry of Health to provide police is very low, community-level redress information, education, and advice on repro- mechanisms may provide an effective viable ductive health.41 alternative to statutory mechanisms.45 Collective action, including by women’s Understanding the interface between groups, often plays an important role in informal and formal justice and how women transmitting information and raising aware- negotiate their way through plural sys- ness about legal and other rights. Information tems is therefore critical. The same barriers and communication technologies are offering can impact their uptake of services in both 38 | Voice and Agency empowering women and girls for shared prosperity settings, including limited resources, cultural for promoting agency through expanding biases among those responsible for uphold- economic opportunities. Research on norms ing the law, and adverse gender norms. Stat- and agency that draw on data from women utory reforms undertaken to improve wom- and men in 20 countries in all world regions, en’s access to informal mechanisms have, so concluded that “women’s ability to work for far, had minimal success,46 although there pay, which most women in the study aspired are some promising directions. In Namibia, to, may be one of the most visible and the involvement of more women leaders and game-changing events in the life of mod- the integration of gender equality provisions ern households and all communities.”49 The into the customary system enhanced the per- study found that women’s agency increased ceived fairness and equity of traditional rule not only from the direct effects of economic and the customary dispute resolution process participation but also as women gained a for women.47 Similarly in Bougainville, Papua greater sense of self-efficacy, broadened New Guinea, women mediators were much their aspirations, and forged new ways to more effective at reaching equitable solu- reconcile their identities as workers with tions in cases of domestic violence, including their identities as mothers. But not all work threatening perpetrators with action in the is equally empowering—working conditions state court if the violence did not stop and matter, as does the type of work that women simultaneously informing the victim of her do. Naila Kabeer concludes from an analysis right to refer the case to the court and provid- of Bangladesh, Egypt, and Ghana that “for- ing information on how to do so.48 mal and semi-formal employment is found to be most likely to contribute to women’s Increasing women’s agency ability to decide on the use of their income, through sectoral policies to make decisions about their own health, to and programs gain respect within the community, to par- ticipate in politics, and to express support We turn now to entry points through selected for a more equitable distribution of unpaid sectoral programs, namely: expanding eco- workloads and, in cultures characterized nomic opportunities and training; design- by son preference, less discriminatory atti- ing gender-responsive social protection; tudes towards their daughters.”50 and increasing access to quality education. Expanding economic opportunities is a Well-designed policies and new approaches major policy challenge. Opportunities for to program implementation in these sectors women to work are opening up slowly at best are proving to have significant positive exter- in most regions—globally, women’s labor nalities for women’s agency, as well as often force participation has actually fallen slightly improving results and overall effectiveness. since 1990, from 57 percent to 55 percent.51 As the World Bank Group’s 2014 report Gender at Expanding economic Work52 shows, compared to men, a wide range opportunities and training of measures show that women are at a disad- One important conclusion emerging from a vantage at work. The measures include gender wide range of work on agency is the potential wage gaps, labor market participation, ability Chapter 2 Enhancing Women’s Agency: A Cross-Cutting Agend | 39 to get work in the most profitable sectors and enterprises, and the balance between paid and Women’s agency cannot increase in unpaid work. Labor market and growth poli- isolation from the wider community. cies for job creation need to include measures Men, boys, community leaders, and to address the gender-specific constraints that family elders who support gender prevent women accessing productive jobs, equality are key allies and stakeholders such as child care responsibilities, lack of skills in changing gender norms. and training, and gender-biased attitudes and expectations about women’s abilities. Further, while jobs can create agency, women’s lack of that provide vocational training, access to agency limits their ability to get good jobs and microfinance, child care services, crisis sup- benefit from them. Gender at Work identifies port for victims of violence, and community policy actions to address women’s agency con- education that targets gender norms and straints across the lifecycle and shows how the women’s health.55 private sector—the main creator of jobs—can Training programs have a mixed record lead and innovate for gender equality. of success and need to be well grounded Multisectoral policy interventions that in the needs of the labor market as well as address agency as part of efforts to pro- the needs and constraints of participants.56 mote economic opportunities can do better Combining work and training increases on both fronts. Indeed, failure to do so can the success rates of training programs, stymie efforts to promote economic oppor- and programs that aim to increase wom- tunities.53 Promising interventions combine en’s job opportunities need to be designed training in vocational business or trade-re- to equip women and girls with the skills lated skills with programmatic elements to needed by employers and enterprises.57 increase aspirations and confidence; social That approach has yielded examples of networks; freedom from violence; and promising programs with positive impacts healthy gender dynamics in the household. for women and girls. To be successful, pro- grams need to incorporate specific design Evidence is emerging that economic features. Box 2.3 shows how the World empowerment programs can reduce wom- Bank’s Adolescent Girls Initiative has delib- en’s exposure to violence when components erately sought to tackle gender biases. geared toward building women’s collective Programs that tackle gender norms and voice and fostering more gender-equitable provide young women and girls with new relationships are included. For example, this information can open up new opportunities. evidence was found for microcredit inter- The following are some examples: ventions in South Africa.54 One-stop shops that support economic empowerment as ■■ Life-skills training and provision of safe well as broader agency-related objectives spaces for girls to discuss their future, are also showing promise. For example, in and the development of support net- El Salvador, Ciudad Mujer establishes com- works, have reduced the risk of early munity-based one-stop centers for women marriage in Ethiopia.59 40 | Voice and Agency empowering women and girls for shared prosperity Box 2.3 How the Adolescent Girls Initiative is helping to overcome agency constraints In 2008, the World Bank launched the Adolescent Girls Initiative (AGI) to generate evidence on how to promote young women’s transition to productive work. Interventions are being implemented and rigorously evaluated in eight countries: Afghanistan, Haiti, Lao People’s Democratic Republic, Liberia, Jordan, Nepal, Rwanda, and South Sudan.58 The box outlines some important lessons on program design, implementation, and monitoring and evaluation to help improve programming decisions. Constraint, Design and Responses ›› Several AGIs use rapid market assessments to identify labor demands and train young women in relevant skills. ›› In Nepal, Liberia, and Haiti, the AGI trains young women to enter more Skills lucrative nontraditional fields. ›› In Liberia, the AGI provides literacy training, which is important for post- conflict settings. ›› AGIs help to build self-confidence and “soft skills” through life skills training. ›› In South Sudan, the AGI created safe-space community clubs and the project Norms goes door-to-door to recruit harder-to-reach younger girls. ›› In Afghanistan, the program recruits participants through the community Shura (elders). ›› In Liberia, classes are held at different times so participants can balance training with other activities. ›› Several AGIs provide participants with stipends to offset costs of attendance, Chores which can be prohibitive for poor young women. ›› AGIs in Nepal, Liberia, and South Sudan offer child care services to young mothers. ›› In Liberia, coaches support and mentor trainees. Service providers are responsible for placement. ›› In Haiti, the AGI offers mentorship services to participants through internships Networks and local nongovernmental organizations. ›› In Nepal, the AGI provides financial incentives to training providers who train and place vulnerable young women. Chapter 2 Enhancing Women’s Agency: A Cross-Cutting Agend | 41 ■■ In Uganda, the BRAC Empowerment and Standards for work are important if work is Livelihoods for Adolescent program tar- to contribute to people’s agency and well- gets girls and women aged 14 to 20 and being. The International Labor Organiza- consists of a social space, where the girls tion’s Decent Work Agenda63 gives guidance and women receive life-skills training for actions that can improve the quality of and local market-informed vocational work and target gender inequalities and training. So far, the program has signifi- provides important benchmarks to ensure cantly improved control over the body that creating jobs goes hand in hand with for adolescent girls, with a 26 percent increased agency. reduction in rates of early childbearing and a 58 percent reduction in rates of Designing gender responsive marriage and cohabitation. A 50 per- social protection cent increase in condom use among the Social protection can be transformative. sexually active was also reported. The Programs that go beyond protection per se share of girls reporting sex against their and include elements to tackle regressive will dropped from 21 percent to almost gender norms by: increasing women’s voice zero.60 and participation; building their aspirations; ■■ A study of graduates of a year-long ICT and, opening up new economic opportuni- course in Mumbai found that women’s ties have had promising results.64 Such ele- self-confidence was boosted. Female and ments have included addressing child care male respondents highlighted different responsibilities; increasing access to finance reasons for encouraging others to learn to and assets and tackling other constraints use computers, including a greater num- that limit work options; increasing skills, ber of girls (56 percent) than boys (40 self-confidence, and aspirations; incentiviz- percent) who noted job-related benefits. 61 ing girls’ schooling; and providing informa- ■■ Similarly, the Intel Learn program, which tion and building awareness about social has provided skills training with a focus issues and rights. on technology literacy, problem solv- The BRAC Challenging the Frontiers of ing, critical thinking, and teamwork to Poverty Reduction–Targeting the Ultra Poor 875,000 girls and young women from (TUP) program channels skills training and 16 countries, has been found to increase assets to ultra-poor women in Bangladesh participants’ self-confidence and sense of power.62 “It was the first time we three had Jobs strategies that pay attention to gen- stepped out of this village on our own. der equality issues, diagnosing the main How could our husbands stop us from constraints that women face in accessing joining Bandhan? They had no financial jobs and identifying and addressing gen- strength to do so.” der-based disparities in access to skills, —Female participant in the Targeting the Ultra Poor pilot, training, and credit, are an important West Bengal, India65 entry point for increasing women’s agency. 42 | Voice and Agency empowering women and girls for shared prosperity and is proving transformational. A Random- Pakistan found that mental health, ized Control Trial (RCT) finds that partici- happiness, and hope rose among partici- pants spend more time in self-employment pants. In one of the India TUP programs, and less time in wage labor, and increase women reported that abuse had stopped their labor market participation, which leads since their economic conditions had sta- to a 36 percent increase in annual income bilized, because men had less “idle” time on average. Benefits spill over to other poor on their hands, reducing alcoholism and women as wages increase at the village subsequently domestic violence.72 level.66 A RCT of a TUP program in rural At the same time, however, there are cases West Bengal, India, found increases in per in which CCTs have reinforced traditional capita household consumption exceeded 25 gender roles and worsened time constraints percent, along with reduced food insecurity, by making mothers primarily responsible for increased assets, and improved emotional various tasks linked to children. This prob- well-being.67 Early findings from pilot pro- lem was found in Bolivia, Ecuador, and Peru,73 grams modeled on TUP in Haiti, Honduras, indicating that context is important. Violence India, and Pakistan suggest these results can against women is not always reduced, as for be replicated.68 example in Mexico and Ecuador, where there is evidence of increased emotional violence “Through experience one learns and triggered by CCTs.74 opens up towards the future. By talking Including elements that increase partici- to others one understands and learns.” pants’ contact with natural leaders from the —Female program participant in Atención a Crisis, community into the design of a social protec- Nicaragua.69 tion program is a promising new approach. A recent experimental study found that social interactions with nearby female lead- In countries worldwide, women report ers substantially increased the program’s a valued sense of security, self-esteem, and impact on human capital investments and enhanced psychological well-being because income diversification, as well as attitudes of cash or asset transfers, including in Ban- about the future. Nine months after the pro- gladesh, Mexico, Nicaragua, Peru, Uganda gram, incomes of beneficiaries who inter- and West Bank and Gaza.70 These benefits acted with leaders increased almost twice can spill over to women’s agency in other as much as the control groups. The effect is areas of their lives. Here are some examples: attributed to the role that increased social ■■ Enrollment in Peru’s Juntos conditional interaction plays in raising aspirations and cash transfer (CCT) program has been changing attitudes.75 associated with a 9 percent decrease In some countries, pensions for women in physical violence and an 11 percent have been shown to increase investments in decrease in emotional violence.71 girls. In South Africa, the effect was to sig- ■■ Early findings from RCTs of TUP pro- nificantly increase their granddaughters’ grams in Haiti, Honduras, India, and educational enrollment and to improve Chapter 2 Enhancing Women’s Agency: A Cross-Cutting Agend | 43 health outcomes for girls. These effects did experienced fewer constraints to their not pertain to boys, and neither boys nor agency and that more educated women girls were affected when men were the ben- have greater sexual autonomy (figure 1.6). eficiaries.76 Similarly, in Brazil, a monthly The impact that education has on women’s old-age pension of R$100 (about US$43.50 agency is illustrated by this example from in 2013) increased girls’ school enrollment rural Bangladesh, where, when girls were rates by 10 percent. Again, no such effects asked how education has made their lives were found for boys, and impacts on girls different from their mothers’, they typically were mostly attributable to female rather replied that it had helped them “find a voice,” than male beneficiaries.77 allowed them to “have a say,” to “speak,” and “to be listened to.”81 Increasing incentives for girls’ school- ing through CCTs can be an effective way Increasing school enrollment and to tackle regressive gender norms. This achieving gender equality in enrollment are approach has been successful in Bangladesh, longstanding development goals. Ensuring Pakistan, and Turkey, for example.78 A recent school enrollment through upper secondary systematic review found that CCTs tend to levels for girls is even more critical. Equally have larger effects on enrollment, especially important is the quality of education that girls’ enrollment, than unconditional cash girls and boys receive in school. Studies also transfers, particularly when explicit school- show that girls are subject to educational ing conditions for payments are monitored streaming and stereotyping that restricts and enforced.79 Benefits to girls of staying their learning and, in turn, contributes to in school can extend beyond the value of gender sorting in the labor market.82 Improv- educational attainment, enhancing their ing the quality of education is likely to be agency in other ways as well. For example, in more central to the post-2015 global frame- Malawi, a CCT targeting 13- to 22-year-old work than access alone—it is estimated that girls and women led to recipients staying 250 million children are currently unable to in school longer and to significant declines read, write, or do basic arithmetic, 130 mil- in early marriage, teenage pregnancy, and lion of whom are in school.83 self-reported sexual activity.80 Education increases women’s agency, and more educated women are often more able to Increasing gender equality take decisions and have choices, even where in education gender norms are restrictive. In South Asia Education’s importance for development and the Middle East and North Africa, women and gender equality is widely recognized— education increases skills and capabilities and allows people to live longer, healthier, “Education is a girl’s best weapon and more productive lives. It can also help to face the world.” shift gender norms that restrict women’s —Girl from West Bank and Gaza84 voice and choice. For example, we saw in figure 1.4 that women with more education 44 | Voice and Agency empowering women and girls for shared prosperity with more education are less likely to have to children; to recognize common illnesses and ask their husbands’ or families’ permission to to treat them; to seek help at the right time; seek medical care; women with more educa- and to use health care services effectively. tion also tend to marry later and have fewer Policies that get more girls and boys into children, as discussed further in chapter 4.85 school are critical but equally important is Among some Roma in Bulgaria, more educated what happens when they get there—we must women are renegotiating gender roles in com- ensure that both girls and boys leave school munities, challenging traditional gender roles, literate and numerate and that the values of and demanding men play a more active role in the school system promote gender equality child care and household chores.86 and protect children from abuse. This requires Enhanced agency is a key reason why building allegiances between adults and children of more educated women are less teachers and boys and girls to provide both likely to have stunted growth: educated safe and supportive school environments free mothers have greater autonomy in making from gender-based violence, bullying (box decisions and more power to act in their 2.4) and textbooks and curricula free from ste- children’s benefit.87 Educated mothers are reotypes and discrimination (box 2.5). It also more likely to take preventative actions, entails training teachers to embody equitable such as purifying water and vaccinating their attitudes and practices and to pay specific Box 2.4. Promoting gender equality in school and improving the learning environment Girls’ and sometimes boys’ ability to acquire an education is tragically disrupted by gender-based violence. In schools where sexual violence against girls is the norm, the education system itself may increase a girl’s chances of dropping out, interrupting her studies, experiencing an unintended pregnancy, or becoming infected with HIV. Addressing harassment and violence in schools is important to promote gender equality in education. South Africa’s Girls’ Education Movement is a civil society organization seeking to promote gender equality and to address the very high levels of violence experienced in schools through student- led clubs and community workshops. It has created a network for improving girls’ education, knowledge, and self-esteem, with boys and adults as allies. The clubs equip children to address issues of education, clean and safe school environments, and good communication among peers, as well as other things. It gives children the opportunity to work together to find solutions to the problems affecting their school lives. The program has seen some progress in reducing aggressive behavior and conflict and in changing attitudes of the students and program leaders who received training. But the results varied significantly by context, and impacts on the broader school environment were not significant. Findings suggest that changes in underlying norms around the acceptability of violence and the unequal treatment of girls is needed to realize the full impact of the program. Sources: Wilson, 2009.88 Chapter 2 Enhancing Women’s Agency: A Cross-Cutting Agend | 45 Box 2.5 Curriculum reform to promote gender equality through education Although global progress has been made in increasing girls’ school enrollment, gender equality in education encompasses much more than access. Social norms influence textbooks, curricular choices, the sex distribution of teachers and administrators, teacher attitudes and behavior, classroom and discipline practices, and the presence of violence—all of which shape opportunities for boys and girls. National curricula can reinforce existing social and gender inequalities by implicitly upholding traditional gender stereotypes or by disregarding the diversity of learning needs and learning styles among girls and boys. At the same time, schools have enormous potential to effect social change and to transform gender relations by expanding the range of possibilities for both boys and girls. One avenue for transformation is the gender sensitization of school curriculum, including revising textbooks. Research suggests that the treatment of girls and women in school textbooks is broadly similar across countries. There are fewer images of women and girls compared to men and boys; men and women are generally portrayed in stereotypical roles and professions; men and women are described as having specific gendered attributes; and contributions of important women are ignored or given less consideration than those of important men. An increasing number of countries are taking action to address these biases. For example, in Thailand, policies have been instituted directing textbooks to be revised and all gender stereotypes to be removed. In Vietnam in 2009, the Vietnamese Ministry of Education, supported by 12 UN agencies, conducted a national textbook review and developed teacher-training modules to promote gender equality, while in Indonesia, teachers are being trained in gender-sensitive methodologies to engage boys and girls. Similarly, the Australian National Curriculum directs teachers to address certain issues within their classes—for example, the Year Seven English curriculum recommends that teachers teach how languages and images create character by identifying and challenging gender stereotypes in girls’ and women’s magazines and popular television programming. Mainstreaming gender into curricula goes beyond revising textbooks and requires incorporating gender equality across the range of studies. Evidence suggests that intensive and long-term attention to gender as part of school coursework has an impact on improving children’s attitudes and behaviors related to gender equality. In Mumbai, India, for example, the Gender Equity Movement in Schools, which includes sessions taught over two years focused on understanding gender roles, power dynamics, violence, and physical and emotional changes related to adolescence, led to participating children reporting more equitable attitudes toward gender. In Tanzania, the national syllabi for secondary schools contain topics related to gender as part of civics education and national examinations. Sources: Connell, 2010; UNGEI, 2012; Levtov, forthcoming; Achyut et al., 2011; and Australian Curriculum, Assess- ment, and Reporting Authority.89 attention to understanding gender relations appropriate disciplinary measures can help and building competencies beyond educa- to support and monitor student and teacher tional basics. behavior. Working with teachers and school staff members is essential, as is engag- Addressing violence in schools requires ing parents and the wider community and a comprehensive, multilevel approach. working directly with students, for example: Establishing school policies and commit- tees to address violence and putting in place ■■ In Tanzania, female teachers were 46 | Voice and Agency empowering women and girls for shared prosperity trained as guardians for female students, Nations Children’s Fund (UNICEF), is with the result that girls felt there was designed to improve the quality of public someone to turn to for advice about primary schools and to create gender-sen- school-based violence and the wider sitive, healthy, safe, and inclusive learning community began to confront sexual vio- environments. A 2009 evaluation across lence more openly.90 six countries found that the approach cre- ■■ The Gender Equality Movement in ated school environments where female Schools program in Indian middle schools students feel included, safe, supported and included discussions of violence in the challenged.92 context of gender relations, emotional ■■ The Forum of African Women Educa- management, and relationships between tionalists Centre of Excellence model boys and girls. Participating students has transformed ordinary schools in 14 reported less tolerance of violence and African countries into gender-respon- greater likelihood of taking positive action sive schools that focus on the physical, in response to peer violence compared to academic, and social dimensions of both students who did not participate.91 girls’ and boys’ education. Elements Promising evidence is emerging on the include gender-responsive training effectiveness of programs that incorporate for teachers; an emphasis on science, elements, such as reform of textbooks and mathematics, and technology for girls; curricula and the provision of safe spaces empowerment training for students; a for girls and boys. Some examples include sexual maturation management program the following: for girls; gender-responsive school infra- ■■ The Child Friendly Schools model, imple- structure; and community involvement in mented in multiple countries by the United school management.93 *** Constraints to agency stem from many sources. Often, deeply rooted social norms are an unrecognized driving factor behind the actions taken by men and women, boys and girls, sometimes entrenched by laws on paper and in practice. The good news outlined in this chapter is that there is a range of polices that can work to promote change across a broad spectrum. We also saw how such programs are working to couple development results in terms of incomes and service access together with increases in women’s agency. Indeed, we saw that improvements in women’s agency are sometimes necessary for other desired development results to be achieved. This is an important stage to set for the examination of major agency deprivations— freedom from violence, control over sexual and reproductive health, land and hous- ing ownership, and participation in politics and collective action—and the more targeted policy and program responses in the chapters that follow. Chapter 2 Enhancing Women’s Agency: A Cross-Cutting Agend | 47 Endnotes Multi-Country Cross-Sectional Study on Men and Violence in Asia and the Pacific,” Lancet 1 Murat Kýrdar, Meltem  Tayfur, and  Ýsmet Global Health 1, no. 4 (2013): e208-e218. Koc,   “The Effect of Compulsory Schooling Laws on Teenage Marriage and Births in Tur- 11 Lisa Mwaikambo, Ilene Speizer, Anne Schur- key,” Munich Personal RePEc Archive Paper mann, Gwen Morgan, and Fariyal Fikree. 38735, Munich University, 2012, http:// “What Works in Family Planning Interven- mpr a .ub.u n i-muenchen .de/38735/1/ tions: A Systematic Review,” Studies in Family MPRA_paper_38735.pdf. Planning 42, no 2 (2011): 67–82. Daniel Ayalew Ali, Klaus Deininger, and 2 12 See chapter 4: Fertility choices. Markus Goldstein, “Environmental and Gen- 13 C. H. Kim and S. J. Lee, “The Role of Husbands der Impacts of Land Tenure Regularization in in Family Planning Behavior,” Psychological Africa: Pilot Evidence from Rwanda,”  Policy Studies in Population/Family Planning 1, no. Research Working Paper Series 5765, World 5 (1973): 1–23, Douglas Storey, Marc Boulay, Bank, Washington, DC, August 2011 Yagya Karki, Karen Heckert, and Dibya Man Karmacha, “Impact of the Integrated Radio Elizaveta Perova and Renos Vakis, “Improving 3 Communication Project in Nepal, 1994– Gender and Development Outcomes through 1997,”  Journal of Health Communication  4, Agency: Policy Lessons from Three Peruvian no 4 (1999): 271–94, Ann E. Biddlecom and Experiences.” (Lima: World Bank, 2013). Bolaji M. Fapohunda, “Covert Contraceptive Paul J. Fleming, Gary Barker, Jennifer 4 Use: Prevalence, Motivations, and Conse- McCleary-Sills, and Matthew Morton, quences,” Studies in Family Planning 29, no 4 “Engaging Men and Boys in Advancing Wom- (1998): 360–72. en’s Agency,” background paper to the report 14 R. Amatya, H. Akhter, J. McMahan, N. William- on Women’s Voice, Agency, and Participation, son, D. Gates, and Y. Ahmed, “The Effect of World Bank, Washington, DC, 2013. Husband Counseling on NORPLANT® Con- Fleming et al., “Engaging men and boys.” 5 traceptive Acceptability in Bangladesh,” Con- See Australian Human Rights Commission, 6 traception  50, no. 3 (1994): 263–73; Alex “Male Champions of Change,” http://www. Chika Ezeh, “The Influence of Spouses Over humanrights.gov.au/male-champions-change. Each Other’s Contraceptive Attitudes in Ghana,” Studies in Family Planning 24, no. For more information see Men Stopping Vio- 7 3 (1993): 163–74; Sitawa R. Kimuna and lence, “Educate and Advocate for Change,” Donald J. Adamchak, “Gender Relations: http://www.menstoppingviolence.org. Husband–Wife Fertility and Family Planning Promundo, CulturaSalud, and REDMAS, 8 Decisions in Kenya,” Journal of Biosocial Sci- Program P: A Manual for Engaging Men in ence 33 (2001): 13–23. Fatherhood, Caregiving, Maternal, and Child 15 German Society for International Cooper- Health (Rio de Janeiro and Washington, DC: ation (GIZ), “Mahallas, Men, and Mothers- Promundo, 2013). in-Law: Community-Based Approaches on Hilde Wallacher, “Engaging Men: The 9 Awareness-Raising in Reproductive Health,” Abatangamuco and Women’s Empowerment Regional Programme Health in Central in Burundi,” Policy Brief 5, Peace Research Asia—Uzbekistan, Gender Pays Off, GIZ Gen- Institute Oslo, Oslo, 2012. der Competition, Bonn, Germany, 2012. 10 Rachel Jewkes, Emma Fulu, Tim Roselli, 16 Gerry Mackie, “Female Genital Cutting: The and Claudia Garcia-Moreno, “Prevalence of Beginning of the End,” in Female Circumci- and Factors Associated with Non-Partner sion: Multidisciplinary Perspectives, eds. B. Rape Perpetration: Findings from the UN Shell-Duncan and Y. Hernlund (Boulder, CO: 48 | Voice and Agency empowering women and girls for shared prosperity Lynne Reiner Publishers, 2000), 245–82. A. An Experimental Intervention Using Mass Malhotra, A. Warner, A. Gonagle, and S. Lee- Media to Change Norms and Behavior in Rife. Solutions to End Child Marriage: What Rwanda.” American Political Science Review. the Evidence Shows (Washington, DC: Interna- 26 B. A. Browne, 1998. “Gender Stereotypes in tional Center for Research on Women, 2011). Advertising on Children’s Television in the Mary Hallward-Driemeier and Tazeen 17 1990s: A Cross-National Analysis,” Journal of Hasan,  Empowering Women: Legal Rights Advertising, 27, no. 1 (1998), 83-96. and Economic Opportunities in Africa (Wash- 27 Stacy, L., Smith, and Crystal Allene Cook. ington DC: World Bank, 2012). 2008. “Gender Stereotypes: An Analysis For more information see Center for Commu- 18 of Popular Films and TV.” The Geena Davis nication Programs, Jordan Health Commu- Institute on Gender in Media. nication Partnership, http://www.jhuccp. org/whatwedo/projects/jordan-health- 28 Kabeer, “Contextualizing the Economic communication-partnership-jhcp. Pathways of Women’s Empowerment.” Tostan, “The Tostan Model,” http://www.tos- 19 29 Penelope Andrews, “Perspectives on Brown: tan.org/tostan-model; Celia W. Dugger, “Sene- The South African Experience,” New York Law gal Curbs a Bloody Rite for Girls and Women,” School Review 49 (2005): 1155 (citing Minis- New York Times, October 15, 2011; and John ter of Finance & Another v. Van Heerden 2004 Le Jeune and Gerry Mackie, “Social Dynamics [11] BGLR 1125 [CC] at para. 26 [S. Afr.]); of Abandonment of Harmful Practices,” United Julia Sloth-Nielsen and Belinda Van Heerden, Nations Children’s Fund (UNICEF) Innocenti “The Constitutional Family: Developments Working Paper, Florence, UNICEF Innocenti in South African Family Law Jurisprudence Research Centre, 2008. under the 1996 Constitution.” International Journal of Law, Policy, and the Family 17 (2): Robert Jensen and Emily Oster. “The Power 20 121–46; Constitution of Kenya, Article 27 (6), of TV: Cable Television and Women’s Status 2010; Kenya Law Reports, Kiogora Rukunga in India.” The Quarterly Journal of Economics v. Zipporah Gaiti Rukunga (2011) eKLR. 124, no. 3(2009): 1057-1094. Naila Kabeer, “Contextualizing the Economic 21 30 Tazeen Hasan and Ziona Tanzer, “Women’s Pathways of Women’s Empowerment: Find- Movements, Plural Legal Systems and the ings from a Multi-Country Research Pro- Botswana Constitution: How Reform Hap- gram,” Pathways Policy Paper, October 2011, pens,” Policy Research Working Paper 6690, Brighton: Pathways of Women’s Empower- World Bank, Washington, DC, 2013. ment RPC 31 Ximena Marinero, “Kuwait Constitutional Lori Heise, “What Works to Prevent Partner 22 Court Rules Women Do Not Need Permis- Violence? An Evidence Overview.” Strive, sion to Get Passport,” Jurist Legal News and 2011, p. 24-26. Research, October 22, 2009. Gertrudes Macueve, et al., 2009. “African 23 32 International Development Law Organi- Women and ICTs: Investigating Technology, zation (IDLO), Accessing Justice: Models, Gender and Empowerment.” International Strategies, and Best Practices on Women’s Development Research Centre, Zed Books, Empowerment (Rome: IDLO, 2013). 2009. 33 Hallward-Driemeier and Hasan, Empowering http://www.equalaccess.org.np/samajhdari 24 Women. “Samajhadari” Radio Program. 34 Paul Prettitore,  Who Needs Legal Aid Ser- Elizabeth Levy Paluk, and Donald P. Green, 25 vices? Addressing Demand in Jordan (Wash- “Deference, Dissent, and Dispute Resolution: ington, DC: World Bank, 2012). Chapter 2 Enhancing Women’s Agency: A Cross-Cutting Agend | 49 35 World Bank, Increasing Access to Justice for 47 IDLO, Accessing Justice, 26–33. Women, the Poor, and Those Living in Remote 48 IDLO, Accessing Justice, 34–39. Areas: An Indonesian Case Study (Washing- ton. DC: World Bank, 2011). 49 Ana Maria Munoz Boudet, Patti Petesch, and Carolyn Turk and Angelica Thumala, On Norms 36 IDLO, Accessing Justice. and Agency: Conversations about Gender Equal- 37 United Nations Women, 2011–2012 Progress ity with Women and Men in 20 Countries (Wash- of the World’s Women: In Pursuit of Justice ington, DC, World Bank, 2012), 145. (New York: UN Women, 2011). 50 Naila Kabeer, Paid Work, Women’s Empow- 38 The eight countries are Bosnia, Brazil, Chile, erment and Inclusive Growth: Transforming Croatia, Democratic Republic of Congo, India, the Structures of Constraint (New York: UN Mexico, and Rwanda. Women, 2013), 80. 39 Huairou Commission Land and Housing 51 World Bank, Gender at Work: A Companion Campaign, http://huairou.org/land-housing. to the World Development Report on Jobs. 40 IDLO, Accessing Justice. (Washington, DC: World Bank, 2014). 41 Republic of Kosovo, Law on Reproduc- 52 World Bank, Gender at Work. tive Health, Law No. 02/L-76, http:// 53 World Bank, Gender at Work. w w w.gazetazyrtare.com/e-gov/index. php?option=com_content&task=view&id=1 54 Paul M. Pronyk, James R. Hargreaves, Julia 28&Itemid=28&lang=en . C. Kim, Linda A. Morison, Godfrey Phetla, Charlotte Watts, Joanna Busza, and John D. World Bank and International Finance Cor- 42 H. Porter, “Effect of a Structural Intervention poration, Women, Business and the Law 2014: Removing Restrictions to Enhance Gender for the Prevention of Intimate-Partner Vio- Equality (London: Bloomsbury Publishing, lence and HIV in Rural South Africa: A Cluster 2013). Randomised Trial,” The Lancet 368, no. 9551 (2006): 1973–83. 43 U.K. Department for International Devel- opment (DFID), “The Policy Statement on 55 Secretary of Social Inclusion, Government Safety, Security and Accessible Justice,” DFID, of El Salvador, “CiudadMujer,” http://www. London, 2003. ciudadmujer.gob.sv. 44 Share of women reporting ever experiencing 56 World Bank, World Development Report physical or sexual partner violence in the 2013: Jobs (Washington, DC: World Bank, 2010 Demographic and Health Survey; Tan- 2012). zania National Bureau of Statistics and ICF 57 World Bank, World Development Report 2013. Macro, Tanzania Demographic and Health Survey 2010 (Dar es Salaam, Tanzania: NBS 58 Rigorous Impact Evaluations are being con- and ICF Macro, 2011). ducted in six of eight pilots (excluding the Lao PDR, and Rwanda). For more informa- 45 IRIN (Humanitarian News and Analysis), tion on AGI, see http://web.worldbank.org/ “Justice without Jail for Tanzanian Wife-Beat- WBSITE/EXTERNAL/TOPICS/EXTGENDER ers,” IRIN, November 1, 2013, http://www. /0,,contentMDK:21914520~pagePK:210058 ir in news .or g/repor t/99049/ju s t ice - ~piPK:210062~theSitePK:336868,00.html. without-jail-for-tanzanian-wife-beaters. 59 Government of Ethiopia, End Child Marriage 46 For a more comprehensive discussion of the Program, http://www.qedgroupllc.com/ challenges associated with women’s access project/end-child-marriage-programme. to formal and informal legal systems and promising approaches see IDLO, Accessing 60 Oriana Bandiera, Niklas Buehren, Robin Bur- Justice, 11–25 gess, Markus Goldstein, Selim Gulesci, Imran 50 | Voice and Agency empowering women and girls for shared prosperity Rasul, and Munshi Sulaiman, “Empowering 69 Karen Macours and Renos Vakis, “Changing Adolescent Girls: Evidence from a Random- Households’ Investments and Aspirations ized Control Trial in Uganda,”  unpublished through Social Interactions,” Policy Research working paper, 2012.  Working Paper Impact Evaluation Series Farida Khan and Rehana Ghadially, “Gen- 61 41:9, World Bank, Washington, DC, 2009 der-Differentiated Impact on Minority Youth 70 Overseas Development Institute (ODI) and of Basic Computer Education in Mumbai U.K. Department for International Develop- City,” Gender Technology and Development ment, “Beneficiary and Community Perspec- 13, no 2 (2009): 245–69. tives on the Palestinian National Cash Trans- Broadband Commission Working Group on 62 fer Programme,” Country Briefing, ODI, 2013; Broadband and Gender, Doubling Digital Elizaveta Perova and Renos Vakis, Improving Opportunities: Enhancing the Inclusion of Gender and Development Outcomes through Women and Girls in the Information Society Agency: Policy Lessons from Three Peruvian (Geneva: International Telecommunications Experiences (Lima: World Bank, 2013); Paul Union and United Nations Educational Scien- Bukuluki and Carol Watson, Transforming tific and Cultural Organization, 2013), 53. Cash Transfers: Beneficiary and Community Perspectives on the Senior Citizen Grant in See World Bank, Gender at Work, 55; and 63 Uganda (London: ODI, 2012); Nicola Jones International Labor Organization, “Decent and Maria Stavropoulou, Resilience for All Work Agenda,” http://www.ilo.org/global/ Towards Gender-Responsive Social Protection about-the-ilo/decent-work-agenda/lang- in South East Asia (New York, UN Women, -en/index.htm. 2013). Nicola Jones and Thea Shahrokh, “Social 64 71 Perova and Vakis,  Improving Gender and Protection Pathways: Shaping Social Justice Development Outcomes through Agency.  Outcomes for the Most Marginalized, Now and Post 2015,” Background Note, Overseas 72 Sengupta, Targeting the Hard Core Poor. Development Institute, London, April 2013. 73 Maxine Molineux and Marilyn Thomson, Anasuya Sengupta, Targeting the Hard Core 65 “Cash transfers, Gender Equity and Wom- Poor: A Qualitative Study on Participants’ en’s Empowerment in Peru, Ecuador, and Ascent Out of Extreme Poverty (Dhaka: BRAC Bolivia,” Gender & Development 19, no. 2 Development Institute, 2013). (2011): 195–212. Narayan C. Das and Raniya Shams, “Asset 66 74 Perova and Vakis,  Improving Gender and Transfer Programs for the Ultra Poor: A Ran- Development Outcomes through Agency.  domized Control Evaluation,” Challenging 75 Macours and Vakis, “Changing Households’ the Frontiers of Poverty Reduction Working Investments and Aspirations through Social Paper 22, BRAC, Dhaka, 2011. Interactions.” Abhijit Banerjee, Esther Duflo, Raghabendra 67 76 Esther Duflo, “Grandmothers and Grand- Chattopadhyay, and Jeremy Shapiro, “Target- daughters: Old‐Age Pensions and Intra- ing the Hard-Core Poor: An Impact Assess- household Allocation in South Africa,” World ment,” draft report, 2011. Bank Economic Review 17, no. (2003): 1–25; Aude de Montesquiou and Syed M. Hashemi, 68 Anne Case and Alicia Menendez, “Does “Creating Pathways for the Poorest: Gradua- Money Empower the Elderly? Evidence tion Model Shows Early Promise,” CCAP blog, from the Agincourt Demographic Surveil- August 1, 2012, http://www.cgap.org/blog/ lance Site, South Africa,” Scandinavian Jour- creating-pathways-poorest-graduation- nal of Public Health 35, suppl. 69 (2007): model-shows-early-promise. 157–64. Chapter 2 Enhancing Women’s Agency: A Cross-Cutting Agend | 51 I. E. Filho, “Household Income as a Determi- 77 Development in Cape Town, South Africa,” nant of Child Labor and School Enrollment (PhD diss. in Educational Leadership and in Brazil: Evidence from a Social Security Comparative and International Education, Reform,” Economic Development and Cultural Pennsylvania State University, State College, Change 60, no. 2 (2012): 399–435. 2009. Ariel Fiszbein, Norbert Rüdiger Schady, and 78 89 Raewyn Connell Education, Change and Soci- Francisco H. G. Ferreira, Conditional Cash ety (Oxford University Press Australia & New Transfers: Reducing Present and Future Pov- Zealand, 2010) UNGEI 2012; Ruti Levtov, erty (Washington, DC: World Bank, 2009). “Addressing Gender Inequalities in Curricu- Michael Woolcock, Berk Ozler, Sarah Baird, 79 lum and Education: Review of Literature and Francisco H. G. Ferreira, “Relative Effec- Promissing Practices. Women’s Voice, Agency tiveness of Conditional and Unconditional and Participation Research series 2014, No. 5, Cash Transfers for Schooling Outcomes in World Bank, Washington, DC 2014. P. Achyut, Developing Countries: A Systematic Review,” N. Bhatla, S. Khandekar, S. Maitra and R. K. Campbell Systematic Review 9, no. 8 (2013): Verma, “Building Support for Gender Equality 1–24. among Young Adolescents in School: Findings Sarah Baird, Eprahim Chirwa, Craig McIntosh, 80 from Mumbai, India,” International Center and Berk Ozler, “The Short-Term Impacts of for Research on Women, New Delhi, India, a Schooling Cash Transfer Program on the 2011; and Australian Curriculum, Assess- Sexual Behavior of Young Women,” Health ment, and Reporting Authority, http://www. Economics 19, suppl. 1 (September 2010): australiancurriculum.edu.au. 55–68. 90 Zaida Mgalla, Dick Schapink, and J. Ties World Bank, “Whispers to Voices: Gender 81 Boerma, “Protecting School Girls against and Social Transformation in Bangladesh,” Sexual Exploitation: A Guardian Programme Bangladesh Development Series 22, World in Mwanza, Tanzania,” Reproductive Health Bank, Washington, DC, 2008. Matters 6, no 12 (1998): 19–30. World Bank, Gender at Work. 82 91 P. Achyut et al., “Building Support for Gender United Nations Educational Scientific and 83 Equality among Young Adolescents in School: Cultural Organization (UNESCO), Teaching Findings from Mumbai, India,” International and Learning: Achieving Quality for All, EFA Center for Research on Women, New Delhi, Global Monitoring Report 2013/4 (Paris: 2011. UNESCO Publishing, 2014), 5. 92 United Nations Children’s Fund, Child World Bank, On Norms and Agency, 77. 84 Friendly Schools Programming: Global Evalu- ation Report (New York: United Nations Chil- World Bank, On Norms and Agency, 77. 85 dren’s Fund, 2009). World Bank, Gender Dimensions of Roma 86 93 Cynthia B. Lloyd, “Priorities for Adolescent Inclusion: Perspectives from Four Roma Com- Girls’ Education,” Washington, DC, Popula- munities in Bulgaria (Washington, DC: World Bank, 2014), p. iii. tion Council, 2012. UNESCO, Teaching and Learning, 167. 87 Felicia Renee Wilson, “Girls’ Education 88 Movement (GEM): Study of Program Imple- mentation and Partnerships for Education 52 | Voice and Agency empowering women and girls for shared prosperity Section Sexual and reproductive health and rights | 53 Chapter 3 Freedom from violence Key messages Box 3.1 > Violence is a major theme of this report because it is among the most egregious and commonly experienced abuses of women’s rights. > The media often present violence as a tragic event or a side effect of war or armed conflict. But around the world, no place is less safe for a woman than her own home—almost one-third (30 percent) of women have experienced physical or sexual violence or both by an intimate partner. > Gender-based violence stems from social norms and expectations that reinforce inequality and place women’s and girls’ choices outside of their realm of control. Major individual risk factors are family history, early marriage, and husband’s alcohol use, while education emerges as the most significant protective factor. > The effects of intimate partner violence (IPV) are felt at the individual, family, and economy levels. > The development costs are substantial. Estimated costs of IPV are close to the average that developing country governments spend on primary education. > Preventing violence relies in part on changing norms and attitudes that perpetuate gender inequalities and sanction gender-based violence. > Treatment and support for victims is critical—but across the globe, the majority of women (6 in 10) who experience violence never seek help or report the violence to anyone. > Economic empowerment interventions can expand women’s agency by adding strategic design features to address IPV—“economic empowerment plus” interventions. > Promising interventions have multiple components and intervene at different levels; address norms; involve the wider community; engage both men and women; and span long periods. Chapter 3 Freedom from violence | 55 Gender-based violence as a its instrumental value in promoting gender development challenge equality in a wide range of outcomes at the individual, family, and society levels. In 2013, tragic episodes of gender-based violence generated a firestorm in the global Because it is more pervasive, our focus media. At center stage were brutal gang rapes is on violence against women, which the in India and the United States and cases of United Nations (UN) defines as “any act of celebrity women being physically assaulted gender-based violence that results in, or is or even murdered by their husbands and likely to result in, physical, sexual or psycho- boyfriends.1 There was also renewed atten- logical harm or suffering to women, includ- tion to the horrors of sexual violence in con- ing threats of such acts, coercion or arbitrary flict affected countries. But gender-based deprivation of liberty, whether occurring in violence goes far beyond those high-profile public or in private life.”7 As the report of the cases, isolated incidents, and specific cul- High-Level Panel of Eminent Persons on the tures or countries. A recent global study Post-2015 Development Agenda notes, such revealed that over 35 percent of women violence takes many forms including acts worldwide have experienced physical or of physical, sexual and economic abuse.8 It sexual partner violence or non-partner sex- takes place in the home, on the streets, in ual violence.2 That is 818 million women3 schools, in the workplace, in farm fields, and —almost the total population of Sub-Saha- in refugee camps, in times of peace as well ran Africa. The most common form is abuse as during armed conflicts and crises.9 Part of by an intimate partner, which has profound the spectrum of violence that women suffer consequences on the health and well-being is described in box 3.2. of women and their families, as well effects Social norms and institutions—both for- on wider communities and development mal and informal—perpetuate such violence outcomes.4 through norms and expectations that rein- Gender-based violence (GBV) is a viola- force inequality and place women’s and girls’ tion of basic human rights and a pervasive decision making about their bodies outside challenge all around the world. GBV is an their realm of control.10 But within and across “umbrella term for any harmful act that is countries, there is no universal agreement perpetrated against a person’s will and that about what constitutes “violence.” Definitions is based on socially ascribed (gender) dif- of the forms of violence along the continuum ferences between males and females.”5 Men are culturally derived. For example, women in and women across all socioeconomic groups Tanzania define forced sex as sexual assault by and development contexts are at risk of GBV. an acquaintance or boyfriend, coercion as the However, women are much more affected use of guilt and emotional manipulation, and because violence both reflects and rein- rape as sexual assault by a stranger.11 Although forces underlying gender-based inequali- such rape is viewed as unacceptable, physical, ties.6 Freedom from violence is an essential emotional, and even sexual violence against a domain of agency for both its intrinsic value wife or partner is frequently viewed as nor- in asserting fundamental human rights and mal, as illustrated by this quote: 56 | Voice and Agency empowering women and girls for shared prosperity violence as a crime has risen from close to Yes, it’s normal, being beaten, yelled at. zero to 76 in just 37 years.16 Combatting vio- If you tell [anyone], your peers will ask lence has emerged as a key priority for the you, is this your first time to be beaten? post-2015 agenda and, in March 2013, more Some of us are used to it, just like the than 100 UN member states pledged to end way we are used to eating ugali. violence against women and girls.17 And since —Female focus group participant Mbeya, Tanzania12 2000, several UN Security Council resolutions have addressed sexual violence and the effect While the acceptance of wife beating of armed conflict on women and girls.18 as part of marital life is commonplace13 recent evidence suggests that rejection The scope and structure of this chap- of violence globally has risen rapidly in ter is as follows. We focus on physical and the past decade.14 As shown in figure 3.1, sexual intimate partner violence (IPV). We do not tackle other forms of violence in many repeat Demographic and Health against women, although child marriage Surveys (DHS), acceptance of spousal vio- is addressed in the next chapter. The next lence dropped or leveled out near zero. At section reviews the prevalence of IPV and the same time our analysis of 55 countries the barriers to reporting. We then highlight for the most recent year shows that 4 in 10 the costs of IPV at the individual, family, women still condone wife beating.15 and economy-wide levels and the key risk The good news is that the momentum to and protective factors. The final section of end gender-based violence is growing, and this chapter lays out promising directions evidence about what works is accumulating. gleaned from evidence about prevention The number of countries recognizing domestic and response efforts. Box 3.2 What is GBV? Stories from survivors in Tonga The Women and Children Crisis Centre in Tonga provides care and support to survivors of violence. The Centre collected stories from survivors describing the range of violence they have suffered at the hands of the men in their lives—husbands, fathers, and fathers-in-law. Women described being pushed and punched. They had their hair pulled and cut. Men cut their bodies with knives and hit them with heavy objects, such as tire jacks and hammers. Men stood and stepped on their faces and beat their backs, burned their bodies with scalding objects, and spat and urinated on them. Women reported being kicked, including being kicked while pregnant. Women and children were forced to have sex. Very young girls were forced to perform fellatio on their fathers. They were raped by their fathers and by their fathers-in-law, some every night. Yet the abuse is not limited to young women by any means. Grandsons attacked their grandmothers, verbally terrorized them, and hit them in the face. Crisis centers around the world record countless similar testimonies every day. And myriad similar stories are never told at all. Source: Ofa Ilikiliki, presentation at the UN Commission on the Status of Women, Washington, DC, March 2014. Chapter 3 Freedom from violence | 57 FIGURE 3.1 Change in percentage of women who believe that wife beating can be justified Argues Goes out without Refuses to have with him telling him sex with him Mali (2001-2006) Burkina Faso (2003-2010) Egypt, Arab Rep. (1999-2011) Mozambique (2003-2011) Zimbabwe (2001-2006) Cambodia (2005-2010) Benin (2001-2012) Haiti (200-2012) Indonesia (2003-2012) Armenia (2000-2010) Nepal (2001-2006) Honduras (2006-2012) Peru (2006-2012) 0 20 40 60 0 20 40 60 80 0 20 40 60 80 Percent Earliest available year Latest available year Source: Voice and Agency 2014 team estimates based on Demographic and Health Surveys for 13 countries for which data over time are available. How large is the challenge? A solid global evidence base about the prevalence and patterns of IPV, the most Around the world, no place is less safe for a common form of gender-based violence, is woman than her own home. Domestic violence now emerging. This path was forged by the is especially pernicious “because it occurs in a landmark 2005 World Health Organization space that is also central to the development of (WHO) Multi-Country Study on Women’s human capabilities—the family.”19 Numerous Health and Domestic Violence against Women, studies show that women are more likely to be physically and sexually assaulted or murdered which systematically documented the mul- by someone they know—often a family mem- tiple forms of violence using a standardized ber or intimate partner.20 A recent systematic questionnaire, training, and procedures.22 review using data from 66 countries found The study revealed the high prevalence of that the share of homicides by an intimate IPV and underscored the repercussions for partner was six times higher for female vic- women’s health and well-being on a global tims compared with male victims (39 percent scale. It should be noted, however, that versus 6 percent, respectively).21 because of the sensitive nature of collecting 58 | Voice and Agency empowering women and girls for shared prosperity such information, all survey estimates are occur within the family or domestic unit or subject to some degree of underreporting between former or current spouses or part- (Box 3.3). The need to develop tools and ners, whether or not the perpetrator shares methodologies to more accurately estimate or has shared the same residence with the prevalence and incidence of violence is dis- victim.”23 The 2005 WHO study documented cussed in chapter 7. wide variation in rates of violence. Among 24,000 women interviewed in 10 countries, Prevalence of intimate partner between 15 percent and 71 percent had violence experienced physical or sexual IPV in their lifetime. Additionally, up to 75 percent had IPV is the focus of this chapter, largely experienced emotionally abusive and con- because it is both the most pervasive form trolling behavior in the preceding year. A of gender-based violence, and one that too preliminary analysis of the 2013 WHO global few governments even recognize as a crime. prevalence dataset (by World Bank regions) IPV includes “all acts of physical, sexual, illustrates the regional variation, as shown in psychological or economic violence that figure 3.2. Among women who had ever been FIGURE 3.2 Share of ever partnered women who have experienced physical or sexual intimate partner violence during their lifetime Europe and Central Asia 29% North America 21% Middle East and North Africa 40% East Asia and the Pacific 30% South Asia 43% Sub-Saharan Africa 40% Latin America and the Caribbean Australia and 33% New Zealand 28% Source: Preliminary analysis of WHO (World Health Organization), global prevalence database (2013) using World Bank regions.24 Note: Areas shaded in grey are not calculated or do not have relevant data. Chapter 3 Freedom from violence | 59 FIGURE 3.3 Share of women experiencing physical or sexual intimate partner violence or both, by ethnic group Honduras Malawi Nepal Nigeria Peru 1 2 3 Share of women 4 experiencing intimate partner violence 5 6 0 - 12% 7 13 - 24% 8 25 - 36% 9 37 - 48% 10 49 - 60% 11 12 Source: Voice and Agency 2014 team estimates based on Demographic and Health Surveys using latest available data from 2008-2012. in a relationship (ever partnered women), emotional, and sexual abuse. Demographic prevalence in developing regions ranges and Health Surveys categorize the severity from 29 percent in Europe and Central Asia of physical violence as follows: to a high of 43 percent in South Asia. ■■ Less severe: pushing, shaking, slapping, These regional averages can mask con- punching, and kicking. siderable variation across and within coun- ■■ Severe: trying to strangle or burn, threats tries, but provide a useful overall snapshot. with a weapon, and attacks with a weapon. Our analysis of DHS data shows the preva- DHS data from some 30 countries show lence of violence across ethnic groups in five that about 3 in 10 women (28 percent) countries (figure 3.3). Although differences report having experienced “less severe vio- exist among the groups in each country, lence” and about 1 in 10 (12 percent) report the data also show that no ethnic group is having experienced “severe” violence. More immune to violence. than 1 in 10 (11 percent) report having expe- The types of partner violence range rienced sexual violence, and one-quarter of from minor to severe and include physical, women report the experience of emotional 60 | Voice and Agency empowering women and girls for shared prosperity violence, including threats of harm and for example, most women who experienced humiliation in front of others. Box 3.3 pro- physical IPV in the past 12 months also vides an overview of the domestic violence reported emotional abuse (ranging from module in the DHS. 61 percent in Colombia to 93 percent in El Yet of course—as we saw from Tonga— Salvador). Women exposed to three or more even these numbers do not give a full pic- controlling behaviors by their partners— ture. Women who experience intimate part- such as efforts to limit her mobility and her ner violence tend to experience multiple contact with friends and family—were also forms. In Latin America and the Caribbean, at much higher risk of experiencing physical Box 3.3 The Domestic Violence Module of the Demographic and Health Surveys The Demographic and Health Surveys (DHS) support nationally representative data collection in the areas of population, health, and nutrition. A standardized domestic violence module was developed in 1998 and has been included in 91 surveys. It collects data on the prevalence of physical, sexual and emotional violence against women since the age of 15, by current or previous partners as well as by non-partners within the household context, and allows the experience of violence to be linked to a range of risk and protective factors.25 In both stand-alone surveys developed specifically to measure violence and broader surveys like the DHS that integrate a violence module, underestimation of the actual prevalence of violence is a concern. The DHS has incorporated enhancements to protect respondents and optimize the accuracy of violence estimates, including offering multiple opportunities for disclosure by asking many different times about any experience of many types of violence, and by generally placing the module toward the end of the questionnaire, which provides the greatest opportunity for developing a rapport between interviewers and respondents. Further, the DHS adheres to strict ethical and safety guidelines aligned with the internationally recognized recommendations of WHO. Key precautions include anonymity of respondents, that is, names are never disclosed and are excluded from all data sets. The guidelines include providing specialized training for interviewers, observing strict informed-consent procedures, assuring privacy during the interview, administering the module to only one eligible woman per household, not asking men in the household questions about violence, and providing referrals to services. Stand-alone surveys like the WHO Multi-Country Study generally report higher prevalence than the DHS, likely due to greater efforts to enhance disclosure in questionnaire design and interviewer training. However, in terms of risk and protective factors and gaps in help-seeking, results from analysis of DHS data are comparable to those from the WHO study. The DHS is largest source for population-based data about violence, and so constitutes a large share of the available data on IPV prevalence. Further detail about the data sources used in this report can be found in chapter 7. Source: Kishor and Johson, 2004; Kishor, 2004; WHO, 2001; Ellsberg and Heise, 2005; Bott et al., 2013.25 Note: The full list of 91 surveys can be accessed at: http://www.dhsprogram.com/What-We-Do/survey-search.cfm. Chapter 3 Freedom from violence | 61 FIGURE 3.4 Share of women experiencing physical or sexual intimate partner violence or both and at least one control on their behavior (percentage) No IPV nor controls 35 Experienced IPV Condones wife beating 26 60 5 21 39 Source: Voice and Agency 2014 team estimates based on Demographic and Health Surveys for 26 countries using data from 2006–12. Note: IPV = intimate partner violence. Controls include limiting a woman’s contact with her family, not permitting her to meet female friends, and insisting on knowing where she is. or sexual partner violence.26 Only 35 per- abuse occurs “often.” Repeated victimiza- cent of women in our sample of 26 countries tion exposes survivors to complex trauma, have experienced neither IPV nor controls with cumulative and devastating effects on on their behavior (figure 3.4). Four out of women’s and girls’ basic functioning across five women who have experienced IPV were a range of activities in day-to-day life.28 also subject to at least one control. Such controlling behaviors are often a risk factor Coercion, forced sex, and rape and precursor to a partner’s use of physical The ability to choose when to have sex and on violence.27 what terms is a critical indicator of agency. Abused women generally experience Social norms dictate women’s perceived emotional violence more frequently than and actual control over their own sexuality, physical violence. In Mali—the only country meaning that power disparities can leave in our DHS data set with frequency data— them unable to control whether, when, how, women report more than twice as many and with whom they have sex.29 There is a occurrences of emotional violence than continuum of volition related to sex, with physical violence. Across our sample, about willful participation at one end and forced one-fifth of women experiencing physi- sex and rape at the other end.30 Even when cal abuse, one-fourth experiencing sexual a young woman chooses her partner, her abuse, and nearly one-third of those expe- choice may be constrained by her economic riencing emotional abuse report that the needs and the social expectations placed on 62 | Voice and Agency empowering women and girls for shared prosperity her.31 Poor women and girls, for example, with a primary education and 69 percent of may engage in transactional or commercial women with no education. Each additional sex for lack of good alternate livelihoods.32 year of education is associated with a 1 Research across Sub-Saharan Africa shows percent increase in the ability to refuse sex that cross-generational sex between girls (figure 4.1). Compared with women with no and older men may involve provision for education, those with a secondary educa- girls’ basic needs, such as school fees and tion have a 14 percent higher probability of food, in exchange for sex.33 being able to refuse sex, all else being equal. These results suggest the importance of Our analysis of 33 developing countries improving girls’ education as an entry point reveals extensive lack of agency: for enhancing their agency. ■■ Almost one-third of women (32 percent) Our analysis of DHS data shows a strong report that they cannot refuse sex with association between early sexual initiation their partners—rising to over 70 percent and the likelihood that sexual initiation was of women in Mali, Niger, and Senegal. forced. In some countries (notably Bangladesh ■■ Over 41 percent say they could not ask and rural Ethiopia), high levels of forced first their partner to use a condom. sex are likely related to early marriage rather Underpinning women’s difficulty in than to violence by acquaintances or strang- refusing sex is the common assumption that ers.37 On average, across 16 countries for consenting to sex at any point in the rela- which data are available, 14 percent of women tionship establishes consent for the rest of report that their first sex was forced; ranging the relationship, which also creates barriers from 1 percent in Timor-Leste to 64 percent to recognizing marital rape. The ability to in Democratic Republic of Congo. Our analysis refuse sex and to ask a partner to use a con- shows that women who report that their first dom are closely correlated,34 but in Sub-Sa- intercourse had been forced are twice as likely haran Africa and Latin America, higher to have experienced IPV. National surveys by shares of women say they can refuse sex Together for Girls found that significant shares than ask a partner to use a condom. These of young women ages 18–24 experienced sex- findings likely reflect attitudes that associ- ual violence before their 18th birthdays—26 ate condom use with illicit sex and promis- percent in Haiti, 27 percent in Tanzania, 32 cuity, and to perceived and real ethnic and percent in Kenya, 33 percent in Zimbabwe, religious prohibitions.35 In many contexts, and 38 percent in Swaziland.38 the higher confidence in refusing sex than in While not the focus of this chapter, it is insisting on condom use may reflect a dislike noted that more than 1 in 20 women globally of condoms among both men and women.36 (7 percent according to WHO) have suffered The strongest correlate of women’s sex- sexual violence by non-partners.39 A recent ual autonomy in a relationship is her level study with more than 10,000 men across Asia of education. Overall, 87 percent of women and the Pacific revealed that prevalence var- with a higher education say they can refuse ies widely—3 percent of men in rural Bangla- sex, compared with 71 percent of women desh and 27 percent in Bougainville, Papua Chapter 3 Freedom from violence | 63 New Guinea, reported having raped a woman 56 percent of women experiencing “severe” who was not their partner.40 Although less physical violence sought help, compared common, men also reported their involve- with 44 percent of those experiencing “less ment in gang rapes; levels ranged from 1 per- severe violence”—and the former were cent of men in urban Bangladesh to 14 per- nearly twice as likely to seek help from the cent of men in Bougainville. police. Across 28 European countries in a recent survey, one-third of women experi- Reporting and responses encing severe IPV sought some form of help, but only 14 percent of women reported their Around the world, the vast majority of most serious incident of IPV to the police. women who experience violence never seek Among European women who did not seek help or report the violence to anyone. Only help for their most severe incident of phys- 2 percent of women in India and East Asia, ical IPV, the most common reasons were 6 percent in Africa, 10 percent in Central choosing to handle it on their own (55 per- Asia, and 14 percent in Latin America and cent), and not feeling it was serious enough the Caribbean made any formal disclosure to warrant reporting (34 percent).45 Even in of their experience of violence.41 cases of severe violence, if help is sought at Our analysis of DHS data shows that all, it is often as a last resort rather than as across 30 countries, on average only 4 in an opportunity for early intervention. 10 women exposed to violence sought any The barriers to reporting and help seek- help, and only 6 percent sought help from ing reflect social norms and systems that are authorities, such as police, lawyers, doctors, not responsive to women’s needs. Figure 3.5 or religious authorities. Similarly, in Bangla- underscores the host of sociocultural and desh two-thirds of abused women did not structural barriers that inhibit women’s abil- seek help at all; only 2 percent ever sought ity to report and seek help of any kind.46 DHS help from institutions; and rural women surveys in six countries ask why women do were even less likely to seek help than not seek help. Many see violence as a “part of urban women.42 Only 3 percent of women life.” But a substantial number did not know experiencing violence sought help from the how or where to report violence. In Burkina police—ranging from 10 percent in Mol- Faso, Côte d’Ivoire, the Dominican Republic, dova and Ukraine, to fewer than 1 percent and Mali, the predominant reason was the in six countries (Bolivia, Burkina Faso, Haiti, belief that that there was no use doing so. Mozambique, Nigeria, and Zimbabwe).43 Studies in Bangladesh, Turkey, and the United States find that women are more likely to seek help from formal agencies as Across 33 developing countries almost domestic violence becomes more severe— one-third of women cannot refuse sex beyond what they can endure or to the point with their partner and over 40 per- at which the woman feels her life or her chil- cent cannot ask their partner to use a dren’s safety is in serious jeopardy.44 Like- condom. wise, our analysis suggests that on average, 64 | Voice and Agency empowering women and girls for shared prosperity FIGURE 3.5 Reasons for not seeking help 45 40 35 30 25 20 15 10 5 0 Bolivia Dominican Peru Burkina Faso Côte de’Ivoire Mali Republic n No use n Part of life n Don’t know n Afraid of divorce, being beaten or trouble n Embarrassed Source: Voice and Agency 2014 team estimates based on Demographic and Health Surveys using latest available data from 2006-2012. Many women also said they were embar- medical and social service providers, as well rassed to tell anyone, which underlines the as police and the justice system. Although the stigma and blame that are often placed on existence and awareness of laws criminaliz- survivors of violence. ing IPV are important, they are not enough. Parallel efforts are needed to increase respon- In Mexico, the most common reasons siveness to women’s needs.49 We turn to that women gave for not reporting included a subject below. Box 3.4 provides an overview perception that the violence was insignifi- of a new World Bank initiative to address cant (29 percent), concern for their children the needs of survivors and other vulnerable (18 percent), embarrassment (14 percent), women through a multipronged approach in and fear of retaliation by their partners (14 the Great Lakes region. percent).47 Only 8 percent of women cited “not knowing they could press charges.” The magnitude of violence against Among a sample of Bangladeshi women, women is a major development challenge. the most common reasons given were high Each IPV incident brings costs at multiple acceptance of violence, fear of stigma, and levels, a theme we will cover in the next fear of greater harm.48 section. The gap in help seeking represents an Costs of violence enormous missed opportunity to enhance women’s agency through entry points in Intimate partner violence has devastating the justice and social service sectors. This consequence for individuals, communi- challenge requires concerted attention from ties, societies, and economies. Tackling this Chapter 3 Freedom from violence | 65 challenge head-on would advance efforts to depression, post-traumatic stress disorder, end extreme poverty and increase prosper- and substance abuse.51 A recent systematic ity for all. This section documents evidence review found that IPV increases a woman’s of the consequences of IPV at the individual, risk of experiencing depression two to three- family, and community levels. fold.52 Survivors of violence are also 2.3 times more likely to have alcohol use disorders. IPV Individual-level effects has also been linked to the risk of contract- ing human immunodeficiency virus (HIV) Exposure to IPV has been linked with a mul- and other sexually transmitted infections, as titude of adverse health outcomes, including well as the risk of attempting or completing acute injuries, chronic pain, gastrointestinal an abortion.53 illness, gynecological problems, depression, and substance abuse.50 Mental health conse- Beyond the damage to health, violence quences include increasing women’s risk of reduces women’s economic opportunities. Box 3. 4 Serving survivors and empowering women in the Great Lakes Poor and vulnerable women in the Great Lakes states (Burundi, the Democratic Republic of Congo and Rwanda) face multiple and mutually reinforcing constraints, including high levels of violence, lack of say over their health, and limited economic opportunities. In the Democratic Republic of Congo alone, nearly three-quarters of women have experienced intimate partner violence. Although women and girls are particularly vulnerable during and after armed conflicts, there is growing recognition that sexual violence has become a wider social phenomenon, including through increasing levels of sexual partner violence. To help respond, the World Bank is preparing a US$75 million operation to address the needs of survivors of violence and of other vulnerable women in targeted communities, with activities spanning several sectors. The project is expected to offer the following support: ›› Improved maternal, reproductive, and obstetric care; ›› Mental health services to reduce post-traumatic stress and to improve daily functioning; ›› Case management, referrals for survivors, and assistance in accessing the judicial system through civil society organizations; ›› Assistance through village savings and loans associations and cooperatives for survivors and other vulnerable groups; and ›› Scale-up of one-stop centers for survivors of violence and provision of emergency contraception kits, antiretroviral drugs, and surgery equipment to boost the technical capacity of facilities to deliver specialized services. Finally, the program aims to build relevant capacities of a full range of professionals—including social workers, the police, and the military—and engage community leaders as agents of change. Source: Great Lakes Emergency Women’s Health and Empowerment Project (P147489). 66 | Voice and Agency empowering women and girls for shared prosperity For example: not experienced IPV. They also face a greater likelihood of being abused themselves, an ■■ Women exposed to partner violence in experience that increases their risk-taking Vietnam have higher work absenteeism, behaviors in adolescence, including drink- lower productivity, and lower earnings ing, drug use, and early initiation of sex.65 than working women who are not beaten.54 Girls who witness their mothers being ■■ In Tanzania, women in formal wage work abused are twice as likely to experience IPV who are exposed to severe partner abuse later, and boys show an increased risk of (both lifetime and current) have 60 per- becoming perpetrators later in life,66 hence, cent lower earnings.55 the much higher risk of a continuing cycle of ■■ In the late 1990s in Nagpur, India, women violence. had to forgo, on average, seven days of Child abuse and partner violence often paid work per violent episode.56 occur in the same home. Evidence from six Accessing treatment and support ser- Latin American countries show that children vices can be expensive. In Uganda, for exam- of women who have experienced IPV are ple, the average direct out-of-pocket expen- more likely to experience beating, spank- diture related to an incident of IPV was ing, or slapping. This ranges from just over estimated to be the equivalent of US$5,57 one-third of children in Paraguay, compared about one-twelfth of average monthly with fewer than one-quarter whose mothers income in rural areas.58 A 2011 household reported no partner violence, to more than survey in Vietnam estimated that out-of- two-thirds of children of abused mothers in pocket expenditures for accessing services Colombia, compared with 58 percent of chil- and replacing damaged property averaged dren of never-abused mothers.67 21 percent of women’s monthly income.59 There are also repercussions for the men who perpetrate IPV. In Vietnam, male Family-level effects perpetrators had higher work absenteeism Violence brings dangers of short-term and following a violent episode.68 A similar study long-term effects on children, who may wit- from Maine in the United States showed that ness frequent abuse. In Monterrey, Mexico, 48 percent of offenders had difficulty con- centrating at work, with 19 percent report- for example, half of abused women reported ing a workplace accident or near miss due to that their children routinely witnessed the preoccupation with their relationship.69 abuse.60 Children exposed to violence at home show impaired socio-emotional func- tioning and educational outcomes in adoles- Economy-wide effects cence and lower job performance, job stabil- Beyond the human costs, violence incurs ity, and earnings into adulthood.61 Research major economy-wide costs. Those costs has consistently found health risks, such as include expenditures on service provision, higher infant mortality rates,62 lower vac- forgone income for women and their fam- cination rates,63 and lower birth weight64 ilies, decreased productivity, and negative compared to children whose mothers have effects on human capital formation. Chapter 3 Freedom from violence | 67 Different models have been developed Figure 3.6 presents estimated costs for to estimate the economy-wide costs of a variety of countries, ranging from 1.2 per- IPV. Although these models vary in their cent to 3.7 percent of gross domestic product, core assumptions, they generally take into equivalent to what many developing country account some combination of the direct and governments spend on primary education. indirect costs that are both tangible (and It should be noted that the estimates are not can be monetized) and intangible (which directly comparable across countries because cannot be readily monetized).70 These esti- the methodologies and data vary. One recent mates typically include costs related to study in the United Kingdom sought to service provision, out-of-pocket expendi- account for the loss of life satisfaction and tures, and lost income and productivity. IPV arrived at estimates of about 10 percent of incurs direct costs on services in the health, gross domestic product.72 social service, justice, and police sectors. In The costs vary with the severity of vio- the United States, health care costs among lence. A recent study found that in Tanza- women experiencing physical abuse have nia—compared with women who had never been estimated to be 42 percent higher been abused by a partner—women experi- than among non-abused women.71 Nota- encing current IPV earned 29 percent less, bly, estimates typically do not include costs and those currently experiencing severe IPV associated with the long-term emotional earned 43 percent less.74 In Chile and Nicara- effects, increased health care needs, and sec- gua, severely abused women earned 61 per- ond-generation consequences. cent and 43 percent less, respectively.75 FIGURE 3.6 Economy-wide costs of intimate partner violence, selected countries 4.0 3.5 3.0 Cost of IPV 2.5 2.0 1.5 1.0 0 Australia United Vietnam Bangladesh Peru Kingdom n Costs of IPV n Spending on primary education Source: Duvvury et al., 2013.73 Note: IPV = intimate partner violence; GDP = gross domestic product. 68 | Voice and Agency empowering women and girls for shared prosperity Who is worst affected? Relationship characteristics determine risk Hundreds of millions of women experience some form of intimate partner violence in Characteristics of the partner and the rela- their lifetimes, and many experience multiple tionship affect a woman’s odds of experienc- forms over the course of their lives. Violence is ing violence. All else being equal, having a not limited to specific regions of the world or husband or partner with some education to socioeconomic, religious, or ethnic groups. reduces the likelihood that a woman will Understanding risk and protective factors experience violence, though the effect is less and how they interact is critical to inform the than that of her own education. Several rela- design of programmatic responses. tionship characteristics affect the odds of experiencing IPV: Our analysis of pooled DHS data from 21 countries explored the risk and protective ■■ Being in a polygamous marriage factors for violence in multivariate mod- increases the risk of IPV by 24 percent. els controlling for individual and relation- ■■ Being married before age 18 increases ship characteristics. Among the strongest the odds by 22 percent. individual-level risk factors are family his- ■■ Women who report their husbands get tory of violence and attitudes toward wife drunk sometimes have an 80 percent beating: higher risk, while having a husband who ■■ Women whose fathers beat their moth- drinks often increases the risk nearly ers have a 2.5 times greater risk of fivefold (4.8 times). experiencing IPV in their adult lives com- Poverty and lack of economic opportuni- pared with women who did not witness ties are associated with increased IPV.76 Such IPV as children. violence may stem from perpetrators’ own ■■ Agreeing with any justification for wife sense of powerlessness or insecurity, espe- beating increases the odds of violence by cially when they feel unable to meet the roles 45 percent. socially assigned to them as men.77 But that sit- uational risk tells only part of the story. Trig- Our analysis also highlights important gers do not create violent behavior, and many protective factors, including education and men who live in poverty or are unemployed wealth: are not violent, even if they use alcohol. ■■ Women with some or completed second- Two recent multi-country studies have ary education have an 11 and 36, percent sought to understand the protective and lower risk of violence respectively, com- risk factors for perpetration of violence. pared with women with no education. The International Men and Gender Equal- ■■ Compared to women living in poorer ity Survey (IMAGES) found that attitudes households, women in wealthier house- toward gender equality are important, holds have a 45 percent lower risk of alongside witnessing parental violence, violence. being depressed, and having been involved Chapter 3 Freedom from violence | 69 in fights.78 Analyses of IMAGES data across reduced risk of IPV, as we saw in chapter six countries undertaken for this report 2. But it is not always the case and results underline this point. We found that men are highly context specific. Here are some had more than 10 percent lower odds of examples: perpetrating partner violence for every one ■■ In Ecuador, unconditional cash transfers standard deviation increase in their Gender to mothers reduced psychological part- Equitable Men score.79 Similar findings from ner violence against women with at least the Partners for Prevention multi-country primary education, but they were found study in Asia highlight that most men do not to increase psychological violence against perpetrate violence and that for those who women with only a primary education or do, the primary risk factors include witness- less, or when their education level was ing or experiencing abuse as children and equal to or more than their partners’.84 having gender-inequitable norms and prac- ■■ A randomized evaluation in Kenya found tices.80 Better understanding of these factors that unconditional cash transfers for the is needed to inform prevention efforts. poor paid through mobile phones reduced IPV against women by 30 - 50 percent and Increased access to income may reduced rape within marriage by 50 - 60 affect risk percent, with larger effects when trans- The relationship between women’s eco- fers were made to women and larger nomic opportunities and exposure to vio- amounts of money were given.85 lence is complex and highly context depen- ■■ A study of Mexico’s Oportunidades pro- dent. Evidence from some studies show gram found that women who received a employment and earnings reduce women’s cash transfer conditional on their chil- risk of violence, whereas others indicate dren’s attendance at school and health that economic participation increases the clinics were 40 percent less likely to risk. A systematic review found mixed evi- experience physical abuse. But they dence as to whether women’s own incomes were more likely to experience threats of reduced or increased the likelihood of expe- physical harm than women who did not riencing violence.81 And several studies have receive a transfer.86 found no statistically significant support for How do we explain these apparent con- either hypothesis.82 In Bangladesh, working tradictions? Conditional cash transfers increased women’s exposure to IPV, but only targeted to women could lessen economic for women who had low education or were stress on the household, increase financial married young.83 Our DHS analyses show independence, and, through conditions, that women’s employment increases the increase women’s exposure to supportive odds of IPV exposure by about 25 percent on services, such as health centers. In some average, though less for women who work cases, women’s newfound resource control for cash than for those who work “in kind.” may reduce violence-inducing economic There is some evidence of social pro- stressors on the household or even open the tection programs being associated with option to leave an abusive relationship.87 70 | Voice and Agency empowering women and girls for shared prosperity On the other hand, women’s increased Fragility and armed conflict control over resources could also challenge increase risk traditional gender roles and aggravate vio- Fragility and armed conflict increase wom- lence, particularly when women are disad- en’s risk of experiencing IPV and other vantaged from the outset.88 Property own- forms of gender-based violence, including ership, particularly land, may give women non-partner rape. Gender-based violence a more stable and tangible exit option than against women and men is a weapon of war work or income—and has a range of other used to humiliate, terrorize, punish, and tor- benefits as we see in chapter 5—but may ture both combatants and civilians. A recent not be sufficient to offset other risk factors. review found that in most conflict settings, The mixed results across studies illustrate between one in four to one in three women the complexity of the relationship between report sexual violence,92 which includes women’s economic opportunities, cash trans- rapes committed systematically by armed fer programs and violence, and the need for groups, rapes that take advantage of weak- careful consideration in program design, a ened policing and governance, and sexual topic to which we return below. exploitation and abuse.93 The incidence of intimate partner vio- Community environment and lence also increases in conflict contexts: norms determine risk ■■ One of the highest rates of recent IPV Various factors at the community level recorded is in the Democratic Republic have been shown to increase individual of Congo in 2007, where 64 percent of women’s risk of experiencing IPV. A system- women reported experiencing physical or atic review of 36 studies from both devel- sexual IPV in the past 12 months.94 oping and developed countries highlights socioeconomic disadvantage and residential ■■ In the West Bank and Gaza, women whose instability as factors that erode the social husbands were exposed to political vio- fabric, thereby reducing the collective social lence were 89 percent more likely to have controls that might otherwise keep violence been physically abused by their partner.95 in check.89 In Colombia, high levels of unem- ■■ In Liberia, men’s perpetration of violence ployment and poverty in a community can against their partners was significantly affect the individual likelihood of violence;90 associated with exposure to traumatic and community-level norms in Jordan were war-related events.96 very influential determinants of individual ■■ In Colombia, the incidence of IPV was risk for IPV, alongside measures of wom- more than 12 percentage points higher en’s empowerment. For example, in Jorda- in areas that had experienced intense nian communities with lower mean age at conflict.97 marriage, higher educational differences between spouses, and higher prevalence of polygyny, women are significantly more likely to experience IPV.91 Chapter 3 Freedom from violence | 71 The state of the evidence: entry point for addressing and challenging What works? the deeply rooted inequalities that perpet- uate IPV. Our systematic review of violence Programmatic and policy approaches to prevention reviews found that positive address intimate partner violence range effects were strongest among interventions from legislative reform and enforcement of with elements targeting attitudes toward protective laws to media campaigns, com- the acceptability of IPV, gendered expecta- munity and school-based programming, tions, and definitions of masculinity.99 This and services for survivors. We undertook requires understanding which attitudes are a systematic review of violence prevention most directly linked with violence and iden- reviews for this report, which concluded tifying potential levers of change, including that the most promising interventions community, opinion, and religious leaders. include multiple components and engage with different audiences and stakeholders.98 In social contexts where IPV is not an Although no single recipe will bring success accepted norm, fewer women report expe- across diverse social and geographic con- riencing violence. As our analyses showed, texts, this review and other recent research women who condoned wife beating were identify core elements that are essential for more likely to be subject to violence. preventing IPV and to alleviate its adverse Although acceptance of wife beating is wide- spread globally, it is not an attitude held by repercussions. Interventions can be broadly the majority of people. Across 66 countries grouped under social norm change, legal for which we have data, including several reform and responses, social support and high-income countries, on average 75 per- services, economic empowerment, and cent of people feel that wife beating cannot integrated approaches. be justified for any of five reasons; namely, if Before we review what is known about a wife goes out without telling her husband, the impacts of selected interventions and if she neglects their children, if she argues what promising approaches these lessons with her husband, if she refuses him sex, offer, it should be noted that the majority of or if she burns the food. However, variation available evaluation evidence comes from exists across countries, and all of the coun- high income countries. While informative, tries surveyed demonstrated some degree greater investment in documenting what of justification, ranging from less than 5 per- works in developing countries is needed, cent in Colombia, the Dominican Republic, particularly with respect to multi-sectoral Nepal, and Ukraine to rates exceeding 75 interventions to prevent and respond to percent in Guinea, Mali, and Timor-Leste. violence. IMAGES data show that in the Demo- cratic Republic of Congo, India, and Rwanda, Boosting positive gender norms countries with high levels of IPV, most men Chapter 2 underscored the importance of agreed with the statement that a woman social norm change for expanding women’s should tolerate violence in order to keep agency. Influencing of norms is a critical the family together. And that attitude is not 72 | Voice and Agency empowering women and girls for shared prosperity isolated to men. Indeed, a recent review Majority opinions and reported attitudes of 23 studies measuring IPV justification about violence reflect social norms, which found that women tended to report a higher are both reflected in and determined by the rate of violence justification than men.100 A legal sanctions against violence. In countries multi-country study of more than 10,000 with legislation against domestic violence, men revealed that their most frequently women’s acceptance of wife beating is lower. cited reasons for committing rape included Among the 31 countries with such legisla- sexual entitlement (73 percent), entertain- tion, 40 percent of women accept domes- ment (59 percent), and as a punishment (38 tic violence compared with 57 percent of percent).101 Younger respondents also tend women in the 12 countries in the sample to report a higher rate of justification than without such legislation.102 older respondents, underscoring that much The need to engage men and boys in more work is needed to change norms and interventions to promote positive norms attitudes among youth. around gender equality, and sensitizing men Figure 3.7 shows country variation in to the repercussions of violence is receiv- women’s acceptance of violence by age, ing greater attention. Some key findings which falls among older age cohorts. In were highlighted in chapter 2. A variety of developing countries, 17 percent of respon- approaches have been used—social mar- dents in Latin American and the Caribbean, keting, awareness campaigns, and edutain- 25 percent in East Asia and the Pacific, and ment (for example, themed television 40 percent in South Asia condone wife beat- shows, radio programs, and theater)—as ing for at least one reason. ways to reach a large number of people with FIGURE 3.7 Share of women who condone wife beating, by age of respondent 48 Share of women who consider wife beating acceptable for 46 at least 1 reason (%) 44 42 40 38 36 34 15 20 25 30 35 40 45 50 Age of respondent (years) Source: Voice and Agency 2014 team estimates based on Demographic and Health Surveys for 55 countries using data from 2001–12. Chapter 3 Freedom from violence | 73 messages about violence. But the message A promising example is Soul City in matters. A quasi-experimental evaluation of South Africa. The initiative included the a mass media campaign in the United States use of prime-time radio and television dra- found mixed results: among women their mas and print materials to portray role perception of violence as severe, awareness models with positive attitudes and behav- of support services, and belief in services’ iors, including help-seeking and help-giv- effectiveness all increased. However, the ing actions. Domestic violence was a major result among men was the inverse: their focus of the campaign. A pretest-posttest perception of violence as severe actually fell evaluation found that the intervention was by a wider margin.103 associated with a 10 percent increase in respondents disagreeing that domestic vio- Male engagement interventions have lence was a private affair and a 41 percent achieved attitudinal changes and decreases increase in awareness of the domestic vio- in self-reported use of violence. These inter- lence hotline. Notably, the intervention was ventions cover a diverse range of settings specifically designed to support implemen- and approaches, from workplace training tation of the national Domestic Violence sessions to community-based dialogues and Act by promoting widespread awareness of campaigns anchored in sports events.104 the protections and consequences defined In Chile, educational workshops for young in the new law, as well as offering greater men on gender equality and the prevention access to necessary support services.107 of violence against women increased their Additionally, research suggests that men knowledge about different forms of violence can often underestimate the disapproval and their intention to reject violence against of both women and men of gender-based women. violent behaviors. Consequently, providing Research suggests that messaging that men with more accurate normative infor- promotes and reinforces positive norms— mation can be more effective in changing such as showing men modeling equitable behaviors and attitudes toward violence and nonviolent behaviors—may be more than in reinforcing notions that men com- effective in reaching men than messages monly condone violence.108 condemning violence.105 Such messages pro- Community mobilization has been used mote the positive aspects of doing “the right to address rigid gender norms and attitudes, thing” rather than asserting fear or shame engaging and educating the wider commu- for doing the wrong thing. And they give vis- nity, including men and boys, on issues such ibility to nonviolent, gender-equitable men as human rights, gendered power dynam- who can be models of positive behavior. Evi- ics, reporting, political participation, and dence from Oxfam’s “We Can” campaign in laws. One prominent example is SASA!, an five countries showed that combined efforts approach developed and implemented in of communication campaigns with activi- Uganda to transform gender relations and ties by community change agents have the power dynamics and to address HIV and greatest success in changing attitudes and violence against women (see box 3.5).109 behaviors.106 Several factors explain the success of SASA!, 74 | Voice and Agency empowering women and girls for shared prosperity Box 3.5 Mobilizing communities against violence: Lessons from SASA! SASA!, which means “Now!” in Kiswahili, is a program developed by Raising Voices and implemented in Uganda by the Center for Domestic Violence Prevention. It is the first community- based violence prevention program in Sub-Saharan Africa to be rigorously evaluated. The program employs multiple strategies to build a critical mass of engaged community members, leaders, and institutions, including local activism, media and advocacy, communication materials, and training. The Activist Kit that is central to SASA! community engagement and mobilization involves four phases: Start, Awareness, Support, and Action. The content evolves with each phase, with power as a central theme. Initial results from a randomized controlled trial show positive effects after almost three years of programming. Compared with control communities, people in SASA! communities have more gender-equitable attitudes and a reduced prevalence of past-year physical violence by an intimate partner. Compared with control communities, SASA! communities report the following striking results: ›› Levels of violence were 52 percent lower. ›› Twenty-eight percent more women and men believe it is acceptable for women to refuse sex. ›› Fifty percent more men and women believe physical violence against a partner is unacceptable. Source: Lori Michau, presentation at the UN Commission on the Status of Women, Washington, DC, March 2014. including the careful theory-based and cul- Legal reform and responses turally tailored design and engagement of Ending gender-based violence depends opinion leaders and the entire community. on the commitment of states to enact and Intervention length is also important. The implement effective and gender-equita- global systematic review undertaken for ble prevention and response mechanisms, this report confirmed that short-term and including prohibitions, policies, and ser- one-off interventions are less likely to show vices. International attention has been positive effects.110 Out of 26 interventions growing since the early 1990s, as evidenced with a less than one-month duration and by several key global norms and move- that included violence prevention as an ments. The 1993 World Conference on outcome, only one showed a statistically Human Rights recognized violence against significant positive result. Although the women as a human rights violation. That duration and intensity of interventions same year, the Declaration on the Elimi- appear to be important to their effective- nation of Violence against Women was the ness, most evaluated interventions are first international instrument to address short-term. These findings emphasize that violence against women explicitly, providing duration and intensity matter, particularly a framework for national and international when it comes to changing deep-seated action. The landmark 1994 International norms and behaviors. Conference on Population and Development Chapter 3 Freedom from violence | 75 called on governments to take legal and pol- removed in 2007, and marital rape can now icy measures to respond to and prevent vio- be prosecuted under the criminal code. Con- lence against women and girls.112 versely, in Kenya and Malawi, marital rape is exempt from criminal rape legislation and The international momentum to address is not even considered a crime. The same is violence has been reflected in national leg- true in India, even after the recent reforms islative reforms, especially over the past to respond to the Delhi bus rape and mur- decade, as shown in figure 3.8. Three-quar- der. The United Kingdom also removed the ters of the 100 countries included in Women, exemption for spousal rape from general Business and the Law 2014 criminalize legislation in 1991, but it has not yet intro- domestic violence and almost four-fifths duced specific provisions or legislation. (79) have legislation addressing sexual harassment in employment. But only 8 out Intimate partner violence remains out- of 100 countries have legislation against side the law in many countries, and in some harassment in public places.113 cases, it is explicitly allowed. The Supreme Court of the United Arab Emirates ruled in However, as shown in figure 3.9, only 2010, for example, that a man has the right 38 of these 100 countries have introduced to physically discipline his wife so long as he specific laws or provisions that explicitly does not leave physical marks.114 criminalize marital rape and sexual assault within marriage. In the 62 countries with Where laws do exist, better enforcement no specific legislation, marital rape can is critical and depends on several factors, still be tried under general rape legisla- including the capacity, training, and atti- tion unless a spousal exemption provision tudes of police and legal personnel. Recent exists. In Ghana, the spousal exemption was research from South Asia casts important FIGURE 3.8 Figure 3.8 Cumulative number of countries with legislation against domestic violence 76 72 74 69 67 61 63 54 56 47 41 35 37 28 30 27 23 19 13 8 3 4 1 2 1976 1984 1989 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 Source: World Bank, Women, Business and the Law 2014: Removing Restrictions to Enhance Gender Equality (Washington, DC: World Bank, 2013) data from 2001–12. 76 | Voice and Agency empowering women and girls for shared prosperity FIGURE 3.9 Countries with specific legislation criminalizing marital rape Is there a specific marital rape law or provision? n Yes n No n No data Source: World Bank, Women, Business and the Law (database), http://wbl.worldbank.org/. Note: Following the WBL methodology which considers the laws governing the main business city of a country, the United States does not meet the criteria for having specific legislation or provisions criminalizing marital rape. New York state removed the marital exemption but has not yet introduced specific legislation criminalizing marital rape. light on some common challenges within the of those that reached the courts resulted in justice system. In India, nearly all police offi- a judgment.116 cers interviewed (94 percent) agreed that A well-functioning justice system, with a husband is allowed to rape his wife. And appropriate sanctions against violence judges in Bangladesh and India typically felt against women, can help deter future vio- women abused by their spouse were partly lence and is an important element of pre- to blame.115 In Nepal, less than 5 percent of vention efforts. the police force is female, and in India, an overwhelming majority of police still view “A lot of men don’t want to get on the domestic violence as a private affair. Failure wrong side of the police. So for a lot of to prosecute remains a significant barrier to them, when there’s a court order, it’s a women accessing justice in many countries, deterrent.” including in Guatemala where of the more —Female community member, Port Moresby, Papua New than 20,000 cases of femicide and domestic Guinea117 violence filed in 2011, fewer than 3 percent Chapter 3 Freedom from violence | 77 While legal reform and responses are a key had domestic violence legislation in place is entry point for increasing women’s agency in associated with a reduced prevalence of about this domain, it is clear that law alone is insuffi- 2 percent. Although this finding underscores cient to prevent violence. Our analysis of DHS the promise of legislative reform as a preven- data points to the existence of national laws as tive measure, alone it is not enough to elimi- an important protective factor: women who nate violence. live in countries with domestic violence legis- In the Pacific, legislative reform and lation have 7 percent lower odds of experienc- investment in response structures are ing violence compared with women in coun- bringing improvements in women’s access tries without such laws.118 Analysis also shows to justice. The recent reforms in Papua New that each additional year that a country has Guinea are promising (box 3.6). Box 3.6 Increasing women’s access to justice in Papua New Guinea In Papua New Guinea, a Family Protection Bill was introduced to Parliament in 2013 after extensive and inclusive stakeholder consultation. This effort was accompanied by the establishment of specialized Family and Sexual Violence Units and survivor-centered training to duty bearers and service providers, and prosecution efforts were strengthened through the creation of a Family and Sexual Offense Unit to support legal staff and to improve case management. The early results are promising. The number of female magistrates rose from 10 in 2004 to more than 900 in 2013. Since 2012, a record number of IPV cases have been heard in Papua New Guinea’s courts, with two out of three resulting in a conviction or guilty plea. Of particular note are two landmark sexual violence cases involving police officers, who received combined sentences of 30 years imprisonment. Source: AusAID, 2012; Mukasa et al., 2014.119 Social support and services ■■ Legal aid—literacy, advocacy, and filing of key documents The needs of women who experience vio- ■■ Economic support—securing of land and lence depend greatly on their circumstances, property, employment, and income-gen- their social and legal context, and their own erating activities120 desired outcomes. Services for survivors broadly fall into four domains: Such services are typically short in dura- tion and are delivered by civil society groups ■■ Physical health—medical care, safe and trained professionals. The primary aim housing, food, and clothing may be to help stabilize survivors so they can ■■ Mental health—therapeutic and trauma report abuse. The evidence that these efforts counseling, building of self-efficacy and improve women’s psychological functioning self-confidence, and reduction of inter- is limited, suggesting that more substan- nalized stigma tial and repeated therapeutic interventions 78 | Voice and Agency empowering women and girls for shared prosperity may be needed.121 Emerging evidence from pair a victim advocate with a police officer trials of longer-term and group-based inter- to meet with a victim and share information ventions with survivors reveals promising on services and legal options, as well as to models for improving coping skills, reduc- warn the perpetrator of the consequences ing symptoms of depression and post-trau- of continued abuse. Eight of 10 experimen- matic stress disorder, and increasing self- tal or quasi-experimental studies, all in the efficacy.122 World Bank funding in the Solo- United States, found that such interventions mon Islands will support improved access to increased the number of subsequent calls to and quality of services for survivors of vio- the police, though there was no evidence of lence, as outlined in box 3.7. actual reductions in of violence.125 And such interventions have limited scope since they As highlighted earlier, most survivors are designed only for survivors who report of violence never report or seek help, a gap violence in the first place. that needs to be overcome. That gap under- scores the importance of efforts to increase Schools and workplaces can disseminate help-seeking behavior for health and other information, in places like public restrooms, social services.123 The limited evidence about by distributing newsletters, and holding what works to increase help-seeking has informational sessions. Awareness-raising identified some promising areas, including campaigns and activities may help break the culture of silence and stigmatization around grassroots outreach, secondary responder violence and urge people to seek help. programs, and mass and interpersonal com- munication efforts. Community outreach Innovative ICT tools can play a role. In workers who share health information can 2013 the World Bank Group supported two build trust with women in the community and Hackathons aimed at developing tools to can encourage help-seeking behaviors.124 To respond to or prevent gender based vio- date, however, evidence about effectiveness lence. The first encompassed countries in is lacking. Secondary responder programs Latin America and the second focused on Box 3.7 Coordinated support for survivors of violence in the Solomon Islands In the Solomon Islands, 64 percent of women have experienced physical and/or sexual partner violence. A World Bank grant supported the government’s efforts to improve access to services for survivors by bringing together specialized organizations in a forum that allowed them to identify gaps and priority actions in supporting victims of violence. As a next step, the grant will deliver a pilot project to strengthen frontline response and help coordinate referral services. Service providers will also be trained in sensitive and timely responses to the needs of survivors of violence. Finally, the project will help the government undertake a diagnostic study on capacity constraints in the institutions and services, and it will support training and data collection. Source: Improving Services for Victims of GBV Project (CGAP13-16). Chapter 3 Freedom from violence | 79 Nepal. A number of different mobile appli- the context. Interventions may be able to cations were developed including an app expand both women’s economic opportu- for women who have experienced violence nities and their agency by adding strategic which provides information and resources design features that intentionally address on obtaining support, and an app that gender-based violence.127 We refer to this allows the user to notify friends or family in approach as “economic empowerment plus” a threatening situation.126 intervention, which have shown success in several African countries. Through Internet-based tools (for example, Internet Resources and Informa- The IMAGE program in rural South Africa tion for Safety [IRIS]) and a smartphone combined microcredit with participatory app (MyPlan), the aim is to increase safe- gender training, social support groups, and ty-seeking behaviors by reducing survi- community mobilization. The results are vors’ decisional conflict about safety in encouraging: their relationship. While the results of the ■■ IPV fell by 55 percent. impact evaluations are not yet available, ■■ Household poverty rates fell. preliminary findings are promising, show- ing that after even just one use, survivors ■■ HIV-related communication improved, as have greater clarity about where and how to did social capital.128 seek help, and less uncertainty about doing In Côte d’Ivoire, a group savings inter- so. Increasing help-seeking is a critical first vention combined with “gender dialogue step toward improving the response mech- groups” led to significant reductions in anisms and ensuring justice for survivors. women’s reports of economic abuse and These broad types of services are essential their acceptance of wife beating compared for women’s well-being and should be inte- with women who participated in only group grated into service provision efforts for sur- savings meetings.129 Women who attended vivors of violence. more than 75 percent of sessions with their male partner were also less likely to report Economic empowerment (plus) physical IPV. In Uganda, a vocational training Increasing women’s economic opportunities program—paired with safe spaces for young women’s interactions and information on can be an important entry point for expand- health and risky behaviors—reduced the ing their ability to prevent and leave violent share of young women reporting forced sex relationships. Yet programs to promote eco- from 21 percent to nearly zero, while also nomic opportunities rarely take the risks of increasing engagement in income-generat- violence systematically into account, and ing activities by 35 percent.130 impact evaluations of effects on violence are rare. But emerging evidence suggests that These evaluations hint at ways to pro- economic empowerment and social protec- mote agency in economic programs, but the tion interventions by themselves can have evidence base is still too thin to determine significant effects on IPV, depending on the the active ingredients of “economic empow- nature of the program, the population, and erment plus” interventions for reducing 80 | Voice and Agency empowering women and girls for shared prosperity violence. Promising components include Preventing childhood exposure to and gender and health training, ongoing groups direct experience of violence is critical to break that facilitate social support, safe spaces the intergenerational cycles of violence.132 to interact with peers, community mobili- Chapters 1 and 2 have already highlighted zation, and, in some cases, elements that the critical role of education. Recent reviews engage male partners to encourage more show some evidence of effective interventions, equitable relationship dynamics as women’s including in developing countries.133 Evidence economic empowerment increases. Further from “girl-friendly” schools shows that adjust- testing of which combinations work best ing the physical school environment can dra- and under what circumstances is needed, as matically improve girls’ attendance, reten- are testing and scaling up. tion, and achievement. Changes included the construction of sex-segregated latrines, clean Integrating violence prevention water supplies, and an array of outreach activ- into other sectoral interventions ities to promote girls’ education. In Burkina Faso, girls’ enrollment increased by 5 percent- Violence is a complex issue, and conse- age points over that of their male classmates.134 quences have adverse ripple effects on other Both girls and boys showed improved test spheres of an individual survivor’s life, as scores in the intervention schools, pointing to well as her family and community. Greater positive effects for boys and increasing their access by women and girls to public spaces likelihood of school completion. can enhance agency by increasing educa- tional attainment, economic opportunities, The fear, threat, and experience of vio- and mobility. However with such opportu- lence are often obstacles to women’s and nities may come risks of being exposed to girls’ engagement in civic, political, and eco- violence. Evidence indicates that structural nomic spheres.135 Efforts to improve urban interventions—which aim to make envi- environments’ safety for women and girls ronments safer and to interrupt the cycle have included enhancements to police pro- of violence before it begins—offer promis- tocols and procedures, help lines, and efforts ing lessons for prevention.131 Among these to promote bystander intervention.136 The lessons are efforts to improve the safety World Bank is also investing in efforts to of public spaces for women and children, prevent and respond to violence through including investments in lighting, safe tran- projects in other sectors, such as an innova- sit, and safer, more gender-equitable school tive transport project in Brazil (box 3.8), to environments. reduce the risk of violence. Chapter 3 Freedom from violence | 81 Box 3.8 Integrating violence: Innovations in a transport project in Brazil As part of a broader effort to develop a regional master plan for the metropolitan transport system, the government of Rio de Janeiro is looking to use trains and cable cars that connect many hillside shantytowns to increase access to job opportunities. With a loan from the World Bank, the government has integrated gender-responsive legal, social, and economic services within this system. The program will support the establishment of Women Reference and Service Centers—one- stop shops that will provide services, referrals, and information for survivors of violence. Electronic public information is used to promote the national law against domestic violence. Sustainable transport links to the Brazilian Women’s House—a federal program that assembles several public services for women—will also be established as part of the effort. In addition, a women’s police station, a women’s clinic, and a child care center will be established. The project will also finance dissemination of information on improved security for women riders. Finally, the program will pilot vocational education and training for women that are at risk of violence. Source: Urban Mass Transport: Gender Agency and Inclusion Project (P147695). *** Violence is one of the most pronounced manifestations of gender inequality across the globe. Addressing violence is critical for promoting women’s agency and for a range of other gender outcomes and development goals. But violence is perpetuated by social norms and institutions that implicitly or explicitly condone this behavior. This means that eliminating violence requires appropriate prevention strategies, as well as response mechanisms that prioritize the needs of survivors. Effective approaches are likely to target multiple entry points within the framework depicted in figure 1.1— improved laws and policies, more gender equitable attitudes, and safer social envi- ronments. This also requires better data for decision making. One recent advance is the international agreement on standard indicators for measuring violence against women, which will provide much-needed data on prevalence as well as monitoring of efforts to reduce violence, as discussed in chapter 7. Preventing violence is critical— and has positive effects in other domains of women’s agency, as will be highlighted in the chapters that follow. 82 | Voice and Agency empowering women and girls for shared prosperity Endnotes 7 UN General Assembly Resolution A/ RES/48/104, “Declaration on the Elimina- Smriti Singh, “Delhi Gang Rape: Case Diary,” 1 tion of Violence against Women,” December Times of India, September 13, 2013, http:// 20, 1993. articles.timesofindia.indiatimes.com/2013- 8 High-Level Panel of Eminent Persons on 09-13/india/41936569_1_mount-eliza - the Post-2015 Development Agenda, A New beth-hospital-delhi-hc-safdurjung-hospital; Global Partnership: Eradicate Poverty and DeNeen L. Brown and Ashley Surdin. “Chris Transform Economies through Sustainable Brown Pleads Guilty to Assault,” Wash- Development (New York: United Nations, ington Post , June 23, 2009, http://www. 2013). washingtonpost.com/wp-dyn/content/ article/2009/06/22/AR2009062200452. 9 Radhika Coomaraswamy, Integration of html?wprss=rss_print/style; and “Saat- Human Rights of Women and the Gender Per- chi Says He Is Divorcing Nigella Lawson,” spective: Violence against Women, Report of Washington Post, July 7, 2013, http:// the Special Rapporteur on Violence against w w w.washing tonpost .com/ lifest yle/ Women, Its Causes and Consequences, style/saatchi-says-he-is-divorcing-nigella- Addendum: Communications to and from lawson/2013/07/07/795ee4d8-e730- Governments (New York: United Nations, 11e2-aa9f-c03a72e2d342 _ stor y.ht ml; 1999). Child marriage is covered in chapter Joanna Pepin, “Cultural Portrayals of Inti- 3. mate Partner Violence: Analyzing Media 10 Jennifer McCleary-Sills, Zayid Douglas, Coverage of Celebrity and Sports’ Figures Annagrace Rwehumbiza, Aziza Hamisi, and Domestic Violence Relationships,” seminar, Richard Mabala, Vijana Tunaweza Newala: University of Maryland, College Park, March Findings from a Participatory Research 6, 2013. and Action Project in Tanzania (Washing- World Health Organization, Global and 2 ton, DC: International Center for Research Regional Estimates of Violence against on Women, 2011); Carol Underwood, Women: Prevalence and Health Effects of Joanna Skinner, Nadia Osman, and Hilary Intimate Partner Violence and Non-Part- HSchwandt, “Structural Determinants of ner Sexual Violence (Geneva: World Health Adolescent Girls’ Vulnerability to HIV: Views from Community Members in Botswana, Organization, 2013). Malawi, and Mozambique,” Social Science & United Nations, World Population Pros- 3 Medicine 73, no. 2 (2011): 343–50. pects: The 2012 Revision (New York: United 11 Jennifer McCleary-Sills, Sophie Namy, Joyce Nations, 2013), DVD. Nyoni, Datius Rweyemamu, Adrophina WHO, Global and Regional Estimates of Vio- 4 Salvatory, and Ester Steven, “Help-Seek- lence against Women. ing Pathways and Barriers for Survivors of Inter-Agency Standing Committee, Guide- 5 Gender-Based Violence in Tanzania: Results lines for Gender-Based Violence Interven- from a Study in Dar es Salaam, Mbeya, and tions in Humanitarian Settings: Focusing on Iringa Regions” (Dar es Salaam: Engender- Prevention of and Response to Sexual Vio- Health/CHAMPION, 2013). lence in Emergencies (Geneva: Inter-Agency 12 McCleary-Sills et al., “Help-Seeking Standing Committee, 2005). Pathways.” United Nations Population Fund, “End- 6 13 Manju Rani, Sekhar Bonu, and Nafissatou ing Widespread Violence against Women,” Diop-Sidibé, “An Empirical Investigation of United Nations Population Fund, New York, Attitudes towards Wife-Beating among Men 2007. and Women in Seven Sub-Saharan African Chapter 3 Freedom from violence | 83 Countries,” African Journal of Reproductive Lori Heise, Mary Carroll Ellsberg, and M. 20 Health 8, no. 3 (2004):116–36; Manju Rani Gottemoeller, Ending Violence against and Sekhar Bonu, “Attitudes toward Wife Women, Population Reports, Series L, no. Beating: A Cross-Country Study in Asia,” 11 (Baltimore: Johns Hopkins University Journal of Interpersonal Violence 24, no. 8 School of Public Health, 1999). (2009):1371–97. Heidi HHHStöckl, Karen Devries, Alexandra 21 Rachael S. Pierotti, “Increasing Rejection 14 Rotstein, Naeemah Abrahams, Jacquelyn of Intimate Partner Violence: Evidence of C. Campbell, Charlotte Watts, and Claudia Global Cultural Diffusion,” American Socio- García-Moreno, “The Global Prevalence logical Review 78, no. 2 (2013): 240–65. of Intimate Partner Homicide: A System- Analysis of DHS data. Our sample covers 15 atic Review,” Lancet 382, no. 9895 (2013): 40 percent of the world’s population; 50 859–65. percent if China is excluded. The variable Claudia García-Moreno, Henrica A. F. M. 22 analyzed here is the share of women who Jansen, Mary Carroll Ellsberg, Lori Heise, accept wife beating for one or more of five Charlotte Watts, WHO Multi-Country Study reasons: (a) going out without permission, on Women’s Health and Domestic Violence (b) burning food, (c) neglecting children, against Women (Geneva: World Health (d) refusing sex, and (e) arguing with a Organization, 2005). husband. Council of Europe Convention on Prevent- 23 World Bank and International Finance Cor- 16 ing and Combating Violence against Women poration, Women, Business and the Law and Domestic Violence, CETS no. 210, (database), http://wbl.worldbank.org/. November 5, 2001, http://conventions.coe. Commission on the Status of Women, 17 int/Treaty/EN/Treaties/Html/210.htm. “Agreed Conclusions on the Elimination WHO, Global and Regional Estimates of Vio- 24 and Prevention of All Forms of Violence lence against Women. against Women and Girls,” March 2013, Sunita Kishor and Kiersten Johnson, Pro- 25 http://www.un.org/womenwatch/daw/ filing Domestic Violence: A Multi-Country csw/csw57/CSW57_Agreed_Conclusions_ Study,” (Columbia, MD: ORC Macro, 2004) (CSW_report_excerpt).pdf. and the related expert paper by Sunita They include (a) Resolution 1325, which 18 Kishor, “Domestic Violence Measurement recognizes the impact of conflict on women in the Demographic and Health Surveys: and girls, particularly their vulnerability to The History and the Challenges.”; World sexual violence; (b) Resolution 1820, calling Health Organisation, Putting Women First: for an end to sexual violence and recogniz- Ethical and Safety Recommendations for ing sexual violence as a weapon of war; (c) Research on Domestic Violence against Resolution 1888, which calls for strength- Women (Geneva: World Health Organiza- ened leadership to end sexual violence, and tion, 2001);ealHH Mary Ellsberg and Lori (d) Regulation 1960, calling for the involve- Heise, Researching Violence against Women: ment of women and civil society organiza- A Practical Guide for Researchers and Activ- tions in implementing UNSCR 1325 and ists, Geneva: World Health Organization establishing tools for action, monitoring, and Program for Appropriate Technology in and reporting. Health (PATH), 2005; Sara Bott, Alessandra Bina Agarwal and Pradeep Panda, “Toward 19 Guedes, Mary Goodwin, and Jennifer Adams Freedom from Domestic Violence: The Mendoza, Violence against Women in Latin Neglected Obvious,” Journal of Human America and the Caribbean: A Comparative Development 8, no. 3 (2007): 362. Analysis of Population-Based Data from 12 84 | Voice and Agency empowering women and girls for shared prosperity Countries (Washington, DC: Pan American and Nadwah Hammoury, “Coerced Sexual Health Organization, 2013). Intercourse within Marriage: A Clinic-Based Study of Pregnant Palestinian Refugees in Bott et al., “Violence against Women in Latin 26 Lebanon,” Journal of Midwifery & Women’s America and the Caribbean.” Health 53, no. 2 (2008):150–54; and Jessica Xiao Xu, Fengchuan Zhu, Patricia O’Campo, 27 D. Gipson and Michelle J. Hindin, “‘Marriage Michael A. Koenig, Victoria Mock, and Jac- Means Having Children and Forming Your quelyn C. Campbell, “Prevalence of and Risk Family, So What is the Need of Discussion?’ Factors for Intimate Partner Violence in Communication and Negotiation of Child- China,” American Journal of Public Health bearing Preferences among Bangladeshi 95, no. 1 (2005): 78–85; Jacquelyn C. Camp- Couples,” Culture, Health & Sexuality 9, no. bell, Daniel Webster, Jane Koziol-McLain, 2 (2007):185–98. Carolyn Block, Doris Campbell, Mary Ann 30 Amy Weissman, Janine Cocker, Lisa Sher- Curry, Faye Gary, Nancy Glass, Judith McFar- burne, Mary Beth Power, Ronnie Lovich, and lane, Carolyn Sachs, Phyllis Sharps, Yvonne Mary Mukaka, “Cross-Generational Rela- Ulrich, Susan A. Wilt, Jennifer Manganello, tionships: Using a ‘Continuum of Volition’ in Xiao Xu, Janet Schollenberger, Victoria Frye, HIV Prevention Work among Young People,” and Kathryn Laughon, “Risk Factors for Gender & Development 14, no. 1 (2006): Femicide in Abusive Relationships: Results 81–94. from a Multisite Case Control Study,” Ameri- can Journal of Public Health 93, no. 7 (2003): 31 Jennifer McCleary-Sills, Zayid Douglas, 1089–97. Annagrace Rwehumbiza, Aziza Hamisi, and Richard Mabala, “Gendered Norms, Trans- Susan Roth, Elena Newman, David Pelcovitz, 28 actional Sex, and Adolescent Pregnancy in and Francine S. Mandel, “Complex PTSD Rural Tanzania,” Reproductive Health Mat- in Victims Expose to Sexual and Physical ters 21, no. 41 (2013): 97–105. Abuse: Results from the DSM-IV Field Trial 32 Ruth Hope, Addressing Cross-Generational for Posttraumatic Stress Disorder,” Journal Sex: A Desk Review of Research and Pro- of Traumatic Stress 10, no. 4 (1997): 539– grams (Washington, DC: Population Refer- 55; Marylene Cloitre, Bradley C. Stolbach, ence Bureau, 2007). Judith L. Herman, Bessel van der Kolk, Rob- ert Pynoos, Jing Wang, and Eva Petkova, “A 33 Audrey E. Pettifor, Helen V. Rees, Immo Developmental Approach to Complex PTSD: Kleinschmidt, Annie E. Steffenson, Cather- Childhood and Adult Cumulative Trauma as ine MacPhail, Lindiwe Hlongwa-Madikizela, Predictors of Symptom Complexity,” Journal Kerry Vermaak, and Nancy S. Padian, “Young of Traumatic Stress 22, no. 5 (2009): 399– People’s Sexual Health in South Africa: HIV 408; Victoria M. Follette, Melissa A. Polusny, Prevalence and Sexual Behaviors from a Anne E. Bechtle, and Amy E. Naugle, “Cumu- Nationally Representative Household Sur- lative Trauma: The Impact of Child Sexual vey,” AIDS 19, no. 14 (2005): 1525–34; and Abuse, Adult Sexual Assault, and Spouse Nancy Luke, “Age and Economic Asymme- Abuse,” Journal of Traumatic Stress 9, no. 1 tries in the Sexual Relationships of Adoles- (1996): 25–35. cent Girls in Sub-Saharan Africa,” Studies in Family Planning 34 no. 2 (2003): 67–86. Rebekah E. Gee, Nandita Mitra, Fei Wan, 29 Diana F. Chavkin, and Judith A. Long, “Power 34 rs = 0.71 (37), p < 0.001. over Parity: Intimate Partner Violence and 35 Geeta Rao Gupta and Ellen Weiss, “Women’s Issues of Fertility Control,” American Jour- Lives and Sex: Implications for AIDS Preven- nal of Obstetrics & Gynecology 201, no. 2 tion,” Culture, Medicine and Psychiatry 17, (2009): 148.e1–148.e7; Marwan Khawaja no. 4 (1993): 399–412. Chapter 3 Freedom from violence | 85 N. N. Sarkar, “Barriers to Condom Use,” 36 45 All sources of support except police. Euro- European Journal of Contraception and pean Union Agency for Fundamental Rights, Reproductive Health Care 13, no. 2 (2008): “Violence against Women.” 114–22. 46 McCleary-Sills et al., “Help-Seeking Path- García-Moreno et al., WHO Multi-Country 37 ways and Barriers.” Study. 47 Sonia M. Frías, “Strategies and Help-Seeking “Building a Safer World for Children: 38 Behavior among Mexican Women Experi- Together for Girls Stakeholder Report encing Partner Violence,” Violence against 2010–2012,” Together for Girls, Washing- Women 19, no. 1 (2013): 24–49. ton, DC, 2013. 48 Naved et al., “Physical Violence by WHO, Global and Regional Estimates of Vio- 39 Husbands.” lence against Women. 49 Mary Ellsberg, Brian Heilman, Sophie Rachel Jewkes, Emma Fulu, Tim Roselli, 40 Namy, Manuel Contreras, and Robin Hayes, and Claudia García-Moreno, “Prevalence of “Violence against Women in Melanesia and Factors Associated with Non-Partner and Timor-Leste: Progress Made since the Rape Perpetration: Findings from the UN 2008 Office of Development Effectiveness Multi-Country Cross-Sectional Study on Men Report,” Australian Agency for International and Violence in Asia and the Pacific,” Lancet Development, Canberra. Global Health 1, no. 4 (2013): e208–e218. 50 Jacquelyn C. Campbell, “Health Conse- Tia Palermo, Jennifer Bleck, and Amber 41 quences of Intimate Partner Violence,” Lan- Peterman, “Tip of the Iceberg: Reporting cet 359, no. 9314 (2002): 1331–36; Ann and Gender-Based Violence in Developing L. Coker, James E. Ferguson, Heather M. Countries,” American Journal of Epidemiol- HBush, Carol Jordan, and Leslie Crofford, ogy 179, no. 5 (2014): 602–12. “Intimate Partner Violence and Gyneco- Ruchira Tabassum Naved, Safia Azim, Abbas 42 logic Health: Focus on Women in Kentucky,” Bhuiya, and Lars Åke Persson, “Physical Vio- paper presented at the American Public lence by Husbands: Magnitude, Disclosure Health Association Annual Meeting and and Help-Seeking Behavior of Women in Expo, San Diego, October 29, 2008. Bangladesh,” Social Science & Medicine 62, 51 Angela Joy Taft and Lyndsey F. Watson, no. 12 (2006): 2917–29. “Depression and Termination of Pregnancy Our analysis of DHS data, available for 24 43 (Induced Abortion) in a National Cohort of countries. Young Australian Women: The Confounding Effect of Women’s Experience of Violence,” Banu Akadli Ergöçmen, İlknur Yüksel-Kap- 44 BMC Public Health 8, no. 75 (2008): 1–8. tanoğlu, and Henrica A. F. M. Jansen, “Inti- mate Partner Violence and the Relation 52 Hind A. Beydoun, May A. Beydoun, Jay S. between Help-Seeking Behavior and the Kaufman, Bruce Lo, and Alan B. Zonderman, Severity and Frequency of Physical Violence “Intimate Partner Violence against Adult among Women in Turkey,” Violence against Women and Its Association with Major Women 19, no. 9 (2013): 1151–74; Naved Depressive Disorder, Depressive Symptoms et al., “Physical Violence by Husbands.”; and Postpartum Depression: A Systematic Michelle Fugate, Leslie Landis, Kim Rior- Review and Meta-Analysis,” Social Science & dan, Sara Naureckas, and Barbara Engel, Medicine 75, no. 6 (2012): 959–75. “Barriers to Domestic Violence Help Seek- Suzanne Maman, Jessie K. Mbwambo, Nora 53 ing: Implications for Intervention,” Violence M. Hogan, Gad P. Kilonzo, Jacquelyn C. Camp- against Women 11, no. 3 (2005): 290–310. bell, Ellen Weiss, and Michael D. Sweat, 86 | Voice and Agency empowering women and girls for shared prosperity “HIV-Positive Women Report More Lifetime health and violence against women: A gen- Partner Violence: Findings from a Volun- der perspective] (Nuevo Léon, Mexico: Aso- tary Counseling and Testing Clinic in Dar ciación Mexicana de Población, Colegio de es Salaam, Tanzania,” American Journal of México, 1996). Public Health 92, no. 8 (2003): 1331–37; 61 Stephenie Holt, Helen Buckley, and Sadhbh Jacquelyn C. Campbell, Anne B. Woods, Kath- Whelan, “The Impact of Exposure to Domes- ryn Laughon Chouaf, and Barbara Parker, tic Violence on Children and Young People: “Reproductive Health Consequences of Inti- A Review of the Literature,” Child Abuse & mate Partner Violence: A Nursing Research Neglect 32, no. 8 (2008): 797–810. Review,” Clinical Nursing Research 9, no. 3 (2000): 217–37; Dan K. Kaye, Florence M. 62 Shireen J. Jejeebhoy, “Associations between Mirembe, Grace Bantebya, Annika Johansson, Wife-Beating and Fetal and Infant Death: and Anna Mia Ekstrom, “Domestic Violence Impressions from a Survey in Rural India,” as Risk Factor for Unwanted Pregnancy and Studies in Family Planning 29, no. 3 (1998): Induced Abortion in Mulago Hospital, Kam- 300–308. pala, Uganda,” Tropical Medicine & Interna- 63 Sunita Kishor and Kiersten Johnson, Pro- tional Health 11, no. 1 (2006): 90–101. filing Domestic Violence: A Multi-Country Nata Duvvury, Nguyen Huu Minh, and Patri- 54 Study (Columbia, MD: ORC Macro, 2004). cia Carney, Estimating the Cost of Domestic 64 Jacquelyn C. Campbell, “Health Conse- Violence against Women in Viet Nam (Hanoi: quences of Intimate Partner Violence,” Lan- UN Women, 2012). cet 359, no. 9314 (2002): 1331–36. Seema Vyas, Estimating the Association 55 65 World Bank, Caribbean Youth Development: between Women’s Earnings and Partner Vio- Issues and Policy Directions (Washington, lence: Evidence from the 2008–2009 Tanza- DC: World Bank, 2003). nia National Panel Survey (Washington, DC: 66 Kishor and Johnson, Profiling Domestic World Bank, 2013). Violence. International Center for Research on 56 67 Gerry Eijkemans, “Breaking the Cycle: Women, Domestic Violence in India: A Sum- Addressing the Intersections between mary Report of a Multi-Site Household Sur- Violence against Women and Violence vey (Washington, DC: International Center against Children,” http://www.unicef. for Research on Women, 2000). org/barbados/ECAO_ Addressing _the_ International Center for Research on 57 intersections_between_Violence_againist_ Women, Intimate Partner Violence: High Women_and_Children.pdf. Costs to Households and Communities 68 Duvvury et al., Estimating the Cost of Domes- (Washington, DC: International Center for tic Violence. Research on Women, 2009). 69 Kim C. Lim, John Rioux, and Ellen Ridley, National Household Survey 2009/2010: 58 “Impact of Domestic Violence Offenders on Socio=Economic Module (Kampala: Uganda Occupational Safety & Health: A Pilot Study,” Bureau of Statistics, 2010), http://www. Family Crisis Services and Maine Depart- ubos.org/UNHS0910/unhs200910.pdf. ment of Labor, Augusta, February 2004. Duvvury et al., Estimating the Cost of 59 70 Tanis Day, Katherine McKenna, and Audra Domestic Violence. Bowlus, “The Economic Costs of Violence M. Granados Shiroma, Salud reproductiva y 60 against Women: An Evaluation of the Litera- violencia contra la mujer: Un análisis desde ture,” expert brief compiled in preparation la perspectiva de género [Reproductive for the Secretary-General’s in-depth study Chapter 3 Freedom from violence | 87 on all forms of violence against women, (Odds Ratio: 0.89, 95 percent Confidence 79 University of Western Ontario, London, Interval: 0.80 - 0.97); Paul J. Fleming, Gary Ontario, 2005. Barker, Jennifer McCleary-Sills, and Mat- Amy E. Bonomi, Melissa L. Anderson, 71 thew Morton,“Engaging Men and Boys in Frederick P. Rivara, and Robert S. Thomp- Advancing Women’s Agency,” Women’s son,“Health Care Utilization and Costs Asso- Voice, Agency, and Participation Research ciated with Physical and Nonphysical-Only Series 2013, no.1, World Bank Washington, Intimate Partner Violence,” Health Services DC, 2013. Research 44, no. 3 (2009): 1054–67. Emma Fulu, Xian Warner, Stephanie Mie- 80 Cristina Santos, “Costs of Domestic Vio- 72 dema, Rachel Jewkes, Tim Roselli, and lence: A Life Satisfaction Approach,” Fiscal James Lang, Why Do Some Men Use Violence Studies 34, no. 3 (2013): 391–409. against Women and How Can We Prevent It? Quantitative Findings from the United Nata Duvvury, Aoife Callan, Patrick Carney, 73 Nations Multi-Country Study on Men and and Srinivas Raghavendra, “Intimate Part- Violence in Asia and the Pacific. (Bangkok: ner Violence: Economic Costs and Impli- UNDP, UNFPA, UN Women, and UNV, 2013). cations for Growth and Development,” Women’s Voice, Agency, and Participation Seema Vyas and Charlotte Watts, “How Does 81 Research Series 2013, no. 3, World Bank, Economic Empowerment Affect Women’s Washington, DC, 2013. Risk of Intimate Partner Violence in Low and Middle Income Countries? A Systematic Vyas, Estimating the Association between 74 Review of Published Evidence,” Journal of Women’s Earnings and Partner Violence. International Development 21, no. 5 (2009): Andrew R. Morrison and Man’a Beatriz 75 577–602. Orlando, “Social and Economic Costs of Jana Lenze and Stephan Klasen, “The 82 Domestic Violence: Chile and Nicaragua,” Impact of Women’s Labour Force Partic- in Too Close to Home: Domestic Violence in ipation on Domestic Violence in Jordan,” the Americas, ed. Andrew R. Morrison and Discussion Paper no. 143, Courant Research Marḯa Loreto Biehl (New York: Inter-Ameri- Centre: Poverty, Equity and Growth, Göttin- can Development Bank, 1999), 51–67. gen, Germany, 2013. Mayra Buvinic, Andrew Morrison, and 76 Rachel Heath, “Women’s Access to Labor 83 Michael Shifter, “Violence in Latin Amer- Market Opportunities, Control of House- ica and the Caribbean: A Framework for hold Resources, and Domestic Violence: Action,” technical study, Sustainable Devel- Evidence from Bangladesh,” World Develop- opment Department, Inter-American Devel- ment 57 (2013): 32–46. opment Bank, Washington, DC, March 1999. Melissa Hidrobo and Lia Fernald, “Cash 84 Rachel Jewkes, “Intimate Partner Violence: 77 Transfers and Domestic Violence,” Journal Causes and Prevention,” Lancet 359, No. of Health Economics 32 (2013): 304. 9315 (2002): 1423–29. Johannes Haushofer and Jeremy Shapiro, 85 Gary Barker, Juan Manuel Contreras, Brian 78 “Policy Brief: Impact of Unconditional Cash Heilman, Ajay Singh, Ravi Verma, and Transfers,” Innovations for Poverty Action, Marcos Nascimento, Evolving Men: Initial New Haven, CT, October 24, 2013. Results from the International Men and Gen- der Equality Survey (IMAGES) (Washington, Gustavo J. Bobonis, Melissa González- 86 DC, and Rio de Janeiro: International Center Brenes, and Roberto Castro, “Public Trans- for Research on Women and Instituto Pro- fers and Domestic Violence: The Roles of mundo, 2011). Private Information and Spousal Control,” 88 | Voice and Agency empowering women and girls for shared prosperity American Economic Journal: Economic Pol- (Calverton, MD: Ministère du Plan et Macro icy 5, no. 1 (2013): 179–205. International, 2008). Amy Farmer and Jill Tiefenthaler, “An Eco- 87 95 Cari Jo Clark, Susan A. Everson-Rose, Sha- nomic Analysis of Domestic Violence,” kira Franco Suglia, Rula Btoush, Alvaro Review of Social Economy 55, no. 3 (1997): Alonzo, and Muhammad M Haj-Yahia, 337–58. “Association between Exposure to Politi- Manasi Bhattacharyam, Arjun S. Bedi, and 88 cal Violence and Intimate-Partner Violence Amrita Chhachhi, “Marital Violence and in the Occupied Palestinian Territory: A Women’s Employment and Property Sta- Cross-Sectional Study,” Lancet 375, no. 9711 tus: Evidence from North Indian Villages,” (2010): 310–16. Discussion Paper no. 4361, Institute for the 96 Patrick Vinck and Phuong N. Pham, “Asso- Study of Labor, Bonn, August 2009. ciation of Exposure to Intimate Partner Douglas D. Perkins and Ralph B. Taylor, 89 Physical Violence and Potentially Traumatic “Ecological Assessments of Community Dis- War-Related Events with Mental Health order: Their Relationship to Fear of Crime in Liberia,” Social Science & Medicine 77 and Theoretical Implications,” American (2013): 41–49. Journal of Community Psychology 24, no. 1 97 Dominik Noe, and Johannes Rieckmann, (1996): 63–107. “Violent Behaviour: The Effect of Civil Con- Christina C. Pallitto and Patricia O’Campo, 90 flict on Domestic Violence in Colombia,” Dis- “Community Level Effects of Gender cussion Paper no. 136, Courant Research Inequality on Intimate Partner Violence and Centre: Poverty, Equity and Growth, Göttin- Unintended Pregnancy in Colombia: Testing gen, Germany, 2013. the Feminist Perspective,” Social Science & 98 Diana J. Arango, Matthew Morton, Floriza Medicine 10, no. 60 (2005): 2205–16. Gennari, Sveinung Kiplesund, and Mary Ell- Jennifer McCleary-Sills, “Jordanian Social 91 sberg, “Interventions to Prevent and Reduce Norms and the Risk of Intimate Partner Violence Against Women and Girls: A Sys- Violence and Limited Reproductive Agency,” tematic Review of Reviews,” Women’s Voice, Journal of International Women’s Studies 2, Agency and Participation Research Series, no. 14 (2013): 12–29. World Bank, Washington, DC, forthcoming. Jo Spangaro, Anthony Zwi, Chinelo Adogu, 92 99 Arango et al., “Systematic Review of Geetha Ranmuthugala, Gawaine Powell Reviews.” Davies, and Léa Steinacker, What Is the Evi- 100 Eve Waltermaurer, “Public Justification of dence of the Impact of Initiatives to Reduce Intimate Partner Violence: A Review of the Risk and Incidence of Sexual Violence in Literature,” Trauma, Violence, & Abuse 13, Conflict and Post-Conflict Zones and Other no. 3 (2012): 167–75. Humanitarian Crises in Lower- and Mid- dle-Income Countries? A Systematic Review 101 Jewkes et al., “Prevalence of and Fac- (London: EPPI-Centre, Social Science tors Associated with Non-partner Rape Research Unit, Institute of Education, Uni- Perpetration.” versity of London, 2013). 102 Population-weighted medians, calculated Elizabeth Jean Wood, “Variation in Sexual 93 from DHS data. Violence in War,” Politics & Society 34, no. 3 103 Sarah N. Keller, Timothy Wilkinson, and A. (2006): 307–42. J. Otgen, “Unintended Effects of a Domestic Enquête démographique et de santé, 94 Violence Campaign,” Journal of Advertising République Démocratique du Congo 2007 39, no. 4 (2010): 53–68. Chapter 3 Freedom from violence | 89 104 “Engaging Men to Prevent Gender-Based Murad H. Elsaidi, “Human Rights and 114 Violence: A Multi-Country Intervention and Islamic Law: A Legal Analysis Challenging Impact Evaluation Study,” report for the UN the Husband’s Authority to Punish ‘Rebel- Trust Fund, Instituto Promundo, Washing- lious’ Wives,” Muslim World Journal of ton, DC, 2012. Human Rights 7, no. 2 (2011): 1–25. 105 “Engaging Men to Prevent Gender-Based M. E. Khan, Ismat Bhuiya, and Aruna Bhat- 115 Violence.” tacharya, “A Situation Analysis of Care and 106 Anuradha Rajlan and Swati Chakraborty, Support for Rape Survivors at First Point “Regional Report of the Assessment of ‘We of Contact in India and Bangladesh,” Injury Can’ Phase II,” draft, Oxfam, UK, December Prevention 16 (2010): A160–A161. 2010. Karen Musalo and Blaine Bookey, “Crimes 116 107 Shereen Usdin, Esca Scheepers, Susan without Punishment: An Update on Vio- Goldsteinb, and Garth Japhet, “Achieving lence against Women and Impunity in Gua- Social Change on Gender-Based Violence: temala,” Hastings Race Poverty and Law A Report on the Impact Evaluation of Soul Journal 10 (2013): 265. City’s Fourth Series,” Social Science & Medi- Stella Mukasa, Jennifer McCleary-Sills, Brian 117 cine 61, no. 11 (2005): 2434–45. Heilman, Sophie Namy, Laura Brady, and 108 Patricia M. Fabiano, H. Wesley Perkins, Alan Shawna Stich, Review of DFAT Programs in Berkowitz, Jeff Linkenbach, and Christo- the Pacific Aimed at Ending Violence against pher Stark, “Engaging Men as Social Justice Women (Washington, DC: International Allies in Ending Violence against Women: Center for Research on Women, 2014). Evidence for a Social Norms Approach,” The countries that had both DHS data on 118 Journal of American College Health 52, no. 3 domestic violence and data on legisla- (2003): 105–12. tion from the World Bank are Azerbaijan, Raising Voices website, http://raisingvoices. 109 Burkina Faso, Cameroon, Colombia, Côte org/sasa/#tabs-419-0-3. d’Ivoire, Ghana, Haiti, Honduras, India, Kenya, Malawi, Mozambique, Nepal, Nige- 110 Arango et al., “Systematic Review of ria, Peru, the Philippines, Tanzania, Uganda, Reviews.” Ukraine, Zambia, and Zimbabwe 111 As quoted in Anna de Costa, “Mobilizing Office of the Public Prosecution, “Strongim 119 Young Men in India to End Violence against Gavman Program Internal Progress Report,” Women,” Christian Science Monitor, Jan- Australian Agency for International Devel- uary 27, 2014, http://www.csmonitor. opment, Port Moresby, Papua New Guinea, com/World/Making-a-difference/Change- 2012; Mukasa et al., Review of DFAT Agent/2014/0127/Mobilizing-young-men- Programs. in-India-to-end-violence-against-women. Mukasa et al., Review of DFAT Programs. 120 112 “Global Norms and Standards: Ending Vio- lence against Women,” UN Women web- Barbara Sims, Berwood Yost, and Christina 121 site, http://www.unwomen.org/en/what- Abbott, “The Efficacy of Victim Services Pro- we-do/ending-violence-against-women/ grams: Alleviating the Psychological Suffer- global-norms-and-standards. ing of Crime Victims?” Criminal Justice Pol- icy Review 17, no. 4 (2006): 387–406. 113 World Bank and International Finance Cor- poration, Women, Business and the Law Judith Bass, “Group Cognitive Processing 122 2014: Removing Restrictions to Enhance Therapy: A Specialized Mental Health Inter- Gender Equality (London: Bloomsbury Pub- vention That Supports Improvements in lishing, 2013). Well-Being for Sexual Violence Survivors,” 90 | Voice and Agency empowering women and girls for shared prosperity paper presented at the Sexual Violence (2006): 1973–83; Paul M. Pronyk, Trudy Research Initiative Forum, Bangkok, Thai- Harpham, Joanna Busza, Godfrey Phetla, land, October 16, 2013; Rachel A. Cohen, Linda A. Morison, James R. Hargreaves, “Common Threads: A Recovery Programme Julia C. Kim, Charlotte H. Watts, and John D. for Survivors of Gender Based Violence,” Porter, “Can Social Capital Be Intentionally Intervention 11, no. 2 (2013): 157–68. Generated? A Randomized Trial from Rural Amelia Gulliver, Kathleen M. Griffiths, Helen 123 South Africa,” Social Science & Medicine 67, Christensen, and Jacqueline L Brewer, “A no. 10 (2008): 1559–70. Systematic Review of Help-Seeking Inter- 129 Jhumka Gupta, Kathryn L Falb, Heidi Leh- ventions for Depression, Anxiety and Gen- mann, Denise Kpebo, Ziming Xuan, Mazeda eral Psychological Distress,” BMC Psychiatry Hossain, Cathy Zimmerman, Charlotte 12, no. 1 (2012): 81. Watts, and Jeannie Annan, “Gender Norms Patricia J. Kelly, Janna Lesser, Esther Per- 124 and Economic Empowerment Intervention alez-Dieckmann, and Martha Castilla, “Com- to Reduce Intimate Partner Violence against munity-Based Violence Awareness,” Issues Women in Rural Côte d’Ivoire: A Random- in Mental Health Nursing 28, no. 3 (2007): ized Controlled Pilot Study,” BMC Interna- 241–53; Jody M. Greene, Susan T. Ennett, tional Health and Human Rights 13, no. 1 and Christopher L. Ringwalt, “Prevalence (2013): 46. and Correlates of Survival Sex among Run- 130 Oriana Bandiera, Niklas Buehren, Robin away and Homeless Youth,” American Burgess, Markus Goldstein, Selim Gulesci, Journal of Public Health 89, no. 9 (1999): Imran Rasul, and Munshi Sulaiman, 1406–9. “Empowering Adolescent Girls: Evidence Robert C. Davis, David Weisburd, and Bruce 125 from a Randomized Control Trial in Uganda,” Taylor, “Effects of Second Responder Pro- World Bank, Washington, DC, 2012. grams on Repeat Incidents of Family Abuse: 131 Kirsten Beyer, Anne Baber Wallis, and L. A Systematic Review,” Australian Institute of Kevin Hamberger, “Neighborhood Envi- Criminology, Canberra, 2008. ronment and Intimate Partner Violence: A World Bank, “Domestic Violence: Can your 126 Systematic Review,” Trauma Violence Abuse, smartphone save your life?” World Bank December 26, 2013. Group blog, 2013, http://www.worldbank. 132 Lynette M. Renner and Kirsten Shook Slack, or g /e n/ne w s/f e a t u r e/ 2013/01/ 22/ “Intimate Partner Violence and Child Mal- domestic-violence-hackathon- treatment: Understanding Intra- and Inter- smartphone-lifesaver. generational Connections,” Child Abuse & Sydney Ruth Schuler, Syed M. Hashemi, 127 Neglect , 30, no. 6 (2006): 599–617, http:// Ann P. Riley, and Shireen Akhter, “Credit dx.doi.org/10.1016/j.chiabu.2005.12.005. Programs, Patriarchy and Men’s Violence 133 Wendy Knerr, Frances Gardner, and against Women in Rural Bangladesh,” Lucie Cluver, “Parenting and the Preven- Social Science & Medicine 43, no. 12 (1996): tion of Child Maltreatment in Low- and 1729–42. Middle-Income Countries: A Systematic Paul M. Pronyk, James R Hargreaves, Julia C. 128 Review of Interventions and a Discussion Kim, Linda A Morison, Godfrey Phetla, Char- of Prevention of the Risks of Future Violent lotte Watts, Joanna Busza, and John D. H. Behaviour among Boys,” Sexual Violence Porter, “Effect of a Structural Intervention Research Initiative, Pretoria, South Africa, for the Prevention of Intimate-Partner Vio- 2011; Christopher Mikton and Alexander lence and HIV in Rural South Africa: A Clus- Butchart, “Child Maltreatment Prevention: ter Randomised Trial,” Lancet 368, no. 9551 A Systematic Review of Reviews,” Bulletin of Chapter 3 Freedom from violence | 91 the World Health Organization, 87 (2009): Shaw, and Kalpana Viswanath (eds.), Build- 353–61. ing Inclusive Cities: Women’s Safety and the Harounan Kazianga, Dan Levy, Leigh L. 134 Right to the City (London: Routledge, 2013). Linden, and Matt Sloan, “The Effect of Nandita Bhatla, Pranita Achyut, Sancheeta 136 ‘Girl-Friendly’ Schools: Evidence from the Ghosh, Abhishek Gautam, and Ravi Verma, BRIGHT School Construction Program in “Safe Cities Free from Violence against Burkina Faso,” American Economic Journal: Women and Girls: Baseline Finding from the Applied Economics 3, no. 5 (2013): 41–62. ‘Safe Cities Delhi Programme’,” UN Women, Carolyn Whitzman, Crystal Legacy, Caro- 135 New Delhi, 2012. line Andrew, Fran Klodawsky, Margaret 92 | Voice and Agency empowering women and girls for shared prosperity Chapter 4 Control over Sexual and Reproductive Health and Rights | 93 Chapter 4 Control over Sexual and Reproductive Health and Rights Key messages Box 4.1 > Sexual and reproductive health and rights are important ends in themselves and can have valuable benefits for women’s own health, nutrition, education, and livelihoods and for the well-being of their children. > Having agency over sexual and reproductive decisions includes being able to choose whether, when and with whom to have sex, to ask a partner to use a condom, and to make decisions about childbearing and ones own health. > Child marriage, high rates of adolescent pregnancy, and women’s limited control over their own sexual and reproductive health decisions often reflect underlying gender inequalities, and carry serious repercussions. > Beyond access to contraceptive methods, promising interventions integrate multi-sectoral actions, such as sexual and reproductive health education, mentoring and peer group training, incentive programs, and activity clubs and sports. > Broader structural and normative changes are critical, which can be achieved by engaging community opinion leaders and gatekeepers to support changing norms. Chapter 4 Control over Sexual and Reproductive Health and Rights | 95 The nature of the challenge financial decisions, as well as have a height- ened risk of violence,3 of contracting HIV/ The World Development Report 2012 placed AIDS or other sexually transmitted infec- being able to decide whether, when and tions (STIs),4 and of early pregnancy. And as with whom to have sex; whether, when, and we saw in chapter 1, the girls most likely to whom to marry; whether or when to have marry early are those with the least educa- children; and how many children to have as tion and lowest economic status.5 expressions of agency. The intrinsic value of agency related to family formation has long Large shares of women and girls are been recognized by governments around the constrained in exercising agency over these world. The Universal Declaration of Human domains. What we see globally, and investi- Rights (1948) asserts that “Marriage shall gate further below, are high levels of unmet be entered into only with the free and full need for contraceptives; high numbers of consent of the intending spouses,” and limits unintended pregnancies; large gaps between marriage to those “of full age.” The right to expressed ideal family size and actual fertil- sexual and reproductive health was recog- ity; and inability to divorce due to financial nized in 1994 by the Programme of Action of dependency or legal restrictions. The stark- the International Conference on Population est violation of girls’ and women’s agency and Development (ICPD), and subsequently addressed in this chapter is early marriage adopted by 179 governments.1 These posi- (coerced sex was discussed in chapter 3). tions stand alongside CEDAW, which was The many other manifestations of women’s discussed in chapter 1. lack of sexual and reproductive agency, such as high levels of maternal mortality and Enabling women’s control over decisions morbidity, HIV/AIDS, and the challenges of about family formation also has instru- accessing health services have recently been mental value, as these decisions can affect explored in greater detail elsewhere6 and investments in their education and eco- are beyond the focus of this chapter. nomic opportunities as well as their health. Delaying marriage is associated with greater educational achievement and lower fertility, and lower fertility can increase women’s “The man brings the woman in his life expectancy and benefit their children’s house [when he marries her]. How could health and education. Conversely, barriers to she be his equal?” girls’ and women’s sexual and reproductive —Men’s focus group, Gliangou, Téra Department, health have adverse repercussions—from Tillabéri Region, Niger7 the individual and family to the community, society, and economy—and on a broad set of development outcomes.2 Moreover, these impacts are often amplified by overlapping As outlined in chapter 1 and explored fur- deprivations. For example, girls who marry ther in this chapter, social norms and laws before their 18th birthday are more likely are important drivers of agency, and this is to face restrictions on their mobility and especially true in the context of sexual and 96 | Voice and Agency empowering women and girls for shared prosperity reproductive decisions (see figure 1.1 where It is also associated with child marriage and social norms and laws, mediated through the early pregnancy, and the ability to negoti- household, impact agency and other gen- ate within a sexual relationship, such as the der-equality outcomes). Interventions and ability to refuse sex with a partner or to ask policies that seek to change norms around a partner to use a condom (figure 4.1). Evi- acceptability of child marriage and modern dence suggests that economic and political contraceptives, for example, will be critical opportunities also play a role. Where women for increasing women’s sexual and repro- are able to support themselves and to be ductive agency. While progressive laws can active members of the community outside of foster women’s greater control over family the home, they have more decision-making formation decisions, in many countries laws power within the relationship itself.11 reinforce constraints, as in the 29 countries The objective of this chapter is three- where men are the legally designated heads fold. First, we highlight how women’s abil- of households. Indeed, 29 of the 188 countries ity to control their reproductive health and that have ratified CEDAW retain exceptions make decisions about family formation to the article requiring countries to eliminate enhances their agency in other aspects of discrimination in all matters related to mar- life, including health, education, and access riage and family relations. Gendered social norms surrounding sexuality are learnt by girls and boys from their families and communities and affect their choices in adult life. In many cultures, it is considered inappropriate for women to learn “too much” about matters related FIGURE 4.1 Education as a key to sexuality. For girls and women, though driver of women’s ability to exercise choice in relationships typically not boys and men, there is stigma around being sexually active.8 Stigma and norms around purity can prevent girls from 100% 100 95 Years of education learning about sex and contraception, giving 80 90 men a prominent role in sex and contracep- 85 60 80 tive choices.9 In Honduras, for example, in 75 40 households where the husband makes the 70 65 20 decisions regarding family size and family 60 planning use alone, women are less likely to 0 55 report that they currently or have ever used 0 5 10 15 20 Years of education modern methods of family planning.10 Ask partner to use condom Which factors affect women’s agency in Refuse sex sexual and reproductive decisions? We saw Source: Voice and Agency 2014 team estimates in chapter 3 that women’s lack of education based on Demographic and Health Surveys for was associated with higher risks of violence. 37 countries, 2006-2012.12 Chapter 4 Control over Sexual and Reproductive Health and Rights | 97 to economic opportunities. Second, the indi- vidual characteristics that affect women’s “Each year, almost one in five women in and girls’ ability to exercise agency in those developing countries becomes pregnant domains are identified, and the impacts of before she turns 18, and 7.3 million girls overlapping disadvantage explored. Third, under age 18 give birth.” evidence about what works in promot- ing women’s exercise of their sexual and reproductive health and rights is reviewed. We build on several recent contributions, six or seven children in 1960, whereas in including those by the International Con- 2009, it was closer to two. In 2005, 78 per- ference on Population and Development cent of Colombian women used contracep- beyond 2014,13 the United Nations Popu- tion, compared to only 21 percent in 1970.17 lation Fund,14 the Council on Foreign Rela- tions,15 Women Deliver, and the Interna- Lower fertility has instrumental value, tional Center for Research on Women.16 increasing women’s life expectancy and enabling them to pursue economic oppor- The next section reviews the magnitude of the challenge and the repercussions for tunities.18 Voluntary use of contraception women’s health and well-being. We then reduces the number of high-risk births highlight lessons learned and promising and lowers maternal mortality rates.19 The approaches. The final section summarizes beneficial impact of smaller family size the key findings and highlights important on children’s health and education is also directions in the way forward. well documented.20 Recent research also suggests that when fertility declines and How great is the challenge? the value of human capital in the economy increases, men are more willing to share This section presents new analyses of wom- household decision-making power with en’s ability to make choices about their women to ensure their children get better sexuality and their sexual and reproduc- education.21 tive health across 55 developing countries, which represent 40 percent of the world’s Yet many women are unable to realize population. their fertility preferences.22 Annually, in developing countries alone, about 80 million Fertility choices women have unintended pregnancies, which In many countries today, women have means that they had wanted to delay that far fewer children than four decades ago. pregnancy or not have it at all.23 Addressing Although many factors have had a role in global unmet contraceptive need—defined this transformation, one of the main mech- as nonuse of contraception by women who anisms through which this change has come do not want to get pregnant for at least two about is women’s use of contraception. For years—could avert 54 million unplanned example, in Colombia, average fertility was pregnancies annually.24 98 | Voice and Agency empowering women and girls for shared prosperity Unmet need may result from inadequate have difficulty communicating with their access to supplies, due to factors such as dis- husbands, or do so only infrequently, are less tance, cost, insufficient supplies, and incon- likely to use contraception or more likely to venient hours. However, beyond ensur- conceal its use.29 In some cases, male part- ing adequate supply, increasing women’s ners restrict contraceptive use altogether.30 agency, education, and access to economic Communication with other family members opportunities all can be important drivers can be important too, for example, women of contraceptive use. A study in South Africa in urban slums in Karachi are more likely found that a woman’s level of education and to use family planning following discussion her control over her earnings are enabling with their mother-in-law.31 factors for the use of modern contraceptive To what extent are women involved in methods.25 In Sub-Saharan Africa women decisions about their fertility? Our analysis in wealthier households and who are more shows that across 55 developing countries, educated are far more likely to use contra- two-thirds of couples make joint decisions ception than their less advantaged peers. regarding contraception, whereas in 15 Only 10 percent of women with no educa- percent of cases, women make this decision tion and 10 percent of those in the poorest themselves. Only a minority of women— households use contraception, compared about 1 in 13—report that their husband to 42 percent of women with secondary or makes this decision alone. However, the higher education and 38 percent of women regional averages vary: up to 1 in 10 women from the wealthiest households.26 Across in South Asia and more than 1 in 5 (21 per- four African countries (Ghana, Namibia, cent) women in Sub-Saharan Africa say Uganda, and Zambia), a woman’s increased their husband makes the decision alone. In participation in household economic deci- Liberia, nearly one-quarter of women say sion-making, her ability to negotiate sexual their husbands make this decision alone. On activity, and a couple’s agreement on fer- average, across 55 countries, 6 percent of tility preferences were all associated with women report that they are not using family increased use of contraceptives that were planning because their husband or others either female-only methods (such as pills, are opposed, with Timor Leste having the intrauterine devices [IUDs], injectables, or highest report of non-use, at 22 percent. implants) or methods that require a certain degree of cooperation from partners (such These patterns help explain why women as male and female condoms).27 who want to delay or prevent pregnancy might choose to use more concealable meth- Open communication and joint deci- ods such as injectables and IUDs. Women’s sion-making about family planning contrib- ute to voluntary contraceptive use and the achievement of fertility intentions.28 Studies “We do not use contraception. Otherwise have shown that communication between people will think we’re sterile.” partners is important for realizing women’s choices about family formation and their —Women’s focus group, Morey, Keita Department, Tahoua Region, Niger32 sexual and reproductive health. Women who Chapter 4 Control over Sexual and Reproductive Health and Rights | 99 covert use is generally motivated by a desire lacking the necessary skills or in an environ- to keep partners, in-laws, or other family ment that lacks minimal medical standards.40 members from finding out that they want These situations account for close to 13 per- to avoid a pregnancy.33 And women who cent of all maternal deaths. Young women are in violent relationships are more likely face particular challenges in accessing abor- to report that their partners have tried to tion care and account for approximately sabotage their use of contraception and that 40 percent, or as many as 3.2 million,41 they have used a method covertly.34 In some of unsafe abortions worldwide.42 contexts, enabling women’s full control over Abortions are illegal or highly restricted family planning shows greater results; for in some 53 countries.43 Such restrictions example, unwanted births in Zambia were have been directly linked to adverse health reduced only when women had individ- consequences.44 Procedural requirements ual control over contraceptive decision- can undermine access to services even making.35 However, enabling women to where they are legal. Other barriers to access and use contraceptives in secret is accessing safe abortion care include social an inadequate and temporary solution, and and religious stigma and travel and dis- one that carries risks. In some cases, women tance-related constraints. fear—or actually face—violence from their partner if the covert use of contraceptives is Early sexual activity and discovered.36 pregnancy It is not just husbands and partners who Early sexual initiation puts girls at greater limit women’s agency over their fertility risk for HIV and other STIs, early pregnancy, decisions. Often family members and com- and early childbearing.45 Every hour, 50 munity leaders exert influence over these young women are newly infected with HIV.46 choices too, as in the following examples: In 2012, more than five million young peo- ple (ages 15-24) were living with HIV, with ■■ In Jordan, women, young brides in par- an estimated 2,400 new infections each ticular, face significant pressure from day.47 In Sub-Saharan Africa, HIV prevalence their mother-in-law to prove their fertil- among young people decreased by almost ity and bear children.37 half between 2001 and 2012, but prevalence ■■ In Tanzania, family planning decisions among young women remains more than are influenced by religious leaders and double that among young men.48 by male dominance over females (espe- Each year, almost one in five women in cially in polygynous relationships).38 developing countries becomes pregnant Satisfying the unmet need for contracep- before the age of 18, and 7.3 million girls tives and improving sexual and reproductive under the age of 18 (2 million of whom are health services would reduce the number of under age 15) give birth.49 This is dangerous. abortions. This is important, because almost In developing countries, pregnancy-related half of all abortions are unsafe,39 mean- causes are the largest contributor to the ing that they are carried out by individuals mortality of girls ages 15–19, killing nearly 100 | Voice and Agency empowering women and girls for shared prosperity 70,000 girls each year.50 Early childbearing pregnancy intention than their counter- also increases the risk of childbirth com- parts in Asia and in Latin America and the plications such as obstetric fistula, which Caribbean. commonly occurs among young girls who Following from our earlier discussion give birth before their bodies are physically about overlapping disadvantage, poor young mature and causes chronic incontinence, women have fewer choices and opportuni- often resulting in, among other things, social ties.58 In several Latin American countries— exclusion. including, Bolivia, Colombia, Dominican Globally, over the past 15 years, adoles- Republic, Haiti, Honduras, and Peru—ado- cent fertility rates have dipped slightly, albeit lescents who have more education, live in with large variation across countries and urban areas, and come from wealthier fami- regions. The average rate in developing coun- lies are less likely to get pregnant.59 Mexican tries is 10 percent, compared to 2 percent in women who gave birth during adolescence developed nations.51 Sub-Saharan Africa has were found to be more likely to depend on the highest regional rate—about 12 percent social assistance. This study also highlights in 2010—and the rate is rising in Burundi, the potential long-term consequences for Chad, The Republic of Congo, Lesotho, and children born to young mothers, including Zimbabwe.52 Following current trends, the low cognitive test scores, poor behavioral region will take nearly 50 years to reach outcomes, grade repetition, and economic Europe’s levels of adolescent fertility. disadvantage.60 Overall, adolescent fertility in Latin Amer- At the same time, adolescents who are ica and the Caribbean (72 births per 1,000 sexually active and wish to avoid preg- women 15–19 years of age) is lower than nancy face a number of barriers to exercis- in Africa.53 The annual reduction in births ing agency over these decisions, including between 1997 and 2010 was 2.7 percent in limited mobility and financial resources, South Asia and 1.6 percent globally, com- lack of accurate information, stigma related pared with just 1.3 percent in Latin America to sexual activity before marriage,61 and and the Caribbean.54 The countries with the restrictive policies. Whether married or highest teenage pregnancy rates (Dominican unmarried, young women and girls around Republic, Guatemala, Honduras, and Nicara- the world have greater difficulty than older gua) reported no change since 2000.55 women in accessing comprehensive contra- ceptive services and care.62 The evidence on Not surprisingly, early marriage is this point is striking: strongly correlated with early childbear- ing.56 The highest rates of adolescent births ■■ Across 22 Sub-Saharan African countries take place in West and Central Africa and for which data are available, adoles- South Asia, where early marriage rates are cent girls have the lowest contraceptive high.57 Figure 4.2 shows the regional share prevalence rate; roughly only 1 in 10 ado- of married adolescents who are, or who lescents who are married or in a union use wish to become, pregnant. Married girls in any form of family planning, compared to Sub-Saharan Africa show higher levels of roughly 1 in 4 women ages 30–44.63 Chapter 4 Control over Sexual and Reproductive Health and Rights | 101 FIGURE 4.2 Pregnancy desires of married adolescent girls 100 90 Share of married adolescent Want to avoid pregnancy, 80 using no method 70 Want to avoid pregnancy, 60 using a traditional girls (%) 50 method 40 Want to avoid pregnancy, 30 using a modern method 20 Want pregnancy or are 10 intentionally pregnant 0 Sub-Saharan South Central Latin America Africa and Southeat Asia and the Caribbean Source: Guttmacher Institute and International Planned Parenthood Federation, “Facts on the Sexual and Repro- ductive Health of Adolescent Women in the Developing World, Guttmacher Institute, New York, 2010. Note: Girls ages 15–19 years. ■■ Young girls are also the least likely group In Indonesia and Malaysia, only married to have reliable information on their women can access comprehensive sexual rights and contraceptive options, they and reproductive services, and in Cambodia are the least likely group to use health parental consent is required for HIV testing facilities, and they generally have an of minors under age 18.67 incomplete understanding of how their Where girls have greater educational reproductive system works.64 and economic opportunities, evidence sug- ■■ Unmarried girls are frequently denied gests that they are more likely to take up care, because providers require parental those opportunities than to have children in consent.65 their teenage years. In Chile, extending the school day reduced teenage motherhood; Research from the East Asia and Pacific an increase in full-day school enrollment of region suggests that restrictive laws—such 20 percent reduced the likelihood of teen as requiring parental consent for HIV or motherhood by 5 percent.68 Beyond the STI testing or permitting parent access formal education system, life skills courses to a minor’s medical records without the also have dramatic impacts; for example, in child’s consent—are a significant barrier to Marathwada, India, attendance in life skills sexual and reproductive health services.66 classes increased the median age of marriage 102 | Voice and Agency empowering women and girls for shared prosperity for girls in the area from 16 to 17 in just two population means that India has the high- years.69 Elsewhere in rural India, a three- est number, accounting for one-third of the year program which provided recruiting world’s child brides. services to women resulted in more women Why is child marriage so common, often choosing to enter the labor market or to despite legal prohibitions? Social norms and obtain more schooling or post school train- expectations are again critical factors. Fam- ing instead of getting married and starting ilies may perceive marriage as a way to pro- families. When given the choice, women vide for their daughter’s future, believing opted for greater economic opportunities.70 it will improve their child’s economic and social circumstances. The reality, however, is Child marriage that girls who marry young are more likely Child marriage in developing countries to remain poor even after marriage.75 remains pervasive, with one-third of girls There is often a large age difference being married before age 18 and 1 in 9 being between child brides and their hus- married before age 15.71 If present trends bands.76 Such age gaps can disempower continue, more than 142 million girls will the wife.77 Research from a number of be married before the age of 18 in the next countries—including Cambodia, Colom- decade, that is, 39,000 girls each day.72 This bia, Haiti, India, Kenya, Peru, South Africa, practice is driven by poverty, social norms, Turkmenistan, and Zambia—suggests and pervasive discrimination against girls. that having a much older husband can CEDAW calls on all countries to take neces- dramatically increase the risk of intimate sary action to eliminate child marriage and partner violence.78 Similarly, in Peru79 encourages lawmakers to set the minimum and India,80 girls who married before the age of marriage at 18 years.73 Early marriage age of 18 were twice as likely to experience forces girls into adulthood and, frequently, violence compared with women who mar- motherhood before they are emotionally or ried later. physically mature, and often before they can complete their education. Growing evidence from Sub-Saharan Africa shows that girls who marry early are Average regional prevalence of child mar- also at greater risk of contracting HIV or riage ranges from 21 percent in Africa and other STIs. Marriage before age 20 is consid- 17 percent in Latin America and the Carib- ered a risk factor for HIV infection; for exam- bean, to 8 percent in Eastern and Southern ple, in Kenya and Zambia, HIV infection rates Europe (figure 4.3).74 These averages mask were found to be higher among married girls considerable variation across countries. than among their unmarried, sexually active Across the 111 countries with data, the counterparts.81 Similarly, in Uganda the HIV prevalence of child marriage ranges from prevalence rate for girls 15–19 years of age 2 percent in Algeria and Libya to as high as was higher for married girls (89 percent) 74 and 75 percent in Niger and Bangladesh. than for unmarried girls (66 percent),82 and While the prevalence in India is not among the age difference between spouses was the highest recorded, the sheer size of its a significant risk factor. Child brides are Chapter 4 Control over Sexual and Reproductive Health and Rights | 103 FIGURE 4.3 Child marriage prevalence in 111 countries Child marriage (Women 20-24) 1-10% 11-20% 21-30% 31-35% 36-40% 41-50% 51-75% Sources: Voice and Agency 2014 team estimates based on Demographic and Health Surveys, International Center for Research on Women, and United Nations Children’s Fund using latest available data from 2001-2012. Note: Countries in gray do not have comparable data available. often unable to ask their husband to get an control and experiences of intimate partner HIV test, to abstain from intercourse, or to violence. demand that he use a condom.83 Our analysis from 55 countries show that Using Demographic and Health Survey girls living in poor households are almost data for 37 developing countries, we find twice as likely to marry before the age of 18 that women who marry after the age of 18 compared with girls in higher-income house- are 2 percent more likely to feel able to refuse holds, as are rural girls compared with those sex than women who marry early. This small from urban areas. A girl may be married off but significant difference increases as age to pay debts, or given away by her family in at marriage increases. Each year marriage exchange for a wife for her brother.84 Dowries is delayed beyond age 18 is associated with are another driving force: in some cases, the higher odds of a woman being able to exer- younger the bride, the cheaper the dowry cise agency to refuse sex. These results illus- her family has to pay. 85 Other local traditions trate the important link between early mar- may be at play; for example, in Bangladesh, riage, adolescents’ sexual and reproductive the Kyrgyz Republic, and Somalia, there is 104 | Voice and Agency empowering women and girls for shared prosperity still a practice of forcing girls who have been Whether divorce is socially acceptable abducted and raped to marry the perpetrator and financially viable can also affect wom- to protect their families’ honor.86 en’s agency. Overall, divorce is still not widely socially acceptable, and in some regions it Beyond poverty, we see other dimen- is even becoming less so, as in Latin Amer- sions of overlapping disadvantage. Child ica and the Caribbean and in South Asia.91 marriage is closely associated with illiteracy, Although divorce is currently illegal only in and more schooling seems to raise the age of the Philippines and Vatican City, in 13 out of marriage: the 70 countries for which data are available, ■■ Evidence from Bangladesh and Sub-Sa- women do not have equal rights to instigate haran Africa suggests that women who a divorce. These countries include some mid- married early are over 5 percentage dle-income and high-income countries, such points less likely to be literate and over 8 as Israel, Jordan, and Malaysia.92 Beyond percentage points less likely to have any what the law says on paper, recent qualita- secondary education.87 tive research across 20 countries found that ■■ In Zimbabwe, only 4 percent of girls aged the unequal division of assets and custody of 15–19 who had attended primary school children are significant obstacles to divorce, were married, compared to 40 percent of alongside opposition by families and commu- those who had received no education; in nities, social isolation, and stigma.93 Haiti the shares were 15 percent and 43 The foregoing has highlighted the nature percent, respectively.88 of constraints on sexual and reproductive ■■ A study across 18 of the 20 countries rights and hinted at some important emerg- with the highest prevalence of child mar- ing opportunities for action. The follow- riage found that girls with no education ing section discusses the evidence on what were up to six times more likely to marry works and identifies key priorities for mov- as children than girls who had received ing forward. secondary education.89 ■■ Across Africa, each additional year a girl is married before age 18 has been found to reduce her probability of literacy by “A man can easily divorce his wife about 6 percentage points, the proba- because his decision is final. [On the bility of having at least some secondary other hand] it is difficult for a woman to schooling by 8 percentage points, and obtain a divorce because people will try to reconcile the couple and, moreover, the probability of secondary school com- if she wants a divorce, she must reim- pletion by almost 7 percentage points.90 burse the bride-wealth even if she had These patterns—of cultural norms, edu- ten children.” cation, and links to poverty—begin to point —Women’s focus group, Toulou Karey, Loga Department, to the layered challenges that policy makers Dosso Region, Niger need to address. Chapter 4 Control over Sexual and Reproductive Health and Rights | 105 Program and policy sexual and reproductive health and contra- evidence: What works? ceptive choices show promise, although they need to be carefully designed to avoid unin- Several types of interventions have been tended consequences.95 In some cases, such shown to expand women’s and girls’ con- interventions have been found to reinforce trol over their sexual and reproductive traditional power dynamics or to produce decisions. These include interventions that only modest behavioral change.96 In Ghana, promote more gender-equitable communi- for example, men who were educated about cation and decision-making around sexual emergency contraception preferred this to and reproductive health and that improve longer-term methods and saw themselves women’s access to and the quality of repro- as the primary decision makers on the issue, ductive health services and information. often without regard for their partners’ pref- Interventions have also tested incentives for erences.97 But evidence of positive impacts preventing child marriage and for support- is also emerging. Mass media campaigns are ing girls and their families. Interventions to associated with higher approval of contra- expand education and economic opportu- ception and better communication between nities offer promise when the need for safe partners about use.98 The mass media cam- spaces, life skills, and job skills are consid- paigns that were evaluated as being most ered. Women’s agency can also be supported effective went beyond providing informa- through more progressive laws, including tion and encouraged boys and men to talk those related to marriage and property. about specific topics, such as gender-based violence.99 As discussed in chapter 2, a common factor across many successful approaches When the Male Motivator project sought is an acknowledgment of the powerful role to address men’s limited engagement in fer- of gender norms. Approaches for effectively tility issues in Malawi, more than one-third of engaging men, boys, communities, and tra- married women reported never having spo- ditional authorities to change norms around ken to their husbands about family planning. sexuality, marriage, and reproduction have Male outreach workers engaged men who been shown to contribute to the strengthen- were not using contraception in discussions ing of women’s agency. 94 about gender roles and norms, especially around the norm of having large families to Engaging men and other demonstrate virility. As a result, contracep- gatekeepers tive use increased significantly, which was attributed to greater ease and frequency of Paying attention to male attitudes and communication between partners, resulting gender dynamics when designing family in more joint decision-making.100 planning interventions is important because responsibility shared between women and Interventions that engage other gate- men ultimately promotes greater agency for keepers of women’s health—older male women in sexual and reproductive health and female family members and commu- decisions. Initiatives that engage men in nity leaders—can improve communication, 106 | Voice and Agency empowering women and girls for shared prosperity especially between young people, couples, However, these may also have the unin- and other family members.101 As highlighted tended consequence of reinforcing gender in box 4.2, working with customary leaders inequalities. For example, in Bangladesh, such as village chiefs can enhance adoles- community-based distribution was found cents’ exercise of agency in their sexual and to reinforce women’s seclusion by limiting reproductive health. interactions in the public space.105 Ensuring universal access to compre- Information and services: hensive adolescent-friendly services can Improving access and quality reduce the likelihood of early pregnancies. Improving the quality of services and Evidence from around the world shows that availability of contraceptive methods can the friendliness of staff, quality of care, cost increase agency. Weak services often have of services, and ease of getting to a facility low usage rates. A recent Egyptian study are important determinants of adolescents’ suggests that certain dimensions of quality access to and use of sexual and reproduc- are particularly important, including privacy tive health services.106 The 47th session of during medical examinations and respect the Commission on Population and Devel- for patient confidentiality, as well as facil- opment in 2014 upheld the responsibility ity cleanliness, friendly staff, shorter wait of governments to safeguard adolescents’ times, and limited staff turnover.103 Com- rights to privacy, confidentiality, respect, and munity-based distribution approaches can informed consent and to address the legal, circumvent women’s mobility constraints.104 regulatory, and social barriers adolescents Box 4.2 Engaging traditional “gatekeepers” in adolescent sexual and repro- ductive health in Malawi Chapsinja, the rural area outside of Lilongwe, is known for customary practices such as initiation ceremonies, during which information is delivered to young people by traditional counselors or akunjira. A study by the Family Planning Association of Malawi (FPAM) documented that traditional counselors were strongly against condom use and that some of the messages they conveyed during the ceremonies likely contributed to the high prevalence of HIV/AIDS, sexually transmitted infections, adolescent pregnancies, and coerced sexual initiation rates for girls. FPAM launched a community mobilization campaign to introduce sexual and reproductive health services and information. They worked with akunjira to transform existing rituals into opportunities for health-promoting messages. Girls and boys were encouraged to go to school and become sexually active only when they are socially, mentally, and physically mature. At the same time, young people were trained as peer educators and volunteer condom distributors in their communities. HIV testing has become more frequent, and according to preliminary results, HIV prevalence is declining. The Chapsinja senior chief now plans to serve as a bridge to chiefs in surrounding areas. Source: Boehmova, 2013.102 Chapter 4 Control over Sexual and Reproductive Health and Rights | 107 face in receiving high-quality reproductive declined by 49 percent in intervention facil- health information and care.107 ities, compared to 25 percent in control facilities.109 Methods for measuring the acceptability of services can be incorporated into perfor- Results-based financing (RBF)—whereby mance monitoring. Such measures can include health service providers are paid accord- client ratings of overall service satisfaction ing to their performance along a previously and such indicators as perceived respect for agreed upon set of outcomes—has been used dignity, provision of satisfactory information to expand women’s options for reproductive and explanation, informed choice and consent, health services and to improve sexual and and availability of preferred contraceptive reproductive health outcomes. In Rwanda, methods. These monitoring indicators link to incentives have been introduced for commu- mechanisms of social accountability, as high- nity health workers based on their perfor- lighted in box 4.3. mance of certain functions, like referrals of Client awareness is important. In Peru, women for antenatal care, institutional deliv- the FEMME Project aimed to standardize ery, referrals of new users of family planning, care with new emergency obstetric guide- and the number of current users.110 Prelimi- lines. Among other things, these guidelines nary results from several countries suggest stressed women’s rights as patients and that efficiency, equity, and accountability can their rights to privacy during care. As a result be strengthened through the results-based of the program, the maternal mortality rate approach, as in the following examples: Box 4.3 Social accountability and service delivery Social accountability mechanisms that help improve the quality of service delivery can improve women’s ability to exercise agency over their own health and their health outcomes, as in the following examples: ›› In Uganda a year after the introduction of a local accountability mechanism which gave clients greater control on the delivery of primary health services, including through provision of information about staff performance, child survival rates improved leading to a 33 percent reduction in under- five mortality. ›› In Peru, a citizen surveillance program increased the number of births in health facilities by almost one-third in just one year and increased access to culturally appropriate delivery options. ›› Recent work from Orissa, India, identified three effective processes that had positive impacts on reproductive and maternal health: generating demand for rights and better services through information campaigns, leveraging intermediaries to legitimize the demands of poor and marginalized women, and sensitizing leaders and health providers to the needs of women. Sources: Grépin and Klugman, 2013; Bjorkman and Svensson, 2009; World Bank, 2011; Papp et al., 2012; and Díaz et al., 2001.108 108 | Voice and Agency empowering women and girls for shared prosperity ■■ In Afghanistan, the number of women Innovations in ICTs show promise in delivering their babies with the support improving quality and access, including of skilled birth attendants more than through enhanced social accountability, doubled from April 2010 to December although more rigorous evaluations are 2012 in RBF-supported facilities.111 needed in developing country settings.116 ■■ In Cameroon, the number of births The following suggest promising directions: attended by skilled professionals more ■■ Text messages can be a feasible method than doubled between 2012 and 2013 in to send information and service refer- the regions involved in an RBF project.112 rals, especially for adolescents.117 ■■ In Nigeria, coverage of institutional deliv- ■■ Reminders by text message have also eries increased by 30 percentage points been used to improve adherence to (to 39 percent), and coverage of antena- antiretroviral treatment and to increase tal care increased from 16 percent to 77 delivery of births by skilled attendants.118 percent in 2012, following the introduc- tion of an RBF approach.113 ■■ Similarly, mobile solutions can improve patient awareness. For example, project More accessible family planning methods Masiluleke’s text message campaign in and information also made a difference in South Africa, which promotes HIV/AIDS Ukraine, where removing barriers to provid- awareness, nearly tripled the volume of ing sexual and reproductive health services calls to a local HIV/AIDS helpline.119 led to a two-thirds decline in abortions since the early 1990s. This decline was attributed New partnerships using ICTs are also to policies and programs that prioritize young emerging. In Nigeria, the World Bank is sup- people’s sexual and reproductive health, for porting an initiative using an SMS message example, through better training of staff and platform, both to elicit citizen feedback on the addition of specialized health facilities, health services and to provide important and through public-private partnerships maternal health information from the Mobile which increased the range of available con- Alliance for Maternal Action.120 In Tanzania, traceptive methods and reduced their cost.114 Vodafone and the Comprehensive Community Based Rehabilitation Hospital in Dar es Salaam Vouchers can cut the costs of accessing have partnered to ensure that poor women can services, which overcomes a critical barrier receive corrective surgery for obstetric fistula. to exercising agency, particularly for young Through this program, women from the poor- people. In Nicaragua, vouchers valid for three est communities receive the funds for trans- months can be used for one free consultation port costs to the hospital through M-PESA, a and a free follow-up visit for services such as mobile-phone-based money transfer.121 counseling on family planning and antenatal care. Vouchers are distributed to adolescents Media can also be used to promote and youth by nongovernmental organiza- changes in perceptions about ideal family tions in markets, outside schools, in clinics, size and contraceptive use. In Brazil, soap on the streets, and door-to-door, and they can operas have successfully promoted lower be transferred to another adolescent.115 fertility by portraying smaller families as a Chapter 4 Control over Sexual and Reproductive Health and Rights | 109 positive norm. Over a 20-year period, fer- that improving access to quality schooling tility rates were significantly lower in com- and retention of girls in school are, among munities exposed to the show than in unex- other things, important routes to prevent- posed communities; the effect was greatest ing child marriage.127 Rising education lev- among women in lower socioeconomic els have been a critical factor in increasing groups and women in the middle and end of the age of marriage in a number of devel- their fertile years.122 In Egypt, community oping countries, including Indonesia, Sri norms—most notably desired family size Lanka, Taiwan, and Thailand.128 and the community level of family planning Given the links to poverty noted above, use—are associated with changes in indi- it is no surprise that financial incentives vidual family planning use over time.123 This can help combat child marriage. Payments link is also reflected in fertility rates falling to girls or their families are sometimes following several national media campaigns tied to investment in daughters’ educa- to promote child spacing and limited fam- tion and to the condition that they do not ily size. As a result, the average number of marry before age 18. In 1994, the Hary- births per woman declined from 5.3 in 1980 ana Government in India launched Apni to 3.1 in 2005.124 Beti Apna Dhan (ABAD, or “Our Daughter, Our Wealth”), which gave poor families an Alternatives to early marriage incentive to keep their daughters in school Programmatic efforts to combat child mar- and unmarried until the age of 18, and riage have gained momentum over the past sought to increase family and community two decades, with some encouraging results, perceptions about the value of girls. An albeit many still on a small scale. Most eval- impact evaluation shows that girls enrolled uated programs (18 out of 23 in a recent in ABAD were significantly more likely to systematic review)125 focus on girls them- be in school and unmarried at age 18 than selves, through life skills training, vocational girls in the control group, although the atti- and livelihood skills training, sexual and tudinal changes by parents about the value reproductive health training, communica- of girls are less clear.129 tion campaigns, mentoring and peer group training, and “safe space” club models. Sev- Other evidence suggests that both con- eral programs have changed knowledge, ditional and unconditional transfers can attitudes, and behavior in just a few years, play an important role: although it is not yet clear which specific ■■ In Brazil, Colombia, and Peru, program components are most important.126 experience suggests that conditional Programs are increasingly addressing cash transfers (CCTs) can be an effec- poverty, education, and lack of economic tive instrument to delay marriage and opportunities as drivers of child marriage childbearing,130 although the underlying (see box 4.4). We have seen clear evidence mechanisms are not fully clear.131 that attending school reduces the likeli- ■■ Evidence from the Punjab Female School hood of child marriage. It follows, then, Stipend Program in Pakistan suggests 110 | Voice and Agency empowering women and girls for shared prosperity Box 4.4 Promising approach to delaying marriage in Ethiopia Ethiopia has high rates of child marriage. In Awi zone, in rural Amhara Region, one in five girls is married before her 15th birthday, and nearly half (44 percent) are married by age 17. Nearly all (92 percent) of these marriages are arranged by their families. In 2004, the regional government partnered with the Population Council on a two-year project to delay marriage and keep girls in school. This program, Berhane Hewan (“Light for Eve”) provided families with cash conditioned on their daughters remaining unmarried and in school for the duration of the program. Other elements included social mobilization of girls led by female mentors, provision of school supplies, livelihood skills for out-of-school girls, and “community conversations” on early marriage and reproductive health. Berhane Hewan was among the first projects targeting child marriage to be rigorously evaluated, and the results are promising. Girls (ages 10–14) enrolled in the program were only one-tenth as likely to be married compared to girls in the control site. Participants were also three times more likely to be in school. Married girls in the project site were three times more likely to be using contraception compared to married girls in the control site. Source: Erulkar and Muthengi, 2009.132 that CCT effects on early childbearing reported increased self-confidence, with were long lasting.133 65 percent feeling “strong and able to face ■■ An evaluation of a cash transfer program any problem.”135 in Malawi found that teenage pregnancy Successfully addressing child marriage and marriage rates were substantially requires locally developed solutions that lower in the unconditional cash trans- are tailored to the context and reflect an fer arm than in the conditional arm, understanding of the social and religious because additional income enabled out- motivations for child marriage. Several pro- of-school girls to delay marriage.134 grams successfully engage community and Policies can also enable adolescents religious leaders in efforts to chip away at to make informed fertility choices and the norms that perpetuate the practice. The broaden their options. In a program in programs engage not only teachers but also rural Egypt, girl-friendly spaces com- community leaders. Others use conditional bined literacy and life-skills training with cash transfers to improve public transpor- sports and changed girls’ perceptions tation to go to schools or improve the qual- about early marriage. Evaluation results ity of schooling.136 An important example show an increase in literacy and school is the Tostan program in Senegal, which enrollment: nearly 70 percent of program mobilized communities to change norms participants entered or reentered school. and expectations around female genital After the project, participants expressed mutilation/cutting, child marriage, and the their intention to marry later, and girls value of the girl child (see box.2.1). Chapter 4 Control over Sexual and Reproductive Health and Rights | 111 Interventions can include one-on-one meet- intertwined—including, for example, lack ings with parents, community meetings and of education or economic opportunities— education sessions, committees and forums to and multiple vulnerabilities that place girls guide services for girls, communication cam- at risk of unintended pregnancies must be paigns and public announcements, and pledges addressed together. Moreover, the prevalence by community gatekeepers. The Maharashtra of adolescent mothers among the poor points Life Skills Program in India engaged parents to the need for interventions to address and adults in committees to develop an adoles- other challenges. To help break an intergen- cent life-skills curriculum. The results show the erational pattern of poverty, improving the power of community engagement in changing socioeconomic conditions of children born to deeply entrenched social norms. In less than young mothers can be critical.139 three years, girls’ age at marriage increased by one year (from 16 to 17), and unmarried Implementation of legal responses girls expressed greater self-confidence and Although not sufficient on their own, legal an increased ability to negotiate with parents. reforms are important. Laws must be backed Among young married women, knowledge by enforcement, including clear mandates, and use of health services increased, and other procedures, and funding and accountabil- decision makers showed greater support for ity mechanisms. Where legal provisions are their reproductive health needs.137 more progressive than social norms, knowl- In Uganda, the Empowerment and Liveli- edge and compliance may remain low. Child hoods for Adolescents (ELA) program illus- marriage has been illegal in Bangladesh since trates the ways in which agency-enhancing the 1920s, yet three-quarters of girls are mar- goals can be included in economic empow- ried by their 18th birthday. In such contexts, a erment. ELA targeted girls aged 14–20 and multipronged approach is needed (box 4.5). provided life-skills training and local market– Even where progressive laws exist, gaps in informed vocational training. The program implementation and procedural requirements significantly improved girls’ agency over sex- render services inaccessible in some coun- ual and reproductive health decisions, with a tries. Take abortion services, for example. In 26 percent reduction in rates of early child- 2009 the Supreme Court of Nepal ruled that bearing and a 58 percent reduction in rates of the country’s abortion law had not been suf- child marriage and cohabitation. A 50 percent ficiently implemented because the cost was increase in condom use was also reported prohibitively high.140 In 2011, the European among girls who were sexually active. The Court of Human Rights found that, for the sec- share of girls who reported engaging in sex ond time, Poland’s lack of a comprehensive against their will dropped from 21 percent to legal framework for implementation violates almost zero.138 its obligation to ensure access to legal abortion Overall, multi-sectoral approaches are services.141 In Ghana, despite having progres- likely to be more effective than a sin- sive abortion laws, unsafe abortion continues gle intervention. This is because the risk to be one of the leading causes of maternal factors for early pregnancy are closely mortality in the country.142 112 | Voice and Agency empowering women and girls for shared prosperity Box 4.5 Egypt: Closing the gap between laws and practice A recent review of laws and the code of medical ethics showed that Egypt recognizes most of the key human rights related to family planning and reproductive health. Reproductive rights are also recognized in the Standards of Practice for Service Providers in Family Planning. Yet the standards and guidelines are not legally binding, and therefore cannot be relied upon to enforce or pursue breaches of rights, such as denial of service, coercion over a woman’s choice of contraceptive method, or lack of informed consent. Women’s lack of agency continues to be a challenge, with data from four regions showing that awareness of the right to autonomy is notably lower than other rights. In some instances, women are asked for spousal consent to receive sexual and reproductive services, although this is not required by law. In 2007, to help close this gap, Mansoura University launched joint medico-legal and ethics training courses for health and law students. Source: Rabie et al., 2013.143 Progressive laws on abortion, when force girls and women into unions. Amend- implemented effectively, can save women’s ments to the penal code of the Kyrgyz Repub- lives by broadening their options in the case lic, for example, increased the penalties for the of an unintended pregnancy and minimize practice of bride kidnapping, in which women their risk of dying from an unsafe proce- and girls are taken against their will for the dure. Legalization in Bangladesh, Romania, purpose of marriage.146 The government can and South Africa led to measurable declines also enforce a minimum age of marriage by in abortion-related mortality.144 In South requiring birth certificates prior to issuing Africa, the annual number of abortion- marriage registrations, although this in turn related deaths fell by 91 percent after liberal- requires a functioning mechanism of compul- ization of the abortion law.145 sory birth registration.147 Employing women Legal interventions can also ensure the in the justice system can also help to increase right to free and full consent to marry. Coun- women’s confidence in the law and make them tries can enhance penalties for those who feel comfortable with seeking help.148 Chapter 4 Control over Sexual and Reproductive Health and Rights | 113 *** Sexual and reproductive autonomy are intertwined with women’s agency and opportunities in life more broadly. Ultimately, fertility decisions should be made on the basis of informed choices and life plans that women and girls themselves value. This opportunity is especially relevant for vulnerable adolescents, who may perceive motherhood as the only option for social mobility and recognition. It is important to engage men, boys, and community and family opinion leaders as informed consumers, responsible decision makers, and advocates for women’s sexual and reproductive health and rights. Programs that target and engage only women will bring only partial results. Endnotes Anita Raj, “When the Mother Is a Child: The 5 Impact of Child Marriage on Health and International Conference on Population 1 Human Rights of Girls,” Archives of Diseases and Development (ICPD), “Programme of in Childhood 95 (2010): 931–35. Action;” http://www.unfpa.org/public/ home/publications/pid/1973. Note that See, for example, Karen A. Grépin and Jeni 6 some of the governments adopted ICPD Klugman, “Maternal Health: A Missed with certain reservations. Opportunity for Development,” Lancet 381, no. 9879 (2013): 191–93; and E. M. Inelmen, Karen A. Grépin and Jeni Klugman, “Mater- 2 G. Sergi, A. Girardi, A. Coin, E. D. Toffanello, F. nal Health: A Missed Opportunity for Devel- Cardin, and E. Manzato, “The Importance of opment,” Lancet 381, no. 9879 (2013): Sexual Health in the Elderly: Breaking Down 1691–93. the Barriers and Taboos,” Aging Clinical and Judith Bruce, “Married Adolescent Girls: 3 Experimental Research 24, no. 3 (2012): Human Rights, Health, and Developmental 31–34. Needs of a Neglected Majority,” Economic and World Bank, Gender, Agency, and Economic 7 Political Weekly 38 (2003): 4378–80; Anita Development in Niger (Washington, DC: Raj, Danielle Lawrence, and Niranjan Saggu- World Bank, 2014). rti, “Association between Adolescent Marriage and Marital Violence in Young Adulthood in K. Fairhurst, S. Ziebland, S. Wyke, P. Seaman, 8 India,” International Journal of Gynecology and and A. Glasier, “Emergency Contraception: Obstetrics 110 (2010): 35–39; Nawal M. Nour, Why Can’t You Give It Away? Qualitative “Child Marriage: A Silent Health and Human Findings from an Evaluation of Advance Rights Issue,” Review of Obstetrics and Gyne- Provision of Emergency Contraception,” cology 2 (2009): 51–55. Contraception 70 (2004): 25–29. Michelle J. Hindin and Adesegun O. Fatusi, 4 Jennifer McCleary-Sills, Alison McGonagle, 9 “Adolescent Sexual and Reproductive Health and Anju Malhotra, Women’s Demand for in Developing Countries: An Overview of Reproductive Control: Understanding and Trends and Interventions,” International Addressing Gender Barriers (Washington, Perspectives on Sexual and Reproductive DC: International Center for Research on Health 35, no. 2 (2009): 58–62. Women, 2012). 114 | Voice and Agency empowering women and girls for shared prosperity Ilene Speizer, Lisa Whittle, and Marion Car- 10 exclude unmarried women because it is dif- ter, “Gender Relations and Reproductive ficult to collect reliable information. Decision Making in Honduras,” Interna- 20 UNFPA, State of the World’s Population: By tional Family Planning Perspectives 31, no. Choice, Not By Chance: Family Planning, 33 (2005):131–39. Human Rights and Development, (New York: Esther Duflo, “Women Empowerment and 11 UNFPA, 2012). Economic Development,” Journal of Eco- 21 Matthias Doepke and Michele Tertlit “Wom- nomic Literature 50, no. 4 (2012): 1051–79. en’s Liberation: What’s in It for Men?” Quar- The countries included are: Albania, Arme- 12 terly Journal of Economics 11, no. 92 (2009): nia, Benin, Burkina Faso, Burundi, Cam- 379–420. bodia, Cameroon, Dem. Rep. Congo, Rep. Congo, Côte d’Ivoire, Dominican Repub- 22 ICPD, Report of the Operational Review of the lic, Ethiopia, Gabon, Ghana, Guyana, Haiti, Implementation of the Programme of Action Honduras, Lesotho, Liberia, Madagascar, of the International Conference on Popu- Malawi, Mali, Mozambique, Namibia, Nepal, lation and Development and Its Follow-up Niger, Nigeria, Rwanda, Sao Tome and Prin- Beyond 2014. cipe, Senegal, Sierra Leone, Swaziland, Tan- 23 Susheela Singh and Jacqueline Darroch, zania, Uganda, Ukraine, Zambia, Zimbabwe. Adding It Up: Costs and Benefits of Contra- International Conference on Population and 13 ceptive Services, Estimates for 2012 (New Development (ICPD), Report of the Opera- York: Guttmacher Institute and UNFPA, tional Review of the Implementation of the 2012). Programme of Action of the International 24 Singh and Darroch, Adding It Up. Conference on Population and Development and Its Follow-up beyond 2014, (New York: 25 Rob Stephenson, Andy Beke, and Delphin ICPD, 2014). Tshibangu, “Contextual Influences on Con- traceptive Use in the Eastern Cape, South United Nations Population Fund (UNFPA), 14 Africa,” Health and Place 14, no. 4 (2008): State of the World’s Population 2013: Moth- 839–50. erhood in Childhood (New York: UNFPA, 2013). 26 UNFPA, How Universal Is Access to Reproduc- tive Health? A Review of the Evidence (New Council on Foreign Relations, Ending Child 15 York : UNFPA, 2010). Marriage: How Elevating the Status of Girls Advances U.S. Foreign Policy Objectives, 27 Mai Do and Nami Kurimoto, “Women’s (New York and Washington, DC: Council on Empowerment and Choice of Contracep- Foreign Relations, 2013). tive Methods in Selected African Coun- tries,” International Perspectives on Sexual Sanyukta Mathur, Margaret Greene, and 16 and Reproductive Health 38, no. 1 (2012): Anju Malhotra, Too Young to Wed: Education 23–33. and Action Toward Ending Child Marriage (Washington, DC: International Center for 28 R. Amatya, H. Akhter, J. McMahan, N. Wil- Research on Women, 2003). liamson, D. Gates, and Y. Ahmed, “The Effect UN Women’s Indicators and Statistics 17 of Husband Counseling on NORPLANT® Database. Contraceptive Acceptability in Bangladesh” Contraception 50, no. 3 (1994): 263–73; UNFPA, State of the World’s Population 2013: 18 Alex Ezeh, The Influence of Spouses over Motherhood in Childhood. Each Other’s Contraceptive Attitudes in Statistics on unmet need understate the 19 Ghana,” Studies in Family Planning 24, no. 3 true demand for family planning. They often (1993): 163–74. Chapter 4 Control over Sexual and Reproductive Health and Rights | 115 29 J. D. Storey, M. Boulay, Y. Karki, K. Heckert, (Working paper, Harvard University, Cam- and D. Karmacharya, “Impact of the Inte- bridge, MA, 2010). grated Radio Communication Project in 36 Ayaga Bawah, Patricia Akweongo, Ruth Sim- Nepal, 1994–1997,” Journal of Health Com- mons, and James Phillips, “Women’s Fears munication 4, no. 4 (1999): 271–94; A. E. and Men’s Anxieties: The Impact of Family Biddlecom and B. M. Fapohunda, “Covert Planning on Gender Relations in Northern Contraceptive Use: Prevalence, Motivations, Ghana,” Studies in Family Planning 30, no. 1 and Consequences,” Studies in Family Plan- (1999): 54–66. ning 29, no. 4 (1998): 360–72. 37 Kay Libbus and Suha Kridli, “Contracep- Nashid Kamal, “The Influence of Husbands 30 tive Decision Making in a Sample of Jorda- on Contraceptive Use by Bangladeshi nian Muslim Women: Delineating Salient Women,” Health Policy and Planning 15, no. Beliefs,” Health Care for Women Interna- 1 (2000): 43; John Casterline, Zeba Sathar, tional 18 (1997): 85–94. and Minhaj ul Haque, “Obstacles to Contra- ceptive Use in Pakistan: A Study in Punjab,” 38 Jeremy Keele, Renata Forste, and Dailan Studies in Family Planning 32, no. 2 (2001): Flake, “Hearing Native Voices: Contracep- 95–110. tive Use in Matemwe Village, East Africa,” African Journal of Reproductive Health 9 Muhammad Masood Kadir, Fariyal Fikree, 31 (2005): 32–41. Amanullah Khan, and Fatima Sajan, “Do Mothers-in-Law Matter? Family Dynam- 39 Gilda Sedgh, Susheela Singh, Isabel Shah, ics and Fertility Decision-Making in Urban Elisabeth Ahman, Stanley Henshaw, and Squatter Settlements of Karachi, Pakistan,” Akinrinola Bankole, “Induced Abortion: Journal of Biosocial Science 35 (2003): Incidence and Trends Worldwide from 1995 545–58. to 2008,” Lancet 379, no. 9816 (2012): 625– World Bank, Gender, Agency, and Economic 32 32. doi: 10.1016/S0140-6736(11)61786-8. Development in Niger. 40 World Health Organization, Prevent- Guttmacher Institute, “Couples’ Reports of 33 ing Unsafe Abortion, http://www. Their Contraceptive Use: Do Husbands in who.i nt/r epr oduc t ivehea lt h/t opic s/ Africa Overstate the Case?” International unsafe_abortion/hrpwork/en/. Family Planning Perspectives 26, no. 4 41 UNFPA, State of the World’s Population 2013. (December 2000): 203–4. 42 Iqbal Shah and Elisabeth Ahman, “Age Pat- Cari Jo Clark, Jay Silverman, Inaam Khalaf, 34 terns of Unsafe Abortion in Developing Basem Abu Ra’ad, Zeinab Abu Al Sha’ar, Country Regions,” Reproductive Health Mat- Abdullah Abu Al Ata, and Anwar Batieha, ters 12, no. 24 (2004): 9–17. ”Intimate Partner Violence and Interference 43 Center for Reproductive Rights, The World’s with Women’s Efforts to Avoid Pregnancy in Abortion Laws, http://worldabortionlaws. Jordan,” Studies in Family Planning 39, no. com/map/ 2 (2008): 123–32; Rebecca Gee, Nandita Mitra, Fei Wan, Diana Chavkin, and Judith 44 Guttmacher Institute, Facts on Induced Long, “Power over Parity: Intimate Partner Abortion Worldwide, http://www. Violence and Issues of Fertility Control,” guttmacher.org/pubs/fb_IAW.html. American Journal of Obstetrics and Gynecol- 45 Michelle Hindin and Adesgun Fatusi, “Ado- ogy 201, no. 2 (2009): 148, e141–47. lescent Sexual and Reproductive Health in Nava Ashraf, Erica Field, and Jean Lee, 35 Developing Countries,” International Per- “Household Bargaining and Excess Fer- spective on Secual and Reproductive Health tility: An Experimental Study in Zambia” 35, no. 2 (2009) 58–62. 116 | Voice and Agency empowering women and girls for shared prosperity Joint United Nations Programme on HIV/ 46 Outcomes among the Children of Young AIDS (UNAIDS), AIDS by the Numbers (New Mothers,” Journal of Marriage and Family York: UNAIDS, 2013). 63, no. 2 (2001): 355–69. Joint United Nations Programme on HIV/ 47 61 Singh and Darroch, Adding It Up. AIDS (UNAIDS), World AIDS Day Report 62 UNFPA, How Universal Is Access to Reproduc- (New York: UNAIDS, 2012). tive Health? Joint United Nations Programme on HIV/ 48 63 UNFPA, How Universal Is Access to Reproduc- AIDS (UNAIDS), Global Report—UNAIDS tive Health? Report on the Global AIDS Epidemic 2013 (New York: UNAIDS, 2013). 64 UNFPA, How Universal Is Access to Reproduc- tive Health? UNFPA, State of the World’s Population 2013. 49 65 UNFPA, Family Planning and Young People: UNFPA, State of the World’s Population 2013. 50 Their Choices Create the Future (New York: Population Reference Bureau, The World’s 51 UNFPA, 2006); Madeline Zavodny, “Fertility Youth 2013 (data set). and Parental Consent for Minors to Receive Carl Haub, “Trends in Adolescent Fertility: 52 Contraceptives,” American Journal of Public A Mixed Picture,” Population Reference Health 94, no. 8 (2004): 1347–51. PMCID: Bureau, Washington, DC, 2013, http://www. PMC1448454. prb.org/Public at ions/A r t icles/2013/ 66 United Nations Educational, Scientific and adolescent-fertility.aspx Cultural Organization (UNESCO), Young Haub, “Trends in Adolescent Fertility.” 53 People and the Law in Asia and the Pacific: A Review of Laws and Policies Affecting Young Joao Pedro Azevedo, Marta Favara, Sarah 54 People’s Access to Sexual and Reproductive E. Haddock, Luis F. Lopez-Calva, Miriam Health and HIV Services (New York: UNE- Muller, and Elizaveta Perova, Teenage Preg- SCO, 2013). nancy Summary 2013, Vol. 1 of Teenage Pregnancy and Opportunities in Latin Amer- 67 UNICEF, “Lost in Transitions: Current Issues ica and the Caribbean on Early Child Bearing, Faced by Adolescents Living with HIV in Poverty and Economic Achievement (Wash- Asia Pacific,” UNICEF, New York, 2013. ington, DC: World Bank Group, 2013). 68 Matias Berthelon and Diana Kruger, “Risky Azevedo et al., Teenage Pregnancy Summary. 55 Behavior among Youth: Incapacitation Effects of School on Adolescent Mother- Hindin and Fatusi, “Adolescent Sexual 56 hood and Crime in Chile,” Journal of Public and Reproductive Health in Developing Economics 95, no. 1–2 (2011): 41–53. Countries”. 69 International Center for Research on UNFPA, State of the World’s Population 2013. 57 Women, Delaying Age at Marriage in Rural Guttmacher Institute, Facts on the Sexual 58 Maharashtra, India. http://www.icrw.org/ and Reproductive Health of Adolescent files/images/Delaying-Age-at-Marriage- Women in the Developing World. in-Rural-Maharashtra-India.pdf. World Bank, Teenage Pregnancy and Oppor- 59 70 Robert Jensen, “Do Labor Market Opportu- tunities in Latin America and the Caribbean nities Affect Young Women’s Work and Fam- on Early Child Bearing, Poverty and Eco- ily Decisions? Experimental Evidence from nomic Achievement (Washington, DC: World India,”  Quarterly Journal of Economics  127, Bank, 2012). no. 2 2012): 753–92. Judith A. Levine, Harold Pollack, and Mau- 60 71 UNICEF, State of the World’s Children (New reen Comfort, “Academic and Behavioral York: UNICEF, 2013). http://www.unicef. Chapter 4 Control over Sexual and Reproductive Health and Rights | 117 org/sowc2013/. 85 SOS Children’s Villages, “Attitudes on child UNFPA, Marrying Too Young: End Child Mar- 72 marriage slow to change in Nepal” June 18, riage (New York: UNFPA, 2012). 2012, accessed on May 2, 2014 at: http:// www.soschildrensvillages.org.uk/news/ UN General Assembly, Declaration on 73 archive/2012/06/at t it udes-on-child- the Elimination of Discrimination against marriage-slow-to-change-in-nepal. Women, 7 November 1967, A/RES/2263. 86 UN Women, “Defining Other Forms of Anita Raj and Ulrike Boehmer, “Girl Child 74 Forced Marriage: Bride Kidnapping,” United Marriage and Its Association with National Nations, New York. Rates of HIV, Maternal Health, and Infant Mortality Across 97 Countries,” Violence 87 Erica Field and Attila Ambrus, “Early Mar- Against Women 19 (2013): 536–51. riage, Age of Menarche, and Female School- ing Attainment in Bangladesh,” Journal Mathur et al., Too Young to Wed: Education 75 of Political Economy 116, no. 5 (2008): and Action Toward Ending Child Marriage. 881–930; Minh Cong Nguyen and Quentin UNICEF, Early Marriage: A Harmful Tradi- 76 Wodon, Estimating the Impact of Child Mar- tional Practice (New York, UNICEF, 2005). riage on Literacy and Education Attainment in Africa (Washington, DC: World Bank, Saranga Jain and Kathleen Kurz, New 77 2013). Insights on Preventing Child Marriage (Washington, DC: ICRW, 2007). 88 UNICEF, Early Marriage: A Harmful Tradi- tional Practice. UNICEF, Early Marriage. 78 89 ICRW, “Child Marriage and Domestic Dallan Flake, “Individual, Family, and Com- 79 Violence”. munity Risk Markers for Domestic Violence in Peru,” Violence Against Women 11, no. 3 90 Quentin Wodon, ed., Child Marriage and (2005): 353–73. Education (Washington, DC: World Bank, forthcoming). ICRW, “Child Marriage and Domestic Vio- 80 lence” (factsheet), 2006. http://www.icrw. 91 World Values Survey, 2010–2014, http:// org/files/images/Child-Marriage-Fact- www.worldvaluessurvey.org. Sheet-Domestic-Violence.pdf. 92 Mala Htun and Laurel Weldon, “Sex Equality Judith Bruce, “Child Marriage in the Con- 81 in Family Law: Historical Legacies, Feminist text of the HIV Epidemic. Population Coun- Activism, and Religious Power in 70 Coun- cil. Promoting Healthy, Safe and Productive tries,” Background paper to World Develop- Transitions to Adulthood,” Brief no. 11, Pop- ment Report 2011. ulation Council, New York, September 2007. 93 World Bank, Defining Gender in the 21st Cen- Robert Kelly, Ronald Gray, Nelson 82 tury: Talking with Women and Men around Sewankambo, David Serwadda, Fred Wab- the World: A Multi-Country Qualitative Study wire-Mangen, and Maria J. Wawer, “Age of Gender and Economic Choice (Washing- Differences in Sexual Partners and Risk of ton, DC: World Bank, 2011). HIV-1 Infection in Rural Uganda,” Journal of 94 Paul Fleming, Gary Barker, Mathew Morton, Acquired Immune Deficiency Syndrome, 32, Jennifer McCleary-Sills, “Engaging Men and no. 4 (2003): 446–51. Boys in Advancing Women’s Agency,” Back- Shelley Clark, “Early Marriage and HIV risks 83 ground paper to Women’s Voice, Agency, and in Sub-Saharan Africa,” Study in Family Plan- Participation (Washington, DC: World Bank, ning 35 (2004): 149–58. 2013). Mathur et al., Too Young to Wed. 84 95 Sylvia Chant and Matthew Guttman, 118 | Voice and Agency empowering women and girls for shared prosperity Mainstreaming Men into Gender and Devel- Planning Outlook and Practice in Egypt: A opment (Oxford, U.K.: Oxfam Great Britain, Rights-Based Approach (Washington, DC: 2000). World Bank, 2013). Elisabeth Rottach, Sidney Ruth Schuler, and 96 104 McCleary-Sills et al., Women’s Demand for Karen Hardee, Gender Perspectives Improve Reproductive Control. Reproductive Health Outcomes: New Evi- 105 McCleary-Sills et al., Women’s Demand for dence (Washington, DC: Population Refer- Reproductive Control. ence Bureau, 2009); McCleary-Sills et al., Women’s Demand for Reproductive Control. 106 Amy J. Kesterton and Meena Cabral de Mello “Generating Demand and Community Sup- Kelly L’Engle, Dawn Chin‐Quee, Michele 97 port for Sexual and Reproductive Health Lanham, Laura Hinson, and Heather Vahdat, Services for Young People: A Review of the “The Good, the Bad and the Ugly: The Role Literature and Programs,” Reproductive of Men in Family Planning Decision Mak- Health 7 (2010): 25; and Annabelle Erulkar, ing, Male Partner’s Roles in Women’s Use Onoka Charles, and Phiri Alford, “What Is of Emergency Contraception,” Paper pre- Youth-Friendly? Adolescents’ Preferences sented at the International Family Planning for Reproductive Health Services in Kenya Conference, Accra, May–June 2009, http:// and Zimbabwe,” African Journal of Repro- www.fpconference2009.org/media/DIR ductive Health 9, no. 3 (2005): 51–58. _169701/15f1ae857ca97193f f f f 82c9f f ffd524.pdf. Commission on Population and Development, 107 Assessment of the Status of Implementation of Gary Barker, Christine Ricardo, and Mar- 98 the Programme of Action of the International cos Nascimento, Engaging Men and Boys in Conference on Population and Development, Changing Gender-Based Inequity in Health: 47th Session, April 12, 2014, http://www. Evidence from Programme Interventions un.org/en/development/desa/population/ (Geneva, World Health Organization, 2007). pdf/commission/2014/documents/CPD47_ UNFPA, State of the World’s Population 2013: 99 Resolution_12042014_cln.pdf. Motherhood in Childhood. 108 Grépin and Klugman, “Maternal Health: 100 D. Shattuck, B. Kerner, K. Gilles, M. Hart- A Missed Opportunity for Development”; mann, T. Ng’ombe, and G. Guest, “Encour- Martina Bjorkman and Jakob Svensson, aging Contraceptive Uptake by Motivating “Power to the People: Evidence from a Men to Communicate about Family Plan- Randomised Field Experiment on Com- ning: The Malawi Male Motivator Project,” munity-Based Monitoring in Uganda, The American Journal of Public Health 101, no. 6 Quarterly Journal of Economics, May 2009: (2011):1089–95. 735-769; World Bank, World Development 101 Lisa Mwaikambo, Ilene Speizer, Anna Schur- Report 2012: Gender Equality and Develop- mann, Gwen Morgan, and Fariyal Fikree, ment (Washington, DC: World Bank, 2012); “What Works in Family Planning Interven- Susan A. Papp, Aparajita Gogoi, and Cather- tions: A Systematic Review” Studies in Fam- ine Campbell, “Improving Maternal Health ily Planning 42, no. 2 (2011): 67–82. through Social Accountability: A Case Study from Orissa, India,” Global Public Health 8, 102 Zuzana Boehmova, “Malawi Family Plan- no. 4 (2012): 449–64; Margarita Díaz, Ruth ning and Nutrition Integration Assessment,” Simmons, Juan Díaz, Francisco Cabral, Deb- World Bank, March 2013. ora Bossemeyer, Maria Y. Makuch, and Laura 103 Tamer Rabie, Zuzana Boehmova, Loraine Ghiron, “Action Research to Enhance Repro- Hawkins, Nahla Abdel Tawab, Sally Saher, ductive Choice in a Brazilian Municipality: and Atef El Shitany, Transforming Family The Santa Barbara Project,” in Responding Chapter 4 Control over Sexual and Reproductive Health and Rights | 119 to Cairo: Case Studies of Changing Practice Phone Technologies Improve Adherence to in Reproductive Health and Family Planning, Antiretroviral Treatment in a Resource-Lim- ed. Nicole Haberland and Diana Measham ited Setting: A Randomized Controlled Trial (New York: Population Council, 2001). of Text Message Reminders,” AIDS 25 , no. 6 (2011): 825–34; Richard Lester, Paul 109 E Rottach et al., Gender Perspectives Improve Ritvo, Edward Mills, Anthony Kariri, Sarah Reproductive Health Outcomes: New Karanja, Michael Chung, William Jack, James Evidence. Habyarimana, Mansen Sadatsafavi, Menal World Bank, “Results-Based Financing for 110 Najafzadeh, “Effects of a Mobile Phone Short Health,” http://www.rbfhealth.org/country/ Message Service on Antiretroviral Treat- rwanda. ment Adherence in Kenya (WelTel Kenya1): 111 Health Results Innovation Trust Fund A Randomised Trial,” Lancet 376, no. 9755 (HRITF), “Initial Results” (press release), (2010): 1838–45. April 17, 2013, http://www.worldbank. 119 United Nations Foundation–Vodafone Foun- org/en/news/press-release/2013/09/23/ dation Partnership, “mHealth for Devel- world-bank-group-invest-700-million- opment: The Opportunity of Mobile Tech- 2015-improve-women-children-health- nology for Healthcare in the Developing poor-countries. World,” United Nations Foundation, Wash- 112 Jacob Robyn, Gaston Sorgho, and Omer ington, DC, 2009. Zang, “Additional Financing: Concept Note 120 Mobile Alliance for Maternal Messages Paper of the Performance Based Financing (MAMA), “Mobile Messages,”  2013. on health in Cameroon.” World Bank, Wash- h t t p : // m o b i l e m a m a a l l i a n c e . o r g / ington DC. mobile-messages. 113 HRITF, “Initial Results.” 121 See Vodafone, Mobile for Good, http://www. 114 UNFPA, State of the World’s Population 2013: voda fone.com/content/index/about/ Motherhood in Childhood. foundation/mobiles_for_good/ccbrt.html and UNFA news, http://www.unfpa.org/ 115 UNFPA, State of the World’s Population 2013: public/home/news/pid/7697. Motherhood in Childhood. 122 La Ferrara et al., “Soap Operas and Fertility: 116 Tilly Gurman, Sara Rubin, and Amira Roess, Evidence from Brazil.” “Effectiveness of mHealth Behavior Change Communication Interventions in Develop- 123 J. Douglas Storey and Esther Kaggwa, “The ing Countries: A Systematic Review of the Influence of Changes in Fertility Related Literature,” Journal of Health Communica- Norms on Contraceptive Use in Egypt, tion 17, Suppl 1 (2012): 82–104. 1995–2005,” Population Review 48, no. 1 (2009): 1–19. 117 Deborah Levine, Jacqueline McCright, Loren Dobkin, Andrew Woodruff, and Jeffrey 124 Fatma El-Zanaty and Ann Way, Egypt Demo- Klausner, “SEXINFO: A Sexual Health Text graphic and Health Survey 2005 (Cairo: Min- Messaging Service for San Francisco Youth,” istry of Health and Population, 2005). American Journal of Public Health 98, no. 3 125 Anju Malhotra, Ann Warner, Allison McGo- (2008): 393–95. nagle, and Susan Lee-Rife, Solutions to End 118 Cristian Pop-Eleches, Harsha Thirumurthy, Child Marriage: What the Evidence Shows James Habyarimana, Joshua Zivin, Markus (Washington, DC: International Center for Goldstein, Damien de Walque, Lesley Research on Women, 2011). MacKeen, Jessica Haberer, Duncan Ngare, 126 Malhotra et al., Solutions to End Child Davide Bangsberg, John Sidle, “Mobile Marriage. 120 | Voice and Agency empowering women and girls for shared prosperity 127 Malhotra et al., Solutions to End Child Martha Brady, Ragui Assaad, Barbara Ibra- 135 Marriage. him, Abeer Salem, Rania Salem, and Nadia ChildMarriageFactsandFigures.http://www. 128 Zibani, Providing New Opportunities to Ado- icrw.org/child-marriage-facts-and-figures. lescent Girls in Socially Conservative Settings: The Ishraq Program in Rural Upper Egypt 129 Annabel Erulkar and Eunice Muthengi, (New York: Population Council, 2007). “Evaluation of Berhane Hewan: A Program to Delay Marriage In Rural Ethiopia,” Inter- Wodon, ed., Child Marriage and Education, 136 national Perspectives on Sexual and Repro- forthcoming. ductive Health 35, no. 1 (2009): 6–14. Rohni Pande, Kathleen Kurz, Sunayana 137 130 Priya Nanda, Nitin Datta, and Priya Das, Walia, Kerry MacQuarrie, and Safanga Jain “Impact of Conditional Cash Transfers on Improving the Reproductive Health of Mar- Girls’ Education: Summary of Research ried and Unmarried Youth in India: Evidence Findings,” ICRW, Washington, DC, 2014. and Effectiveness and Costs from Commu- nity-Based Interventions (Washington, DC: 131 Luis Felipe Lopez-Calva and Elizaveta Per- ICRW, 2006). ova, “The Impact of Conditional Cash Trans- fers on the Incidence of Teenage Pregnan- Oriana Bandiera, Niklas Buehren, Robin 138 cies: Evidence from Peru,” Background Burgess, Markus Goldstein, Selim Gulesci, Paper for Teenage Pregnancy and Opportu- Imran Rasul, and Munshi Sulaiman, nities in Latin America and the Caribbean On “Empowering Adolescent Girls: Evi- Early Child Bearing, Poverty and Economic dence from a Randomized Control Trial in Achievement, World Bank, Washington, DC, Uganda,” London School of Economics and 2012; J. P. Azevedo and Marta Favara, “The Political Science, London, 2012. Impact of Bolsa Familia on the Incidence of World Bank, Teenage Pregnancy and Oppor- 139 Teenage Pregnancies in Brazil,” Background tunities in Latin America and the Caribbean. Paper for Teenage Pregnancy and Opportu- Tamer Rabie, Zuzana Boehmova, Loraine 140 nities in Latin America and the Caribbean, Hawkins, Nahla Abdel Tawab, Sally Saher, World Bank, Washington, DC, 2012. and Atef El Shitany, Transforming Family 132 Teenage fertility may decrease through (a) Planning Outlook and Practice in Egypt: A increased knowledge about contraception Rights-Based Approach (Washington, DC: and access to family planning services; World Bank, 2013). or (b) increased school attendance. Con- Lakshmi Dhikta v. Government of 141 versely, CCTs may increase adolescent fer- Nepal , Writ No. 0757, Jestha, 2066 (2009) tility—just as with overall fertility—if the (Supreme Court of Nepal). Cited from the subsidies lower the costs of bringing up a Center for Reproductive Rights, http:// child, at least in the short run. worldabortionlaws.com/about.html. 133 Andaleeb Alam, Javier Baez, and Ximena R. R. v. Palando , App. No. 27617/04, 142 Del Carpio, “Does Cash for School Influ- Eur. Ct. H.R., para. 267 (2011). Cited from ence Young Women’s Behavior in the Lon- the Center for Reproductive Rights, http:// ger Term? Evidence from Pakistan,” Policy Research Working Paper 5669, World Bank, worldabortionlaws.com/about.html. Washington, DC, 2011.  See http://www.ipas.org/en/Where-We- 143 134 Sarah Baird, Craig McIntosh, and Berk Ozler Work/Africa/Ghana.aspx. “Cash or Condition? Evidence from a Cash Janie Benson, Kathryn Andersen, and Ghaza- 144 Transfer Experiment,” Quarterly Journal of len Samandari, “Reductions in Abortion-Re- Economics (Oxford University Press) 126, lated Mortality Following Policy Reform: no. 4 (2011): 1709–53. Evidence from Romania, South Africa and Chapter 4 Control over Sexual and Reproductive Health and Rights | 121 Bangladesh,” Reproductive Health 8, no. 39 147 Center for Reproductive Rights, “Child Mar- (2011), doi:10.1186/1742-4755-8-39. riage in South Asia: Stop the Impunity,” Rachel Jewkes and Helen Rees, “Dramatic 145 press release, November 10, 2013. Decline in Abortion Mortality Due to the 148 See, for example, World Bank, Gender and Choice on Termination of Pregnancy Act,” Justice in Afghanistan (Washington, DC: South African Medical Journal 95, no. 4 World Bank, forthcoming). In the report, (2005): 250. women mention the lack of female justice UN Women, “Marriages: Free and Full Con- 146 providers as a key obstacle hindering wom- sent in Entering a Marriage, http://www. en’s right to due process. endvawnow.org/en/articles/765-mar - riages-.html?next=766. 122 | Voice and Agency empowering women and girls for shared prosperity Section Sexual and reproductive health and rights | 123 Chapter 5 Control over land and housing Key messages Box 5.1 > Access to and control over land can expand women’s agency, increasing self-esteem, economic opportunities, mobility outside of the home, and decision-making power. > Fewer women than men report owning land or housing. More women who are married, divorced, or widowed report owning land than do single women. Women are more likely to report they own land jointly than individually. > Social norms, customary practices, inaccessible and weak institutions, and, in many cases, women’s lack of awareness of their rights are important barriers to the full realization of women’s land rights. > Three key areas of legislation—family law, inheritance law, and land law—affect the right to own and control property. Complex and sometimes contradictory sets of statutory laws, customs, and norms affect women’s land ownership. > Law reform and improved implementation, better gender-sensitive administration, and richer sex-disaggregated data are all needed to ensure women’s land rights are fully realized. Chapter 5 Control over land and housing | 125 Women’s control over the interplay of laws-including statutory, land and housing as a customary and religious laws—and social development challenge norms. Where good laws do exist, social norms that dictate men are the sole decision Access to and control over housing and land makers in the household may mean that can expand women’s agency and increase those laws are not implemented. We focus their access to a range of opportunities. The on three key areas of legislation—family World Development Report 20121 empha- law, inheritance law, and land law—which sized that assets are instrumental in increas- affect the right to own and control property.4 ing agency through boosting voice and Together these govern the distribution of bargaining power in household decision- household property and assets and discrim- making, access to capital, and overall eco- ination against women in these areas can nomic independence. This chapter explores result in unequal property rights. Today, 37 the benefits for women of owning and con- of 143, countries included in Women, Busi- trolling land, presents new analysis about ness and the Law 2014 still have discrimina- ownership patterns, and highlights evidence tory laws in place.5 about what work to increase women’s own- ership. Access to movable assets and credit Potentially transformative are clearly important for women’s economic effects empowerment but are outside the scope of this report. “When I have the feeling of security, that my land will not be taken away, I am A growing global consensus recognizes able to grow food on it to feed my family the intrinsic and instrumental importance for and support my community. My commu- both women and men of secure rights to land nity is more secure when I am secure.” and other productive resources in eradicat- ing poverty and reducing gender inequality.2 —Parents of Children with Disabilities (ZPCD), (Zimbabwe) Huairou Commission member6 Equal rights to access, use, and control land are enshrined in international agreements such as the Universal Declaration of Human Rights and CEDAW and regional agreements Control over land and housing has instru- mental value. Women who have more con- such as the Arab Charter on Human Rights. trol over land- whether through inheritance, These rights have become increasingly land titling, improved documentation, or important with the advent of large-scale stronger communal rights- tend to have land deals in developing countries. Women greater self-esteem, respect from other fam- are often marginalized in such processes ily members, economic opportunities, mobil- because they lack formal land rights and are ity outside of the home, and decision-making not included in negotiations leading up to the power.7 Constraints on their land ownership leasing or sale of land.3 can reduce the efficiency of land use8 and Women’s ability to exercise agency women’s economic opportunities,9 and can over land and housing is determined by exacerbate land conflict.10 Studies document 126 | Voice and Agency empowering women and girls for shared prosperity that women’s access to land and housing can seek paid work, instead of spending time affect girls’ survival rates,11 their nutritional safeguarding their land against property status12 and investment in girls’ schooling,13 invasion.18 suggesting far-reaching benefits. ■■ In Nepal, women who own land are sig- Amartya Sen has noted that female land nificantly more likely to have a final say ownership can increase a woman’s contri- in household decisions, and children of bution to the common good of the house- mothers who own land are less likely to hold, giving her a stronger voice.14 A broad be underweight,19 with associated ben- selection of literature affirms this view, efits for almost all maternal and child showing that enhanced ownership of assets, nutritional outcomes. particularly land, raises household wellbe- ■■ In Ecuador, joint land ownership ing through increased female bargaining increased women’s participation in power.15 This increase can translate into household decisions about crop cultiva- greater participation in household spending tion.20 decisions and stronger, more realistic exit ■■ In rural Karnataka, India, ownership of options in the case of an unhappy or abusive land and housing improved women’s relationship. Twenty years ago, Bina Argaw- mobility outside the home and their abil- al’s study of rural South Asia identified gaps ity to make decisions about their work, in land ownership and control as the most health, and household spending.21 important contributor to disparities in eco- nomic well-being, social status, and empow- Our analysis of 15 countries suggests the erment.16 More recent studies reinforce overall level of gender equality in a country and extend this finding, with the following is correlated with the share of women who associations: report owning housing (figure 5.1). ■■ In Vietnam, women with a joint title are For single women, land ownership can more aware of legal issues, are more provide independence. It may allow them likely to proactively seek a Land Tenure to postpone marriage or to stay in school Certificate, have more say in the use and longer. In India, for example, changes in the disposition of land, and are more likely inheritance laws allowing unmarried daugh- to earn independent incomes than those ters to inherit ancestral land delayed the age who are not on the title.17 of marriage and increased investment in education for girls (see also box 5.2).22 ■■ In Peru, squatter households who were given property titles experienced a 22 Expanding women’s ownership of land percent reduction in fertility rates, and and housing is not a panacea, nor is land females who received a joint title reduced legislation alone. Access to credit, markets, their probability of having a child twofold education, extension services, technology, compared to those in families where the personal mobility, and public voice all influ- title was in the male partner’s name only. ence women’s ability to claim and make use Receipt of titles also allowed women to of property rights.24 Chapter 5 Control over land and housing | 127 FIGURE 5.1 Higher shares of women’s housing ownership are associated with more gender equality Gender Inequality Index – 2012 0.7 CIV COG BFA 0.6 HTI MOZ SEN ZWE UGA 0.5 GAB IND NPL HND BDI RWA 0.4 ARM 0.3 0 20 40 60 80 100 Share of women who own housing Sources: Demographic and Health Surveys data 2010-2012; United Nations Development Programme Gender Inequality Index. Box 5.2 Can control over land reduce gender-based violence? Chapter 3 reviewed the extent and gravity of gender-based violence, focused primarily on violence in the home. Women’s asset ownership may provide exit options from unhappy or abusive relationships and can reduce vulnerability to domestic violence. In Kerala, India, a panel survey of ever-married women (that is, married women, widows, or divorcees) found that more than 70 percent of the women who owned land or housing and who had experienced long-term physical violence had left their husband and the matrimonial home, and few had returned. In contrast, almost all of the 20 percent of women who did not own such property went to live with their parents, and, ultimately, half returned to their husbands. Land and housing ownership also demonstrated a more secure exit option than employment: only one-third of women were employed, and the majority were in low-paying and irregular employment. In West Bengal, India, women who did not own land or housing were more likely to report domestic violence than those who did—57 percent of those without property compared to 35 percent of women property owners. In contrast, in Sri Lanka, women who owned property did not report lower rates of intimate partner violence than women who did not own property. The mixed nature of these findings may be due in part to women acquiring property after marriage, when a pattern of violence may already be entrenched in the relationship, but further exploration is needed. The introduction of joint titling in Ethiopia’s Amhara region in 2000, supported by strong implementation, has meant that men are more reluctant to ask for a divorce because they risk the division of their property. However, whether this trend is beneficial for women or ties them to abusive relationships is unclear. It is important to bear in mind that most of the studies do not show causal relationships but instead demonstrate associations. Sources: Agarwal and Panda, 2007; Teklu, 2005.23 128 | Voice and Agency empowering women and girls for shared prosperity What does ownership and access and control may differ depending on control mean? the purpose of the survey. The Demographic and Health Surveys, for example, ask men Ownership and control are defined in box and women individually about ownership of 5.3. Generally, survey measures of owner- land and housing. ship use self-reported ownership, which is obtained simply by asking respondents The different types of ownership and use whether they own the land. This means that rights outlined in box 5.3 affect the degree land may be reported as “owned” even when of women’s control over land or housing, this is not strictly the case legally. Questions which in turn can have important bearing on about asset ownership more generally typ- household decision-making and well-being. ically refer to the household and are only It is also important to recognize that joint sometimes followed by inquiries about ownership does not necessarily mean that individual ownership, which is needed to women and men have equal management conduct gender analysis. Questions about rights over the land or housing they own. Box 5.3 What do we mean by ownership and control? Clarifying ownership and control in relation to land and housing is important. Reported ownership is where a respondent reports that they own land or housing (usually posed at the household level). However, even when women self-report as joint owners of land parcels with their husband, often only his name is on the documentation. Documented ownership indicates that an individual’s name is on the relevant document, either alone or jointly with someone else. This document may be a formal land title or a customary certificate, certificate of sale, or inheritance documentation. Inclusion on a document provides a verifiable form of ownership, and important for women in the wake of a divorce or a husband’s death and also in the case of sale or transfer of land. Effective ownership is about decision-making power over how to use and dispose of property. Individuals may have partial ownership rights, such as user rights to cultivate, rent, and even bequeath land, but lack authority to sell the land. In general, women’s documented ownership of land is lower than self-reported ownership. For example, 8 percent of women and 15 percent of men are self-reported landowners in Ghana, which falls to 1 percent and 2 percent, respectively, among those with a formal ownership document. Individuals can have control over land under customary law in a number of countries, including Malawi and Rwanda. Under customary law individuals may have rights of use, with differing degrees of freedom to lease out, mortgage, bequeath, or sell. Land rights also may have a temporal or locational dimension: they may accrue only for a person’s lifetime or some lesser period, and may be conditional on the person residing on the land or locally. The extent to which an individual perceives herself to be an owner may affect her choices about the land and decision-making more broadly. Sources: Doss et al., 2013; Doss et al., 2011; Agarwal, 1994; and Doss, 2012.25 Chapter 5 Control over land and housing | 129 Land and housing rights may differ in authority over land. In Africa, the size of the urban as compared to rural areas. Increased gender gap varies by country and type of urbanization means that owning housing land and by type of landholding.27 in urban areas will become increasingly Our analysis of Demographic and Health important for providing a place for women Survey data suggests that women are less and their families to live as well as a poten- likely to report owning land or housing than tial source of income. In rural areas, agricul- are men in most of the 13 countries cov- tural land may be more important for men’s ered.28 In some cases, the gaps are striking and women’s livelihoods and have greater (figure 5.2). In several countries, the share benefits for women’s agency than house of male landowners exceeds that of women ownership.26 This is an area where data and by a large margin, but the reverse is never evidence is limited and further research is true; where women report owning more needed. land than do men, as in Rwanda, the dif- ference tends to be slight. In Burkina Faso How large is the challenge? more than twice as many men than women Data on land ownership are scarce and often (65 percent and 31 percent, respectively) not comparable across countries but avail- report owning housing. In a few countries, able data suggest that women are disad- women’s reported ownership is equal to or vantaged. Women who own land are more slightly greater than men’s; in Mozambique, likely to own it jointly, whereas men are 64 percent of women and 59 percent of men more likely to own land alone. Often women report owning housing for example. In some can access land only through male relatives. countries, very few women or men own A women’s ability to inherit or hold onto property; in Nepal, roughly one-quarter land in the case of divorce is often limited, of men and less than one-tenth of women and daughters may not be entitled to inherit report owning housing. land. Our analysis of Demographic and Men’s reporting of sole ownership is Health Survey data finds that women, who higher in all 13 countries, most notably in are younger, have less education, and single Burkina Faso, Burundi, Côte d’Ivoire, Hon- are more disadvantaged than others and duras, Nepal, Senegal, and Uganda. Women that certain characteristics, such as being tend to report owning land jointly rather married or working, can increase women’s than owning it alone, with the exceptions of likelihood of owning land and housing. Honduras and Nepal (figure 5.3). In Armenia, Honduras, Indonesia, and Senegal, similar Existing disparities proportions of men and women report own- ing land jointly, but in the other countries, the A recent review of 17 micro studies from share of women reporting joint ownership of Sub-Saharan Africa suggests that regardless land is much higher than that of men. of indicator and country, women are disad- vantaged relative to men in self-reported Widespread debate surrounds the mer- land ownership, documented ownership, its of joint versus individual titles, but rig- management control, and decision-making orous evidence is limited and the patterns FIGURE 5.2 Share of women and men who report owning housing or land Nepal, 2011 Senegal, 2010 Cote d'Ivoire, 2011-12 Burkina Faso, 2010 Women and men Honduras, 2012 who own housing Zimbabwe, 2010 Uganda, 2011 Indonesia, 2012 Rwanda, 2010 Ethiopia, 2011 Burundi, 2010 Mozambique, 2011 Armenia, 2010 0 10 20 30 40 50 60 70 80 Share of population (%) Nepal, 2011 Senegal, 2010 Honduras, 2012 Women and men Cote d'Ivoire, 2011-12 who own land Burkina Faso, 2010 Zimbabwe, 2010 Uganda, 2011 Armenia, 2010 Indonesia, 2012 Source: Voice and Agency Ethiopia, 2011 2014 team estimates based on Mozambique, 2011 Demographic and Health Surveys data for 13 countries, latest year Rwanda, 2010 available. Burundi, 2010 0 10 20 30 40 50 60 70 80 Share of population (%) FIGURE 5.3 Joint and sole reported ownership of land by gender 50 45 Sole Joint 40 Share of population who own land (%) 35 30 25 20 15 10 5 0 Armenia Burkina Faso Burundi Côte de’Ivoire Ethiopia Honduras Indonesia Mozambique Rwanda Senegal Uganda Zimbabwe Nepal Armenia Burkina Faso Burundi Côte de’Ivoire Ethiopia Honduras Indonesia Mozambique Nepal Rwanda Senegal Uganda Zimbabwe Source: Voice and Agency 2014 team estimates based on Demographic and Health Surveys data, latest year available, 2010–12. Chapter 5 Control over land and housing | 131 are complex and differ widely within and Post conflict reconstruction may provide between countries. Social norms are opportunities for reinforcing women’s land important; contravening social norms by rights. In the aftermath of the conflict in obtaining individual property rights may Rwanda for example, the government initi- carry high social costs, making joint prop- ated a policy of inclusive land reform to avert erty rights more attractive.29 In Chandigarh, the possibility of future disputes over land. India, where joint titling was introduced in The resulting Matrimonial Regimes, Liber- 2000, women reported valuing joint tenure ties and Succession Law (2000) and Organic because their husbands could not sell the Land Law (2005) have improved inheritance property without their consent, and joint rights and joint titling for women.33 titling improved their self-esteem and access to financial and economic information and Marital status and work increased their involvement in household Our analysis of DHS data for 15 countries decision-making.30 In Heredia, Costa Rica, identifies some of the characteristics associ- joint titles gave bargaining power to women ated with women’s land and housing own- during divorce proceedings, but also meant ership. Marital status, participation in labor that couples were more likely to stay in markets, and geographical location influ- unhappy or violent relationships because ence whether women are likely to report neither could afford to buy out the other.31 owning property. Figure 5.4 shows that The effects on women’s decision-making in rural areas, married women (including power vary. In Mali, Malawi, and Tanzania, women who are formally married or in de women’s individual property ownership facto partnerships) have the highest prob- increased their agricultural decision-making ability of owning land and are about five power compared to women with joint own- times more likely to own land than women ership, but had little effect on non-agricul- who have never married. Employment tural decision-making. In India, individual increases the likelihood of land ownership property ownership had little influence on too.34 In some countries, the effect is par- agricultural and household decision-making ticularly notable. Women who work in the and joint ownership had a negative correla- Republic of Congo, for example, are 24 per- tion with women’s inputs into household cent more likely to report owning land than decision-making. Social norms in India may women who do not work. Evidence from mean that women with higher social status India also suggests that land ownership is who are property owners prefer to say that they are not involved in agricultural deci- sion-making. More data and evidence are Social norms in India may mean that needed (for example, on individual own- women with higher social status who ership at plot level) to enhance our under- are property owners prefer to say that standing of how individual and joint own- they are not involved in agricultural ership affects women’s role in household decision-making. decision-making.32 132 | Voice and Agency empowering women and girls for shared prosperity FIGURE 5.4 Probability of rural land ownership, by marital and working status 50 45 40 Predicted probability of owning land (%) 35 30 25 20 15 10 5 0 Living Never Married Widowed Formerly Working Not together married married working Source: Voice and Agency 2014 team estimates using Demographic and Health Surveys data for 15 countries, latest year available, 2010–12. Note: Figure shows marginal effects on land ownership in rural areas controlling for age, work status, number of children, husband’s education, marital status, various household characteristics, property regime (separate or community of property), and country fixed effects.Note: IPV = intimate partner violence; GDP = gross domestic product. a strong predictor of self-employment for in constitutions; in Cambodia, the Land Law rural women.35 In urban areas, married (2001) grants collective ownership rights to women are three times more likely to report indigenous communities,38 and in Rwanda, owning housing than never married women. the Organic Land Law (2005) recognizes customarily acquired land.39 Laws and norms Inheritance regimes reflect norms Statutory laws, customs and norms affect around women’s property ownership and land ownership. Customary and local ten- are often very complex. Key aspects include ure systems are widespread. Worldwide, whether sons and daughters are treated as many as two billion people live under equally and whether spouses have equal customary tenure regimes,36 and less than inheritance rights to each other’s estates, 10 percent of land in Africa is held under in addition to provisions governing those statutory land tenure.37 These sets of laws who pass away without a will. In 28 of the and norms can be complex and sometimes 143 countries in the WBL database, statu- contradictory. Most developing countries tory inheritance laws differentiate between have hybrid or plural land tenure systems, women and men.40 These countries include where statutory property tenure rights exist all those covered in the Middle East and alongside customary regimes. Customary North Africa and nine in Sub-Saharan Africa tenure is often recognized under statute or (Burundi, Guinea, Lesotho, Mali, Mauritania, Chapter 5 Control over land and housing | 133 Senegal, Sudan, Tanzania, and Uganda). Ten and families may feel they have fulfilled countries (Cameroon, Chile, Democratic their financial obligations to their daughters Republic of Congo, Republic of Congo, Côte by providing a dowry on marriage.45 D’Ivoire, Ecuador, Gabon, Haiti, Mauritania, Across plural legal systems, there is no and the Philippines) limit married women’s clear pattern as to which source of law is rights over property by requiring women to more advantageous for women; this will have their husband’s permission to enter depend on local context. Where discrimina- into a transaction concerning land (but lack tory practices prevail, statutory laws may a similar requirement for men). provide women with more secure rights than Customary inheritance laws prevail under custom. In other cases, women’s rights constitutionally in more than one-quarter may be better protected under customary of countries in Sub-Saharan Africa, and are tenure.46 Statutory titling systems introduced often biased against women.41 Customary in Kenya resulted in titles being held mostly land in Ghana is specifically excluded from in men’s names only and failed to recognize statutory inheritance laws and, instead, usu- women’s customary rights to use the land.47 ally devolves to a male heir in accordance Statutory laws can reinforce norms around with custom.42 In some instances, judges women’s traditional roles in land cultiva- may have a choice as to which law to apply. tion. In India, the Odisha Land Reforms Act This is the case in Sri Lanka, where inher- allows only female heads of households to itance questions may be answered by the lease their land for cultivation on the prem- body of case law, Roman-Dutch law, Shari’a ise that women should not have to engage in law, or customary law, depending on the eth- such activities directly.48 In many countries nic group.43 in Europe and Central Asia, privatization of farmland meant that the male head of house- The reform of discriminatory inheri- hold is the person listed formally on the title tance laws can have broad positive impacts. or deed, thereby reducing the de facto rights Females whose father died after the 1994 of other family members.49 reform in India were 22 percentage points more likely to inherit land than those In Rwanda, a land tenure regularization whose father died before the reform came pilot found that legally married women into effect. The reform also had additional were more likely to have their informal land benefits for women’s agency, including an rights documented and secured and to be increase in women’s age at marriage relative regarded as joint owners than women in to men and increased educational attain- customary unions. Analysis of the pilot pro- ment.44 However, implementation of the law gram’s effect showed the probability of hav- is still weak, suggesting potentially more far ing documented land ownership fell by nine reaching results. For example, public offi- percentage points for women in customary cials may be ill informed, and women may unions.50 In response to these findings, the lack awareness of their rights. Social norms government changed the program to enable can lead to women giving up their rights in women in all types of unions to register return for protection from male relatives, land. This result suggests the wider social 134 | Voice and Agency empowering women and girls for shared prosperity context, including social norms and custom- Observations on Sri Lanka, for example, noted ary practices, needs to be explicitly consid- “discriminatory practices prevent women ered when introducing land tenure reform. from acquiring ownership of land since only the ‘head of household’ is authorized to sign While the lack of statutory rights can official documentation such as land owner- leave women vulnerable to displacement ship certificates and receive pieces of land or land grabbing in the event of widowhood from Government,” because social norms or divorce, social norms can limit statutory mean that the head of the household was laws effectiveness, depending on the wom- most often deemed to be male.55 an’s social position, education, and residency in a rural or urban area.51 In some places, longstanding social norms shape women’s Default marital property regimes access to rural land. It has been found that The default marital property regime will plough-intensive farming communities that automatically apply to distribution of assets were historically more dependent on male upon death or divorce, unless the spouses labor, limit women’s role in agricultural choose to opt out in. These laws and the activity and their access to land.52 manner in which they are implemented Discriminatory family laws can place often reflect social norms around the alloca- control in the hands of men even in cases tion and control of household property and where joint ownership is formally recog- directly affect access to land. In some coun- nized, reflecting patriarchal social norms tries, no alternative to the default regime around men’s and women’s role within the exists. The main types of marital regimes are household. In 29 countries, married women outlined in box 5.4. cannot be the head of household or head of Our analysis across 15 countries suggests the family in the same way as a man. In the that the type of marital property regime Democratic Republic of Congo, a married affects the reported levels of women’s prop- woman must obtain her husband’s permis- erty ownership; women who live in coun- sion for all legal acts in which she incurs a tries with community of property regimes personal obligation, including registering are more likely to report owning land and land in her name, and the husband has the housing than those who live in countries right to administer joint marital property.53 with separation of property regimes. This In Cameroon, a husband can legally dispose is illustrated in figure 5.5, with the size of joint property without his wife’s consent of the circles reflecting the share of prop- and can even administer his wife’s personal erty owned by women relative to men. In property.54 These examples indicate the pos- Burundi, Ethiopia, and Mozambique, coun- sible need for wide-reaching reforms of fam- tries with community of property regimes, ily laws as part of the process of land reform more than three-quarters of women report in order to achieve the intended effects. owning property. In contrast, Nepal and Head-of-household provisions can also Senegal, countries with separation of prop- limit the effect of land titling programs. The erty regimes, have the smallest share—16 CEDAW Committee in its recent Concluding percent and 20 percent, respectively. Chapter 5 Control over land and housing | 135 Box 5.4 Main types of marital property regimes Community of property regimes treat all assets, including land and housing, acquired during the marriage as the joint property of the couple, with the exception of inheritance or gifts specifically earmarked for one spouse. These regimes implicitly recognize nonmonetary contributions to the household, including domestic labor and child care. About 87 countries have community of property regimes. Community of property regimes can be divided into three categories: ›› Full: all property acquired before marriage is regarded as joint property. ›› Partial: all property acquired before marriage remains personal property of each spouse. ›› Deferred: all property acquired during marriage is treated as individually owned during the marriage but is divided equally on divorce or death. Separation of property regimes provide that all property is individually owned unless specified as jointly owned. In the event of marital dissolution, each spouse leaves with the property they brought into or acquired during the marriage. Specific legislation can be enacted to recognize nonmonetary contributions, but these provisions are rare. Of the 46 countries that have separation of property regimes, only nine recognize nonmonetary contribution. Sources: World Bank, 2014.56 FIGURE 5.5 Women’s reported ownership of housing or land, by type of marital property regime Nepal 13% Haiti 36% Burkina Faso 40% Ethiopia 61% Honduras 40% Senegal 18% Côte d’Ivoire 34% Congo. Rep 21% Indonesia 53% Uganda 51% Gabon 29% Rwanda 59% Burundi 62% Property Regime Mozambique 68% Community Deferred Separate Zimbabwe 44% Share of women who report owning housing or land 10-39% Source: Voice and Agency 2014 team estimates based on Demographic and 40-60% Health Surveys data for 15 countries, latest year available, 2010–12; and World Bank, Women, Business and the Law 2014: Removing Restrictions to Enhance 61-99% Gender Equality (Washington, DC: World Bank, 2013). 136 | Voice and Agency empowering women and girls for shared prosperity FIGURE 5.6 Probability that a woman will report owning housing or land in rural areas, conditional on type of property regime and marital status Never married Owning land Separation of property regime Community of property regime Married Owning housing Never married Married 0 10 20 30 40 50 60 Predicted probability (%) Source: Voice and Agency 2014 team estimates based on Demographic and Health Surveys data, latest year available. Note: Figure shows marginal effects on housing and land ownership in rural areas controlling for age, work status, number of children, husband’s education, marital status, various household characteristics, property regime (separation or community of property), and country fixed effects. For rural married women, the probabil- provide for all children to inherit equally, ity of reporting land and housing owner- women’s share of a couple’s wealth is ship is 17 and 29 percentage points higher, around 44 percent. In comparison in Ghana respectively, if they live in a country with a and Karnataka, which both have separation community of property regime compared of property regimes and inheritance prac- to a country with a separation of property tices that are male biased, women’s share of regime (figure 5.6). And married women a couple’s wealth is much lower, 19 percent in urban areas are nearly three times more and 9 percent, respectively.58 likely to own housing in those countries.57 The probability of land ownership for Weak implementation of unmarried women in a country with a com- laws munity of property regime is higher too, Social norms, customary practices, the most likely because such counties also have inaccessibility and weak capacity of institu- more gender-equal inheritance laws. tions and women’s lack of awareness of their Research in Ecuador, Ghana and in Kar- rights all pose important barriers to the real- nataka, India, supports these findings. In ization of women’s land rights. Qualitative Ecuador, which has a partial community of work undertaken in Niger as background for property regime and inheritance laws that this report found that women tend to accept Chapter 5 Control over land and housing | 137 a range of normative constraints on their rights. Legally, they can buy land, but the “Women cannot buy land because of few women who did this were strongly crit- social norms that forbid it.” icized for not following customary ways and —Men’s focus group, Alpha Koura, Département of told they were dishonoring their families.60 Dosso, Région of Dosso, Niger59 In Tanzania, the Land Act (1999) abolishes customary discriminatory practices, and with an analysis of how existing legislation makes local land authorities responsible for and social norms affect the distribution of protecting women but implementation of the property between husbands and wives, and law has been slow and uneven.61 Most coun- sons and daughters. We identify two broad tries in Europe and Central Asia have laws areas of focus—ensuring gender equality that require property to be divided equally under the law and ensuring effective imple- among heirs regardless of gender; however mentation of laws and land policies. some groups require women to relinquish their inheritance in favor of male relatives.62 Ensuring gender equality under In Honduras, the World Bank is working with the law the government to overcome implementa- Gender equality in the distribution and own- tion challenges to enable women’s access to ership of land should be clearly specified in equal property rights (box 5.5). country constitutions and land administration Land administration systems can help to laws. Constitutional protections can provide ensure women’s land rights are effectively oversight of all sources of law and for redress implemented. Special provisions can ensure against discrimination. In 2014, for example, women and other family members enjoy the Nigeria Supreme Court voided Igbo cus- equal rights under land reforms where men tomary law, which prevented girls from inher- are assumed to be the head of the household. iting their father’s estate, as unconstitutional.64 Monitoring sex-disaggregated land admin- Any divergence between different sources of istrative data can help flag implementation law can be addressed under the constitution. issues, but often these data are difficult to The Constitution of Uganda, Article 33, for access and not well maintained. example prohibits “laws, cultures, customs or traditions which are against the dignity, The state of the evidence: welfare or interest of women…” Beyond this, What works? strengthening family, inheritance, and land laws for women; reforming discriminatory Improving women’s ownership and control laws; and harmonizing statute, customary, and of land and housing requires institutional religious laws are all critical. change, as shown by the cogs in our frame- work in figure 1.1, where social norms, laws Promoting gender equality in and household decisions interact to impact family, inheritance, and land laws gender equality outcomes. Any intervention intended to improve women’s access to and Community of property regimes can provide control over land therefore needs to begin solid legal ground for advancing married 138 | Voice and Agency empowering women and girls for shared prosperity Box 5.5 Expanding women’s access to land rights in Honduras In 2003, the Honduras Land Administration Program (PATH) was launched to address gaps in formalization of property rights, promote a more dynamic land market, increase investments in land, and reduce social instability resulting from conflict over land. Increasing formal property ownership for women was also seen as critical for improving rural productivity. At that time, about 30 percent of the country’s land was registered, of which women owned less than 13 percent. A World Bank gender audit revealed that despite provisions allowing for joint titling, few women were named on a land title. The audit identified three primary constraints, namely limited awareness of legal rights among women and land administration personnel, land registries and other municipal offices who lacked capacity and procedural guidance on ways to implement the laws, and social norms which dictated that men were the head of household, resulting in titling being in their names alone. To address these gaps, PATH held almost a dozen stakeholder workshops in different municipalities. The participants included local authorities, community leaders, and indigenous peoples’ organizations, and the aim was to facilitate better understanding of the program among beneficiaries and the importance of gender-responsive design. The resulting gender strategy employed the following methods to improve implementation: ›› The guiding documents and instruments were revised so that implementation included specific objectives for strengthening women’s access to land. ›› A media campaign, including a radio program on “Gender situation of women in indigenous communications,” was initiated, with materials targeted at men and women of all ages to convey clear messages on the importance of land titling and ways to use titles to advance economic opportunities. ›› Project indicators were strengthened to include joint titling. The revised results and monitoring framework includes such targets as 30 percent of new titles to include women, and 25 percent of those receiving training in alternative dispute resolution, and in territorial planning and natural resource management are women. Source: World Bank, 2012.63 women’s ownership of property and for inheritance and land laws all interact. We securing women’s property rights upon have seen that reforms to marital property dissolution of a marriage. Clear consent laws and land laws can be undermined by requirements should be in place for transfer discriminatory inheritance laws. Similarly or sale, requiring the informed written con- head of household laws, which limit women’s sent of both spouses. Separation of property rights across a range of spheres including regimes should recognize non monetary con- property ownership, should be repealed.65 tributions to family wealth and provide for Laws in other areas can also address wom- equal division upon divorce or death. Mar- en’s property rights. In Brazil, India, and ital property regimes should also extend to Serbia, for example, national domestic vio- cover those in de facto relationships.Family, lence legislation now provides that victims Chapter 5 Control over land and housing | 139 of domestic violence may stay in the marital occupancy and use rights over the marital home, regardless of who owns it.66 home as well as to movable and immov- able property. Namibia’s Communal Land Inheritance laws should ensure that Reform Act (2002) gives surviving spouses property rights are transmitted equita- who reside in rural areas the right to remain bly across generations. As the example of on communal land that had been allocated India demonstrates, equalizing entitlements to the deceased. This right is not affected by under inheritance laws can dramatically improve girls and women’s lives and poten- remarriage. tially transform social norms. Discretion to circumvent women’s inheritance rights Harmonizing statutory, customary, through a will should be restricted. In Ethio- and religious regimes pia, the Land Use and Administration Proc- Experience from more than 40 World Bank lamation (2000) requires that the transfer of land reform projects in Europe and Central land through inheritance is legal only if the Asia over the past two decades has demon- wife has signed the husband’s will, which strated that challenges associated with serves to protect the widow’s interests.67 social norms and culture, cannot be solved Social norms mean that widows are by legal reform alone.71 This underlines that often pressured to waive their rights in the importance of understanding local con- return for protection from male family text and the realities on the ground, includ- members. Laws that prevent women from ing customs and traditions and the way giving up their rights for a specified period statutory rights play out, is essential for the of time following a spouse’s death can help design of effective land reform policies and offset such familial or community pressure. interventions. Country gender assessments In Jordan for example, a recent law prevents are a useful tool to advance this understand- women from handing over their inheri- ing (see box 5.5). tance for three months following receipt Where constitutional and legal reforms and requires them to formally register the strengthen the rights of women but conflict housing and land in their name.68 The legal with norms and custom, changes in cus- grounds for divorce and allocation of prop- tomary tenure systems need to be accom- erty rights upon divorce in family laws modated.73 As far as possible, protections should be reviewed so that joint titling does afforded under constitutional and statutory not become a constraint to exit. This is espe- law should be extended to include situations cially important for women living in abusive covered by custom. Some examples: relationships. Reform of land administration laws can ■■ In South Africa, the community of also have positive impacts. Mandatory joint property regime applies to customary titling has led to a marked increase in mar- marriages as well as civil marriages.74 ried women’s land ownership in Rwanda, ■■ In the Lao People’s Democratic Repub- for instance.70 Laws should ensure that lic, where strong matrilineal rights exist, the surviving spouse has at a minimum the World Bank worked with the Lao 140 | Voice and Agency empowering women and girls for shared prosperity Box 5.6 Tools to guide gender land assessments The World Bank’s “Toolkit for Integrating Gender-Related Issues in Land Policy and Administration Projects” provides guidelines for a gender analysis of the socioeconomic and cultural conditions in the project area, including with regard to statutory and customary property rights, land policies and legislation, land administration institutions, and land market transactions. In the West Bank and Gaza, for example, such an analysis identified that inheritance is determined according to Shari’a law and proof of ownership is issued by the Shari’a court. However, customary practice has encouraged women to cede their shares to their brothers. A gender analysis of the local conditions led the project design to counteract this type of pressure by issuing titles that include the names of all rightful heirs and specify their shares. Landesa’s “Women’s Land Tenure Framework for Analysis: Inheritance” provides a framework for assessing women’s ability to inherit land in a specific country, state, or community. Collaboration between the state and progressive traditional authorities can identify options for upholding women’s land rights in customary tenure areas. In Foshan City, Guangdong Province, China, for example, the local government took action to address land rights of women who married someone from another village. Using clear guidance and principles based on relevant Chinese laws, the district government set up a working group of government officials to review village rules and work together with villagers to change the local provisions that discriminate against women’s equal rights to land. Judicial procedures were applied to enforce compliance. After one year of the administrative and judicial intervention, 95 percent of married women in the district (about 18,000 women) were granted equal land rights. Sources: World Bank, 2009; Landesa, 2013; PLAAS (Institute for Poverty, Land and Agrarian Studies), 2011; UN Women (United Nations Entity for Gender Equality and Empowerment of Women), 2013.72 Women’s Union to ensure that existing (including customary and religious leaders) customary rights were reflected in the on women’s rights as well as on the benefits new land registration systems.75 of women’s land ownership. In Aceh, Indo- ■■ The Uganda National Land Policy com- nesia, for example, the RALAS (Reconstruc- mits to reform customary law, to modify tion of Aceh Land Administration System) the rules of transmission under custom- project worked with local Shari’a courts to ary land tenure, to guarantee gender produce a manual that provided guidance equality and equity, and to ensure on inheritance rights and helped protect that the decisions of traditional land widows from dispossession.77 management institutions uphold consti- An initial gender assessment can also tutional rights and obligations on gender help identify potential implementation chal- equality.76 lenges associated with social norms. In some cases, implementation challenges can be as Improving implementation obvious as allowing space for a second name Policy reforms and programmatic interven- on a land title, or adjustment of the ways in tions need to be coupled with awareness which acreage is distributed, for example to raising for women, men, and local leaders address inequalities that may result from Chapter 5 Control over land and housing | 141 gender differences in seemingly unrelated of indigenous lands included participatory legislation, such as retirement age, as in workshops to identify the most import- Vietnam (box 5.6). ant factors affecting women and engaged women in the cadaster process. The proj- An obvious key to the successful imple- ect includes targets for the number of new mentation of land rights is the commitment, titles given to women in the monitoring and willingness, and capacity of government results framework. Project results suggest staff and agencies. Establishing gender the gender strategy has raised awareness of units within land administration units and gender inequality in land access across all appointing a lead gender focal point within agencies involved and has increased wom- land registries can help.79 Female represen- en’s access to services through the provision tation in land administration institutions of additional opening hours and hotlines for can also be mandated by law, as in the women.82 Uganda Land Commission, in district land boards and in parish level committees.80 Awareness building and training pro- Women should also be involved in the design grams are important for ensuring women and men, as well as land administration offi- of national land policies and in the monitor- cials, are aware of what rights women and ing of their implementation. In Pernambuco, men have. Some examples: Brazil, women formed local committees to monitor the land titling program and make ■■ In Tanzania, a World Bank study explor- their needs known.81 Similarly, in Nicaragua ing the barriers to formalization of land a World Bank project involving demarcation title in Dar es Salaam found that demand Box 5.7 Lessons from Vietnam’s land reform process Vietnam’s 1993 Land Law did not appear to discriminate in granting land rights because it used neutral language such as “individuals” and “users” when referring to the targeted beneficiaries of the reforms. The 2000 Family and Marriage Law also provided for equal spousal rights to assets and property, including land. In practice however, gender disparities resulted. Initially, the Land Use Certificates that were used had space for only one name, which was to be filled in by the household head (generally the husband). The unintended consequence was that fewer women had their names on certificates. Gender disparities also resulted from the allocation of acreage based on the ages of household members, whereby working-age individuals received the largest shares. Because female households tended to have fewer adults of working age, they received on average less land than male-headed households. The legal retirement age for women was also five years earlier than for men. In practice, this led to women ages 55–59 being allocated half the amount of land allocated to men of the same age. Vietnam has made progress in remedying these disparities. For example, a 2001 government decree stipulated that the names of both husband and wife should be included on the certificate if the land was jointly owned, and the 2003 Land Law mandates joint titling. Sources: Menon et. al, 2013; Food and Agriculture Organization of the United Nations, Gender and Land Rights Database.78 142 | Voice and Agency empowering women and girls for shared prosperity for land titles, when offered at affordable for women’s land rights. There are several prices, was very high. When combined examples from Europe and Central Asia: with a campaign emphasizing the impor- ■■ In the Kyrgyz Republic, men and women tance of women being registered as joint were trained as community-based advis- landowners and discounts for female ers to provide free advice to villagers registration, the share of households on the process of applying for land that indicated they would include both ownership.86 husband and wife on the title increased from 24 percent to 89 percent.83 ■■ A land reform project in Tajikistan in 2007 sponsored by the U.S. Agency for ■■ In Romania, the World Bank’s Comple- International Development included menting EU Support for Agricultural supporting a group of women to mount Restructuring Project included a com- a legal challenge to the decision of a local munication and mediation specialist in official to revoke their land rights. The field teams to provide information on women were successful, and the court’s social issues. Local leaders were also decision was used around the country to involved in public awareness campaigns, educate communities and local officials and special meetings were organized just about respecting women’s land rights.87 for Roma women.84 ■■ In Moldova, Joint Information and Ser- ■■ In Rwanda, the National Land Centre has vices Bureaus offer a one-stop shop for undertaken training of local land com- advice on the job market, health care, mittees across the country, including agriculture, and land laws. The bureaus making a video showing how women’s centralize at least nine key service pro- rights should be recorded.85 viders in one office, reducing time spent Local awareness-raising campaigns and visiting multiple locations, and have community dialogue can also be effective helped more than 10,000 women, pri- ways to build understanding and support marily those from remote rural areas.88 *** Control over land and housing affects women’s agency across a range of domains, including household decision making and access to economic opportunities. Improv- ing women’s property rights can have transformative effects on social norms and women’s status within the household and the community. More and better sex– disaggregated data are needed to highlight gaps in ownership and access. This will help inform policy interventions to ensure the inclusion of women in land titling programs. Data availability and gaps are reviewed in chapter 7. A broad approach needs to ensure that beneficial customary and communal rights are recognized and respected, coupled with reforms to promote gender equality and provide effective implementation. Chapter 5 Control over land and housing | 143 Endnotes Allocation: Evidence from Bangladesh, Ethi- opia, Indonesia, and South Africa,” Oxford World Bank, World Development Report 1 Bulletin of Economics and Statistics 65 2012: Gender Equality and Development (2003): 283–327. (Washington, DC: World Bank, 2011). Klaus Deininger and Raffaella Castagnini, 10 Tim Hanstad, D. Hien Tran, and Matt Ban- 2 “Incidence and Impact of Land Conflict in nick, “Why Land Should be Part of the Post Uganda,” Policy Research Working Paper, 2015 Development Agenda,” Landesa, New World Bank, Washington, DC, March 2004; Delhi, 2013. and Sandra Joireman, “The Mystery of Julia Behrman, Ruth Meinzen-Dick, and 3 Capital Formation in Sub-Saharan Africa: Agnes R. Quisumbing, “The Gender Implica- Women, Property Rights, and Customary tions of Large-Scale Land Deals,” IFPRI Pol- Law,” World Development 36, no. 7 (2008): icy Brief 17, April 2011, International Food 1233–46. Policy Research Institute, Washington, DC, Nancy Qian, “Missing Women and the Price 11 2011. of Tea in China: The Effect of Sex-Spe- Mary Hallward-Dreimeier and Tazeen 4 cific Earning on Sex Imbalance,” Quarterly Hasan, Empowering Women: Legal Rights Journal of Economics 123, no. 3(2008): and Economic Opportunities in Africa 1251–85. (Washington, DC: World Bank, 2013). Esther Duflo, “Grandmothers and Grand- 12 World Bank and International Finance Cor- 5 daughters: Old-Age Pensions and Intra- poration, Women, Business and the Law household Allocation in South Africa,” The 2014: Removing Restrictions to Enhance World Bank Economic Review, 17, no. 1 Gender Equality (London: Bloomsbury Pub- (2003), 1-25. lishing, 2013). Nancy Luke and Kaivan Munshi, “Women 13 Birte Scholz, Patricia Chaves, Fati Al Hassan, 6 as Agents of Change: Female Income and and Jacqueline Leavitt, “Grassroots Women Mobility in India,” Journal of Development and Tenure Security: Key to Empowerment Economics 94 (2011): 1–17. and Resilient Communities,” Huairou Com- Amartya K. Sen, “Gender and Cooperative 14 mission, New York, 2013. Conflict,” in Persistent Inequalities, ed. Irene Yana Rodgers and Nidhiya Menon, “A 7 Tinker, 123–149 (Oxford, U.K: Oxford Uni- Meta-Analysis of Land Rights and Women’s versity Press, 1990). Economic Well-Being,” Brandeis University, Lisa C. Smith, Usha Ramakrishnan, Aida 15 Waltham, MA, August 2012; and Nandita Ndiaye, Lawrence Haddad, and Reynaldo Bhatla, Swati Chakraborty, and Nata Duv- Martorell, The Importance of Women’s Sta- vury, Property Ownership and Inheritance tus for Child Nutrition in Developing Coun- Rights of Women for Social Protection: The tries, Research Report 131 (Washington, DC: South Asia Experience—Synthesis Report International Food Policy Research Insti- of Three Studies (Washington, DC: Interna- tute, 2003); and Cheryl Doss, “The Effects tional Center for Research on Women, 2006). of Intrahousehold Property Ownership on Christopher Udry, “Gender, Agricultural 8 Expenditure Patterns in Ghana,” Journal of Production and the Theory of the House- African Economies 15, no. 1 (2005): 149–80. hold,” Journal of Political Economy 104, no. Bina Agarwal, “Gender and Command over 16 5 (1996): 1010–46. Property: A Critical Gap in Economic Analy- Agnes R. Quisumbing and John A. Maluccio, 9 sis and Policy in South Asia,” World Develop- “Resources at Marriage and Intrahousehold ment 22, no. 10 (1994): 1455–78. 144 | Voice and Agency empowering women and girls for shared prosperity World Bank, Analysis of the Impact of Land 17 Munk Ravnborg, “Land Administration, Tenure Certificates with both the Names of Gender Equality, and Development Coop- Wife and Husband in Vietnam (Hanoi: World eration: Lessons Learned and Challenges Bank, 2008). Ahead,” Danish Institute for International Erica Field, “Entitled to Work: Urban Prop- 18 Studies, Copenhagen, 2013. erty Rights and Labor Supply in Peru,” Quar- 25 Doss et al., “Gender Inequalities in Owner- terly Journal of Economics 122, no. 4 (2007): ship and Control of Land in Africa: Myths 1561–602. versus reality.” International Food Policy Research Institute Discussion Paper 01261. Keera Allendorf, “Do Women’s Land Rights 19 Washington, DC. 2013.; Cheryl Doss, Car- Promote Empowerment and Child Health in men D. Deere, Abena D. Oduro, Hema Swa- Nepal?” World Development , Vol. 35, no. 11 minathan, Suchitra J. Y., Rahul Lahoti, W. (2007): 1975–88. Baah-Boateng, L. Boakye-Yiadom, Jackeline Carmen Diana Deere and Jennifer Twyman, 20 Contreras, Jennifer Twyman, Zachary Catan- “Asset Ownership and Egalitarian Decision zarite, Caren Grown, and Marya Hillesland, Making in Dual-Headed Households in The Gender Asset and Wealth Gaps: Evidence Ecuador,” Review of Radical Political Eco- from Ecuador, Ghana, and Karnataka, India nomics 44, no. 3 (2012): 1–8; and Carmen (Bangalore: Indian Institute of Management Diana Deere and Jennifer Twyman, “Land Bangalore, 2011); Agarwal, “Gender and Ownership and Farm Management in Ecua- Command Over Property”; and Cheryl Doss, dor: Egalitarian Family Farming Systems “Intrahousehold Bargaining and Resource and Gendered Constraints,” University of Allocation in Developing Countries,” World Florida, Gainesville, 2012. Bank, Washington, DC, 2012. Hema Swaminathan, Rahul Lahoti, and 21 26 Carole Rakodi, “Expanding Women’s Access Suchitra J. Y., “Women’s Property, Mobility, to Land and Housing in Urban Areas,” and Decision Making: Evidence from Rural Women’s Voice, Agency and Participation Karnataka, India,” IFPRI Discussion Paper Research Series 2014, no. 8, World Bank, 01188, International Food Policy Research Washington, DC, 2014. Institute, Washington, DC, 2012. 27 Doss et al., “Gender Inequalities in Owner- Klaus Deininger, Aparajita Goyal, and 22 ship and Control of Land in Africa.” Hari Nagarajan, “Inheritance Law Reform and Women’s Access to Capital: Evidence 28 DHS data is limited to the women between from India’s Hindu Succession Act,” Policy 15 and 49 and does not cover older women. Research Working Paper 5338, World Bank, 29 Cecile Jackson, “Gender Analysis of Land: Washington, DC, 2010. Beyond Land Rights for Women?” Journal of Bina Agarwal and Pradeep Panda, “Toward 23 Agrarian Change, 3, no. 4 (2003): 453–80. Freedom From Domestic Violence: The 30 Namita Datta, “Joint Titling: A Win-Win Pol- Neglected Obvious,” Journal of Human icy? Gender and Property Rights in Urban Development 8, no. 3 (2007); and Askale Informal Settlements in India,” Feminist Teklu, “Land Registration And Women’s Economics 12, nos. 1–2 (2006): 271–98. Land Rights In Amhara Region, Ethiopia,” 31 L. Blanco Rothe, A. V. Naranjo Porras, Y. IIED Research Report 4, International Insti- Ugalde Benavente, and F. Ramírez Agüero, tute for Environment and Development, “Equidad de género y derechos de propie- Addis Ababa, Ethiopia, 2005. dad: Una investigación exploratoria sobre Rachel Spichiger, Rikke Brandt Broegaard, 24 el impacto genérico de programas de titu- Rasmus Hundsbæk Pedersen, and Helle lación conjunta en Costa Rica—El caso de Chapter 5 Control over land and housing | 145 Guararí, Heredia,” Fundación Arias para la World Bank and International Finance Corpo- 40 Paz y el Progreso Humano, San José, 2002. ration, Women, Business and the Law 2014. Cheryl Doss, Sung Mi Kim, Jemimah Njuki, 32 Andrew Harrington and Tanja Chopra, 41 Emily Hillenbrand, and Maureen Miruka, “Arguing Traditions: Denying Kenya’s “Women’s Individual and Joint Property Women Access to Land Rights,” Justice for Ownership: Effects on Household Deci- the Poor Research Report 2, World Bank, sionmaking,” IFPRI Discussion Paper 1347, Washington, DC, 2010. International Food Policy Research Insti- Hallward-Dreimeier and Hasan, Empower- 42 tute, Washington, DC, 2014. ing Women. UN Women, Realizing Women’s Rights to 33 Elisa Scalise, “Women’s Inheritance Rights 43 Land and Other Productive Resources (New to Land and Property in South Asia: A Study York: United Nations Women, 2013). of Afghanistan, Bangladesh, India, Nepal, The probabilities are married women, 32 34 Pakistan, and Sri Lanka,” Rural Develop- percent; 10 percent, respectively; living ment Institute, 2009. together: 21 percent; widowed: 25 percent; Deininger, Klaus, Aparajita Goyal, and 44 formerly married: 14 percent. Hari Nagarajan, “Inheritance Law Reform Nidhiya Menon and Yana Rodgers, “How 35 and Women’s Access to Capital: Evidence Access to Credit Affects Self-Employment: from India’s Hindu Succession Act,” Policy Differences by Gender during India’s Rural Research Working Paper 5338, World Bank, Banking Reform,” Journal of Development Washington, DC, 2010. Studies 47, no. 1 (2011): 48–69. Landesa, “The Formal and Informal Barriers 45 U.S. Agency for International Development 36 in the Implementation of the Hindu Succes- (USAID), “The Future of Customary Tenure: sion (Amendment) Act 2005 in the Context Options for Policymakers,” Property Rights of Women Agricultural Producers of Andhra and Resources Governance Briefing Paper Pradesh, Bihar, and Madhya Pradesh,” UN 8, USAID, Washington, DC, 2013. Women, New York, 2013. Lorenzo Cotula, Sonja Vermeulen, Rebeca 37 Rachael S. Knight, “Statutory Recognition of 46 Leonard, and James Keeley, Land Grab or Customary Land Rights in Africa: An Inves- Development Opportunity? Agricultural tigation into Best Practices for Lawmaking Investment and International Land Deals and Implementation,” FAO Legislative Study in Africa (London: International Institute 105, Food and Agriculture Organization of for Environment and Development, 2010); the United Nations, Rome. Klaus Deininger, “Land Policies for Growth Harrington and Chopra, “Arguing Traditions.” 47 and Poverty Reduction,” World Bank, Wash- ington DC, 2003. Cheryl Doss et al., “Women’s Individual and 48 Joint Property Ownership.” Alexandra Xanthaki, “Land Rights of Indige- 38 nous Peoples in South-East Asia,” Melbourne Victoria Stanley, Tony Lamb, and Samantha 49 Journal of International Law 4, no. 2 (2003): De Martino, “Property Rights for Women in 467–96. the ECA Region: Results from Recent World Bank Projects,” Agriculture and Environ- Daniel Ayalew Ali, Klaus Deininger, and 39 mental Services Department Note 1, World Markus Goldstein, “Environmental and Gen- der Impacts of Land Tenure Regularization Bank, Washington, DC, 2013. in Africa: Pilot Evidence from Rwanda,” Pol- Ali, Deininger, and Goldstein, “Environmen- 50 icy Research Working Paper 5765, World tal and Gender Impacts of Land Tenure Reg- Bank, Washington, DC, 2011. ularization in Africa.” 146 | Voice and Agency empowering women and girls for shared prosperity Bina Agarwal, “Bargaining and Gender Rela- 51 63 “Women’s Economic Empowerment in Latin tions: Within and Beyond the Household,” America and the Caribbean: Policy Lessons Feminist Economics 3, no. 1 (1997): 1–51. from the World Bank Gender Action Plan,” Alberto Alesina, Paola Guilano, and Nathan 52 Working Paper 76098, World Bank, Wash- ington, DC, 2012. Nunn, “On the Origins of Gender Roles: Women and the Plough, Quarterly Journal of 64 Vanguard, “Nigeria: Inheritance Rights Economics 128, no. 2 (2013): 469–530. for Girl Child,” All Africa Reports, April 24, 2014. World Bank and International Finance Cor- 53 poration, Women, Business and the Law 65 UN Women, Realizing Women’s Rights to 2014. Land and Other Productive Resources. World Bank and International Finance Cor- 54 66 Raquel Rolnik, Report of the Special Rap- poration, Women, Business and the Law porteur on Adequate Housing as a Compo- 2014. nent of the Right to an Adequate Standard of Living, and on the Right to Non-discrimina- UN Committee on the Elimination of Dis- 55 tion in this context, A/HRC/19/53, United crimination against Women, “Concluding Nations, New York, December 30, 2013. Observations of the Committee on the Elim- ination of Discrimination against Women: 67 Askale Teklu, “Land Registration and Wom- Sri Lanka,” 48th session, January 17–Febru- en’s Land Rights in Amhara Region, Ethio- ary 4, 2011, CEDAW/C/LKA/CO/7, 38–39, pia,” IIED Research Report 4, International February 4, 2011. Institute for Environment and Develop- ment, Addis Ababa, 2005. World Bank and International Finance Cor- 56 poration, Women, Business and the Law 68 World Bank, Opening Doors: Gender Equal- 2014. ity and Development in the Middle East and North Africa (Washington, DC: World Bank, For married women, the probability of own- 57 2013). ing housing if she lives in a country with community of property regime is 43 percent 69 “Setting the Post-2015 Development Com- compared to 26 percent in a country with a pass: Voices from the Ground,” Cafod, Lon- separation of property regime. The equiv- don, 2013. alent figures for never married women are 70 Daniel Ayalew Ali, Klaus Deininger, and 16 percent in a country with a community Markus Goldstein, “Can Land Tenure Reg- of property regime compared to 4 percent ularization Empower Women? Evidence in a country with a separation of property from Rwanda’s National Program,” World regime. Bank, Washington, DC, forthcoming. Cheryl Doss et al., “The Rural Gender Asset 58 71 Stanley et al., “Property Rights for Women and Wealth Gaps.” in the ECA Region.” World Bank, Republic of Niger: Gender, 59 72 World Bank, “Toolkit for Integrating Gen- Agency and Economic Development in Niger der-Related Issues in Land Policy and (Washington, DC: World Bank, forthcoming). Administration Projects,” World Bank, World Bank, Republic of Niger: Gender, 60 Washington, DC, 2010; Landesa, “Wom- Agency and Economic Development in Niger. en’s Land Tenure Framework for Analysis: Inheritance,” Landesa, Seattle, 2013; PLAAS Spichiger et al., “Land Administration, Gen- 61 (Institute for Poverty, Land and Agrarian der Equality and Development Cooperation.” Studies), “Securing Women’s Access to Land: Stanley et al., “Property Rights for Women 62 Linking Research and Action—An Overview in the ECA Region.” of Action-Research Projects in Southern Chapter 5 Control over land and housing | 147 Africa,” Synthesis Report 15, PLAAS, Bell- World Bank, “Toolkit for Integrating Gen- 79 ville, South Africa, March 2011; UN Women, der-Related Issues in Land Policy and Realizing Women’s Rights to Land and Other Administration Projects,” World Bank, Productive Resources. Washington, DC, 2009. This is consistent with the Committee on 73 UN Women, Realizing Women’s Rights to 80 Food Security’s “Voluntary Guidelines on Land and Other Productive Resources. the Responsible Governance of Tenure of Hallward-Dreimeier and Hasan, Empower- 81 Land, Fisheries and Forests in the Context ing Women. of National Food Security,” Food and Agri- Stanley, V., E. Weiss, and M Vyzaki. “Support- 82 culture Organization of the United Nations, ing Women Through Agriculture Projects In Rome. The Latin America And Caribbean Region,” Hallward-Dreimeier and Hasan, Empower- 74 World Bank, Washington, DC, 2013. ing Women. Daniel Alayew Ali, Matthew Collin, Klaus 83 Lao PDR, Land Titling Project (LTP, 1997- 75 Deininger, Stefan Dercon, Justin Sandefur, 2005) and Land Titling Project II (LTP II, and Andrew Zeitlin, “Are Poor Slum-Dwell- 2004-2009). ers Willing to Pay for Formal Land Title? Evidence from Dar es Salaam,” World Bank, UN Women, Realizing Women’s Rights to 76 Washington, DC, October 1, 2013. Land and Other Productive Resources. Stanley et al., “Property Rights for Women 84 Keith Bell, “Lessons from the Reconstruc- 77 in the ECA Region.” tion of Post-Tsunami Aceh: Build Back Bet- ter through Ensuring Women Are at the UN Women, “Progress of the World’s 85 Women 2011–2012: In Pursuit of Justice,” Center of Reconstruction of Land and Prop- United Nations, New York, 2012. erty,” Smart Lessons, 2011. UN Women, Realizing Women’s Rights to 86 Nidhiya Menon, Yana Rodgers, and Alexis 78 Land and Other Productive Resources. Kennedy, “Land Reform and Welfare in Viet- nam: Why Gender of the Land-rights Holder World Bank, “Toolkit for Integrating Gen- 87 Matters,” (Washington, DC: World Bank, der-Related Issues in Land Policy and 2013); Food and Agriculture Organization Administration Projects.” of the United Nations, Gender and Land UN Women, “Reaching Out with Public Ser- 88 Rights Database, http://www.fao.org/ vices in Rural Moldova,” United Nations, gender/landrights/home/en/. New York, 2013. Chapter 6 Amplifying voices Key messages Box 6.1 > Having a voice is a critical dimension of agency and gender equality, but women’s voices are too often denied by a combination of adverse attitudes and norms towards women’s leadership. > The landscape for engaging in politics and collective action is being rapidly transformed by new technologies such as the internet and social media, although gaps in access must be overcome. > Women’s combined strength, through collective action and women’s movements, can play a central role in building momentum for progressive legal reform, changing adverse social norms and promoting accountability. > Women’s political participation can positively impact the range of policy issues considered and prioritized, the types of solutions proposed, and can enhance perceptions of government legitimacy. > Women still represent less than one-quarter of parliamentarians worldwide. However major gains in women’s political participation have been made in some regions and countries, such as Nicaragua, Rwanda and South Africa. > Promising directions for enhancing women’s political participation are emerging, including quotas which can help to change stereotypes and providing financing and training to help level the playing field. Chapter 6 Amplifying Voices | 151 Why women’s voice and We identify three main routes for ampli- participation matters fying women’s voices. Starting with women themselves, we explore the role of the media To have a voice is to be a citizen.1 Having a and new information and communication voice means having the capacity to speak technologies (ICTs) for expanding voice up and be heard and being present to and participation. Second, we consider not shape and share in discussions, discourse, only women’s combined strength but also and decisions. Full and equal participation the central role that women’s groups can requires that everyone has a voice. Partici- pation in decision making enables women play in promoting gender equality, poverty to voice their needs and challenge gender reduction, and shared prosperity. Finally, norms in their community—individually we explore women’s representation in for- and collectively.2 If women’s participation is mal institutions, where some, albeit slow, to be transformative, their voices need to be progress has been made in most countries heard in a broad range of decision-making and regions. Collective action and women’s forums, from households to national par- participation in community-level decision liaments. Women’s movements can play a making can also lead to greater accountabil- pivotal role in building the momentum for ity, an area we touch on briefly in the policy progressive policy and legal reform. Digital section of this chapter. technologies such as mobile phones and the Internet provide new avenues for greater “Nothing, arguably, is as important voice, increased access to knowledge, and today in the political economy of devel- the potential for developing wider social opment as an adequate recognition of and professional networks. political, economic and social participa- The extent to which women are able to tion and leadership of women.3” participate in public decision-making and —Amartya Sen, 1999 make their voices heard is shaped by social norms, the legal framework, and by the nature of formal political institutions such as political parties and parliamentary struc- Tackling poverty and boosting shared tures. This is reflected in the interaction prosperity demand that everyone has the of formal and informal institutions in our opportunity to participate fully in all aspects framework in figure 1.1. At the same time, of life. Public discussion is an important women’s greater voice can seek to change vehicle for social change and economic discriminatory social norms, as in the case progress. At the same time, political and of Malala Yousafzai, from Pakistan (see box social participation has intrinsic value for 6.2) and can be a force for progressive law individual well-being. reform (see box 6.3). Women’s greater voice and participation can also affect other areas Evidence about the instrumental value of of women’s lives, including economic oppor- women’s participation in public discourse is tunities and access to services. growing. Autonomous women’s movements, 152 | Voice and Agency empowering women and girls for shared prosperity for example, have been influential in the the specific needs of women and girls.9 To passing of legislation on gender-based vio- be prevented or stymied from participat- lence at the national level, as explored in ing in the political life of the community is chapter 3. Women’s political participation a major deprivation,10 and, as noted in chap- can positively impact the range of policy ter 1, voicelessness is a social dimension of issues considered and prioritized, and the poverty. types of solutions proposed. Some dimen- sions are clear: Driving social change for women’s agency: the role Women’s participation results in greater ■■ of ICTs responsiveness to citizen needs, often increasing cooperation across party and ICTs, such as the Internet and mobile ethnic lines, and delivering more sustain- phones, are potentially powerful tools for able peace.4 increasing women’s voice and participation ■■ When more women are elected to office, in both formal and informal public spaces. policy making increasingly reflects the ICTs can increase women’s and girls’ access to knowledge and information beyond their priorities of families and women.5 In immediate environment and enable wider Rwanda and South Africa, an increase in social and professional networks. By pro- the number of female lawmakers led to viding new ways of connecting people and progressive legislation on land inheri- creating new ways for opinions to be heard, tance and reproductive rights.6 ICTs can be used to overcome the restricted ■■ Recent cross-country analysis found mobility and social exclusion from public that when at least one-quarter of mem- spaces that is often experienced by women bers of parliament are women, laws and girls. They can also provide spaces for that discriminate against women, such collective action to attract supporters and as laws that restrict women from being build momentum. ICTs can help shape the head of household, are more likely to be aspirations and hopes of the next genera- repealed.7 tion of women and girls, including their eco- ■■ Around the world, women lawmakers nomic opportunities, expectations of gender are often perceived as more honest and roles, and leadership abilities. more responsive than their male counter- The rapid expansion of ICTs globally parts; a study of 39 countries found that is well known, indeed, today more people a higher female presence in legislatures can access a mobile network than energy is correlated with higher perceptions of and clean water.11 There are currently an government legitimacy among men and estimated 6.8 billion active mobile cellu- women.8 lar subscriptions worldwide, and almost Conversely, as the World Bank (2013) 40 percent of people globally had access to among others has shown, exclusion and lack the Internet in 2013, compared with only of voice can reinforce disparities and lead 10 percent a decade ago. And yet women to fewer or poorer services that address in developing countries are less likely than Chapter 6 Amplifying Voices | 153 men to have access to the Internet, and 21 a game park in Tanzania, for example, gar- percent (or 300 million) fewer women than nered 1.7 million signatures and resulted in men in developing countries have access to the government changing its policy.15 a mobile phone.12 ICTs are being used to address threats to Using ICTs to increase voice and women’s security by gathering information participation and providing women with a means to take preventative actions. In some large cities The landscape for engagement in politics innovative crowdsourcing tools have been and public action is rapidly being trans- developed. HarassMap in Cairo and Mumbai formed by new technologies, as the case of allow incidents of harassment to be reported Malala Yousafzai illustrates (box 6.2). Inter- anonymously online or through mobile net and mobile technology have the poten- devices and then linked to city maps, giving tial to mobilize people around an issue on women an overview of potential hotspots a scale previously unimaginable.13 Collec- for sexual harassment. Hollaback!, which tive action online can range from petitions now operates in 70 cities across 24 coun- or educational forums to antiharassment tries, provides women and girls with a public initiatives and flash mobs.14 People with forum in which to voice their experiences of Internet access can gather voices and sup- public harassment. Through the exchange of port globally for local causes. Websites such experiences, participants reported a change as avaaz.org and change.org allow users to in their previous acceptance of such behav- formulate and circulate their own online ior.16 As more people access the Internet petitions globally. A petition opposing the and smart phones, such crowdsourcing and eviction of Masai villagers to make way for online socialization tools hold promise.17 Box 6.2 I believe in the power of the voice of women “When I was a girl in Swat, only a few of us were speaking, but our voice had an impact. And now not only I but millions of girls are speaking through our voice, through raising our books and pens. My father said, ‘Do not do one thing with your daughters—do not clip their wings. Give them the same right the boys have.’ Bad people came, and they took all normal life from us … they had so many rules as to what we could do. We could not go to market. We fought for our rights. I wrote the blog for the BBC and wrote my diary. Not only was I speaking, many of my friends were talking to the media and to news channels. Media was like a messenger, telling what we are saying to the world. We spoke, we wrote and raised our voices. We spoke, and we achieved our goal. Now Swat is now a peaceful place. Girls are going back to school and are allowed to go to the market.” Source: Malala Yousafzai in conversation with Jim Yong Kim, World Bank, Washington, DC, December 2013. 154 | Voice and Agency empowering women and girls for shared prosperity Addressing the digital divide Collective action as a As a conduit to greater political participa- catalyst for change tion and public action, ICTs have limitations; Beyond individual actions and informal net- some key facts need to be borne in mind. In works, collective action mobilizes people 2013, nearly 8 in 10 Internet users (77 per- around shared concerns, which can include cent) lived in developed countries.18 And, a combating discrimination. Mobilization large gender divide in access, use, owner- can take a variety of forms. It can be regu- ship, and development of ICT products per- lar or sporadic and can take place through sists, especially in Sub-Saharan Africa (43 an organization, government entities, or percent fewer women and girls have access entirely outside formal structures. It can be than men), the Middle East and North Africa localized or transnational; it can be induced (34 percent fewer women and girls have from outside or evolve organically.24 access), and South Asia (one-third fewer women and girls have access).19 Collective action can enhance women’s voice and agency and reduce gender dis- In developing countries, both men and parities through a number of channels. It women face barriers to ICT access, notably can be used to increase accountability, by inadequate infrastructure and high costs. providing a platform for participatory bud- However, social norms can further restrict geting, expenditure tracking and community women’s access.20 In the Arab Republic of Egypt and India, for example, 12 percent of scorecards, thereby giving citizens voice and women stated that they did not access the helping claim resources through bottom-up Internet more often because they did not pressure.25 think it was appropriate, and more than 8 Other actions may involve informing and percent did not access it more often because educating people about civil rights, provid- family or friends would disapprove.21 At ing spaces for women’s and girls’ groups or home, husbands might regulate the family’s peer support networks, or campaigning for use of radio, mobile phone, Internet access reforms of discriminatory laws or practices, and television. In some cases, husbands as in the following examples: determine whether and how wives can use a mobile phone. In Afghanistan, more than ■■ In Nigeria, the Legislative Advocacy half the women who did not have access Coalition on Violence Against Women to a phone cited lack of permission from campaign contributed to the passage of family members as a major obstacle.22 Ten- the Violence Against Persons (Prohibi- sions can arise and result in physical or ver- tion) Bill in 2013. The new law includes bal abuse. In Zambia for example, a recent a more comprehensive definition of rape, study found that many women report social stricter sentences, compensation for vic- and economic benefits from mobile phones, tims of rape and other sexual offenses, but phones also created conflict between protection from further abuse through spouses and in some cases reinforced tradi- restraining orders, and a fund to support tional gender power differences.23 victim rehabilitation.26 Chapter 6 Amplifying Voices | 155 ■■ In Malawi, Let Girls Lead’s Adolescent forest groups in South Asia showed that Girls’ Advocacy and Leadership Ini- low participation by women was driven tiative significantly contributed to the by long-standing beliefs about women’s drafting and enactment of local bylaws roles and women sometimes lost out as a to eradicate child marriage. The iniative result of decisions taken by the groups.30 In included advocacy with village chiefs Bangladesh, studies of collective action in and traditional leaders. Adolescent girls health and education have shown that elite interviewed after the bylaws came into groups were most motivated to participate effect reported cases of girls leaving and enjoyed the most benefits.31 Programs marriages and returning to school, and aimed at increasing civic participation can they noted that the new penalties and inadvertently exclude less networked and associated community disapproval were resourced women. A project in Kenya that deterring child marriage. aimed to address low levels of community A growing body of evidence suggests that participation among women through lead- when women participate in self-help groups ership, management, and agricultural train- and other participatory development pro- ing was found to mostly benefit women of grams, increased agency accompanies eco- higher socioeconomic status who enrolled nomic outcomes,27 in Africa for example: in higher numbers. The project failed to reach more disadvantaged women, espe- ■■ In Mali and Tanzania, women members cially older women.32 of agricultural producer self-help groups benefit from increased mobility, and in Even when income gains occur as a result Ethiopia, members reported enhanced of collective action or self-help groups, the control of household expenditures. In link between material gain and agency is Mali, along with greater autonomy over not always straightforward. For example, the use of agricultural incomes, group an impact assessment of a women’s income members were consulted more on com- generation support project in Uganda found munity and organizational decision that while household incomes increased, making.28 there was little evidence that women ■■ In western Uganda, members of a joint became more empowered in terms of par- microfinance and coffee cooperative ticipating in household decision making or generally have higher incomes and have avoiding intimate partner violence.33 used their group status to affect broader A recurring theme in this report is the change, including around the tolerance importance of normative change, and col- of gender-based violence and joint land lective action can be a tool for driving such titling.29 wider social and political changes. There is But the effects of self-help groups and a long history of mobilization led by women other participatory development projects for public action to change gender-based can be limited. Groups can at times exclude norms, combat discrimination, and secure poor, less educated, or otherwise more mar- women’s rights. Collective action may ginalized women. A study of community begin with a narrow focus, but then evolve 156 | Voice and Agency empowering women and girls for shared prosperity to address broader issues. For example, the on violence against women than those with Self Employed Women’s Association (SEWA) weaker or nonexistent movements (box in India began focusing on labor regulations 6.3). Autonomous women’s movements and expanded into women’s political leader- were found to be more important than left- ship. In Botswana, the women’s group Emang wing parties, the numbers of women legisla- Basadi (Stand Up Women) initially mobilized tors, or even national wealth.36 to oppose discriminatory citizenship laws and Women’s groups can be particularly evolved into a broader campaign to reform important in conflict-affected states to help family laws and introduce domestic violence secure access to justice. Umoja Wa Akina legislation. But there are cases where wom- Mama Fizi (United Women of Fizi) pro- en’s collective action seeks to entrench exist- motes women’s rights and raises women’s ing gender biases. For example, conservative consciousness in the Democratic Repub- women’s networks in Mali opposed progres- lic of Congo, training members to docu- sive provisions in the new Family Code, which ment human rights abuses and conduct- would have benefited many women by out- ing workshops on citizenship, rights, and lawing polygamy and requiring all marriages gender-based violence for rural women.38 to be registered. They were unsuccessful how- Women are also increasingly playing active ever and a new more progressive Family Code roles in wider social movements, as in was passed in 2012.34 Egypt, where female political activists and Women’s movements have generally reporters contributed to the revolution side played a pivotal role in progressive policy and by side with men. In 2007, the International legal reform. Autonomous women’s move- Civil Society Action Network (ICAN) worked ments together with women’s machineries with the Association of War Affected Women (women’s ministries and their equivalents) in Sri Lanka to build a ground-breaking can effectively translate international norms cross-party coalition of female politicians into institutional reforms and are associated to ensure participation in the peace process with more progressive policy regimes to in Sri Lanka. Group members visited South address violence against women.35 The fact Africa to share experiences on constitution of being autonomous allows movements to drafting and conflict resolution and par- take on issues that political parties and other ticipated in a Harvard University training associations may find too politically sensitive. course on negotiation skills and coalition For example, political parties often have little building.39 appetite for challenging norms and behaviors perceived as private—including those affect- Women’s political voice ing marriages or personal relationships, or and participation can be concerning morality. transformative Recent research across 70 countries Women remain underrepresented in key found that countries with the strongest fem- public institutions in most of the world. inist movements tend, other things being Greater representation of women in local equal, to have more comprehensive policies and national government can influence Chapter 6 Amplifying Voices | 157 Box 6.3 Women’s movements drive reforms In India, the women’s movement helped secure wide-ranging reforms in the aftermath of the heinous attack of a young student on a bus in Delhi in December 2012. India’s women’s groups and their male allies used the attack and the inadequate response from India’s authorities to highlight the customary violence against women that has been ignored by a culture of impunity and silence. Galvanized by these mass protests, a committee was set up and headed by a former Chief Justice to review antirape provisions and make recommendations for state action. Previously in India, women’s networks had used CEDAW to press for sexual harassment legislation. In 1997, the Indian Supreme Court relied on the Constitution and India’s obligations under CEDAW to recognize the right to gender equality and a working environment free from sexual harassment and abuse. The court produced the first guidelines on sexual harassment and employment, which ultimately led in 2007 to legislation prohibiting sexual harassment in the workplace. Women’s movements also use litigation as a powerful tool to challenge unconstitutional discriminatory laws and to address gaps in legal frameworks, such as the successful challenge to discriminatory customary inheritance laws in Botswana in 2013. In Egypt, tireless mobilization by women’s groups in the wake of the revolution led to a new decree that established the right of Egyptian women whose husbands are from West Bank and Gaza to pass on her citizenship to their children. In El Salvador, Asociación de Madres Demandantes helps women enforce court-ordered child support payments, and its work has attracted national attention and led to several measures guaranteeing greater support for children. Source: Hasan and Tanzer, 2013; Joan M. Caivano et al., 2008; UN Women, 2011.37 both policy considerations and budget women in Parliament led to a rise in bud- allocations: get allocations to education.42 ■■ In a poll of members of parliament from Women’s parliamentary caucuses and 110 countries conducted between 2006 committees can build collaboration on gen- and 2008, female parliamentarians were der issues across political parties: more likely to prioritize social issues such as child care, equal pay, parental ■■ In Uganda, the Women’s Parliamen- leave, and pensions; physical concerns tary Association engages directly in the such as reproductive rights, physical legislative process, creating awareness safety, and gender-based violence; and campaigns, sharing information, and development matters such as poverty building networks with nongovernmen- reduction and service delivery.40 tal organizations and women’s groups.43 ■■ Data from 19 member countries of the ■■ In Pakistan, the women’s parliamentary Organisation for Economic Co-operation caucus contributed to police reforms and Development (OECD) found that by visiting women’s police stations and having more women legislators resulted reporting on findings. The caucus also in higher total educational spending.41 meets regularly with the Afghan wom- In Sweden, an increase in the number of en’s parliamentary caucus and works on 158 | Voice and Agency empowering women and girls for shared prosperity joint implementation of a peace process ■■ More investment in drinking water infra- between the two countries.44 structure and better availability of public ■■ In Brazil the women’s caucus sponsored goods.48 legislation to establish a gender quota, a ■■ Less corruption, with men and women law on violence against women, and the reporting they are less likely to pay a inclusion of funds for social programs bribe when the gram panchayat presi- and gender equality initiatives into the dent is a woman.49 budget.45 ■■ Positive effects on parents’ aspirations Although evidence on the effect of wom- for their daughters, narrowing of the en’s participation in local government is gap between fathers’ and mothers’ sparser, Mansuri and Rao (2013) find a aspirations for their daughters, and growing body of evidence from India that simultaneous positive effects on the shows women’s participation has positively aspirations of girls themselves.50 affected social norms, as well as investments ■■ Increased reporting of crimes against in public services.46 In India, since 1992, one- women and police responsiveness.51 third of all seats on village councils (gram panchayat) and one-third of all presidencies Requiring female participation in deci- are reserved for women.47 The documented sion making can help counter exclusion and effect of these quotas includes the following: combat stereotypes (box 6.4).52 Evidence Box 6.4 “If anyone listens, I have a lot of plans”: Women’s participation in Afghanistan Afghanistan’s National Solidarity Programme (NSP) mandates participation of women in community development councils. Evidence of the mandate’s effect suggests that increasing women’s participation in community development processes can improve women’s mobility and enable access to a range of information, including legal rights. In most places, the NSP increased women’s mobility by giving them socially acceptable places to visit on a regular basis and by showing men that women’s mobility has a positive effect on women’s personal growth. A woman from Parwan province reported: “NSP allowed women to come out of their houses and gather in groups. This has helped a great deal. The more women come out the less people will talk.” At the same time, men’s perceptions have changed: “After we allowed the women’s shura (council) to become active, people do not make negative comments anymore, and if they do, nobody pays attention.” Mandating female participation improved both men’s and women’s attitudes toward female participation in community affairs (by 20 percent and 8 percent, respectively) and increased the likelihood that women would be involved in income-generating activities by some 13 percent. Sources: Sippi Azarbaijani-Moghaddam, 2010; World Bank, 2013.54 Chapter 6 Amplifying Voices | 159 FIGURE 6.1 Share of women in parliament by region and world, 1997–2013 25 Share of parliamentary seats occupied 20 by women (%) 15 10 5 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 Region East Asia and Pacific Middle East and North Africa Europe and Central Asia South Asia High income Sub-Saharan Africa Latin America and the Caribbean World Source: Voice and Agency 2014 team estimates based on World Bank, World Development Indicators 2013. from the United States, for example, sug- public decision-making forums.55 This sec- gests that living in a state with competitive tion examines participation rates in national female political candidates can increase parliaments and the judiciary and highlights women’s self-efficacy.53 the links with underlying social norms and perceptions. Underrepresentation and biased As of 2014, women hold only 22 per- gender norms cent of parliamentary seats globally. This Women remain underrepresented in key is almost double the rate in 1997, at just 12 public institutions in most of the world. percent, but still far from parity (figure 6.1). Discriminatory social norms can inhibit As of January 2014, there were 18 women women’s effective participation in poli- heads of state and government, which is tics at national and local levels. Weak net- about 10 percent.56 However, significant works, limited access to campaign financing, regional and country variation exists, and lower levels of education, greater family progress is evident: in the Nordic countries, responsibilities, and fewer opportunities women hold 42 percent of parliamentary for acquiring political experience can all seats; in Rwanda, the share is close to two- also hamper women’s participation in thirds (64 percent); and in Andorra, Cuba, 160 | Voice and Agency empowering women and girls for shared prosperity Box 6.5 Women are underrepresented at senior levels in the justice sector Globally, women account for 27 percent of judges. In some countries, women judges are restricted in their judicial duties. In the Islamic Republic of Iran, for example, women judges are unable to issue and sign final verdicts. In nearly two-thirds of the 123 countries for which there are data, women comprise less than 25 percent of justices in constitutional courts (figure 6.2). But there is significant country and regional variation. In Rwanda, Sierra Leone, and Zambia more than half of justices are women. At 60 percent, Sierra Leone has the highest share of female constitutional judges in the world, and the chief justice is a woman. With six female chief justices—in Gabon, Ghana, Niger, Nigeria, Sierra Leone, and Zambia—Sub-Saharan Africa has more female chief justices than any other region. Note: Of the 600 judges in Jordan’s court system, about 40 are women. Nicaragua, Senegal, the Seychelles, and in Northern Africa to a high of 30 percent in South Africa, the share exceeds 40 percent.57 Sub-Saharan Africa.61 Fewer than 5 percent of mayors globally are women, and many Overall across developing regions, the countries, including Afghanistan, Egypt, and Middle East and North Africa has the low- Morocco, have only one or two women serv- est levels of representation, at 16 percent, ing in this position nationwide.62 In the 77 and Latin America and the Caribbean has countries with data, fewer than one in five the highest, at 25 percent (the same as the mayors are women, with the exceptions of high-income country average). Mauritius, New Zealand, Serbia, and Latvia Even when women enter parliament, with 40 percent, 26 percent, 26 percent, and they are less likely to hold ministerial posts, 25 percent, respectively. 63 are mostly confined to social portfolios Similarly, women’s participation in local when they do so, and rarely hold high offi- groups, such as community management cial positions.58 In 2013, only about 14 per- groups, may also be more limited than men’s. cent of all presiding officers of parliament or Although group membership is common in one of its houses were women.59 Women are some developing countries, this does not also underrepresented in the justice sector, necessarily translate into effective partici- including in constitutional courts (box 6.5). pation.64 Globally, women are also less likely Data on women’s participation at the than men to participate in informal political local level, although scarce, indicate even discussions. Data gathered on engagement lower representation. The proportion of in political discussions with friends shows women among locally elected councilors, that across all regions, more men report for example, ranges from a low of 8 percent having such conversations (see box 6.6). Chapter 6 Amplifying Voices | 161 FIGURE 6.2 Percentage of female justices in constitutional courts Percentage of female justices in constitutional courts No information 0-15% Is the Chief Justice 16-30% a woman? 31-60% Yes Sources: Voice and Agency 2014 team estimates for 123 countries based on World Bank, Women, Business and the Law 2014. Mehrangiz Kar, 2008; Organisation for Economic Co-operation and Development (OECD), 2012; United Nations Children’s Fund (UNICEF), 2011; UN Women, 2012.60 Box 6.6 Differing levels of FIGURE 6.3 Rate of participation in political political engagement discussion with friends The World Values Survey asks women and men in 86 countries (representing nearly 90 East Asia and Pacific percent of the world’s population) about Sub -Saharan Africa how often they discuss political matters South Asia with friends. Across all regions of the world, Europe and Central Asia reports of such discussions were higher among men than women (figure 6.3). People Latin America and the n Male were most engaged in the East Asia and Caribbean n Female Pacific region, where one in five reported Middle East and North Africa “frequent” conversations about politics, but 0 5 10 15 20 25 30 the share of men (30 percent) was more than Percent double that of women (13 percent). Source: Voice and Agency 2014 team estimates based on World Values Survey data, latest years available, 1996–2012. 162 | Voice and Agency empowering women and girls for shared prosperity FIGURE 6.4 Attitudes toward women leaders reflected in the share of women in parliament 70 Share of women in lower house of Rwanda 60 50 Andorra Sweden South Africa parliament (%) 40 Uganda 30 Iraq 20 10 0 Qatar 0 10 20 30 40 50 60 70 80 90 100 Share of the population who agrees that men make better political leaders than women (%) Sources: Voice and Agency 2014 team estimates for 87 countries based on World Values Survey data, latest years available (1996-2012); Inter-Parliamentary Union, 2013. Attitudes toward female less likely to believe they meet the criteria to leadership affect women’s run for office than men.67 political participation But attitudes vary, and people in some Social norms often reinforce the notion regions seem to place particular value on that men are better leaders than women female politicians. In Latin America and the and limit the capacity for women to aspire Caribbean for example, 54 percent of men to public roles. On average across 86 coun- and 70 percent of women in 2007 agreed tries, most men (53 percent) and a sizable women had done a better job than men in proportion of women (41 percent) assert politics. The belief that women make equally that men make better political leaders than good leaders as men is correlated with the women.65 An earlier survey of 46 coun- percentage of female representation in par- tries found that most respondents in North liament (figure 6.4). Countries where fewer America, Latin America and the Carib- respondents say that men make better polit- bean, and Europe rated men and women as ical leaders tend to have the highest share equally good leaders. In contrast, the major- of women parliamentarians─including ity of participants in Bangladesh, Ethiopia, Andorra, Argentina, Canada, Finland, the Kuwait, Mali, Pakistan and West Bank and Netherlands, and Sweden─ while countries Gaza, and Mali preferred men as leaders.66 where about 80 percent of respondents These attitudes can limit women’s political believe men make better political leaders aspirations. For example, evidence from the have lower shares of women parliamentari- United States shows that despite compara- ans, including Egypt, the Islamic Republic of ble backgrounds, accomplished women are Iran, Iraq, Jordan, Ghana, Mali, and Qatar.68 Chapter 6 Amplifying Voices | 163 Quotas can reduce barriers to having had women leaders and increased women’s participation further after five years.72 This suggests that quotas can take time to work. Quotas may help change stereotypes and attitudes regarding women as leaders and Quotas can also be used to increase increase women’s overall engagement in women’s representation in the judiciary. For politics and civic life. In Rwanda and South example, the “Future Judges Program” in Africa, political quotas have had a major Jordon has a minimum quota of 15 percent effect on increasing women’s representa- female participants and aims to triple the tion. The Rwandan constitution (article 76, number of female judges by 2014. By 2013, section 2) guarantees women 30 percent women were 17 percent of all judges, com- of all government seats, and women now pared to 7 percent in 2010. compose 64 percent of parliamentarians.69 But positive effects are not assured and Before the African National Congress estab- will depend on the experience of the women lished a 30 percent quota for female candi- leaders and the nature of local hierarchies. dates in 1994, South Africa ranked 141st For example, women in panchayats were in the world in the percentage of legislative less effective in locations where upper seats held by women. Just six years later, the castes owned most of the land.73 Like- country ranked 10th and currently ranks wise, in Burundi in 2010, legislative quotas 8th in the world, with women holding 42 increased the share of women parliamentar- percent of parliamentary seats and 40 per- ians to 30 percent, but women continued to cent of ministerial positions.70 align themselves across partisan lines and At the gram panchayat (village council) failed to address institutional or legal gen- level in India, quotas for women and sched- der discrimination.74 uled castes have weakened prevailing ste- reotypes around women’s fitness to partic- Implications for policy ipate as political leaders and shifted social As this chapter has demonstrated, it is norms. After just seven years’ exposure, important to include women in decision men in villages with women leaders were making at all levels—national and local pol- no longer biased against them and parents’ itics, community-level decision making, and aspirations for their daughters were higher. collective action processes. ICTs can facili- After the quotas ended, women continued to tate this inclusion, but there are challenges run for and, in many cases, win elections.71 and barriers and much that we still do not know. Similarly, a recent study across 32 vil- lages in India assessing trends in women’s Increasing access to ICTs and political involvement (including voting making content relevant patterns, knowledge about rights, and par- ticipation in local-level political and social ICTs can be a driving force in providing new activities) finds that women’s involvement spaces and mediums for expanded par- is significantly higher after three years of ticipation of women in the public sphere, 164 | Voice and Agency empowering women and girls for shared prosperity but some limitations must be overcome. services that meet the needs of women and A broad policy approach that focuses on provide new spaces for online or mobile col- access to ICTs—cell phones, Internet, and laboration. So far little evidence exists on social media—and the quality of content is successful approaches and more research is required. needed to fully understand what interven- tions work and how such partnerships can Greater access to ICTs can be achieved best be achieved. in part by reducing costs and creating safe spaces where women can access hardware Greater use of ICTs for participatory and training. Ownership of computers and mapping can provide valuable information smart phones for both women and men on mobility and access to public spaces in many parts of the world remains a long by highlighting specific constraints facing way off; alternative public access points can women and girls. Using ICTs in this way can help fill this gap. This approach can include provide a low-cost alternative for monitor- leveraging existing public domains such as ing and assessment, including assessments libraries, health centers, schools, and com- on the perceptions of violence. For example, munity centers. Public libraries, for exam- mapping applications in urban townships ple, are often safe, trusted places for women in South Africa provide information about and girls to visit, and many already have travel distances and safe spaces for girls computer access points and training pro- and boys. The mapping was subsequently grams that could be leveraged to meet gen- used to provide recommendations for pub- der access objectives.75 In Copán, Honduras, lic services provision, including the creation for example, a local library provides technol- of safe spaces where adolescent girls could ogy training and digital literacy programs meet and build networks (see box 6.7). to local girls and women.76 Many of the girls trained in the program subsequently Supporting collective action become library volunteers, training others Women’s movements play a pivotal role in the community. in building the necessary momentum and Leveraging both public and private sec- consensus for progressive policy and legal tors is integral for creating a media network reform. The appropriate role for devel- that is relevant and accessible to women and opment agencies in supporting collective girls. This requires supporting the creation action processes is not always clear, how- and the development of online content that ever some evidence suggests that such pro- meets the needs and interests of women cesses are more effective when they develop and girls, especially content developed by organically rather than as the result of top- women and girls. Fostering collaboration down support. Nonetheless, development between technology providers, manufac- agencies and partners can help support turers, content producers, and end users policy reforms that create the space for col- to provide women-tailored content will be lective action to operate and can help foster critical.77 Service providers can team up stronger dialogue among collective action with women’s groups to produce tools and groups and policy makers. Chapter 6 Amplifying Voices | 165 Box 6.7 Mapping and creating safe spaces for girls in South Africa The daily activities of adolescents in South Africa are overshadowed by social, economic, and physical risks that constrain their access to the public sphere and limit their economic opportunities. In particular, the threat of gender-based violence looms large for girls. As a result, they are encouraged to stay close to their home and limit their movements and activities to safe spaces. In 2004–05, the Population Council conducted a participatory mapping exercise in an urban township and rural community. Groups of students from both primary and secondary schools, roughly 11 to 15 years of age, were asked to draw the area that represented their community and rate its level of safety. Their drawings were compared with Google satellite images for each locality (figure 6.5). FIGURE 6.5 Urban spatial area considered safe by sex and grade Grades 8–9, girls Grade 5, boys Grade 5, girls Grade 8, boys Source: Hallman et al., 2013.78 In urban areas, the space mapped by younger girls (6.33 square miles) was 2.5 times larger than the area mapped by older girls (2.62 square miles). In contrast, older boys mapped an area twice as large as younger boys, 7.81 square miles compared to 3.79 miles. The results show that at age 11, girls’ physical space was larger than that of boys of the same age but then fell dramatically after puberty. The same was true in rural areas. Boys reported a number of spaces as “very” or “extremely” safe, including primary and secondary schools, but girls did not use these categories to describe any space. In fact, urban secondary schools were ranked as very unsafe by adolescent girls. Libraries were rated as somewhat safe by younger and older urban girls, whereas older boys described them as extremely safe. Older girls in particular reported most spaces in their restricted navigable areas as unsafe. The findings suggest that participatory mapping tools can provide a useful instrument for assessing perceptions of violence and creating targeted programs for adolescents. For example, local libraries and youth centers can be transformed into safe spaces for girls by designating girls- or boys-only spaces at different times and by providing social support networks for boys and girls to negotiate insecurity and dangers in their immediate environments. Source: Hallman et al., 2013. 166 | Voice and Agency empowering women and girls for shared prosperity Development agencies can support Increasing accountability knowledge exchange among groups within Social accountability mechanisms dissem- and across countries and can provide finan- inate information, promote dialogue and cial support to pilot innovative and locally driven programs that use collective action negotiation, and provide forums through processes to help shift norms and behav- which the voices of previously excluded iors.79 Support to women’s groups for spe- groups, such as women and girls, can be cific interventions should embody large ele- heard.82 Such mechanisms are typically ments of local problem solving and learning designed with the aim of improving ser- by doing, which can be achieved in part by vice delivery and government performance, devolving responsibility to and empowering which can align with the interests of wom- the groups with local legitimacy as frontline en’s groups and thus be an important chan- implementers (box 6.8). nel through which collective action operates. Formal quotas or inclusion mandates Varying means exist for engaging in and are often needed in self-help groups and supporting social accountability mechanisms. other collective participatory processes to Channels of direct influence include gender counter exclusion and prevailing stereo- audits, which focus on systems and processes types.81 Such mandates can ensure broader within institutions, and public interest leg- representation within groups and enable islation such as right-to-information acts. greater access by those who may otherwise Some initiatives, such as the municipal gov- be marginalized. ernment of Recife, Brazil’s partnership with Box 6.8 Working with civil society to prevent gender-based violence in Haiti In the aftermath of the 2010 earthquake in Haiti, as people crowded into camps, women and children became increasingly vulnerable to violence and assault. To help address this risk, the World Bank Group joined forces with the Commission of Women Victims for Victims (KOFAVIV), a grassroots Haitian organization comprised of female survivors of sexual gender-based violence, and their international partner organization, MADRE, to address and prevent violence in five of Haiti’s internally displaced persons camps. The project worked with KOFAVIV’s community outreach workers and peer counselors—often rape survivors themselves living in the camps—and provided health and safety kits to vulnerable women and girls. This support was coupled with media outreach campaigns to promote awareness and prevention of violence as well as technical assistance to KOFAVIV from MADRE to improve project coordination, financial management, communications, long-term planning, and monitoring and evaluation. The initiative enhanced women’s civic participation by building coalitions and networks with government, international institutions, media, and other women’s rights nongovernmental organizations in Haiti. KOFAVIV also became a lead convener in working with civil society groups, police, hospitals, and the government to address gender-based violence in Haiti. The capacity at KOFAVIV increased as more empowered community agents became involved in the initiative. Source: World Bank, 2013.80 Chapter 6 Amplifying Voices | 167 the Women’s Coordinating Group (Coorde- 1. Quotas. Quota systems can help over- nadoria da Mulher) have emphasized wom- come barriers and change attitudes around en’s direct participation in budgeting.83 women’s political participation (box 6.9). Quotas work if women have a prominent Other social accountability initiatives place on party lists and the lists are enforced, that tackle gender issues include specific components enabling women’s voices to be preferably with sanctions for noncompliance. heard in ways that challenge social norms. Women will benefit from a quota only if they For example, in the Indian state of Odisha are placed in winnable positions on a party in 2006, the White Ribbon Alliance for Safe list, rather than being at the bottom with lit- Motherhood84 embarked on a program to tle chance of success.86 Quotas are unlikely address high maternal mortality and mater- to be a panacea, however: social hierarchies, nal health programs using three tools: mater- resistance from incumbents, and entrenched nal death audits via verbal autopsies; health gender norms can limit their effectiveness. facility checklists; and public hearings and Whether quotas work depends on the context rallies. Public hearings provide new ways for and care must be taken so that thresholds do women to collectively voice their concerns not become a ceiling rather than a minimum and demands in a safe space. Their demands for women’s participation.87 were reinforced and legitimized by local offi- There is now strong evidence about the cials and the media, contributing to leaders’ effectiveness of quotas and inclusion man- enhanced receptivity to women’s needs. This new understanding opened opportunities dates at the national level and in collective for improved service delivery. The program action groups and participatory projects results reflected a general recognition among and programs. Less is known globally about informants that ‘‘subtle mind-sets—among their effectiveness in local-level political both marginalized women as well as leaders institutions, and more trials and impact and service providers—play as much of a role evaluation across a larger number of coun- in the success or failure of social accountabil- tries and regions are needed. ity as any manifest factors and structural 2. Leveling the playing field. Greater barriers” and that accountability tools are access to financial resources and leadership needed to challenge “the ingrained socio-cul- training opportunities are needed. Promis- tural norms that perpetuate health dispari- ing approaches are underway in a number ties in the first place.”85 of countries, including: Opportunities to expand ■■ In Mexico, a federal law requires that women’s political participation parties spend 2 percent of their funding on women’s leadership initiatives, and Although women’s political participation remains low overall, the good news is that strong lobbying by a coalition of wom- attitudes toward female leaders are chang- en’s groups ensured strong enforcement ing and a body of experience is accumulat- mechanisms. ing. Promising approaches to increase wom- ■■ In Burkina Faso, the quota law is tied en’s participation include the following. to federal campaign funding and offers 168 | Voice and Agency empowering women and girls for shared prosperity extra funding incentives to parties who 4. Increase women’s participation fill the 30 percent benchmark.88 throughout the electoral process. Elec- ■■ In Croatia, the Gender Equality Act toral commissions and other bodies that requires special measures be taken to conduct elections can be important entry promote gender balance in all branches points for increasing women’s participation of government and that registered polit- in national and local politics. This includes ical parties must adopt action plans ensuring that women have the documen- to achieve gender balance. The Social tation needed to register to vote and that Democratic Party established a women’s electoral processes are inclusive. Examples wing that provided training on leader- include the following: ship and political communications. This ■■ The Democracy Monitoring Group in training helped develop a network of Uganda developed women’s participa- experienced women politicians, and fol- lowing the 2007 elections, women made tion check lists and reporting templates up 32 percent of the Party’s parliamen- for its observers, conducted focus tary group and held influential positions groups, analyzed voter register data, and within parliament.89 produced a comprehensive gender anal- ysis for the 2011 elections.92 The Women in Public Service Project, launched in 2011 by the US State Depart- ■■ In parts of the Middle East and North ment and the Woodrow Wilson Center, part- Africa, young men and women are ners with academic institutions around the brought together as activists and party world to empower the next generation of members to become accustomed to female leaders though training and men- working side by side as equals. In the toring. The aim is reach a minimum of 50 Republic of Yemen, youth councils have percent representation of women in public taught young women and men critical service by 2050. conflict prevention and mitigation skills, encouraging them to work together effec- 3. Civil society activism. Strategies to tively to resolve community disputes and increase women’s participation in politics advocate for local youth issues.93 and public life, led by civil society groups or individuals, can be an important source of 5. Consider proportional representation. support for women candidates. For example, Proportional representation systems out- the Elect Haitian Women campaign on tele- perform other systems in getting women vision and radio was run by a local women’s into parliament and are more effective at group throughout the country to encourage implementing quotas.94 In 2012, women voters to support women candidates.90 In accounted for 25 percent of members of par- the United States, EMILY’s List recruits and liament in proportional representation sys- trains women candidates, introduces them tems in contrast to 14 percent in first-past- to key donors and the media, and helps them the-post systems, and 18 percent in mixed raise funds for their campaigns.91 proportional representation and first-past- the-post systems.95 Chapter 6 Amplifying Voices | 169 *** Women’s voices can be transformative, helping to highlight and combat discrimina- tion, focusing attention on policies that can reduce gender inequality, and directing resources to programs that can enhance opportunities for women, girls, and their families. Making real advances in women’s political participation requires interven- tions and support on several fronts to foster positive changes in norms and practice. Collective action is an important vehicle for women’s voice, and autonomous wom- en’s movements play a critical role in driving positive changes that can shift social norms over the longer term. In many parts of the world, attitudes are changing, leading to increased recognition of women’s leadership ability at national and local levels. 170 | Voice and Agency empowering women and girls for shared prosperity Endnotes 10 Amartya Sen, “Democracy as a Univer- sal Value,” Journal of Democracy 10, no. 3 Jean Drèze and Amartya Sen, India: Develop- 1 (1999): 3–17. ment and Participation (New York: Oxford University Press, 2002). 11 Michael Nique and Firas Arab, “Sustainable Energy and Water Access through M2M Naila Kabeer, Paid Work, Women’s Empow- 2 Connectivity,” GSMA, London, 2013. erment, and Inclusive Growth: Transforming the Structures of Constraint (New York: UN 12 GSMA and Cherie Blair Foundation for Women, 2013). Women, “Women and Mobile: A Global Opportunity,” GSMA, London, 2010. Amartya Sen, Development as Freedom 3 (New York: Knopf, 1999). 13 Alison Evans and Divya Nambiar, “Collective Action and Women’s Agency: A Background Susan Markham, “Women as Agents of 4 Paper,” Women’s Voice, Agency and Partic- Change: Having Voice in Society and Influ- ipation Research Series 2013 no. 4, World encing Policy,” Women’s Voice, Agency and Bank, Washington, DC, 2013. Participation Research Series 2013, no. 5, World Bank, Washington, DC, 2013. 14 Tom Postmes and Suzanne Brunsting, “Col- lective Action in the Age of the Internet: Mark P. Jones, “Legislator Gender and Leg- 5 Mass Communication and Online Mobili- islator Policy Priorities in the Argentine zation,” Social Science Computer Review 20 Chamber of Deputies and the United States (2002): 290–301. House of Representatives,” Policy Studies Journal 25, no. 4 (2005): 613–29; Leslie A. 15 Craig Kielburger and Mark Kielburger, “Can Schwindt-Bayer, “Female Legislators and Online Petitions Change the World?” Huff- the Promotion of Women, Children, and ington Post , October 22, 2013. Family Policies in Latin America,” back- 16 Jill P. Dimond, Michaelanne Dye, Daphne ground paper for The State of the World’s LaRose, and Amy S. Bruckman, “Hollaback! Children 2007 (New York: United Nations The Role of Collective Storytelling Online Children’s Fund, 2006). in a Social Movement Organization,” in Pro- International IDEA (International Institute 6 ceedings of the 2013 Conference on Com- for Democracy and Electoral Assistance), puter Supported Cooperative Work, 477–90 Women in Parliament: Beyond Numbers (New York: ACM, 2013). (Stockholm: International IDEA, 2005). 17 It is worth mentioning that the HarassMaps Mary Hallward-Driemier, Tazeen Hasan, and 7 in Cairo and Mumbai have not lived up Anca Bogdana Rusu, “Women’s Legal Rights to their full potential yet, with 97 reports over 50 Years: Progress, Stagnation, or recorded in Mumbai since October 2012 and Regression?” Policy Research Working Paper a little more than 1,200 reports in all of Egypt 6616, World Bank, Washington, DC, 2013. since 2010. See http://harassmap.org/en/ what-we-do/the-map/ and https://akshara. Leslie A. Schwindt-Bayer and William Mish- 8 crowdmap.com/main for more information. ler, “An Integrated Model of Women’s Rep- resentation,” Journal of Politics 67, no. 2 18 International Telecommunication Union (2005): 407–28. Cited in Markham, “Women (ITU), “The World in 2013: ICT Facts and as Agents of Change.” Figures,” ITU, Geneva, 2013. World Bank, Inclusion Matters: The Founda- 9 19 Intel and Dalberg Global Development Advi- tion for Shared Prosperity (Washington, DC: sors, Women and the Web (Washington, DC: World Bank, 2013). Intel, 2012). Chapter 6 Amplifying Voices | 171 Simone Cecchini and Chris D. Scott, “Can 20 “Randomized Impact Evaluation of Phase-II Information and Communications Tech- of Afghanistan’s National Solidarity Pro- nology Applications Contribute to Poverty gramme (NSP): Estimates of Interim Pro- Reduction? Lessons from Rural India,” gram Impact from First Follow-Up Survey,” Information Technology for Development 10 World Bank, Washington, DC, 2010; Chris- (2003): 73–84. topher Blattman, Eric Green, Jeannie Annan, and Julian Jamison, “Building Women’s Eco- Intel and Dalberg Global Development Advi- 21 nomic and Social Empowerment through sors, Women and the Web. Enterprise: An Experimental Assessment U.S. Agency for International Development 22 of the Women’s Income Generating Support (USAID), “Connecting to Opportunity: A (WINGS) Program in Uganda,” Logica Study Survey of Afghan Women’s Access to Mobile 1, World Bank, Washington, DC, April 2013; Technology,” USAID, Washington, DC, 2013. Oxfam, “Women’s Collective Action in the Honey Sector in Ethiopia: Involving Margin- Kutoma J. Wakunuma. “Implicating Mobile 23 alized Women in Collective Action—Making Phones in Violence against Women: What’s a Difference through NGO Interventions,” Gender Got to Do with It?” GenderIT Oxfam, London, February 2013. blog, 2012, http://www.genderit.org/ resources/implicating-mobile-phones- Evans and Nambiar, “Collective Action and 28 violence-against-women-what-s-gender- Women’s Agency.” got-do-it. Meier zu Selhalsen, “Does Agency Mat- 29 Evans and Nambiar, “Collective Action and 24 ter and Do Microfinance Self-Help Groups Women’s Agency.” Empower Women?” Evans and Nambiar, “Collective Action and 25 Bina Agarwal, “Pparticipatory Exclusions, 30 Women’s Agency.” Community Forestry, and Gender: An Analysis for South Asia and a Conceptual Brenda Kombo, Rainatou Sow, and Faiza 26 Framework,” World Development 29, no. 10 Jama Mohamed (eds.), Journey to Equality: (2001): 1623–48. 10 Years of the Protocol on the Rights of Women in Africa (New York: Equality Now, Simeen Mahmud, “Making Rights Real 31 2013). in Bangladesh through Collective Citizen Action,” IDS Bulletin 33, no. 2 (2002): 31–39. Felix Meier zu Selhausen, “Does Agency 27 Matter and Do Microfinance Self-Help Mary Kay Gugerty and Michael Kremer, 32 Groups Empower Women? A Case Study “Outside Funding and the Dynamics of of a Joint-Microfinance and Coffee Coop- Participation in Community Associations,” erative from the Mountains of the Moon in American Journal of Political Science 52, no. Uganda,” presented at the conference on 3 (2008): 585–602. Design and Dynamics of Institutions for Col- Evans and Nambiar, “Collective Action and 33 lective Action, Utrecht University, Utrecht, Women’s Agency.” Netherlands, November 29–December 1, Mary Hallward-Dreimeier and Tazeen 34 2012; Eeshani Kandpal and Kathy Baylis, Hasan, Empowering Women: Legal Rights “Expanding Horizons: Can Women’s Sup- and Opportunities in Africa (Washington, port Groups Diversify Peer Networks in DC: World Bank, 2013). Rural India?” American Journal of Agricul- tural Economics 95, no. 2, (2013): 360–67; Laura S. Weldon and Mala Htun, “Femi- 35 Andrew Beath, Fontini Christia, Ruben nist Mobilization and Progressive Policy Enikolopov, and Shahim Ahmad Kabuli, Change: Why Governments Take Action to 172 | Voice and Agency empowering women and girls for shared prosperity Combat Violence Against Women,” Gender 46 Ghazala Mansuri and Vijayendra Rao, Local- and Development 21, no.2 (2013), 231-247. izing Development: Does Participation Work? (Washington, DC: World Bank, 2013). Weldon and Htun, “Feminist Mobilization 36 and Progressive Policy Change.” 47 Mansuri and Rao, Localizing Development . Tazeen Hasan and Zioner Tanzer, “Wom- 37 48 Raghabendra Chattopadhyay and Esther en’s Movements, Plural Legal Systems and Duflo, “Women as Policy Makers: Evidence the Botswana Constitution: How Reform from a Randomized Policy Experiment in Happens,” World Bank, Washington, DC, India,” Econometrica 72(5 (2004): 1409– 2013; UN Women, “Progress of the World’s 33; and Esther Duflo and Petia Topalova, Women: In Pursuit of Justice.” (New York: “Unappreciated Service: Performance, Per- United Nations, 2011); Joan M. Caivano and ceptions, and Women Leaders in India,” Thayer Hardwick, “Civil Society and Social Working Paper, Department of Econom- Movements; Building Sustainable Democ- ics, Massachusetts Institute of Technology, racies in Latin America,” Inter-American Cambridge, 2004. Development Bank, 2008. 49 Chattopadhyay and Duflo, “Women as Policy Evans and Nambiar, “Collective Action and 38 Makers”  ; and Duflo and Topalova, “Unap- Women’s Agency.” preciated Service.” Sanam Anderlini and others. 2010. “What 39 50 Lori Beaman, Esther Duflo, Rohini Pande, the Women Say: Participation and UNSCR and Petia Topalova, “Female Leadership 1325: A Case Study Assessment by the Raises Aspirations and Educational Attain- International Civil Society Action Network ment for Girls: A Policy Experiment in India,” and the MIT Center for International Stud- Science 335, no. 6068 (2012): 582–86. ies?” International Civil Society Action Net- work (ICAN) and Center for International 51 Lakshmi Iyer,  Anandi Mani, Prachi Mishra, Studies and Petia Topalova, “The Power of Political Voice: Women’s Political Representation Inter-Parliamentary Union (IPU), Equality 40 and Crime in India,” American Economic in Politics: A Survey of Men and Women in Journal: Applied Economics 4, no. 4 (2012): Parliaments (Geneva: IPU, 2008). 165–93. Li-Ju Chen, “Female Policy Maker and Edu- 41 cational Expenditure: Cross-Country Evi- 52 Mansuri and Rao, Localizing Development . dence,” Research Paper in Economics 1, 53 Lonna Rae Atkeson, “Not All Cues Are Cre- Department of Economics, Stockholm Uni- ated Equal: The Conditional Impact of versity, 2008. Female Candidates on Political Engage- Markham, “Women as Agents of Change.” 42 ment,” Journal of Politics 65, no. 4 (2003): 1041–61. Evans and Nambiar, “Collective Action and 43 Women’s Agency.” Sippi Azarbaijani-Moghaddam, “A Study of 54 Gender Equity through the National Solidarity South Asian Regional Secretariat of Women 44 Programme’s Community Development Coun- Parliamentarians, “Report on the First cils: ‘If Anyone Listens, I Have a Lot of Plans,” Meeting of the Network of South Asian Danish Committee for Aid to Afghan Refugees, Women Parliamentarians,” Colombo, Febru- Kabul, 2010; World Bank, “Randomized Impact ary 25–27, 2013, 12. Evaluation of Afghanistan’s National Solidar- Evans and Nambiar, “Collective Action and 45 ity Programme,” Report 81107, World Bank, Women’s Agency.” Washington, DC, July 2013. Chapter 6 Amplifying Voices | 173 55 UN Women, “Women and Elections: Guide org/2009/06/29/morocco-celebrat - to Promoting the Participation of Women in ing-the-first-female-mayor-of-marrakesh/; Elections,” United Nations, New York, 2005. and Jeffrey Fleishman, “In Egypt, a Village Boasts the Nation’s First Female Mayor,” 56 UN Women, “Women in Politics,” United Los Angeles Times, March 8, 2009, http:// Nations, New York, 2014. a r t ic le s . l at i me s . c om/20 0 9/m a r/0 8/ 57 Inter-Parliamentarian Union, “Women in world/fg-egypt-mayor8. National Parliaments,” February 20, 2014, United Nations Statistics Division (New 63 http://www.ipu.org/wmn-e/classif.htm. York: United Nations, 2010), 120. 58 World Bank, World Development Report Agarwal, “Participatory Exclusions, Com- 64 2012: Gender Equality and Development munity Forestry, and Gender.” (Washington, DC: World Bank, 2012). Latest data are in UN Women, “Gender Justice: Key Nate, Breznau, Valerie A. Lykes, Jonathan 65 to Achieving the Millennium Development Kelley, and M. D. R. Evans, “A Clash of Civi- Goals,” United Nations, New York, 2010. lizations? Preferences for Religious Political Leaders in 86 Nations World Values,” Jour- 59 Inter-Parliamentary Union, “Women Speak- nal for the Scientific Study of Religion 50, no. ers of National Parliaments,” 2014, http:// 4 (2011): 671–91. www.ipu.org/wmn-e/speakers.htm. Juliana Menasce Horowitz, “How the World 66 UN Women, In Pursuit of Justice: 2011–12 60 Rates Women as Leaders,” Pew Research Progress of the World’s Women Report. (New Center, Washington, DC, 2007. York: United Nations, 2012); Mehrangiz Kar, “Discrimination against Women under Richard L. Fox and Jennifer L. Lawless, “Gen- 67 Iranian Law,” Mehrangiz Kar blog, 2008, dered Perceptions and Political Candida- http://www.mehrangizkar.net/english/ cies: A Central Barrier to Women’s Equality archives/000416.php; Organisation for Eco- in Electoral Politics,” American Journal of nomic Co-operation and Development, Social Political Science, 55, no. 1 (2011): 59–73. Institutions and Gender Index, 2012, http:// World Bank, Inclusion Matters. 68 www.oecd.org/dev/poverty/theoecdsoci alinstitutionsandgenderindex.htm; United Inter-Parliamentarian Union, “Women in 69 Nations Children’s Fund (UNICEF), “Jordan: National Parliaments.” MENA Gender Equality Profile—Status of Richard Hausmann, Laura D. Tyson, and 70 Girls and Women in the Middle East and Saadia Zahidi, The Global Gender Gap Report North Africa,” UNICEF, New York, 2011, 4. 2012 (Geneva: World Economic Forum, 61 United Nations Statistics Division, World’s 2012). Women 2010: Trends and Statistics (New Beaman et al., “Female Leadership Raises 71 York: United Nations, 2010), 119. Aspirations and Educational Attainment for 62 United Nations Statistics Division, 2010; Girls.” see also Golnar Motevalli, “Afghanistan’s Dhanmanjiri Sathe, Stephan Klasen, Jan 72 First Female Mayor Proves Critics Wrong,” Priebe, and Mithila Biniwale, “Can the Guardian, February 24, 2013, http://www. Female Sarpanch Deliver: Evidence from g ua rdia n .co.u k/world/2013/feb/24/ Maharashtra,” Economic and Political afghanistan-first-female-mayor; Jillian Weekly 48, no. 11 (2013): 50–57. York, “Morocco: Celebrating the First Female Mayor of Marrakesh,” Global Voices, Radu Ban and Vijayendra Rao, “Is Delibera- 73 June 29, 2009, http://globalvoicesonline. tion Equitable: Evidence from Transcripts of 174 | Voice and Agency empowering women and girls for shared prosperity Village Meetings in India,” Policy Research Institute of Management and Harvard Cen- Working Paper 4928, World Bank, Washing- ter for Population and Development Stud- ton, DC, 2009. ies, Working Paper 13, no. 1, 2003. Evans and Nambiar, Collective Action and 74 83 BRIDGE, “Budgets and Gender,” Gender and Women’s Agency. Development In Brief 12, Institute of Devel- opment Studies, Brighton, UK, 2002. Ricardo Gomez, Libraries, Telecenters, 75 Cybercafés and Public Access to ICT: Interna- 84 Susan A. Papp, Aparajita Gogoi, and Cath- tional Comparisons (Hershey, PA: IGI Global, erine Campbell, “Can Social Accountabil- 2012). ity Improve Maternal Health in India?” London School of Economics and Political Beyond Access, “Empowering Women and 76 Science, November 6, 2013, http://blogs. Girls through ICT at Libraries,” Issue Brief, lse.ac.uk/indiaatlse/2013/11/06/can- October 2012, http://www.intgovforum. social-accountability-initiatives-improve- or g /c m s/w k s2013/wor k shop _ maternal-health-in-india/. background_paper/268_1367875859.pdf. Papp et al., “Can Social Accountability Improve 85 Broadband Commission Group on Broad- 77 Maternal Health in India?”;and Tamara band and Gender, “Doubling Digital Oppor- Windau-Melmer, “A Guide for Advocating for tunities: Enhancing the Inclusion of Women Respectful Maternity Care,” Futures Group, and Girls in the Information Society,” Inter- Health Policy Project, Washington, DC, 2013. national Telecommunication Union and United Nations Educational, Scientific, and 86 Drude Dahlerup, “About Quotas,” Quota Cultural Organization, Geneva, 2013, 41. Project, International Institute for Democ- racy and Electoral Assistance and Stock- Kelly K. Hallman, Nora J. Kenworthy, Judth 78 holm University, Stockholm, 2009. Diers, Nick Swan, and Bashi Devnarain, “The Contracting World of Girls at Puberty: Vio- 87 United Nations Focal Point for Electoral lence and Gender-Divergent Access to the Assistance, “Promoting Women’s Electoral Public Sphere among Adolescents in South and Political Participation through UN Elec- Africa,” Poverty, Gender and Youth Working toral Assistance,” Policy Directive, United Paper no. 25, Population Council, New York, Nations, New York, December 24, 2013. 2013. 88 Julie Ballington, Randi Davis, Mireya Reith, Evans and Nambiar, “Collective Action and 79 Lincoln Mitchell, Carole Njoki, Alyson Women’s Agency.” Kozma, and Elizabeth Powley, Empower- ing Women for Stronger Political Parties: World Bank, 2013, “Women and Girls in 80 A Guidebook to Promote Women’s Political Haiti’s Reconstruction: Addressing and Pre- Participation (New York: United Nations venting Gender-Based Violence,” Evaluation Development Programme and National Report AUS2645, World Bank, Washington, Democratic Institute for International DC, 2013. Affairs, 2012), 31. Mansuri and Rao, Localizing Development . 81 89 Ballington et al., Empowering Women for World Bank, “Social Accountability: What 82 Stronger Political Parties. Does It Mean for the World Bank?,” in Social 90 Markham, “Women as Agents of Change.” Accountability Sourcebook (Washington, DC: World Bank, 2006); and Asha George, 91 For more information about EMILY’s List, “Accountability in Health Services Trans- see the organization’s website at http:// forming Relationships and Contexts,” Indian www.emilyslist.org/. Chapter 6 Amplifying Voices | 175 Markham, “Women as Agents of Change.” 92 Inter-Parliamentary Union, “Increased 95 Women’s Political Participation still Depen- Markham, “Women as Agents of Change.” 93 dent on Quotas, 2012 Elections Show,” press Pippa Norris, “Women’s Legislative Par- 94 release, March 5, 2013, http://www.ipu. ticipation in Western Europe,” Western org/press-e/pressrelease20130305.htm. European Politics 8, no. 4 (1985): 90–101; Richard E. Matland, “Enhancing Women’s Political Participation: Legislative Recruit- ment and Electoral Systems,’’ in Women in Parliament: Beyond Numbers, edited by Julie Ballington and Azza Karam, 93–111 (Stock- holm: International IDEA); Krook, “Reform- ing Representation.” Chapter 7 Closing gaps in data and evidence Key messages Box 7.1 > Lack of data on women’s agency is a challenge to be overcome, but it should not be an excuse for inaction in the meantime. > To measure progress and compare women’s agency across countries, more and better data and new measures are needed. > The internationally agreed minimum set of 52 gender indicators and the statistical indicators and guidelines for measuring violence against women provide a strong basis for improving the availability of data and evidence on women’s agency > Promising efforts to produce more national-level data on women’s agency includes Data2X, the Evidence on Data and Gender Equality initiative, and the Living Standards Measurement Study–Integrated Surveys on Agriculture. > Quantitative data need to be complemented by qualitative information. > Several promising new initiatives are combining data in composite indices to measure multiple agency deprivations across different domains. > At the operational level, project and program indicators need to establish baselines and more systematically track agency in key domains. > More rigorous evaluations of what works are needed, particularly around collective voice, normative change and the design of multi-sectoral programs. Chapter 7 Closing Gaps in Data and Evidence | 179 Monitoring change and into agency and agency constraints and assessing progress deprivations at the country level. Qualitative approaches can be designed and implemented across countries—as used by the World Bank’s On Norms and Agency and Voices of the “Data not only measures progress, it Poor studies, for example—or developed and inspires it.” applied to a particular country as in Niger. 2 —Hillary Rodham Clinton1 The focus of this chapter is more on the quan- titative indicators used to track progress and hold decision makers to account. Often, how- We wrap up by focusing explicitly on data ever, combining qualitative and quantitative and evidence. Progress in promoting wom- methods will provide the best way to measure en’s voice and agency needs to be captured agency and to provide the benchmarks that and monitored. Overcoming lack of data is can be used to track progress in increasing clearly a challenge, but it is not an excuse for agency, as, for example, in the Pathways to inaction. To measure progress and compare Empowerment Project.3 at the national level and across countries, more and better data must be produced for We first take stock of current efforts to some areas, and new measures are needed measure women’s agency. The chapter then for others. Efforts must also focus on increas- highlights key gaps that must be addressed to ing data quality, raising standards of data ensure a robust evidence base to inform pol- collection, and promoting a common under- icy making and finally outlines key priorities standing of agency measures and their defi- for future investment to address these gaps. nitions. Policy research efforts are needed to broaden the evidence base through rigorous Country-level data testing of interventions that explicitly aim to A range of data exists at the country level that enhance women’s agency. casts light on voice and agency. This section As chapter 1 highlighted, agency is a com- begins by giving an overview of internation- plex construct that is inherently difficult to ally agreed gender indicators and commonly measure. The data used in this report reflect available sources before highlighting gaps what people say about what they think and and then turning to some new and compos- do in different domains of their lives—their ite measures that can be used to provide a expressions of agency. While all data derived fuller picture. from self-reporting are subject to a degree of bias, the advantage of using reported behav- Internationally agreed gender iors rather than personal perceptions is the indicators greater objectivity and comparability across The 2013 United Nations Statistics Division samples and contexts. (UNSD) guidance on gender indicators gives This report has shown that both qualita- important new impetus, founded on broad- tive and quantitative data can provide insights based government agreement. The minimum 180 | Voice and Agency empowering women and girls for shared prosperity set of 52 agreed gender indicators cover eco- women, including physical and sexual vio- nomic participation, education, health, polit- lence as well as child marriage and female ical participation, and human rights. These genital cutting, and guidelines for producing quantitative measures are complemented by statistics on these and other measures of vio- 11 qualitative indicators covering national lence against women have been developed.5 norms and laws on gender equality that are A selection of the indicators most relevant to aimed at monitoring how effectively national measuring the agency deprivations covered legislation ensures gender equality.4 Specific in this report, including these violence indi- standardized measures of violence against cators, are in box 7.2. Box 7.2 Selected internationally-agreed indicators on agency Economic structures, participation in Health and related services, includes: productive activities, and access to ›› Contraceptive prevalence among women resources, includes: who are married or in a union, aged 15-49 ›› Proportion of population with access to credit, by sex Human rights of women and girl ›› Proportion of adult population owning land, children, includes: by sex ›› Proportion of women aged 15-49 subjected to physical or sexual violence in the last 12 ›› Proportion of individuals using the Internet, months by an intimate partner by sex ›› Proportion of women aged 15-49 subjected to ›› Proportion of individuals using mobile/ physical or sexual violence in the last 12 months cellular telephones, by sex by persons other than an intimate partner ›› Proportion of households with access to ›› Prevalence of female genital mutilation/ mass media (radio, TV, Internet), by sex of cutting (for relevant countries only) household head ›› Percentage of women aged 20-24 years old who were married or in a union before age 18 Education, includes: ›› Adjusted net enrolment in primary education ›› Adolescent birth rate by sex ›› Gross enrolment ratio in secondary education Public life and decision-making, by sex includes: ›› Women’s share of government ministerial ›› Gross enrolment ratio in tertiary education positions by sex ›› Proportion of seats held by women in ›› Share of female science, engineering, national parliament manufacturing, and construction graduates at tertiary level ›› Women’s share of managerial positions ›› Educational attainment (primary/secondary/ ›› Share of female police officers post-secondary/tertiary) of the population ›› Share of female judges age 25 years and older, by sex Source: UN Department of Economic and Social Affairs, Statistics Division, http://genderstats.org/. Chapter 7 Closing Gaps in Data and Evidence | 181 Collecting data on violence brings unique three optional topics (FGM/C, attitudes methodological challenges to ensuring con- towards VAW, and reporting to authorities/ sistency, accuracy, and quality while also help-seeking). This guidance includes the adhering to established ethical standards. nine internationally agreed core indicators Following an earlier request from the United on VAW in box 7.3. Nations General Assembly, in 2009, the UN The World Bank Gender Data Portal Statistics Commission (UNSC) adopted a includes comparable data for 144 develop- proposed core list of violence against women ing and 31 developed countries.6 This draws (VAW) indicators and requested UNSD and on sources that provide systematically and other stakeholders draw upon and further consistently compiled data, and shows elaborate existing methodological guide- that many of these internationally-agreed lines. In 2013, the UN published the Guide- and recommended indicators are not cur- lines for Producing Statistics on VAW, which rently collected and reported by countries. include four core topics (physical, sexual, Whereas most of the countries (136) in this psychological and economic violence) and sample collect sex-disaggregated data on Box 7.3 Internationally-agreed indicators for measuring violence against women 1. Total and age specific rate of women subjected to physical violence in the last 12 months by severity of violence, relationship to the perpetrator and frequency 2. Total and age specific rate of women subjected to physical violence during lifetime by severity of violence, relationship to the perpetrator and frequency 3. Total and age specific rate of women subjected to sexual violence in the last 12 months by severity of violence, relationship to the perpetrator and frequency 4. Total and age specific rate of women subjected to sexual violence during lifetime by severity of violence, relationship to the perpetrator and frequency 5. Total and age specific rate of ever-partnered women subjected to sexual and/or physical violence by current or former intimate partner in the last 12 months by frequency 6. Total and age specific rate of ever-partnered women subjected to sexual and/or physical violence by current or former intimate partner during lifetime by frequency 7. Total and age specific rate of ever-partnered women subjected to psychological violence in the past 12 months by the intimate partner 8. Total and age specific rate of ever-partnered women subjected to economic violence in the past 12 months by the intimate partner  9. Total and age specific rate of women subjected to female genital mutilation Source: United Nations, Guidelines for Producing Statistics on Violence against Women.6 182 | Voice and Agency empowering women and girls for shared prosperity access to credit, none have this data on the patterns of women’s agency at the country share of women who own land.8 Only 22 level and to compare outcomes across coun- developing countries collect disaggregated tries. Numerous surveys have incorporated data on Internet use, and none collect data relevant questions into their standard mod- on the female share of mobile phone or mass ules, particularly in the domains of intimate media users. On the other hand, collection partner violence, sexual and reproductive of health data is comparatively high—95 health, asset ownership, and economic developing countries report on contracep- activity. While these indicators typically tive prevalence. capture only proxy measures rather than a direct measure of agency, these surveys To date there is also very low uptake of offer a validated set of measures that have the recommended indicators of women’s been instrumental in providing a clearer and girls’ human rights. No countries sys- picture of women’s status across the globe. tematically collect data on the prevalence Box 7.4 presents a selection of key primary of physical and sexual violence, and only and secondary data sources, including those 19 developing countries report on female used for this report, and examples of the genital mutilation/cutting. Data collection agency measures they contain. on child marriage is also limited, although nearly all countries report on adolescent Together these represent a wealth of fertility (only six developing countries do publicly available data sources. However, not). In the domain of women’s voice, cover- gaps remain and the need for more and bet- age is mixed. Nearly all countries report on ter data is critical, as we discuss next. women’s share of ministerial positions (171 out of 175) and seats held in parliament Closing data and evidence gaps (170 out of 175). Yet fewer than half (70 out More investment is needed to produce of 175) report on women’s representation national-level data that can measure levels, in managerial positions. Data on the pro- patterns and changes in women’s agency. portion of judges and police officers who are Among the key gaps are measures of agency female are missing from the World Develop- that capture aspirations and self-efficacy, ment Indicators. The good news, however, is measures of control over household spend- that international efforts are under way to ing and investment decisions, measures of support countries to collect a wider range mobility, and measures of participation in of key gender-relevant data in the future, politics at the local level. The World Bank including those outlined in this chapter. Group, together with governments and a number of partner organizations, is working Available data and indicators to close these data gaps. We do not have a direct measure of agency, Data2X, a partnership among the United but rather a valuable set of proxy indica- Nations (UN) Foundation, the William and tors that measure reported experiences, Flora Hewlett Foundation, and the United attitudes and behaviors. Standardized mea- States government, is producing a Gender sures can be used to establish levels and Data Blue Print to prioritize gender data Chapter 7 Closing Gaps in Data and Evidence | 183 Box 7.4 Selected data sources and examples of measures of agency Country Data source coverage Frequency Measures of agency, examples Demographic and 90+ Typically Unmet need for contraceptives Health Surveys every five Experience of physical and sexual violence years Gallup World Poll 160 Varies by Support for women’s work and rights country (every Support for women in leadership positions one to two years) World Values Survey 75 Typically Perceptions about severity of discrimination every five Justifications for wife beating years Regional Barometer 88 Every two to Support for women in leadership positions Surveys three years Support for women’s rights Women, Business, 143 Every two Legislation to addresses domestic violence and the Law years Equal weight of woman’s testimony in court Inter-Parliamentary 186 Updated Women in elected and ministerial positions Union regularly Portfolios held by women ministers WHO Multi-Country 10 One time Experience of physical and sexual intimate Study of Violence partner violence Against Women Reporting of experiences of intimate partner violence Women’s belief in right to refuse sex WomanStats 175 Updated Women’s property rights Database regularly Presence of laws against domestic violence Multiple Indicator 90 Typically Unmet need for contraceptives Cluster Surveys every five Justification for wife beating years Female genital mutilation prevalence OECD Social 102 Every two to Child marriage Institution and three years Women’s legal rights to land and other Gender Index property Reproductive Health 33 Country Unmet need for contraceptives Survey dependent Fertility preferences Gender and Land 80 Updated Land owned by women Rights Database regularly Women’s property rights 50 Years of Women’s 100 Annually Equal property rights (married and unmarried) Legal Rights (1960–2010) Equal rights to get a job or pursue a profession Database IFC Enterprise 135 Varies by Female participation in firm ownership Surveys country Female managers LSMS–ISA 7 Varies by Female access to credit country Household decision-making power distribution 184 | Voice and Agency empowering women and girls for shared prosperity Box 7.5 Analysis of current data gaps and possible ways forward Expression of agency Gap Possible way forward Freedom from Current prevalence Create a global initiative to operationalize violence estimates are the UNSD guidelines to collect survey data underestimates systematically at the country level and the IAEG- and are not GS-endorsed minimum standard indicators. collected through Expand the number of countries collecting the standardized minimum indicators (through stand-alone surveys methods. like the WHO study or integrated modules as in the DHS). Invest in panel data to track progress in reducing violence and measuring related consequences. Integrate violence modules or dedicated surveys in core post-conflict programming and research. Sexual and Indicators are limited Expand existing surveys (DHS, RHS) to cover reproductive to maternal and services other than reproductive and MCH health and rights reproductive health services (for example, STI/HIV screening and services. Reliable prevention, treatment of obstetric fistula) and more data on many critical systematically gather and automate health service aspects of health provider data. are not available Expand coverage of existing survey instruments because of weak to include SRH indicators for adolescents, statistical and civil complemented by data from mobile phones and registration systems. other new technology. Collect data on factors that shape demand: accessibility, affordability, and appropriateness of services. Access to land Asset ownership is Include measures of ownership at the individual typically collected level that also consider how assets are actually at household (not controlled or shared within the household. individual) level. Track both official and customary property laws. Where data do exist (such as LSMS), they may not be consistent across countries. Voice Comparable sex- Expand data capabilities to get accurate disaggregated information on women’s representation at data on women’s subnational levels and in political party leadership. representation at Advocate to election management bodies for the subnational data on voter registration and turnout to be level and voter sex-disaggregated. registration and turnout are not widely available. Chapter 7 Closing Gaps in Data and Evidence | 185 gaps.9 To date, the partnership has identified Women and other agencies to pilot the nine 26 gaps across five domains, based on the internationally agreed core VAW indicators need, coverage, and policy relevance. Pri- in several countries using the stand-alone mary data gaps identified include violence survey promoted by the UNSD guidelines. against women, sexual and reproductive The Evidence on Data and Gender Equal- health, access to land, and voice.10 ity (EDGE) Initiative aims to accelerate Drawing on this initial work by Data2X, efforts to collect comparable gender indi- Box 7.5 highlights the gaps and opportuni- cators on health, education, employment, ties for improved data collection across var- entrepreneurship, and asset ownership. ious expressions of women’s agency. This three-year project—a collaboration between UN Women and UNSD with the Efforts are already underway to fill some OECD and the World Bank—is building on of the identified gaps. More than 70 coun- the work of the Inter-agency and Expert tries have conducted studies specifically to Group on Gender Statistics (IAEG-GS) to measure the prevalence and nature of var- develop methodological guidelines to mea- ious forms of violence against women, at sure asset ownership and entrepreneurship least 40 of these at the national level. Some from a gender perspective. Improving ques- focus on IPV, while others were broader and tionnaires and surveys can help countries to encompassed a wide range of physical and collate data at an individual level. For exam- sexual violence and threats of violence per- ple, determining which household members petrated by partners, other family members, are going to be the subject of the survey can other known men, and strangers. A num- produce more nuanced data on control over ber of cross-country efforts have also been assets. To fill the gaps, EDGE will pilot sur- carried out, including the landmark World vey modules as part of existing household Health Organization (WHO) Multi-country questionnaires.11 EDGE is also working on Study on Women’s Health and Domestic the standardization of surveys, with some Violence against Women. Of note also are regional and country specificity.12 the Demographic and Health Surveys (DHS) implemented by ICF International in part- The Living Standards Measurement nership with national governments and with Study—Integrated Surveys on Agriculture funding from U.S. Agency for International (LSMS–ISA), a collaboration between the Development (USAID), the International World Bank and the Bill and Melinda Gates VAW Survey coordinated by the European Foundation, is improving the type and qual- Institute for Crime Prevention and Control, ity of household data collected by statistical and the European Union Agency for Funda- offices and is currently being implemented mental Rights EU-Wide Study on Violence in seven Sub-Saharan African countries.13 Against Women Study. However, a gap still Several existing survey instruments could remains in the collection of internationally be improved by adding questions to yield comparable data using standardized mea- richer sex-disaggregated data. Beyond surements. To help close this gap, the World asking if anyone in the household owns Bank is exploring partnerships with UN any agricultural land, which is common in 186 | Voice and Agency empowering women and girls for shared prosperity surveys, the next question could ask who the Continual improvement in the capacity of owners are. The names of all persons on the statistical agencies in developing countries is ownership document could be listed. Ques- essential to ensure sustainability and scaling tions on the respondent’s right to sell, rent, up of data collection. The World Bank has a or bequeath land, and their decision-making number of financial instruments that offer power over its use, provide data on the level potential support for improved measures of of individual control.14 voice and agency. In particular, the Trust Fund for Statistical Capacity Building, which primar- Innovative ways of collecting and using ily supports targeted capacity improvements sex-disaggregated data can help to illumi- at the national level for low income countries, nate key challenges and priority areas for will devote more focus and attention to filling action. Across seven countries in Europe critical gaps in gender statistics. and Central Asia for example, the World Bank is working with the Food and Agri- Additionally, new indicators on agency culture Organization to implement the are needed. In order to select new indica- Voluntary Guidelines on the Responsible tors, consultation with a broad range of Governance of Tenure of Land, Fisheries experts, stakeholders and national statisti- and Forests in the Context of National Food cal offices and agreement between govern- Security through gender action plans.15 ments and the relevant intergovernmental These include collecting sex-disaggregated agencies will be needed to select, test, and data (a) by recording an applicant’s gender evaluate options and methodologies. The at the time of submission, as in Kosovo and development and testing process can begin Albania; (b) as part of other data collected at the project or service–provider level. For by the land agencies, such as personal iden- example, work for this report suggests new tifier, tax number, and social security num- areas to consider for measuring freedom of ber; and (c) by linking the property register violence. Those areas include (a) measures with other government registers contain- of violence against children and adolescents ing gender information, such as the civil or (b) consistent and culturally appropriate population register. Innovative uses include measures of attitudes toward violence, linking sex-disaggregated land data to spa- including masculinity and justifications for tial data to promote more targeted policy different forms of violence; (c) measures making and implementation.16 that capture access to justice and protective services for survivors of violence; and (d) UN Women has partnered with United measures of access to and quality of social Cities and Local Governments to launch data services such as alternative accommodation gathering in the area of women’s represen- and livelihood support. tation at subnational levels and in political party leadership.17 This effort will provide Greater understanding of agency over more accurate information on women’s sexual and reproductive health decisions political participation and the barriers could be gained from knowing more about women face to exercising agency in this sexual autonomy, including experiences domain. of forced or unwanted sex and pregnancy Chapter 7 Closing Gaps in Data and Evidence | 187 intentions, including whether respondents and leadership positions in the public and believe they have a right to determine the private sectors would help to better under- timing and number of pregnancies they stand changes in women’s voice and influ- have and what the reasons are for unin- ence in specific national contexts. tended and mistimed pregnancies. Control of land and housing could be better mea- Policy research efforts are needed to sured with more data on women’s registra- deepen the evidence base through rigorous tion of usufruct or ownership rights to land testing of interventions that explicitly aim to and on women’s use of land administration enhance women’s agency. Efforts are under- services. Specific measures are also needed way, for example, through the Social Obser- that can track implementation of laws to vatory in India, which documents innova- make it possible for women to register land tions and promising practices18 and through and property in their own names or jointly. work supported by the World Bank Group’s Tracking women’s appointments to high- Umbrella Facility for Gender Equality (UFGE) level positions in national organizations in Latin America and the Caribbean (box 7.6). Box 7.6 Measuring and expanding agency in Latin America and the Caribbean In Latin America and the Caribbean ”(LAC), as in other regions, lack of rigorous measures for agency and low awareness of the importance of agency create serious impediments to its advancement through programs and policy initiatives. To fill this gap, the World Bank Group’s Umbrella Facility for Gender Equality (UFGE) supports initiatives aimed at 1) expanding data, indicators, and evidence on agency and its effects on other dimensions of gender equality and 2) road testing evidence on how to expand agency. In all projects supported by LAC UFGE, agency is either one of the outcomes or plays an instrumental role in achieving other project objectives. Some examples are as follow: ›› In Ecuador, the Text Me Maybe pilot intervention aims to increase the evidence-base on the role of agency in policy interventions that address teen pregnancy prevalence and risk factors and road test the effectiveness of text message reminders in changing behaviors. ›› In St. Lucia, UFGE funds support efforts to reduce women’s vulnerability in natural disasters by enhancing their agency, especially for female heads of households, through climate adaptation activities. ›› In Northeast Brazil, a pilot intervention to enhance women’s agency in agricultural production integrates tools to measure their empowerment relative to productivity, income-generation, access to markets, and gender roles and decision-making power within the household, producer groups, and the broader community. LAC UFGE also supports capacity building for project teams in conceptualizing and measuring agency, including through a recent workshop to review existing measures of various dimensions of agency and resources for addressing empirical challenges. The first results from this work will be available in September 2014. For more information about the UFGE, see http://www.worldbank.org/en/topic/gender/publication/ umbrella-facility-for-gender-equality. 188 | Voice and Agency empowering women and girls for shared prosperity Composite indices Oxford Poverty and Human Development Initiative to measure women’s own sense of We do not have a single measure that effi- agency.21 Box 7.7 outlines the approach and ciently captures what is meant by agency shows some early findings from Chad where and the expansion of choice. Choice is quite the index is being tested. different from income poverty, for exam- ple, where information about the amount Another important contribution is the of money that an individual or family has is Social Institutions and Gender Index (SIGI) compared to some threshold. Indeed, as our database, developed by the OECD. The SIGI figure 1.3 in chapter 1 showed, the ability to focuses on the underlying social institu- make choices often varies across the differ- tions that influence gender roles and that ent expressions of agency. Combining data can limit or enable individual or collective in an index allows different and sometimes agency. The SIGI is also a composite index, overlapping deprivations to be brought composed of five equally weighted sub-in- together to produce a composite and stan- dices that measure discriminatory family dardized value. However, selecting the most code; restricted physical integrity; son bias; restricted resources and entitlements; and appropriate items to capture these syner- restricted civil liberties.22 gies in a single measure without losing the meaning of the underlying measures is a sig- These types of composite measures can nificant challenge. offer important contributions to under- standing the status, nature, and constraints Several recent initiatives seek to meet on women’s agency at the national level. these challenges. The Women’s Empower- The next section explores recent efforts to ment in Agriculture Index (WEAI), which measure progress in women’s agency at has been piloted in Bangladesh, Guatemala, the program and project levels and pres- and Uganda, measures women’s empower- ents suggestions for innovations on these ment, agency, and the inclusion of women measures. in the agriculture sector.19 It comprises two sub-indices: the first measures five domains Program and project-level of empowerment—namely, agricultural pro- indicators and emerging duction, resources, incomes, leadership and good practice time. The second, the Gender Parity Index, At the operational level, indicators are measures the relative inequality between needed to establish baselines and track the primary adult male and female in each progress. Many governments and interna- household. The results can be used to tional development agencies have devel- increase understanding of the connections oped frameworks and guidelines to mon- between women’s empowerment, agricul- itor gender results in project activities. tural growth, and food security, and is being The voice and agency front is one in which rolled out in 19 countries.20 experience and measurement guidance are The Relative Autonomy Index (RAI) is now beginning to emerge. For example, another new approach, developed by the USAID has issued a checklist, “Assessing Chapter 7 Closing Gaps in Data and Evidence | 189 Box 7.7 Using new measures of women’s autonomy in Chad: The Relative Autonomy Index Oxford Poverty and Human Development Initiative “We know about malnutrition, but if (OPHI) is testing the Relative Autonomy Index the meat doesn’t go mostly to the man, (RAI) with men and women and between different there is trouble in the house.” women in the same household in Chad, which has some of the largest gendered economic and health —A woman from Loumia, Chad inequalities in the world. OPHI, in collaboration with United Nations Children’s Fund–Chad and the Indic Society for Education and Development, used the RAI within a nationally representative survey, which addressed topics such as consumption, work, living standards, education, and subjective well- being, with a focus on health behaviors and outcomes. To measure autonomy, people were asked whether their actions were coerced to some extent, or done to please others, and whether their actions embodied the respondents’ own values. Their degree of autonomy was analyzed for their actions in eight domains—including domestic work, market work, making major household purchases, participating in groups, and feeding young children. The most striking finding is that women are less autonomously motivated than men across every single domain. Women ages 15-26 had the lowest autonomy, in sharp contrast to men of the same age. Uneducated men had greater autonomy than uneducated women in all domains (except group membership), but men and women who participated in a group had higher autonomy in all other domains without exception. However, results show that autonomy is not necessarily associated with more life satisfaction or happiness. Levels of education and household income—commonly used indicators of agency—are not strongly associated with autonomy levels either. Given these additional insights, the results suggest that the RAI could be used to enhance or replace other indirect measures of agency. These early findings suggest the RAI yields new information that can contribute to understanding the linkages between women’s autonomy, agency, and development outcomes. Further analysis will explore the relationship between the individual’s autonomy and specific behavior such as hand washing, exclusive breastfeeding, and child marriage. In Chad, early marriage is widely accepted— and 1 in 10 girls has had a child before age 15.b Measuring autonomy in these ways, therefore, could help to identify high-impact pathways that empower multiple generations within the same household. Source: Kristof (2013); Alkire et al., forthcoming; Ford, (2013).23 Achievement of Gender Objectives,” which and measurement of results of programs to includes women’s self-efficacy, attitudes address deprivations of agency, including toward gender-based violence and women’s violence against women.25 access to opportunities, and laws promot- ing gender equality.24 The United Kingdom’s The World Bank Group monitors key gen- Department for International Development der indicators as part of its corporate score- has published guidance notes on the design card and is increasingly monitoring gender 190 | Voice and Agency empowering women and girls for shared prosperity results in project activities. Some useful RMS (covering fiscal years 2015 to 2017) proxy measures of agency are included. This includes the number of women receiving section briefly reviews recent Bank experi- antenatal care as a key health sector indica- ence and highlights areas for improvement. tor in relevant client countries. Since 2009, standardized Core Sector Indicators (CSIs) Experience at the World Bank gather data on a uniform set of indicators Group at the project level, to enable results to be reported at the corporate level. Among the In the Corporate Score Card and the Inter- more than 140 CSIs used across the Bank, national Development Association (IDA) about 30 are gender specific, tracking Results Measurement System (RMS), the female beneficiaries of projects; gender-spe- World Bank Group monitors key gender cific results track agriculture, conflict pre- indicators and outcomes in areas where vention, civic engagement, social inclusion, client country policies and operations are access to finance and health, among others being supported. For example, the IDA’s 17 (see examples in box 7.8). However, as CSIs Box 7.8 Selected World Bank Group Gender Core Sector Indicators related to agency Agency dimension Gender Core Sector Indicators Freedom from gender-based Beneficiaries who experience a feeling of greater security violence attributable to the project in the project areas —female (number) Control over reproductive People with access to a basic package of health, nutrition, or health and rights reproductive health services—(number) Ownership and control over Land parcels with use or ownership rights of females recorded as a land result of the project—(number) Representatives in community-based decision making and Voice and influence management structures that are from the vulnerable or marginalized beneficiary population—female (number) Vulnerable and marginalized people who participate in non-project Voice and influence consultations and decision-making forums—female (number) Targeted clients who are members of an association—female Voice and influence (number) Women’s employment figures and corporate board positions Voice and influence awarded to women— (percentage) (IFC indicator) Small and micro-enterprise capital given to women Voice and influence entrepreneurs— (percentage) (IFC indicator) All dimensions depending Direct project beneficiaries—(number), of which (percentage) are on the project female Chapter 7 Closing Gaps in Data and Evidence | 191 are not yet widely incorporated across Bank More generally, World Bank Group proj- projects,26 additional efforts are needed ects include specific results indicators. to increase their uptake and expand their Efforts to capture agency in such projects scope for measuring agency. are illustrated in box 7.9. Box 7.9 Examples of World Bank Group project indicators related to women’s agency Gender-based violence ›› The 2008 Protection from gender-based violence in Côte d’Ivoire Project measured the percentage of the surveyed population that listened to the project’s radio shows about GBV at least two times and the number of women experiencing physical intimate partner violence. ›› The 2012 Breaking the Cycle of Violence Project in Honduras included measures of the number of youth whose awareness about GBV increased over the course of the project. Sexual and reproductive health and rights ›› The 2010 Nepal Second Health Nutrition and Population and HIV/AIDS Project included the following agency-related indicators: the current use of modern contraceptives among women ages 15–49 and the percentage of married women ages 15–49 with unmet need for family planning. ›› The 2014 Tajikistan Health Services Improvement Project included contraceptive prevalence rates as a results indicator, measured as the share of women in project districts ages 15–49 using modern methods of family planning. Control over land and housing ›› The 2010 Kosovo Real Estate Cadastre and Registration Project incorporated an indicator on the percentage of women with use of or ownership rights registered (both joint and individual). ›› The 2007 Vietnam Poverty Reduction and Support Operation Credit (PRSC-6) measured asset ownership among women. Voice and influence ›› The 2010 Afghanistan Third Emergency National Solidarity Project included targets for the proportion of female representatives in the Community Development Councils taking active part in making decisions for community development. ›› The 2010 Bangladesh Empowerment and Livelihood Improvement “Nuton Jibon” Project monitored the percentage of village-level Executive Committees (Gram Samitis) where the chronically poor hold decision-making positions and the percentage of these who are women. 192 | Voice and Agency empowering women and girls for shared prosperity *** Capturing and monitoring efforts to promote women’s voice and agency requires high quality data, but data gaps remain significant. Overcoming these gaps will require greater investment and concerted effort from governments, international agencies, and local and national implementing partners. As highlighted in this chap- ter, several initiatives are already underway to develop and standardize indicators that better measure agency and allow for more direct comparison across countries. These include Data2X, EDGE, LSMS–ISA, in addition to recent international agree- ment on minimum gender indicators and guidelines for collecting data on violence against women. Some research efforts are underway to design and test new mea- sures of agency and deepen the evidence base for policy through rigorous test- ing of interventions that explicitly aim to enhance women’s agency, but more are needed. Together these efforts will contribute to closing important data and knowl- edge gaps on agency and will serve the priorities of the post-2015 agenda, which will increase demand for data to measure advancements in women’s empowerment and our progress towards gender equality. Endnotes Women (New York: United Nations, 2013), http://unstats.un.org/unsd/gender/docs/ 1 Hillary Rodham Clinton, remarks at “Evi- Guidelines_Statistics_VAW.pdf. dence and Impact: Closing the Gender Data 7 Analysis is based on data available on the Gap,” hosted by the U.S. Department of State World Bank Gender Data Portal and includes and Gallup, Washington, DC, July 19, 2012. gender datasets from the United Nations com- 2 World Bank, Gender, Agency, and Economic piled by its regional commissions and sectoral Development in Niger (Washington, DC: agencies, as well as surveys and reports con- World Bank, forthcoming). ducted or funded by the World Bank Group, 3 See Naila Kabeer, “Contextualising the such as World Development Report 2012: Economic Pathways of Women’s Empow- Gender Equality and Development (Washing- erment: Findings from a Multi-country ton, DC: World Bank, 2012). Other in-country Research Programme,” Pathways Policy sources may also be available. Users are rec- Paper, Institute of Development Studies, ommended to check data sources. University of Sussex, Brighton, U.K., 2011. 8 For data availability of the core list of gen- der indicators by International Devel- 4 The full listing of indicators can be found at opment Association, International Bank http://genderstats.org/. for Reconstruction and Development, 5 United Nations Department of Economic and Organisation for Economic Co-op- and Social Affairs, Guidelines for Produc- eration and Development countries, see ing Statistics on Violence against Women http://datatopics.worldbank.org/gender/ (New York: United Nations), http://unstats. monitoring-progress. un.org/unsd/gender/docs/Guidelines_ 9 For more information about Data2X, see the Statistics_VAW.pdf. description on the UN Foundation’s website 6 Please add new endnote which has the full at http://www.unfoundation.org/what- reference: United Nations, Guidelines for we-do/issues/women-and-population/ Producing Statistics on Violence against data2x.html. Chapter 7 Closing Gaps in Data and Evidence | 193 Marya Buvinic, Rebecca Furst-Nichols, and 10 WEAI was developed by the U.S. Agency 19 Gayatri Koowal, “Data 2X Mapping Gen- for International Development, Inter- der Data Gaps,” UN Foundation, New York, national Food Policy Research Institute, October 2013, https://app.box.com/s/ and Oxford Poverty and Human Devel- amtbqh6a99ywzyjxub9c. opment Initiative. WEAI is available Cheryl Doss, ”Technical Report: Measuring 11 at http://www.ifpri.org/publication/ Individual Level Asset Ownership and Con- womens-empowerment-agriculture-index. trol,” draft report, EDGE Initiative, Gronin- Feed the Future, “Women’s Empower- 20 gen, Netherlands, 2013. ment in Agriculture Index,” International United Nations Statistical Division (UNSD) 12 Food Policy Research Institute, Washing- and UN Women, “Evidence and Data for ton, DC, 2012, http://feedthefuture.gov/ Gender Equality (EDGE) Project: Report of sites/default/files/resource/files/weai_ the Follow-up Meeting on Measuring Asset brochure_2012.pdf. Ownership from a Gender Perspective,” The RAI was first designed by Richard Ryan 21 UNSD and UN Women, New York. and Ed Deci. The countries are Ethiopia, Malawi, Mali, 13 Quotation is from Nicholas D. Kristof, 22 Niger, Nigeria, Tanzania, and Uganda. “Women as a Force for Change,” New York Cheryl Doss, Carmen Diana Deere, Abena D. 14 Times, July 31, 2013, http://www.nytimes. Oduro, and Hema Swaminathan, 2013. “Col- c o m / 2 0 1 3/0 8/0 1/o p i n i o n / k r i s t o f- lecting Sex-Disaggregated Asset Data,” Gen- women-as-a-force-for-change.html. See der Asset Gap Project Policy Brief 4, Centre also Johnny Cruz, “WV in Chad: NYT’s Nich- for Public Policy, Indian Institute of Manage- olas Kristof, Babies, and Changing Lives,” ment, Bangalore, 2013. World Vision Magazine, July 19, 2013, http://worldvisionmagazine.org/story/ Food and Agriculture Organization of the 15 w v-chad-ny ts-nicholas-kristof-babies- United Nations (FAO), Governing Land for and-changing-lives. Sabina Alkire, Pierre Women and Men: A Technical Guide to Sup- Pratley, and Ana Vaz, Women’s Autonomy port the Achievement of Responsible Gen- in Chad: Measurement and Distinctiveness, der-Equitable Governance Land Tenure, Women’s Voice, Agency, and Participation Governance of Tenure Technical Guide 1 Research Series (Washington, DC: World (Rome: FAO, 2013), http://www.fao.org/ Bank, forthcoming); Liz Ford, “Child Mar- fileadmin/user_upload/landright/docs/ riage Could Trigger Surge in Africa’s Under- Technical_Guide.pdf. 15s Pregnancy Rate: UN,” Guardian, Octo- World Bank, “Land and Gender: Improving 16 ber 30, 2013, http://www.theguardian. Data Availability and Use in the Western com/global-development/2013/oct/30/ Balkans,” background paper for the Annual child-marriage-africa-teenage-pregnan - World Bank Conference on Land and Pov- cy-un. erty, Washington DC, March 25, 2014. Organisation for Economic Co-opera- 22 Buvinic et al., “Data 2X Mapping Gender 17 tion and Development (OECD) Develop- Data Gaps.” ment Centre, “2012 SIGI: Social Institu- Details about the Social Observatory can be 18 tions and Gender Index—Understanding found in the World Bank brochure, “Social the Drivers of Gender Inequality,” Paris: Observatory Rural Livelihoods Project in OECD Development Centre, 2012, http:// India,” at http://www.worldbank.org/ w w w.gender index .org/sites/default/ content/dam/ Worldbank/doc ument/ files/2012SIGIsummaryresults.pdf. SAR/SO.pdf. U.S. Agency for International Development 23 194 | Voice and Agency empowering women and girls for shared prosperity (USAID), “Gender Equality and Female 26 For example, a recent analysis of the port- Empowerment: Policy,” USAID, Washington, folio in the World Bank Sustainable Devel- DC, 2012. opment Network shows that CSI uptake U.K. Department for International Develop- 25 is found in less than half (46 percent) of ment (DFID), “A Theory of Change for Tack- projects. ling Violence against Women and Girls,” CHASE Guidance Note 1, Conflict, Humani- tarian, and Security Department, DFID, Lon- don, 2012, http://www.gadnetwork.org. uk/storage/VAWG_guidance1_toc1.pdf. Annex I Bckground Papers | 195 Annex I: Background Papers Alison Evans and Divya Nambiar, “Collective Action and Women’s Agency: A Background Paper,” Women’s Voice, Agency and Participation Research Series 2013, No. 4, World Bank, Washington, DC, 2013. Carole Rakodi, “Expanding Women’s Access to Land and Housing in Urban Areas,” Women’s Voice, Agency and Participation Research Series 2014, No. 8, World Bank, Washington, DC, 2014. Diana J. Arango, Matthew Morton, Floriza Gennari, Sveinung Kiplesund, and Mary Ellsberg, “Interventions to Prevent and Reduce Violence Against Women and Girls: A Systematic Review of Reviews,” Women’s Voice, Agency and Participation Research Series, World Bank, Washing- ton, DC, forthcoming. Evelyn Namubiru-Mwaura, “Land Tenure and Gender: Approaches and Challenges for Strength- ening Rural Women’s Land Rights,” Women’s Voice, Agency and Participation Research Series 2014, No. 6, World Bank, Washington, DC, 2014. Karen A. Grepin and Jeni Klugman, Closing the Deadly Gap between What We Know and What We Do (Washington, DC: World Bank, 2013). Nata Duvvury, Aoife Callan, Patricia Carney, and Srinivas Raghavendra, “Intimate Partner Vio- lence: Economic Costs and Implications for Growth and Development,” Women’s Voice, Agency and Participation Research Series 2013, No. 3, World Bank, Washington, DC, 2013. Paul J. Fleming, Gary Barker, Jennifer McCleary-Sills, and Matthew Morton, “Engaging Men and Boys in Advancing Women’s Agency: Where We Stand and New Directions,” Women’s Voice, Agency and Participation Research Series 2013, No. 1, World Bank, Washington, DC, 2013. Rangita de Silva de Alwis, “Women’s Voice and Agency: The Role of Legal Institutions and Women’s Movements,” Women’s Voice, Agency and Participation Research Series 2014, No. 7, World Bank, Washington, DC, 2014. Ruti Levtov, “Addressing Gender Inequalities in Curriculum and Education: Review of Liter- ature and Promising Practices,” Women’s Voice, Agency and Participation Research Series, World Bank, Washington, DC, forthcoming. Sabina Alkire, Pierre Pratley, and Ana Vaz, “Women’s Autonomy in Chad: Measurement and Distinctiveness,” Women’s Voice, Agency and Participation Research Series, World Bank, Washington, DC, forthcoming. Seema Vyas “Estimating the Association between Women’s Earnings and Partner Violence: Evidence from the 2008–2009 Tanzania National Panel Survey” Women’s Voice, Agency and Participation Research Series 2014, No. 2, World Bank, Washington, DC, 2013. 196 | Voice and Agency empowering women and girls for shared prosperity Shaha Riza, “Women’s Collective Action in the Middle East and North Africa,” mimeo, 2013. Susan Markham, “Women as Agents of Change: Having Voice in Society and Influencing Policy,” Women’s Voice, Agency and Participation Research Series 2013, No. 5, World Bank, Washing- ton, DC, 2013. World Bank, Republic of Niger: Gender, Agency and Economic Development in Niger (Washing- ton, DC: World Bank, forthcoming). World Bank and TrustLaw Connect, Women and Land Rights: Legal Barriers Impede Wom- en’s Access to Resources, 2013, available at: http://www.trust.org/contentAsset/raw- data/1440891e-13ac-434a-bc73-9264e9aabbbf/file Voice Agency and “The World Bank Group’s twin goals of ending extreme poverty and boosting shared prosperity demand no less than the full and equal participation of women and men, girls and boys, around the world.” —World Bank Group President Jim Yong Kim “The world is changing beneath our feet and it is past time to embrace a 21st Century approach to advancing the rights and opportunities of women and girls at home and across the globe.” —Former US Secretary of State Hillary Clinton, remarks at 2013 Women in the World Summit www.worldbank.org/gender/agency