74191 CONFERENCE EDITION On Norms and Agency Conversations about Gender Equality with Women and Men in 20 Countries Ana Maria Munoz Boudet, Patti Petesch, and Carolyn Turk with Angelica Thumala Supported by © 2012 International Bank for Reconstruction and Development / The World Bank 1818 H Street NW Washington DC 20433 Telephone: 202-473-1000 Internet: www.worldbank.org This work is a product of the staff of The World Bank with external contributions. The findings, interpretations, and conclusions expressed in this work do not necessarily reflect the views of The World Bank, its Board of Executive Directors, or the governments they represent. Rights and Permissions The material in this work is subject to copyright. Because The World Bank encourages dissemination of its knowledge, this work may be reproduced, in whole or in part, for non-commercial purposes as long as full attribution to this work is given. Any queries on rights and licenses, including subsidiary rights, should be addressed to the Office of the Publisher, The World Bank, 1818 H Street NW, Washington, DC 20433, USA; fax: 202-522-2422; e-mail: pubrights@ worldbank.org. Contents FOREWORD 5 ACKNOWLEDGMENTS 6 ABOUT THE AUTHORS 9 ABSTRACT 11 INTRODUCTION: THE NORMS OF POWER AND THE POWER OF NORMS 12 1. The study approach 14 2. Methodology of the study 16 3. Discussing and researching gender equality: A brief introduction to the primary study concepts 19 a. Power, empowerment, and agency 20 b. Gender inequality in agency 23 4. Creating and enforcing gender through norms, roles and beliefs 24 a. Norms and roles 24 b. Reinforcing norms 25 c. Changing norms 26 5. Overview of chapters 26 References 28 PART I: GENDER NORMS 32 CHAPTER 1 THE RULES WE LIVE BY: GENDER NORMS AND IDEAL IMAGES 36 1. Normative frameworks for household gender inequalities 36 a. The good wife 37 b. The good husband 41 2. The good girl, the good boy 42 3. Community-level views of gender norms 44 a. Views on working mothers 45 b. The good woman (not) in public 47 CHAPTER 2 NEGOTIATING THE NORMS THAT BIND: A WINDING ROAD 50 1. The quiet relaxing and changing of norms 51 2. Timing is all: Negotiating opportunities and gender-specific responsibilities 52 a. Incessant activity 53 b. The rise of the Swedish husband: An emerging male double-burden? 54 c. Open dialogue and emergence of household power sharing 56 3. Intergenerational transmission of the possibility of change 57 4. Gender norms in transition 59 5. “A woman should be beaten if she deserves punishment� 62 a. Extent and forms of domestic violence 63 b. Causes and consequences of domestic violence 66 References 68 PART II: HAVING AND MAKING CHOICES 71 CHAPTER 3 STRATEGIC LIFE DECISIONS: WHO HAS THE FINAL SAY? 73 1. Investing in education: Why should girls and boys go to school? 75 2. Why should I leave school? Not my choice! 78 3. From school to work: Getting the first job 81 a. You say, I say: The weight of adult voices on job decisions 82 4. “First comes love, then comes marriage, then comes baby in a baby carriage� 83 a. So he proposed... 85 b. How can a child take care of another child? 87 c. How many children is too many? 88 d. Pink or blue? Girl or boy? 91 5. What is mine is yours: Asset control and decision-making 92 a. Pocket money 95 6. When does choice mean agency? 96 References 98 PART III: EMPOWERMENT 100 CHAPTER 4 WHAT DRIVES AGENCY? WHAT CRUSHES IT? 103 1. Step by step: Climbing the “Ladder of Freedom and Power� 103 a. Defining the steps on the ladder 104 b. A representative ladder 106 2. Perceptions of factors shaping agency 109 3. Combinations of mobility factors 113 a. Moving up the ladder in the city 115 b. Moving up the ladder in the countryside 117 c. Women climbing, men tumbling in the city 119 4. Men’s and women’s interdependent agency and gender norm change 121 References 123 CHAPTER 5 STRUCTURES OF OPPORTUNITY AND STRUCTURES OF CONSTRAINT 125 1. Community factors that fuel agency 126 2. Whose jobs? 128 3. It takes a village: Local economic dynamism and empowerment 131 a. When choices disappear or stay poor 132 b. When choices grow 135 4. Impact of laws and local civic action on empowerment 137 a. Equal rights for women? 138 b. Do institutions work? Where to turn for help 139 c. Whose voice counts? Local political leadership and civic action 141 5. Change women need 143 References 146 FINAL THOUGHTS 148 APPENDIX 152 Foreword The 2012 World Development Report: Gender Equality and Development offered a critical message: that ef- fective policy making and unwavering focus on progress and persistence in achieving gender equality matter greatly for beneficial development outcomes. In the past quarter century, we have seen remarkable gains for women. Women now represent 40 percent of the global labor force. Women are living longer than men all around the world. And gender gaps in education, once prevalent, are reversing with increasing enrollment of girls and young women. But even with this progress, gender disparities still persist in access to opportunity and resources, and in terms of individual agency. This World Bank report On Norms and Agency: Conversations about Gender Equal- ity with Women and Men in 20 Countries, provides tremendous insight on gender norms - an area that has been resistant to change, and which constrains achievement of gender equality across many diverse cultures. The report synthesizes data collected from more than 4000 women and men in 93 communities across 20 countries. It is the largest data set ever collected on the topic of gender and development, providing an unprecedented opportunity to examine potential patterns across communities on social norms and gender roles, pathways of empowerment, and factors that drive acute inequalities. The analysis raises the profile of persistent social norms and their impact on agency, and catalyzes discourse on the many pathways that create opportunities for women and men to negotiate transformative change. The report is underpinned by the fact that arguably the single most important contribution to development is to unleash the full power of half the people on the planet – women. It underscores how crucial making invest- ments in learning, supporting innovations that reduce the time costs of women’s mobility, and developing a critical mass of women and men pushing the boundaries of entrenched social norms are in enhancing women’s agency and capacity to aspire. We know that women need the tools of development, but development also needs women. All the disadvan- tages that women experience around the world, from poverty to violence, from ill health to illiteracy, also limit the advancement of families, communities, and entire nations. The Rockefeller Foundation is pleased to continue our collaboration with The World Bank through this report, and proud to have supported its research and production. We commend it to all who believe in building a more equitable and resilient world for the well-being of humanity. Acknowledgments T his research draws on the contributions of m any people who supported a 20-country rapid qualitative assessment entitled “Defining Gender in the 21st Century Talking with Women and Men around the World, A Multi-Country Qualitative Study of Gender and Economic Choice.� First and foremost we need to thank the 4,000 women and men and boys and girls who joined in nearly 500 focus groups. We asked very much of them, and they graciously shared their time and opinions with us. The project was led and managed by Carolyn Turk. The core team also included Ana María Muñoz Boudet, Patti Petesch, Angelica Thumala and Maria Beatriz Orlando. Valuable research assistance was provided by Paula Barros, Greta Gober, Gwendolyn Heaner, Rudy Herrera Marmol, Bethany Timmons, and Roberto Miranda. We gratefully acknowledge the World Development Report 2012 team that initiated and supported this project throughout, including the WDR co-directors Ana Revenga and Sudhir Shetty, and the team –Luis Benveniste, Aline Coudouel, Jishnu Das, Markus Goldstein, and Carolina Sanchez Paramo. We also extend appreciation to Elisabeth Huybens of the Social Development Unit for Europe and Central Asia and Cyprian Fisiy with the Social Development Network team for hosting this publication project. The “Defining Gender� data collection effort included national research teams from around the world led by: Chona Echavez and Pierre Fallavier (AREU, Afghanistan); Ugyen Lhamo (Druk Associates, Bhutan); Jean-Fran- çois Kobiané (Institut Supérieur des Sciences de la Population (ISSP) Université de Ouagadougou, Burkina Faso); Magaly Pineda and Sergia Galvan (Centro de Investigacion para la Accion Femenina (CIPAF), Santo Do- mingo, the Dominican Republic); Priya Chattier (The University of South Pacific, Suva, Fiji); Sanjeev Sasmal and Sulbha Khanna (Sutra Consulting, India); Rizki Fillaili (SMERU, Jakarta, Indonesia); Gwendolyn Heaner (GK Con- sulting, research on Liberia); Dumitru Slonovschi (Magenda Consulting, Moldova); Patricia Zárate (Instituto de Estudios Peruanos, Peru); Samia M. Al-Botmeh and Lamis Abunahleh (Centre for Development Studies, Birzeit University, Ramallah, West Bank and Gaza); Paul Barker, Marjorie Andrew and Almah Tararia (Institute of Na- tional Affairs, Papua New Guinea); Greta Gober (Centre for Gender Research, University of Oslo, research on Poland); Hana Baronijan and Sasa Jovancevic (IPSOS, Serbia); Imraan Valodia and Kruschen Govender (School of Development Studies, University of KwaZulu-Natal, South Africa); Mohamed Braima (Sudanese Organization for Education, Sudan); Adalbertus Kamanzi (CORDEMA, Tanzania); Giovanna Declich (Togo), Hhuat Tha Hong and Linh Tran (Institute for Social Development Studies, Hanoi Vietnam); and Ramzia Aleryani, Sabria Al-Thwar, and Mai Abdulmalik (Yemeni Women Union, Sana’a, the Republic of Yemen). Khalil Al Medani initially led the Sudan study, but unfortunately died soon after completion of the fieldwork. The Institute for Women’s Policy Research (IWPR) team, led by Jane Henrici and Allison Helmuth, helped with the original data coding and analysis. Amanda Lubold and Charles Ragin contributed with the qualitative com- parative analysis. Jeni Klugman, Director of PREM Gender, provided valuable support. Peer reviewers included Kathleen Beegle, Maitreyi Das, Aline Coudouel, Karla Hoff, Naila Kabeer, and Deepa Narayan. We also appreciated comments from Sarah Haddock, Dominique van de Walle, Rasmus Heltberg and Elizaveta Perova, among others. Getting this large research initiative off the ground in time to meet the WDR’s production schedule was a major task. World Bank staff from across regions and sectors responded quickly and helpfully with guidance on re- search design, local research partners, data analysis, and myriad technical and administrative needs. Our very special thanks to Dean Joliffe and Andy Kotikula, Tara Vishwanath, Nandini Krishnan, Abdalwahab Khatib, Andy Mason, Trang Nguyen, Shubha Chakravarty, Erol Graham, Iris Boutros, Mia Hyun, Yulia Immajati, Hesti Marsono, Dan Mont, Nicholas Menzies, Nora Dudwick, Owen Ozier, Andrea Gallina, Valery Vega, Roby Senderowitsch, Dan Owen, Sophia Georgieva, Hadyiat El- Tayeb Alyn, Trine Lunde, Arun Joshi, Adama Ouedrago, Sophia Geor- gieva, Liz Ninan, Chris Thomas, Maria Elena Garcia Mora, Elena Bardasi, Arun Joshi, Vivek Suri and Michael Woolcock. The team also consulted experts outside the World Bank, including Lori Heise, David Crocker, Van- essa Gray, and Janice Newberry. This publication is due to the support of the Rockefeller Foundation. The study benefited greatly from a re- search workshop conducted with the lead researchers from 18 countries at the Rockefeller’s Foundation Bel- lagio Center. A special thanks for making this possible to the Center and Foundation teams, in particular to Heather Grady and Sundaa Bridget-Jones. We are also grateful for relevant discussions with Rockefeller’s team, their relevant comments, support and patience through the preparation of this book. The team is also grateful for the financial support provided to the World Development Report 2012 that made possible the collection of our unique dataset, including the Government of Norway through its Royal Ministry of Foreign Affairs, AUSAID, CIDA, the Government of Sweden through its Ministry for Foreign Affairs, the Nike Foundation, and the Fast Track Initiative Education Program Development Fund. Our great thanks to Kristin Hunter’s editorial work on successive drafts, and the Directions in Development pro- duction team. We also thank the resource management team of Sonia Joseph and Evangeline Santo Domingo, and Cecile Wodon, Rebecca Sugui and Mihaela Stangu for ongoing help with coordination. Despite efforts to compile a comprehensive list, some who contributed may have been inadvertently omitted. The team apologizes for any oversights and reiterates its gratitude to all who contributed to this research. About the Authors Ana Maria Munoz Boudet is a World Bank consultant. She was a core team member of the World Devel- opment Report 2012 and co-author of the Latin America and Central America gender reports. She has worked on gender and poverty issues in the Latin America and Caribbean, Europe and Central Asia and Africa regions. Prior to joining the Bank, she worked for Inter American Development Bank, the United Nations Development Program (UNDP) and was a researcher at the Latin American Faculty of Social Sciences (FLACSO). Ana Maria holds a Master degree from the London School of Economics and is in the process of completing her Ph.D. at the University College of London. Patti Petesch is a World Bank consultant, specializes in qualitative field research on poverty, gender, conflict, and participatory development. Her recent research explores factors that enlarge individual and collective empowerment at the local level, and the contribution of these processes to local democracy, poverty reduc- tion, gender equality, and more secure and prosperous communities and nations. She was study coordinator and co-author of the World Bank’s Voices of the Poor and Moving Out of Poverty global research programs, and contributor of On Norms and Agency companion reports on West Bank and Gaza and Yemen. She recently published “Reflections on Global and Local Pathways to Women’s Empowerment and Gender Equality: The Good, the Bad, and the Sticky� (Ethics and Social Welfare). Carolyn Turk is the World Bank’s country manager for Rwanda and was the Lead Social Development Spe- cialist and Acting Sector Manager in the World Bank’s Europe and Central Asia Social Development Depart- ment when this research started. She is an expert in poverty policy analysis, including quantitative and qualita- tive instruments; statistical capacity building; national strategic planning and budgeting processes; and design and implementation of social accountability tools. Prior to joining the Bank she was senior planning officer in the Ministry of Finance in Papua New Guinea, Deputy Director of Action Aid, and Social Development adviser at the DFID. She has earned undergraduate and postgraduate degrees from Cambridge University in the Econom- ics and Politics of Development. Maria Angelica Thumala is a Lecturer at the Sociology Department of the Catholic University of Chile and Research Associate at the Centre for Criminology of the University of Oxford. She currently teaches Sociol- ogy of Gender and Ethnographic and Qualitative Research Methods. As a consultant for the World Bank she has contributed to the West Bank and Gaza, and Yemen gender reports for the Middle East and North Africa Region. She has also published on consumption, cultural change, development, and religion in Latin America. Angelica holds a PhD in Sociology from the University of Cambridge and an MA in Philosophy and Social Theory from the University of Warwick. Abstract Social norms, gender roles, beliefs about one’s own capacity, and assets, as well as communities and countries, determine the opportunities available to women and men—and their ability to take advantage of them. World Development Report 2012 shows significant progress in many areas, but gender disparities still persist. Our study covered 20 countries in all world regions, where over 4,000 women and men, in remote and tradi- tional villages and dense urban neighborhoods, in more than 500 focus groups, discussed the effects of gender differences and inequalities on their lives. Despite diverse social and cultural settings, traits and expectations of the ideal “good� woman and “good� man were remarkably similar across all sample urban and rural communi- ties. Participants acknowledged that women are actively seeking equal power and freedom, but must constantly negotiate and resist traditional expectations about what they are to do and who they are to be. When women achieve the freedom to work for pay or get more education, they must still accommodate their gains to these expectations, especially on household responsibilities. Girls’ desire for education, which nurtures their aspirations for greater agency, exceeded that of boys in rural and urban communities. Both young women and men wished for more education and better jobs than are com- mon in their communities and strikingly wanted to marry later, bear children later, and have more autonomy in choosing their partners than traditional community norms dictated. The main pathways for women to gain agency are education, employment, and decreased risk of domestic vio- lence. A safer space encourages women to negotiate for more participation and equality in household discus- sions and decisions. Women’s ability to contribute to family finances and control (even partially) major or minor assets helps them gain more voice at home and in public spheres. Women’s aspirations and empowerment to break gender barriers occur regardless of dynamic or poor economies, while men’s perceived gain in agency— and their identity as breadwinner—largely depends on economic conditions. When only a few women manage to break with established norms—without a critical mass—traditional norms are not contested and may be reinforced. The process of gender norm change thus appears to be uneven and challenging, lagging behind topical conditions. The easy co-existence of new and old norms means that house- holds in the same community can vary markedly in how much agency women can exercise, and women feel less empowered when opinions and values of families and communities stay with traditional norms. INTRODUCTION The norms of power and the power of norms “ Equality means that both the husband and wife have equal rights to make choices in their lives. — Urban woman, Fiji � “ [Equality for my daughter allows her] to have power, an education, and … more opportunities. — Rural woman, Peru � T wo of the many questions asked at the earliest fect on their sense of agency and empowerment, and to stages of preparing the World Bank’s (2012) World learn about the changes in women’s and men’s lives as Development Report 2012: Gender Equality and these gender norms flexed or persisted. Development were how do women and men make deci- sions about their economic participation, and how do we Gender equality in these 20 countries has increased in learn about this. To try to answer them, the World Bank many domains. Like changes documented for most of the launched a small qualitative data-collection pilot study in world, girls are staying in school longer than their moth- four countries. The objective was to find out what women ers did. More women are economically active and their and men saw as the main forces driving their decisions on participation in local networks and civic organizations has economic participation—from how they used their time increased. And many women feel that they have more to their ability to start a business. The exercise quickly control over their lives. Yet, significant gender dispari- expanded to an unprecedented “bottom-up� exploration ties are still evident: intrahousehold alloc ations of time, of how women and men make decisions in all dimensions responsibilities, and power are unequally distributed of life, how gender differences are experienced, and how among men and women. Almost everywhere, men remain On Norms and Agency these differences, dictated by social norms, shape wom- the primary income earners in their households, as well en’s and men’s everyday lives. The research covered 20 as the main decision-makers. And there are countries and countries from all world regions and more than 4,000 communities where income poverty, conflict situations, participants in 97 communities—from remote and tradi- rurality, or distance increase these existing gender gaps.2 tional villages in Papua New Guinea, Yemen, and Liberia, to urban neighborhoods in Vietnam, Poland, and Peru Our study findings reveal that behind the progress to- (see map I.1).1 In each country, local researchers orga- wards gender equality and persistent gender gaps lies nized about 500 focus groups to elicit information about an almost universal set of factors embedded in social 12 the impact of gender norms on women and men, the ef- and gender norms, as heard in the experiences related by focus group participants. Women’s and men’s oppor- Social norms play a central role in the relation between tunities and actions are determined as much by social people’s agency and the opportunities that their com- norms—including gender roles and beliefs about their munities provide. Social norms can either help or hinder abilities and capacities—as by the conditions of the com- an individual’s capacity to take advantage of available munities and countries they live in. opportunities. Certain ideas or images that reflect ideal behaviors for men and women are remarkably similar The narratives from the sample communities show many across countries and locations within countries. Adoles- commonalities across countries and cultures in how gen- cents participating in the study reported little variation der differences define women’s and men’s roles, and dic- in the different tasks and behaviors demanded in order tate responsibilities in households, markets, and public to be seen as a “good girl� or �good boy�—whether they life in their communities. They also reveal how innumera- live in a remote highlands village in Papua New Guinea or ble social and cultural norms, traditions, beliefs, and gen- in a busy city in the Dominican Republic. Likewise, adult eral perceptions of the appropriate place and behavior views of a “good wife� or a “good husband� reiterate a for women and men permeate all aspects of community clear distinction between productive and reproductive and individual life. These informal institutions (so named gender roles, as well as expected feminine or masculine in World Development Report 2012) interlock with civic behaviors (loving and caring versus having authority and institutions, the institutions of the state, the market, and providing well). intrahousehold bargaining dynamics to shape and some- times reinforce the gender inequities of power—and im- Yet everyday practices also include different forms of pact the choices and freedom of women and girls (and resistance to—and flexibility about—ideal gender roles. men and boys). Negotiation and resistance to gender norms are evident MAP I.1: COUNTRIES INCLUDED IN THE QUALITATIVE ASSESSMENT OF GENDER DIFFERENCES IBRD 39636 October 2012 Poland Moldova Serbia Afghanistan West Bank and Gaza Dominican Bhutan Republic Vietnam Sudan Burkina Faso (Northern) India Rep. of (Orissa, Togo Yemen Andhra Liberia Pradesh) Papua New Guinea Tanzania Indonesia Peru (Jakarta,Banten, Sumatera Barat) Fiji On Norms and Agency South Africa (Kwazulu-Natal) Source: World Bank. Countries and regions where WDR qualitative assessments were carried out 1 The countries included in our study are Afghanistan, Bhutan, Burkina Faso, Dominican Republic, Fiji, India, Indonesia, Liberia, Moldova, Papua New Guinea, Peru, Poland, Serbia, South Africa, Sudan, Tanzania, Togo, Vietnam, West Bank and Gaza, and Yemen. 2 See World Development Report 2012: Gender Equality and Development (World Bank 2012) for more detail and data on these trends. 13 throughout the study sample. Inasmuch as they imply 1. The study approach a challenge to the sexual division of power, departures from the norm can sometimes be harshly punished. This study is based on the assumption that gender equal- Among the consequences of conflict over gender roles ity is a development objective in its own right as much as or norm abiding, the most disempowering one is violence it is instrumental to the achievement of such develop- against women. ment. Following Amartya Sen’s (2002) notion of devel- opment as expanding freedoms equally for all people, The change in women’s ability to participate and have our study assumes that the freedom to pursue a life a voice in strategic life choices, especially in education of one’s own choosing is a key component of develop- and reproduction, is reflected both in women’s achieve- ment. In other words, we see development as connected ments and aspirations. Education, employment, and to the freedom to enjoy a genuine set of opportunities family formation are the primary areas where women and choices. In a similar vein, Nussbaum (1999) frames see their agency and ability to decide expanding. The the challenge for development around liberty, but also autonomy of young girls and boys appears to be greater notes that skewed preferences due to persistent gender than in previous generations, and their ambitions differ inequalities impact girls’ and women’s liberty. Particularly from current practices in their communities, from age at in poor countries, this shows up in the gap between for- marriage to number of children to level of schooling. But mal rights and the absence of basic material conditions it is among girls and young women where these change necessary to realize those rights.3 The intrinsic value is most evident. of gender equality lies in increasing both women’s and men’s choices, autonomy, and self-efficacy, as well as Increased agency allows women to move from enduring their exercise and use of equal rights. complete compliance to constraining and unequal gen- der norms, to questioning those norms in face of poten- The instrumental value of gender equality—the benefits tial opportunities, to changing their aspirations, as well that a more equal society obtains in terms of the produc- as their ability to seek and achieve desired outcomes. tivity, inclusive institutions, and well-being of future gen- While women have increased their perceived empower- erations, among others—is rigorously explored in World ment and freedom in many countries, more so than men, Development Report 2012. Empowering women does this change does not always alter constraining norms. indeed provide benefits for the well-being of societies. However, as Duflo (2011) notes, the relationship between Inequalities derived from gender norms and lack of ca- economic development and women’s empowerment is pacity to decide (agency) affect perceptions of power not always a virtuous one. Empowering women does in- and freedom. The main pathways to increased percep- deed change society’s and households’ choices in ways tion of empowerment we can identify from the focus that are beneficial for their members, but not in all cases: groups narratives, combine control over material and it is not always women who make the best decisions for personal life conditions with a favorable structure of long-term development. opportunities. While these are equally relevant to men and women, men depend largely on the economic con- If we think of gender equality as a result of gains in three ditions of their communities to feel empowered, more dimensions—endowments, economic opportunities, and so than women. agency—then this equality is largely dependent on the in- On Norms and Agency teractions between four institutions: households, formal In a more enabling environment, which not only creates state institutions, markets, and informal institutions. Fol- more opportunities but also changes the individual’s lowing a graphic representation of this conceptual frame- capacity to aspire to access them, normative change is work from World Development Report 2012 (figure I.1), our more likely. For example, women’s economic participa- study zooms in on the specific interactions between so- tion has the potential to alter traditional definitions of cial norms and agency with a focus on the household. gender roles, duties, and responsibilities, but it can also change the main components of both men’s and women’s Women’s agency, while a central element of gender 14 identities. equality, is an area where more research is needed and FIGURE I.1: WORLD DEVELOPMENT REPORT 2012 ANALYTICAL FRAMEWORK QUALITY Po ER E INFORMAL ND lici es MAL INFOR IONS GE INSTITUTIONS IT U T AND SOCIAL INST D S O CIAL ECONOMIC NORMS AN M S N O R OPPORTUNITIES HOUSE- ETS HOUSE- OUR FOCUS HOLDS MARK HOLDS AGENCY ENDOW- MENTS F INS ORMA AGENCY TITU L TIO NS Figure I.1. GR OW T H Source: World Bank (2012, 9). where less information is available. Several studies have This study deals with these difficulties within the scope been conducted on empowerment and on some agency allowed by its cross-country sample and methodology. As components, but many questions remain.4 The analysis presented in more detail later, we provide a foundation in our report seeks to contribute to this body of litera- for a systematic exploration of agency by looking at the ture by looking at agency and social norms together. Of structures of constraint, or the norms that underpin gen- all factors driving gender inequalities, these two seem der inequalities, and the negotiations that surround these to be the most elusive in helping direct policy interven- norms (Part I). We also look at different life choices where tions and measurement. Our findings align with Kabeer changes in the capacity to decide are reflected (Part II). (2001) and the difficulties that appear when attempting And finally, we attempt to offer a more dynamic and com- to measure agency. First, it is not sufficient to learn about plete view of the process and determinants of changes in women’s ability to make choices without looking at the power and agency, as they are perceived by individuals extent their agency is reflected in their life choices and within their specific community setting (Part III). the conditions under which they exercise their agency. Second, context matters: without looking at context, By exploring how gender norms and roles shape women’s it is not possible to assess the extent their agency has (and men’s) agency and empowerment, their decision- increased or not. The need to focus on context makes making at critical junctures in their lives, their perceived On Norms and Agency cross-country analysis more difficult. Finally, changes in ability to gain power, and their economic opportunities, agency are not clear predictors of processes of norma- new entry points for policy design can be found, as well tive change if the structures of opportunities and con- as ways to recraft existing development policies to be- straints are not taken into account. come more effective and better serve women’s needs. 3 Alkire (2002) makes similar arguments, but from a perspective of gaps in human development rather than lack of rights, and calls attention to the need to assess well-being at the individual and wider collective or societal levels, as well as the capabilities and assets needed for exercising agency. 4 Kabeer (2001), Ibrahim and Alkire (2007), and Samman and Santos (2009) provide surveys of the different studies and perspectives under which they were conducted. 15 The common patterns we found across countries have was transcribed into text documents following a template important implications for policy design and action. provided by the global team and analyzed by a mix of techniques including coding and interpretative analysis. To ensure validity, we verified conclusions (as suggested 2. Methodology of the study by Miles and Huberman 1994) and cross-checked them with the national reports by the local teams. However, it Our study assesses qualitatively the dynamics of gender is important to note that this is, first and foremost, a sub- norms and agency in the construction of gender equality. jective exploration; the samples are small and not statisti- The research was designed to capture men’s and wom- cally representative of each country or region. en’s perspectives and their own accounts on how they experience gender differences in their communities. We chose 97 communities in the 20 countries to contrib- ute to a unique dataset made up of men’s and women’s The methodology we chose builds directly on two ma- focus groups with three different age groups (more than jor global studies at the World Bank, Voices of the Poor 500 focus groups), pulled from remote mountain top (Narayan et al. 2000; Narayan and Petesch 2002) and villages in Bhutan to refugee camps in Sudan to urban Moving Out of Poverty (Narayan and Petesch 2007, 2010; neighborhoods in Vietnam and Poland (see table I.1 be- Narayan, Pritchett, and Kapoor 2009; Narayan 2009). low). Sample countries were chosen opportunistically These works apply primarily qualitative techniques, such from all world regions and, when possible, from different as focus groups and individual interviews, to examine realities within each region.7 However, the identification questions of poverty and how people move out of pov- of the sample was also dependent on the availability of erty across diverse contexts in the developing world. A local research teams, funding, and time constraints de- guiding principle for these studies, as with this one, is termined by the production cycle of World Development the focus on learning inductively from local individuals’ Report 2012. experiences and interpretations of their own reality. We wanted to work from a vantage point that gives primacy The local research teams in each study country consisted to local people’s own perceptions and interpretations of of lead researchers with extensive country knowledge their experiences. For this study, we aimed to capture and qualitative field experience, plus experienced focus local narratives of different situations where gender dif- groups facilitators who received training and followed a ferences come into play without imposing pre-conceived detailed methodology guide to conduct the fieldwork. In concepts and models. each country, the research teams identified the commu- nities to survey, following the study guidelines that in- The research was conducted in 20 different countries, us- cluded sampling communities from rural and urban areas, ing the same data collection instruments and the same from different socioeconomic situations, and represent- set of questions for all cases, which permitted a multi- ing, when possible, different realities within the country country assessment of similarities, trends, and patterns. (see appendix A). The teams were asked to sample a A set of research instruments was developed including minimum of four communities to capture a better off and three focus group guidelines –one for each of the three poorer urban community, and a better off and poorer ru- different age groups included in the study, as well as a ral community, which we hypothesized would provide a separate questionnaire for the key informant in each range of experiences that reflected the average situation On Norms and Agency community; the same set was used in all countries to en- of the country. In some countries, this was done based sure comparability.5 Changes to adapt language or make on household survey data; in others, it was based on rep- additions that were more appropriate for local conditions resentation of different country regions. The choices of were discussed between the local and global research regions and geographic areas were discussed with the teams to ensure comparability was respected. The re- global team and sampling selection approved to ensure search strategy was flexible enough to capture bottom- consistency with the global sample. up data from very different places and also provide a reasonably adequate means for comparative analysis of Within the communities, five different data collection 16 the large volume of data collected.6 The data collected tools were used: three structured focus group discus- sions, a key informant interview in the form of a com- module charted how young adult women and men spend munity questionnaire with closed and open-ended ques- their days, and another explored different levels of pow- tions, and a mini case study (see table I.2 below). The er and freedom that adult women or men might have in three focus groups were structured by age: adolescents their communities. Some questions were posed to all (12–17 years), young adults (18–24 years), and adults (25– three age groups; others were specific to one group. 60 years). Each age group was then divided into men’s Table I.2 summarizes the main topics that were asked to and women’s groups. Field teams also received instruc- the different groups. Each topic was covered by a set of tions to construct the groups, as much as possible, to questions and exercises. reflect the range of educational and livelihood experi- ences common for each age group in the community. The In order to limit bias, which can be introduced by focus research teams invited individuals to participate in the group dynamics, facilitators received training in addition- exercise through household visits, postings, and informa- al measures to foster inclusive discussions that would tion given to community leaders among others. capture a range of attitudes and experiences common in the specific communities. For some key questions, Prior to initiating the focus groups, facilitators inter- for instance, focus group members had opportunities to viewed local key informants, identified during earlier respond by “voting� in private and then volunteering to community visits by asking local authorities and people discuss their responses. from the community. These informants completed a com- munity questionnaire to provide extensive background We designed the study methodology to account for information about the sample communities. A key infor- the dynamics of gender relations and social norms in mant might be a community leader, government official, the study communities. Understanding that gender politician, an important local employer, a business or fi- norms influence everyone’s behaviors as much as their nancial leader, teacher, or healthcare worker. At the end expectations about how the opposite sex behaves, of their time in the communities, the research teams also we kept groups separated by sex. Likewise, different collected “mini case-studies,� which were unstructured age groups were assessed separately to account for interviews with a focus group participant or someone generational differences and avoid power imbalances. else in the community who might understand the gen- We hoped to give all participants a safe environment der gains or inequalities in the communities. Local teams where they felt free to express their thoughts and in- were free to choose their case studies based on their teract openly about life situations that they may not knowledge of the community and the country. normally reflect upon8. For example, when we asked women in Afghanistan to describe their preferences Each focus group, organized by sex and age, met sepa- and interests regarding marriage or childbearing deci- rately. While the focus groups of young adults and adults sions, the research format first captured their initial were conducted in all 20 countries, only a sub-sample accounts. Then discussion leaders posed further ques- of 9 countries included focus groups with adolescents. tions to encourage them to probe beneath the face Focus groups discussed a wide range of topics, includ- value of their accounts—for instance, from a power ing reasons for happiness and favorite free-time activi- perspective—so that they could begin to identify the ties; decisions surrounding when to leave school, where set of values and other norms affecting their decisions. to work, and family formation; and gender differences in In many cases, what was accepted as the “norm� was On Norms and Agency accumulating savings and controlling major assets. Ques- far from what the women desired or what they consid- tions also explored issues of domestic violence, public ered right. Focus group participants were also invited safety, and women’s physical mobility. One research to corroborate or refute each other’s views. 5 The methodological note in appendix 1 includes more details on the data collection tools, as well as on the analysis techniques used. 6 At the end of the data collection effort, the team gathered about 7,000 pages of narrative text of transcripts of focus groups and interviews. These data were structured, cleaned, and imported into qualitative analysis software. 7 Based on the World Bank’s classification of regions, which includes Latin America and the Caribbean, South Asia, Sub-Saharan Africa, Middle East and North Africa, Europe and Central Asia, and East Asia and the Pacific. 8 All research team members participating in each discussion were the same sex as the focus group participants. 17 Table I.1: Qualitative assessment sample Focus groups Total no. Of individuals Young (est. 8 Per In-depth Country Communities Adults adults Adolescents Total group) cases Afghanistan 4 8 8 N.A 16 128 8 Bhutan 4 8 8 8 24 192 4 Burkina Faso 4 8 8 8 24 192 4 Dominican Republic 4 8 8 8 24 192 4 Fiji 6 12 12 12 36 288 6 India 8 16 16 16 48 384 8 Indonesia 4 8 8 N.A 16 128 14 Liberia 9 18 18 N.A 36 288 12 Moldova 4 8 8 N.A 16 128 4 North sudan 5 10 10 10 30 240 4 Papua new guinea 6 12 12 N.A 24 192 6 Peru 4 8 8 N.A 16 128 5 Poland 4 8 8 N.A 16 128 4 Serbia 5 10 10 N.A 20 160 4 South africa 4 8 8 N.A 16 128 4 Tanzania 4 8 8 N.A 16 128 4 Togo 4 8 8 8 24 192 4 Vietnam 4 8 8 N.A 16 128 4 West Bank and Gaza 6 12 12 12 36 288 6 Yemen 4 8 8 8 24 128 4 Totals 97 194 194 90 478 3760 113 In order to move beyond a static view, or a single mo- portunities or limit equality between men and women in ment in time, and capture dynamics of change, all groups the communities studied, shed light on similar inequali- were asked at different stages to compare conditions ties in other contexts and the processes behind them. today on key study topics with conditions 10 years ago or between the current and previous generations. They Parallel to quantitative analysis of gender dimensions in also reported on their aspirations for their own future development, the insights derived from qualitative meth- and the future of their children. ods expand the information available on questions relat- On Norms and Agency ed to norms and to intrahousehold and community-level The study findings reflect the range of norms possible in dynamics. In particular, contextual factors and their inter- the 20 countries rather than the average situation in each actions with the deeper influences of power relations and individual country case. However, the global findings of norms on women’s and men’s decisions are difficult topics the research are more telling and consistent, which in for even well-designed household surveys to explore ef- many areas can be extended to other settings. The con- fectively. Yet, the scarcity of information on the role of sistency of the descriptions of gender norms and associ- these complex factors limits our understanding of these ated behaviors, and the relationship between norms and issues and possible levers for policy action. This is the 18 agency, and how these elements interact to generate op- area where we see our research contributing the most. Table I.2: Summary of methodology Data collection method Themes Respondents Community Information on local context and changes in the 1 –2 key informants questionnaire structure of opportunities. Focus group discussion • Happiness 2 groups (ages 18–24): with young adults • Daily time use (included hourly time use reporting by 3–5 focus group • 8 –12 young adult participants) women • Decisions: transitions from school to work and family formation • 8 –12 young adult men • Independence, cooperation, and obligations in economic decision- making processes • Divorce, family dispute resolution mechanisms • Local economic opportunities • Savings practices • Community participation • Knowledge of gender-related rights • Role models • Hopes for the future Focus group discussion • Happiness 2 groups (ages 25–60): with adults • Differences in the exercise of power and freedom, with a focus on • 8 –12 women economic decisions (via exercise creating a “ladder of power and • 8 –12 men freedom�) • Local economic opportunities • Independence, cooperation, and obligations in economic decision- making processes • Divorce, family-dispute resolution mechanisms • Sources of economic support • Household gender relations • General patterns of domestic and community violence • Hopes for the future Focus group discussion • Happiness 2 groups (ages 12–17): with adolescents • Daily time use • 8 –12 adolescent girls • Value of education • 8 –12 adolescent boys • Aspirations for the future • Local economic opportunities • Savings, assets, and control of assets • Formation of families • Norms surrounding adolescent girls and boys • Domestic violence and public safety • Social networks Mini case study Detailed story of a finding that emerges as important for understanding 1–2 key informants gender norms or structures shaping economic decisions in that locality. On Norms and Agency 3. Discussing and researching gender their different communities. However, there is not only equality: A brief introduction to the one way to understand these concepts. We briefly re- primary study concepts view different views of norms, agency, and power, and the reasons gender norms have such a decisive hold on Throughout this report, certain concepts—social norms, women, men, and the societies where they live. agency, empowerment, and structure of opportunities, among others—appear over and over. We explore their The powerful influence of gender norms on an indi- interrelation by using the voices of the participants in vidual’s actions—a central area of concern in gender the study as they reflect on the contexts and realities of research—is one of the foundations of gender inequality. 19 As Ridgeway (2009) notes, gender is a core frame for or- behavior deemed appropriate for women and men—are ganizing social relations and, as such, it depends on com- constructed and defined.10 They define the norm. mon knowledge (i.e., cultural knowledge) that guides and coordinates individuals’ actions in a given situation. But a. Power, empowerment, and agency these frameworks deem women and men unequal, based on their perceived differences. Power in its different expressions has always been behind gender inequalities. In fact, empowerment and agency Inequality is a feature of all societies, whether it is un- are not only highly relevant to gender research but are equal power, opportunities, outcomes, or justice. Most essential to the questions we posed during the field work. societies have structures and institutions whose role Inasmuch as no social system exists without some divi- is to preserve the prevalent social order or organizing sions by gender, gender is co-substantial to the structure framework. Gender inequality is no exception. The in- of power in all its forms (Caramazza and Vianello 2005): equalities that arise from the different roles played by women and men, the unequal power relationships be- • Power over, or domination, the ability to influence tween them, and the consequences of this inequality on someone else’s actions and thus determine their in- their lives are visible in all societies. The problem is that terests and preferences (Lukes 1974) these inequalities all too frequently pose disadvantages • Power to, or agency, people’s enduring capacity to to women. Women face consistent differences between act (Sen 1985; Isaac 1987) their opportunities and outcomes and the opportunities • Power with, or solidarity, and power within, or con- and outcomes of men. sciousness (Rowlands 1997, Ibrahim and Alkire 2007)11 The point of departure for gender inequality is our bio- While we saw all these different forms in the research, logical difference, which is visible and in most cases easi- our focus is on power as agency and, as such, as the ca- ly distinguishable. But it is less easy to find a cut-off point pacity to act to achieve desired objectives. between the biological and the social distinction as a basis for gender inequality. Benhabib et al. (1995) rightly Our intent is to reveal men’s and women’s accounts of notes that, while equality of condition seems to be the their gains in autonomy as gains in power, capacity, and ideal, in many societies today, the more equal condi- potential to act, even when actions fail or are never tak- tions are, the less explanation there is for the remaining en. Power is not only the ability to make people do what differences—to the point that inequality may end up be- they would not otherwise do but also the ability to enable ing mistaken and merged with innate or natural qualities people to do what they could not otherwise do (Hartsock of men or women.9 Preferences, needs, and constraints 1996). It is this positive, creative notion of power that we can differ systematically between men and women, and aimed to capture in the focus groups, as well as what we this may reflect both biological sexual factors, as well as stress in our analysis: a view of power as agency with learned gender behaviors (see box I.1). individuals gaining the ability to act and decide. Learned gender attributes make up gender identity and Before proceeding further, it is important to visit the re- determine gender roles; they also may be valued differ- lated elements of empowerment, agency, and opportu- ently, generating a power imbalance. These “gender sys- nity structure. Probably the term most associated with On Norms and Agency tems� (Ridgeway and Correll 2004; Ridgeway and Smith- gender equality is empowerment, the expansion of free- Lovin 1999), where gender is seen as an institutionalized dom of choice and action as a result of a process of gain- system of social practices and organized social relations ing power (Narayan 2002; Narayan and Petesch 2005). It of inequality, are based on the different attributes associ- refers to the process of gaining control over resources— ated with our biological differences. Gender systems are material and non-material—in order for individuals to gain embedded in all societal institutions, from formal legal the capacity to exercise the right to determine their own frameworks (such as family law or labor regulations) to choices. Empowerment also refers to the way individuals religions and traditional culture. Through these systems, acquire the ability to influence change in their lives (Mos- 20 femininity and masculinity—the roles and patterns of er 1989), and are able to take advantage of opportunities. Box I.1: It’s not sex, it’s gender: From biology to learned behaviors Researchers disagree on where gender differences come from. The observable differences between men and women, in areas such as risk aversion, trust, leadership, moral behavior, attitudes about competition, and compassion, have been attributed to biological factors, learned preferences and behaviors, and consistent differences in opportunities.a Opportunities have not been equally distributed among women and men. For example, the fact that girls have achieved so much progress in education is as much a shift in the distribution of opportunities as a change in society’s view of what women and men are able and capable of doing. Most societies at different stages have resisted educating women. For some, educating women was not “natural�: the reasons have ranged from ideas that women’s nature does not include the ability to learn, that women do not need education to secure their future, to that there is no need for incentives for educating women. Teaching women to read and write was considered wrong because “a learned lady threatened male pride.�b But today, most societies agree on the value of education for both girls and boys. In school, differences in performance between girls and boys have been explained by differences in their cognitive abilities, in forms of learning, in their aspirations, in their views on the value of education, and in teacher performance, among others.c For example, Hoff and Pandey (2006) look at how learned discrimination, in their study of Indian students of different castes, may affect performance on tests when cast is made salient, vis-à-vis when it is not. The authors find that when caste is identified or emphasized in a given setting or situation, low-caste students perform worse, reproducing the caste gap and hierarchy. Similar studies, where race, ethnic background, and gender have been used to trigger expected response in an experimental setting, show similar results.d Gender equality, even if for the benefit of everyone’s well-being, challenges the social foundation of inequality, as well as its “natural�—or biological—foundation. In the case of education, it not only contradicts the notion of who has the right to education but also challenges ideas of who can join the qualified labor force (which now includes men and women) and what constitutes women’s and men’s appropriate place in society. a. Gender differences have been analyzed experimentally in different areas of economics and under very different settings. Recent reviews of this literature include Ergun, Garcia-Munoz, and Rivas (2011), Croson and Gneezy (2009), and Eckel and Grossman (2008). Lippa (2005) provides a good summary of findings from the psychological and behavioral studies field. b. Labalme (1980, 4). c. The Young Lives study (Dercon 2011) shows that parents have different aspirations for their children’s educations than their children, and that the parents’ aspirations are transmitted and adopted by children. World Development Report 2012 cites the example of some English subject text- books, currently in use in Australia and Hong Kong SAR, China, that tend to depict women in a limited range of social roles and present stereotyped images of women as weaker and operating primarily in domestic domains, and may impact girls’ aspirations. d. Among others, see Steele and Aronson (1995), Shih, Pittinsky, and Ambady (1999) and Krendl et al. (2008). Empowerment’s departing point is a condition of disem- ity to make strategic life decisions in a context where powerment or the inability to exercise any influence. this was previously denied to them. As with many other concepts, there is no single definition of empowerment. On Norms and Agency Hence, empowerment is contingent on agency as much Ibrahim and Alkire (2007) identified over 30 different as it is on available resources and opportunities. In other definitions of empowerment in the literature and a num- words, it is the expansion of agency (Ibrahim and Alkire ber of approaches to measuring it empirically. 2007; Kabeer 2001) or the expansion of people’s abil- 9 Arendt’s (1979) work on the origins of totalitarianism attributes this change to the appearance of the social sphere, where nothing is private or an object of political debate. 10 For example, gender differences in participation in the labor force may derive from a history of specialization due to our biological differences. As Alesina et al. (2010) and Alesina, Guiliano, and Nunn (2011) argue, based on the evolution of labor-intensive agriculture (before mechanization) that re- quired physical strength, which is more common in men than women, women specialized in home production due to their lack of physical strength. 11 Robeyns (2005) also presents a detailed account of all the different approaches to power that have been used in development practice. 21 With a development approach, agency as the expan- to make one’s own choices and having the self-confidence sion of an individual’s own power and freedom is strongly to do so are not matched by any real opportunities to linked to the capabilities approach.12 According to Sen choose from (Alkire 2002; Robeyns 2003).� Opportuni- (2002), agency is part of a person’s development free- ty structures—by which we mean resources, institutions, doms, where one is able choose how to use entitlements established processes (traditions, moral codes, gender in pursuit of self-determined goals.13 Agency, together norms), and other enabling factors—foster the empower- with opportunities, is central to development. Alkire ment process and are necessary to create an enabling (2009) adds agency’s constructive role in the creation context for agency to manifest (Alsop, Bertelsen, and of values and norms to its intrinsic and instrumental rel- Holland 2006; Narayan 2005). Formal and informal rules, evance. And Nussbaum (1998, 1999), who looks more at state and local institutions, the market, and civil society legal frameworks and rights, specifically addresses the are elements of the social structures within which choice gender dimensions of agency. Nussbaum argues that takes place. As the context for action, the structure of women’s agency is different from men’s due to unequal opportunities for gender equality is where agency can social and political circumstances that give women un- be realized (Kabeer 2001). equal capabilities. For Kabeer (2001), agency is a di- mension of empowerment, together with resources and Because of their characteristics, agency and empower- achievements, without which the process of women’s ment are more difficult to measure than the structural empowerment is not possible. Agency thus is the ability factors that determine them. Research tends to focus to make one’s own choices and act upon them. on the opportunity structure or the prerequisites for agency (such as literacy, access to information, access to Whether agency is seen as the ability to formulate strate- land, rights, etc., which are also opportunity structures), gic life choices or the ability to control the resources that instead of the components of agency itself (Alkire 2009). come to bear, its relation to empowerment and decision- The problem is that these prerequisites do not always making (as the capacity to act and bring about change) translate into agency nor are they the same for all indi- is clear.14 Benhabib et al. (1995) and Fraser (1997) include viduals.16 Alsop (2005) argues for going beyond the mere in their definitions of agency the subjective capacity existence of an opportunity, for a research focus on the for choice and also the capacity for self-determination, necessary conditions of agency:17 where women—and men—get to play an active role in the formation of their identity and do not passively absorb • Existence of choice: whether an opportunity to make external determinations or constraints. As such, agency a choice exists turns subjects into autonomous, purposive actors, ca- • Use of choice: whether a person or group actually pable of choice and self-definition, able (to attempt) to uses the opportunity to choose become the individual they have chosen to be through • Achievement of choice: whether the choice brings the actions that express it (Lister 1997). about the desired result Autonomy and agency are inter-related. As a necessary World Development Report 2012 opts for exploring dif- condition for any action, autonomy allows individuals to ferent manifestations of agency (or lack of it), where question the social norms, rules, and practices impacting decision-making by women can be identified,18 similar to their choices; to reflect upon these rules; and, if needed, many of the studies surveyed by Kabeer (2001). On Norms and Agency to take action to change them (Doyal and Gough 1991; Dworkin 1988).15 Self-efficacy, or people’s belief in their In our study, we attempt to look at agency and gender ability to mobilize available resources—human, material, inequality in agency. For men and women, agency dif- or social—to make their choices a reality, is autonomy, fers: they have different degrees of empowerment, dif- which is an essential component of agency. ferent sets of choices, different opportunities, different capacities to exercise their choices, and different levels Agency without access to resources is a somewhat pas- of achievement of chosen outcomes. This disparity in sive capacity. As van Staveren (2011, 1) notes, “agency agency usually plays to women’s disadvantage. Hence, 22 without resources is rather meaningless when being able we want to understand “inequality of agency� and its central role in perpetuating gender inequality (Rao and opportunities—is treated separately to better acknowl- Walton 2004). To do this, we tried to look at agency from edge their role in promoting or restraining agency. three angles: the ability to act, self-efficacy or the belief that acting is possible, and the ability to control the re- The background conditions of society governing women sources to make a choice possible. and men vary. Not all societies are the same, nor are the economic, cultural, social, political, religious, security, and b. Gender inequality in agency other conditions of the 20 countries visited in the study. Within each country, communities are highly heteroge- We designed the fieldwork methods to capture differ- neous. Local conditions matter; they have an impact on ent manifestations of gender inequality in agency. First, women’s choices and preferences. Women and men con- we looked at decision-making processes in households stantly adapt their choices to what is happening around as expressions of agency and autonomy in strategic life them. If the context does not provide fair conditions for decisions, such as marriage, childbearing, education, action, this inequality is registered by households and in- and job choices. Second, we assessed the dynamics dividuals, and shapes their preferences in ways that may of changes in perceptions of power and freedom, as be detrimental (particularly women). accounted by men and women in the 97 sample com- munities, and the many dimensions they identified. These “adaptive preferences� have an impact on agency. Third, we looked for the necessary conditions or fac- What you do not see, you do not know and you can- tors determining the ability of women to feel empow- not aspire to. For example, many of the women we in- ered: education, aspirations, income, lack of threat of terviewed reported a preference for flexible work ar- violence, and more. We do not try to cover everything, rangements, such as part-time work, informal sector or but attempt to show how interconnected agency and non-regulated work, and self-employment. It is worth empowerment are with social norms around gender, asking, and we do so, if such preferences are shaped by the structure of opportunities, and the community women’s prescribed (gender) role as mothers and the contexts of the study participants. opportunities available for working mothers provided by local markets. Some women have the perception that In the same vein, the opportunities presented in the dif- the employers prefer workers without care responsibili- ferent communities are not equally distributed or open to ties; others do not have public provision of childcare both sexes. In fact, a community’s opportunity structure in their communities. And still others do not have the may include elements that reproduce gender inequal- qualifications to get a job. ity or women’s subordination, as noted by G. Sen and Grown (1987), Elson (1999), and Nussbaum (2000). For This is an example of the material and contextual precon- the purpose of our analysis, the structure of opportuni- ditions to agency, in whose absence there is no real exer- ties is comprised of the formal and informal institutions, cise of agency, merely a simulacrum of choice (Nussbaum the market, and the household. For analytical purposes, 2006). This adaptation may lead to an inequality trap, social norms—normally part of the overall structure of where women’s muted preferences affect their capacity 12 As defined by Amartya Sen. For an application of the approach to gender inequality, see Robeyns (2003) and Nussbaum (2001). 13 A person’s agency freedom, for Sen, should be understood as including the individual’s aims, objectives, allegiances, obligations, and—in a broad On Norms and Agency sense—the person’s concept of the good. Also see Sen (1985). 14 Samman and Santos (2009) provide a good survey and summary of these two different positions. 15 Doyal and Gough (1991, 53) define autonomy as “the ability to make informed choices about what should be done and how to go about doing it. This entails being able to formulate aims and beliefs about how to achieve them, along with the ability to evaluate the success of those beliefs in the light of empirical evidence.� 16 See also Ibrahim and Alkire (2007) and Kabeer (1999; 2001), who refer to studies that use measures of access to land as an indicator of empower- ment. They argue that these types of studies, by focusing only on land ownership or legal capacity to own, forget the pathways by which such access translates into agency and achievements in women’s lives. 17 How these three conditions are measured, however, is not clear. 18 The World Development Report 2012 identifies freedom of movement, fertility control, freedom from domestic violence, and the ability to have a voice in society as the main components of agency. Ibrahim and Alkire (2007) propose certain “exercises of agency� areas with specific sets of indicators, including control over personal decisions, choice in household decision-making, domain-specific autonomy, power to change aspects in one’s life at the individual level, and power to change aspects in one’s life at the community level. 23 to aspire (Appadurai 2004). In the long run, it reduces dividual behavior in a society. (ii) They specifically pre- their agency because of a context that affects their ability scribe what behavior is expected and what is not al- to see the pathway to achieve their desired goals. lowed in specific circumstances. (iii)They tell a person what to believe others expect of her behavior and tell others what to expect from that person. (iv) There is an 4. Creating and enforcing gender expected agreement, or belief that the agreement ex- through norms, roles and beliefs ists, on the content of the norm and an enforcement of such agreement or belief by whoever holds power. So how can we understand gender constructions of indi- viduals and the strong hold of social norms on our behav- Social norms are powerful forces; they are prescrip- iors and beliefs? Social norms are difficult to measure. If tions or dictates reflected in the formal structures of so- they appear as clear and concrete directives for actions, ciety, in its informal rules, its gender role divisions, and deriving from a given society’s values, they are easily cap- permeating beliefs, attitudes, and behaviors. They hold tured in laws and formal rules. If they refer to informal, power via emotional control (Elster 1989), social expec- implicit rules that govern what a person can and cannot tations (Bicchieri 2006), and prescription (Akerlof and do in the pursuit of daily life, they are more elusive (Hech- Kranton 2000), as well as internal commitment (Alex- ter and Opp 2001; Portes 2006; Fehr and Gachter 2000). ander 2003). In many cases, particularly with gender Regardless of their form, compliance with the norms—and norms, the joint presence of at least two of these forces sanctions for breaking them—are to be expected. makes the norm more binding. As Bicchieri (2006) sug- gests, individuals prefer to conform to the norm due to Norms around gender stem from a society’s ideals val- the belief that other people will also conform, to the ues of what it means to be a woman or a man. Failure point that a collective agreement is created between to conform to these dictates can trigger strong social normative beliefs and behavior.20 How people believe sanctions, such as ridiculing men for being emotional or they should behave, what their behavior is, and how so- scorning women who dress inappropriately. These norms ciety expects them to behave are all faces of the same include everything from cultural beliefs to expected be- system that enforces a norm. haviors and practices. Gender norms, in particular, have not changed greatly partly because they are widely held Although being a mother, a husband, or a student can be and practiced in daily life, because they often represent performed differently by different people, specific be- the interests of power holders, and because they instill haviors are associated with each. We expect mothers to unconscious learned biases about gender differences care for their children and students to attend school and that make it easier to conform to long-standing norms take exams. These behavioral regularities are what make than to new ones.19 them social roles. Social norms of gender are in constant dialogue with Gender roles are part of these expected behaviors women’s agency and may determine women’s capacity and, particularly, are sex-typed behaviors (Eagly, Beall, to act. As such, they operate as social determinants that and Sternberg 2004). Gender-ascribed roles define interact with an individual’s will in the form of a belief the ideal expected behaviors for men and women in system around women and men. It is with this under- any position they occupy in society or in any activity, On Norms and Agency standing that we observed social norms in the data col- overlapping with other expected role behavior. In oth- lected in the 20 countries, as they appeared over and er words, gender roles define what is deemed appro- over again in women’s and men’s accounts of their daily priate for women and men, and define what attributes lives in their communities. men and women should have and display in any situa- tion. As such, gender roles are norms that women and a. Norms and roles men comply with all the time, whether in the house- hold or the street, in private or in public. They perme- For the purpose of this study the main characteristics ate daily life and are the basis of self-regulation, hence 24 that define social norms are that: (i) They regulate in- affecting individual agency. The constant presence of gender as a backdrop to all Supporting the acquisition of gender norms is a pro- other roles makes gender roles unlike others. Gender cess of social punishments for transgressing the norms. has no specific site and it is not constrained to a physi- While in some cases this policing takes the form of so- cal space, such as a household. Gender is constructed cial sanctions, such as bullying, social ostracism and even in relation to the opposite sex’s attributes; as men and violence, there are also more subtle strategies operat- women are always present in society, so is gender. This ing over everyone. Parents and schoolteachers, for ex- is understood as “doing gender� (West and Zimmerman ample, perceive future disadvantages and emphasize 1987; West and Fenstermaker 1995). Doing gender means compensating behaviors, such as teaching girls to find being permanently accountable to what is expected that good husbands or boys to behave in a more masculine men and women will do—basically replicating and repro- fashion. Adolescent girls rapidly learn the limits, such as ducing the markers of what is considered the essential intuiting norms about their newly developed bodies and differences between the sexes. experimenting with new ways to walk, sit, or dress. And adult women negotiate daily with different sanctions and b. Reinforcing norms expected behaviors. If gender differences are translated into gender inequali- Social norms are enforced via different mechanisms, ties, the constant reinforcement of these differences such as coercion, overt punishment, institutional meth- (when we “do gender�) may lead to accepting these in- ods of control (e.g., the police), the power of the media, equalities as the norm. If women have more disadvan- and more covert expectations and rules transmitted in tages with respect to men, they will reproduce them to everyday interactions. Among the covert are two power- the point that both women and men believe that such ful concepts: the normal and the deviant. Deviant is any disadvantages are not only normal but how things should behavior that threatens expectations and norms of indi- be. And even when women’s opportunities and resourc- vidual behavior or that may challenge power. Ideas about es change—women earn income, acquire assets, etc.— what is normal and what is deviant are constructed by the belief may not change or may change more slowly those who have the power to impose their views and (Ridgeway 1997). The beliefs that underpin these norms have them accepted. may even persist by adapting to new conditions. Gender inequality has prevented women from partici- Part of the explanation for why these beliefs are so en- pating in key domains of society that define and gener- trenched comes from learning what it is to be a girl or ate the rules and definition of what is normal. The notion a boy, or a man or a woman, from very early in life.21 We of the normal has been monopolized by men. The overall learn the rules of the game and we then continuously re- societal norm is male; moreover, it is a particular sort of produce them, almost as if we were following an “ethics� masculinity (“hegemonic� as Connell 1987 terms it)23 that of gender behavior that controls our self-judging process- is regarded as normal. For example, in the labor market, es, as well as our awareness of judgment by others.22 the “ideal� employee is free from the time constraints of 19 See World Development Report 2012 (World Bank 2012, 174, box 4.7) for an explanation of processes that make social norms very difficult to dislodge, even when the conditions that gave rise to them no longer make objective sense. 20 Bicchieri (2006) defines the expectations that underlie norm compliance as: 1) empirical expectations, where individuals believe that a suf- ficiently large subset of the relevant group or population conforms to the norm in a given situation; 2) normative expectations, where individuals believe that a sufficiently large subset of the relevant group or population expects them to conform to the norm in a given situation situations; (3) On Norms and Agency normative expectations with sanctions, where individuals believe that a sufficiently large subset of the relevant group or population expects them to conform to the norm in a given situation, prefers them to conform, and may sanction behavior. 21 Socialization is the process by which prevailing social and cultural norms of what constitutes appropriate gender behavior is transmitted to children. 22 Following the categories created by Garfinkel (1967), West and Zimmerman (1987) call this process “accountability� of our gender practice. Our everyday behavior, according to Garfinkel, is “accountable� in the sense that it is intelligible and legitimate, and observed as fitting a specific pattern so it doesn’t need to be explained to anyone in order to be identified and accepted. 23 Connell (1987) uses the term hegemonic masculinity to describe ways that some forms of masculinity are more culturally exalted and socially dominant than others. “Hegemonic masculinity is constructed in relation to women and to subordinated masculinities. The other masculinities need not be as clearly defined—indeed, achieving hegemony may consist precisely in preventing alternatives gaining cultural definition and recognition as alternatives, confining them to ghettos, to privacy, to unconsciousness� (Connell 1987, 186). Connell also talks of “emphasized femininity� to describe patterns of femininity that have more cultural and ideological support than others. None can be hegemonic in a social context where women are 25 in an overall subordinate position in relation to men—where women are not in the positions of power that enable a definition of femininity in a way that serves women’s interests. running a home and caring for children. The definition of includes women in productive decisions over land and a “normal� employee enables the reproduction of gen- property, and goes further. der inequality. Furthermore, this ideal is reproduced in labor codes, which established the 8-hour working day, Our analysis clearly shows how normative frameworks the notion of a “family wage,� and more. around gender are changing—albeit slowly—and opening space for new practices and producing more opportuni- c. Changing norms ties for women and men. However, this change is being contested: backlashes are common and the change is But there are also challenges to the norms, and norms uneven. Movement in one area does not always mean bend, relax, evolve, and change. The communities in our movement in other areas or for everyone. research show that what was improbable 10 years ago is now possible—men help with housework and take care of children, women work for pay and manage their own 5. Overview of chapters money. Rather than “undoing gender� (as suggested by Butler 2004 and Deutsch 2007), it seems that change How do agency and norms work together to increase has come through modifications in the normative frame- women’s and men’s ability to change their lives? What works associated with gender. The powerful grip of gen- do women and men need to realize the power and free- der norms rests upon other social norms that organize dom to make choices—even when facing material or nor- society and help us live together. The collective ability to mative constraints—in their communities? Agency is as articulate alternative, oppositional norms is part of the much about choice and the power to act as it is about agency of the individuals inhabiting society. voice. And the decisions that women and men make, their ability to act, and their voices are dependent on Norms are negotiated and change through a variety of both the opportunities and constraints typifying the so- channels. Ridgeway and Correll (2004) suggest that ex- cieties they live in. posure to counter-stereotypical images, such as a work- ing mother or a female politician, and the delinking of Women’s preferences are not independent of the social negative associations with these images can change and material environment they inhabit; they are contin- the status of expected behavior to the point that the gent on it, as economists increasingly recognize (Fehr gender norm varies or becomes irrelevant. Our analy- and Hoff 2011). The sample communities in our research sis explores the appearance of conflicting norms, for showcase different enabling environments, such as differ- example, that arise from different role demands, social ent market dynamics and different normative prescrip- and technological changes that affect the cost-benefit tions on women’s actions, in which women make choices. balance of enforcing the norm, and willingness of media For example, to some women in rural and isolated com- or information outlets (Jensen and Oster 2009, Chong, munities or other restricted environments, access to a Duryea, and La Ferrara 2008) to show that other norma- road is highly strategic and liberating. For other women tive arrangements are possible—which are all means of in a large city with public transportation, a new road may negotiating norms. simply be practical and reduce the time or cost of their commute. For both sets of women, however, the road is Change can happen when the normative frameworks be- a route to (gain) power. On Norms and Agency come less strict, allowing gender differences to be less of a determinant. For example, when governments revise But sometimes a road is not enough. Social norms are the legislation on inheritance rights to allow daughters major factors affecting women’s agency. Their strength and sons alike to inherit their parents’ property, it not and ability to permeate all areas of individual action only weakens the social norm that says that male heirs make them determinants of the context in which agency should be given preference but also introduces variabil- can be exercised. Due to the presence of norms, agen- ity in inheritance practices. Such legislation changes the cy is not evenly distributed across spheres of life. The ownership of assets in society, questions marital prac- same road that can increase women’s opportunities to 26 tices based on men being the sole land proprietors, engage in labor and retail markets—through which they Box 1.2: Quick glossary – Agency is the ability to make meaningful choices and act upon them. – Empowerment, the expansion of power and freedom to use resources and take advantage of available opportuni- ties, comes as a result of gains in agency. Hence, empowerment is contingent on agency as much as it is on available resources and opportunities. – Structure of opportunities means the resources, institutions, established processes (traditions, moral codes, gen- der norms), and other enabling factors that foster the empowerment process and are necessary for agency to manifest (Alsop et al. 2006; Narayan 2005). – Social norms, the gender-ascribed formal structures, informal rules, gender role divisions, and permeating be- liefs, attitudes and behaviors, are treated as a separate element, although they are a part of the structure of opportunities. can increase their agency—does not always translate di- behaviors that do not conform to these norms, including rectly into a greater voice inside their homes. Variations domestic violence. in agency depend on the strength of the norms and roles that regulate each sphere. Part II probes how norms interact with agency in strate- gic life choices. Chapter 3 covers the effects of gender Conversely, even with existing normative frameworks, differences in making life-defining decisions that shape women are seeing their power surge. They told us in the young women’s and men’s futures, such as why adoles- focus groups that they see themselves gaining power cent boys or girls decide to leave school and how they and freedom, more so than the men. And women’s gains choose their first job. It also looks at decisions on fam- in economic empowerment are underpinned by more ily formation against a backdrop of reported expecta- active participation in society, both of which are driving tions and actual practices. The discussion includes the their empowerment. local norms that impinge on each decision, as well as the sense of agency and power to make those decisions in This report is structured in three sections. Part I focuses the words of the focus group participants. on gender norms and the dynamics of negotiation, ac- ceptance, and resistance around them. Chapter 1 sets Moving more directly into the dynamics of empow- the stage by synthesizing the range of views on gender erment, Part III examines the factors that individuals roles found in the 20 countries. By focusing on gender identify as primarily increasing their power and free- ideals—the good wife and the good husband, the good dom, as well as the opportunity structures associated girl and the good boy—the chapter reveals how little with them. Chapter 4 considers the dynamics of em- variation exists around the world in expected behavior powerment and agency, the dynamics of individual by the sexes. Looking at different generations, however, gains in power, and the explanations behind them. It On Norms and Agency shows that the slow change that is occurring is due more presents evidence that women see gains in their ca- to relaxation of norms than radical upheaval. pacities to shape their lives, while men report that they are stagnating or sliding backwards. Chapter 5 looks at Continuing with prevailing norms, chapter 2 presents dif- the community conditions associated with power loss ferent ways in which norms are negotiated and resisted. or gain, particularly markets, formal institutions of rep- First, it looks at non-conflictive and negotiated practices, resentation, conflict resolution, and legal regulations. including incipient change to what is deemed possible The chapter especially considers how market dynamics or tolerable for masculine and feminine behavior. Then, interact with changing gender norms to shape women’s the chapter moves to accepted practices for disciplining perceptions of empowerment. 27 The concluding remarks summarize the main findings and Bicchieri, C. 2006. The Grammar of Society: The Nature outline future areas for research and policy action. and Dynamics of Social Norms. New York: Cambridge University Press. Bicchieri, C., and A. Chavez. 2010. Behaving as Expect- References ed: Public Information and Fairness Norms. Journal of Behavioral Decision Making 23 (2): 161–78. Agarwal, B. 1997. “Bargaining� and Gender Relations: Bicchieri, C., and R. Muldon. 2011. Social Norms. Stanford Within and Beyond the Household. Feminist Econom- Encyclopedia of Philosophy (Spring 2011 edition), ed. ics 3 (1): 1–51. E.N. Zalta. Stanford, CA, USA: Stanford University, ———. 2003. Gender and Land Rights Revisited: Explor- Center for the Study of Language and Information, ing New Prospects via the State, Family, and Market. Metaphysics Research Lab. http://plato.stanford.edu/ Journal of Agrarian Change 3 (1 and 2): 184–224. archives/spr2011/entries/social-norms/. Alesina, A.F., Y. Algan, P. Cahuc, and P. Giuliano. 2010. Bourdieu, P. 1977. Outline of a Theory of Practice. Cam- Family Values and the Regulation of Labor. NBER bridge University Press. Working Paper, no. 15747. Cambridge, MA, USA: Na- Brickell, C. 2006. The Sociological Construction of Gen- tional Bureau of Economic Research. der and Sexuality. Sociological Review 54: 87–113. Alesina, A.F., P. Giuliano, and N. Nunn. 2011. On the Ori- Butler, J. 2004. Undoing Gender. New York: Routledge. gins of Gender Roles: Women and the Plough. NBER Caramazza, E., and M. Vianello. 2005. Gender, Space, Working Paper, no. 17098. Cambridge, MA, USA: Na- and Power. London: Free Association Books. tional Bureau of Economic Research. Chong, A., S. Duryea, and E. La Ferrara. 2008. Soap Op- Alexander, Jeffrey C. 2003. The Meanings of Social Life: A eras and Fertility: Evidence from Brazil. CEPR Discus- Cultural Sociology. New York: Oxford University Press. sion Paper, no. 6785. London: Centre for Economic Alkire, S. 2002. Dimensions of Human Development. Policy Research. World Development (30) 2: 181–205. Connell, R.W. 1987. Gender and Power. Oxford: Polity ———. 2009. Concepts and Measures of Agency. In Ethics, Press. Welfare, and Measurement. Vol. 1 of Arguments for a Connell, R.W., and J.W. Messerschmidt. 2005. Hegemon- Better World: Essays in Honor of Amartya Sen, ed. K. ic Masculinity: Rethinking the Concept. Gender and Basu and R. Kanbur. Oxford University Press. Society 19 (6): 829–59. Akerlof, G., and R. Kranton. 2000. Economics and Iden- Croson, R., and Gneezy, U. 2009. Gender Differences in tity. Quarterly Journal of Economics 115 (3): 715–53. Preferences. Journal of Economic Literature 47 (2): 1–27. Alsop, R., ed. 2005. Power, Rights, and Poverty: Concepts Deutsch, F. 2007. Undoing Gender. Gender and Society and Connections. Washington, DC: World Bank. 21: 106–127. Alsop, R., M. Bertelsen, and J. Holland, eds. 2006. Em- Doyal, L., and Gough, I. 1991. A Theory of Human Need. powerment in Practice: From Analysis to Implemen- London: Macmillan. tation. Washington, DC: World Bank. Duflo, E. 2011. Women’s Empowerment and Economic Appadurai, A. 2004. The Capacity to Aspire: Culture Development. NBER Working Paper, no. 17702. Cam- and Terms of Recognition. In Culture and Public Ac- bridge, MA, USA: National Bureau of Economic Re- On Norms and Agency tion, ed. V. Rao and M. Walton, 59–84. Stanford, CA, search. USA: Stanford University Press. Dworkin, G. 1988. The Theory and Practice of Autonomy. Arendt, H. 1979. The Origins of Totalitarianism. San Di- Cambridge University Press. ego, CA, USA: Harcourt Brace Jovanovich. Eckel, C.C, and P.J. Grossman, 2008. Sex and Risk: Ex- Benhabib, S., J. Butler, D. Cornell, and N. Fraser, eds. 1995. perimental Evidence. Monash Economics Working Feminist Contentions: A Philosophical Exchange. Paper, no. 09. Victoria, Australia: Monash University, New York: Routledge. Department of Economics. 28 Eagly, A.H., A. Beall, and R.S. Sternberg, eds. 2004. The Psy- 1998. The Male in the Head: Young People, Hetero- chology of Gender. 2nd ed. New York: Guilford Press. sexuality, and Power. London: Tufnell Press. Elson, D. 1999. Labour Markets as Gendered Institutions: Ibrahim, S., and S. Alkire. 2007. Agency and Empower- Equality, Efficiency, and Empowerment Issues. World ment: A Proposal for Internationally Comparable Development 27 (3): 611–27. Indicators. OPHI Working Paper 04, Oxford: Oxford Elster, J. 1989. Social Norms and Economic Theory. Jour- Poverty and Human Development Initiative. nal of Economic Perspectives 3 (4): 99–117. Isaac, J. 1987. Beyond the Three Faces of Power: A Real- Ergun, S., T. Garcia-Munoz, and M.F. Rivas. 2012. Gender ist Critique. Polity 20 (1): 4–31. Differences in Economic Experiments. Special Issue Jejeebhoy, S. 1995. Women’s Education, Autonomy, and on Behavioural and Experimental Economics, Revista Reproductive Behaviour: Experience from Develop- Internacional de Sociologia 70 (extra 1): 15–26. ing Countries. Oxford: Clarendon Press. Fehr, E., and S. Gachter. 2000. Fairness and Retaliation: Jensen, R., and E. Oster. 2009. The Power of TV: Cable The Economics of Reciprocity. IEER Working Paper, Television and Women’s Status in India. Quarterly no. 40. Zurich: Institute for Empirical Economic Re- Journal of Economics 24 (3): 1057–1094. search. Kabeer, N. 1999. Resources, Agency, Achievements: Re- Fehr, E., and K. Hoff. 2011. Tastes, Castes, and Culture: flections on the Measurement of Women’s Empower- The Influence of Society on Preferences. Policy Re- ment. Development and Change 30: 435–64. search Working Paper, no. 5760. Washington, DC: ———. 2000. The Power to Choose: Bangladeshi Women World Bank. and Labour Market Decisions in London and Dhaka. Foucault, M. 1995. Discipline and Punish: The Birth of the London: Verso. Prison. 2nd ed. New York: Vintage Books. ———. 2001. Reflections on the Measurement of Women’s Fraser, N. 1997. Justice Interruptus: Critical Reflections Empowerment. In Discussing Women’s Empowerment: on the “Postsocialist� Condition. New York and Lon- Theory and Practice. Sidastudies No. 3. Stockholm, don: Routledge. Sweden: Novum Grafiska AB and Sida. Garfinkel, H. 1967. Studies in Ethnomethodology. Engle- ———. 2005. Gender Equality and Women’s Empower- wood Cliffs, NJ, USA: Prentice-Hall. As cited in C. ment: A Critical Analysis of the Third Millennium West and D. Zimmerman. 1987. Doing Gender. Gen- Development Goal. Gender and Development 13 (1): der & Society 1 (2): 125–51. 13–24. Ginsburg, F., and R. Rapp. 1991. The Politics of Reproduc- Krendl, A.C., J.A. Richeson, W.M. Kelley, and T.F. Heath- tion. Annual Review of Anthropology 20: 311–43. erton. 2008. The Negative Consequences of Threat: Hartsock, N. 1996. Community/Sexuality/Gender: Re- A Functional Magnetic Resonance Imaging Investiga- thinking Power. In Revisioning the Political: Feminist tion of the Neural Mechanisms Underlying Women’s Reconstructions of Traditional Concepts in Western Underperformance in Math. Psychological Science 19 Political Theory, ed. N. J. Hirschmann and C. Di Ste- (2): 168–75. fano. Boulder, CO, USA: Westview Press. Labalme, P.H., ed. 1980. Beyond Their Sex: Learned Hechter, M., and K.-D. Opp, eds. 2001. Social Norms. New Women of the European Past. New York: New York York: Russell Sage Foundation. University Press. On Norms and Agency Heise, L. 2012. Draft. Background report prepared for Lewis, O. 1961. The Children of Sánchez: Autobiography World Development Report 2012: Gender Equality of a Mexican Family. New York: Random House. and Development, World Bank, Washington, DC. ———. 1966. La Vida: A Puerto Rican Family in the Culture Hoff, K., and P. Pandey. 2006. Discrimination, Social Iden- of Poverty—San Juan and New York. New York: Ran- tity, and Durable Inequalities. American Economic Re- dom House. view 96 (2): 206–211. Lippa, R.A. 2005. Gender, Nature, and Nurture. 2nd ed. Holland, J., C. Ramazanoglu, S. Sharpe, and R. Thomson. Mahwah, NJ. USA: Lawrence Erlbaum Associates. 29 Lister, R. 1997. Citizenship: Feminist Perspectives. Bas- ———. 2006. Capabilities as Fundamental Entitlements: ingstoke, UK: Macmillan. Sen and Social Justice. In Amartya Sen’s Work and Miles, M.B., and A.M. Huberman. 1994. Qualitative Data Ideas: A Gender Perspective, ed. B. Agarwal, J. Analysis: An Expanded Sourcebook. 2nd ed. Thou- Humphries, and I. Robeyns. Abingdon, UK, and New sand Oaks, CA, USA: Sage. York: Routledge. Moser, C. 1989. Gender Planning in the Third World: Portes, A. 2006. Institutions and Development: A Con- Meeting Practical and Strategic Gender Needs. ceptual Re-Analysis. Population and Development World Development 17 (2): 1799–825. Review 32: 233–62. Narayan, D., ed. 2005. Measuring Empowerment: Cross- Rao, V., and M. Walton. 2004. Culture and Public Action: Disciplinary Perspectives. Washington, DC: World Relationality, Equality of Agency, and Development. Bank. In Culture and Public Action, ed. Rao and Walton. ———., ed. 2009. Moving Out of Poverty: The Promise of Stanford, CA, USA: Stanford University Press. Empowerment and Democracy in India. Vol. 3. New Ray, D. 2002. Aspirations, Poverty, and Economic Change. York: Palgrave Macmillan; Washington, DC: World Photocopy. Department of Economics, New York Bank. University. http://www.nyu.edu/econ/user/debraj/ Narayan, D., with R. Patel, K. Schafft, A. Rademacher, and Courses/Readings/povasp01.pdf. S. Koch-Schulte. 2000. Voices of the Poor: Can Any- Reyben, R. 2005. Linking Power and Poverty Reduction. one Hear Us? New York: Oxford University Press for In Power, Rights, and Poverty: Concepts and Connec- World Bank. tions, ed. R. Alsop. Washington, DC: World Bank. Narayan, D., R. Chambers, M.K. Shah, and P. Petesch. Ridgeway, C.L. 2009. Framed Before We Know It: How 2000. Voices of the Poor: Crying Out for Change. Gender Shapes Social Relations. Gender & Society New York: Oxford University Press for World Bank. 23 (2): 145–60. Narayan, D., and P. Petesch. 2002. Voices of the Poor: Ridgeway, C.L., and S.J. Correl. 2004. Unpacking the From Many Lands. New York: Oxford University Gender System: A Theoretical Perspective on Gen- Press for World Bank. der Beliefs and Social Relations. Gender and Society ———. 2005. Moving Out of Poverty: Understanding Free- 18 (4): 510–31. dom, Democracy, and Growth from the Bottom Up— Ridgeway, C.L., and L. Smith-Lovin. 1999. The Gender Methodology Guide. Washington, DC: World Bank, System and Interaction. Annual Review of Sociology Poverty Reduction Group, Poverty Reduction and 25: 191–216. Economic Management Network. Robeyns, I. 2003. Sen’s Capability Approach and Gender ———., eds. 2007. Moving Out of Poverty: Cross-Disciplin- Inequality: Selecting Relevant Capabilities. Feminist ary Perspectives on Mobility. Vol. 1. New York: Pal- Economics 9 (2–3): 61–92. grave Macmillan; Washington, DC: World Bank. Rowlands, J. 1997. Questioning Empowerment: Working ———. 2010. Moving Out of Poverty: Rising from the Ashes with Women in Honduras. Oxford, UK: Oxfam. of Conflict. New York: Palgrave Macmillan; Washing- Samman, E., and M.E. Santos. 2009. Agency and Empow- ton, DC: World Bank. erment: A Review of Concepts, Indicators, and Em- Narayan D., L. Pritchett, and S. Kapoor. 2009. Moving Out pirical Evidence. OPHI Research Paper 10a. Oxford: On Norms and Agency of Poverty: Success from the Bottom Up. New York: Oxford Poverty and Human Development Initiative. Palgrave Macmillan; Washington, DC: World Bank. Schuler, S.R., and E. Rottach. 2010. Women’s Empower- Nussbaum, M.C. 1998. Public Philosophy and Internation- ment across Generations in Bangladesh. Journal of al Feminism. Ethics 108: 762–96. Development Studies 46 (3): 379–96. ——— . 1999. Sex and Social Justice. Oxford University Press. Scott, J.C. 1985. Weapons of the Weak: Everyday Forms ———. 2001. Women and Human Development: The Capa- of Resistance. New Haven, CT, USA: Yale University bilities Approach. Cambridge University Press. Press. 30 Sen, A. 1985. Well-Being, Agency, and Freedom: The Dewey Lectures. Journal of Philosophy 82 (4): 169– 221. ———. 2002. Development as Freedom. New York: An- chor. Sen, G., and C. Grown. 1987. Development, Crisis, and Al- ternative Visions. New York: Monthly Review Press. Shih, M., T.L. Pittinsky, and N. Ambady. 1999. Stereotype Susceptibility: Identity Salience and Shifts in Quanti- tative Performance. Psychological Science 10: 80–83. Steele, C.M., and J. Aronson. 1995. Stereotype Threat and the Intellectual Test-Performance of African- Americans. Journal of Personality and Social Psy- chology 69 (5): 797–811. van Staveren, I. 2012. An Exploratory Cross-Country Analysis of Gendered Institutions. Journal of In- ternational Development (forthcoming). Published electronically December 9, 2011. doi: 10.1002/jid.1850. Walby, S. 1990. Theorizing Patriarchy. Oxford: Black- well Publishers. West, C., and S. Fenstermaker. 1995. Doing Difference. Gender & Society 9 (1): 8–37. West, C., and D. Zimmerman. 1987. Doing Gender. Gen- der & Society 1 (2): 125–51. World Bank. 2012. World Development Report 2012: Gender Equality and Development. Washington, DC: World Bank. On Norms and Agency 31 One of Sisum’s most unforgettable childhood memories was an angry outburst by her father when she asked why men and boys always dined first in Samtse, her small vil- lage in southern Bhutan.24 Sometimes this meant that Sisum had to wait to eat until late in the afternoon if her PART I: father had business in neighboring towns. “I am not so On Norms and Agency used to this,� she recounted of her hunger and frustration with the delays, “because in my uncle’s house [in the city] Gender norms such practices are not followed. They are all educated and they feel it is not right.� Sisum lives with her uncle’s family in Thimphu, Bhutan’s capital, during the months she is in school. She is also “sad for mother, who is always working so hard in the house and she does not even get to eat a meal together with the rest of the family.� 32 This eventful day, Sisum found the courage to ask her closely adhere to norms prescribing what women and mother about the mealtime tradition, but her mother only men are expected to do, particularly when it comes to the explained that it “has been followed since our ancestors’ division of domestic and breadwinner roles in the house- time.� Searching for a reason that made more sense to her, hold. And these productive and reproductive gender Sisum then sought out her father for an explanation. Her roles differences are mirrored and replicated at the wider father, however, responded by completely losing his tem- community level. Change is happening, but at a very slow per over her question, directing his rage at her mother: pace. The data shows incremental and uneven changes in gender roles and norms, and a diversity of forces driving Before our marriage, you were nothing. Your family these transitions. On one hand, norms are being modi- members were poor and we always had to support fied by negotiation and adaptation by men and women them. I am fed up with your family members and this is in response to new, emerging views on gender equality the third time that you have gone against the culture (box PI.1). Wider forces are also driving change, including and tradition that we follow in this house. It is you who new legislation, education achievement, communications have instigated Sisum to question these things. How technology, and many others. On the other hand, change does a girl of her age learn how to talk like this? ... In is resisted in both discourses and practices. As discussed this house, you all are supposed to do what I say. in chapter 2, domestic violence, in many cases seen as I am the head of the family. Without me, you would not a man’s right over his wife, occurs as a reaction to chal- have proper shelter or even meals to eat. How dare you lenges to the norm; but widely held beliefs and everyday complain and question why women have to eat after practices like dinner rules also change. men. It is up to us whether we want you to eat after men or not eat at all. If young Sisum is confused by certain traditions in her fam- ily or how she ought to behave with her father, she is not Sisum’s brother had to step in during their father’s furi- alone. Sometimes uncertainty can be an advantage for ous eruption to prevent her mother from being beaten in flouting traditions that no longer make sense. And some- front of the family and servants. times uncertainty about acceptable conducts creates space for disagreement and violent enforcement of the norm. Sisum, now age 26, comes from a wealthy family, but Samtse’s 500 or so residents are mostly illiterate and Ridgeway and Correll (2004) note that beliefs in gender poor. Her education and exposure to new norms in rap- stereotypes are so resilient such that descriptive attri- idly urbanizing Thimphu have clashed with her father’s butes of the “typical� man or woman have remained sta- expectation of keeping traditional village practices. In ble since the 1970s.25 Chapter 1 reveals a similar finding: one world, Sisum was raised to conform to strict gender the focus groups’ reports of the traits associated with codes of subordination and respect for her father’s au- the ideal “good wives,� “good husbands,� “good girls,� and thority; yet, simultaneously in the city, her other world, “good boys� are remarkably constant across countries she was exposed to changing expectations about the and locations, and have remained largely unchanged, proper roles and conduct for a girl and her father. Sisum compared to previous generations. However, as shown will shortly complete her engineering degree and is de- throughout this report (particularly chapter 2), in every- termined to find a way to help change the traditions in day practices, there is more margin for negotiation, de- her village. She knows about organizations that work on spite adherence to the ideal, and change is inescapable On Norms and Agency women’s and children’s rights in Thimphu, but they have as more women participate in labor markets and more yet to reach places like Samtse. husbands help out at home. In some respects, Sisum’s life straddles the 97 communi- By definition, social norms are accompanied by surveil- ties in our dataset. Overall, the communities in our study lance and sanctioning practices to insure compliance, 24 Pseudonyms are used in place of particular individuals or communities named in this study. In some cases, community names have been re- placed with references to districts or municipalities. 25 Ridgeway and Correll (2004, 526–28) also cite a set of studies that looks empirically at the resilience of gender beliefs: Fiske et al. 2002; Luep- tow, Garovich-Szabo, and Lueptow 2001; and Spence and Buckner 2000. 33 Box PI.1: What is gender equality? Views from the ground When we asked the adult focus groups about the desirability of gender equality, their views differed strongly (Figure BPI.1.1 ). Below is a flavor of the perceptions, ranging from the large majority who were favorably disposed to the notion of gender equality to those with decidedly mixed views. Mutual respect, understanding, consultation, harmony, freedom, less stress and violence – “Equality between men and women means that they have a happy relationship and are comfortable talking to each other about their problems.� (Adult man, Labasa, Fiji) – “They should be able to do whatever they really want to do.� (Adult woman, Hyderabad, Andhra Pradesh, India) Sharing of work and household responsibilities – “Equality for me means that all of us should work and should enjoy the fruit of our work. I should not work alone while the man is just sitting there.� (Adult woman, urban Nsenene village, Tanzania) – “Happiness and equality are related. If the husband understands that happiness is supporting and helping his wife in housework and in taking care of children, the happiness of the family will be reinforced.� (Adult man, Ba Dinh district, Hanoi, Vietnam) – “Before a woman had no opportunity to work, and now she does. If a woman cooks, the man should wash and change a child’s diapers.� (Adult woman, Santiago de los Caballeros, Dominican Republic) Equal rights and nondiscrimination – “Equality between a man and a women means there should not be gender discrimination and there should be equal opportunities for both.� (Urban women, Nellore, Andhra Pradesh, India) About half of the rural men either qualified or flatly rejected gender equality as a goal for their society (see figure 3). In the village of Levuka, Fiji, a man insisted, “There cannot be any equality between a man and a woman because men make all the decisions.� A village man from Malangachilima, Tanzania, declared, “I do not think it means a lot to me. [Women] go to the office or into politics, but look at how our families are dying. They [women] do everything, but do you see the immoralities around? ... That is equality. It is a disaster.� In the interviews in the Roma community in Serbia, the defense of highly unequal gender relations and the institu- tions that support these relations was very strong. The young Roma men did not appear to be open to questioning their privileges or to seeing any value in more equal relations: “A wife does not matter the least bit in making deci- sions, it is my good will�; “who even asks a woman about anything�; and “it all depends on what the husband wants. FIGURE BPI 1.1: EQUALITY BETWEEN A If he wants to get rid of his wife and children, it will be WOMAN AND A MAN? as he wishes. If he wants to get rid of her, but keep the children, it will again be as he wishes.� A mixture of domi- Rural nant cultural views about gender roles and the need to Women protect a permanently threatened culture seem to lurk Urban behind these voices. Women In the Muslim communities sampled, women and men On Norms and Agency Urban Men sometimes quoted religious scripture and defined equal- ity in relation to “rights and duties�: men and women have Rural responsibilities that accord with the gender-ascribed Men roles of male breadwinners and female caretakers. In rural 0 20 40 60 80 100 Shirabad Ulya, Afghanistan, the men disagreed with “the Frequency of mentions present policy of the government and other non-Muslim Unfavorable view Mixed View Favorable view people [that] women [of Afghanistan] should be free like Note: Data from 194 adult focus groups. the women of Europe or America. But we don’t like this equality and it is not good.� 34 ranging from community pressure for expected behav- iors to explicit enforcement by violence. Women’s and men’s constant accountability to conform to norms has implications for their agency and ability to take action. The evidence from the research data strongly suggests that when households and communities find ways to relax and change inequitable gender norms, men’s and women’s individual and collective agency increase and reinforce one another. The analysis indicates that the interplay between more equitable gender norms and more widely shared voice and power is often accompa- nied by more inclusive and effective local-level institu- tions, which are embedded in and reproduce existing normative frameworks. In particular, greater gender equality in communities makes it more likely that claims by weaker groups will be heard, deemed legitimate, and addressed—even a small, inconsequential question by a young girl who is hungry. Recognition is the first step of the process.26 References Fiske, S. T., A.J. Cuddy, P. Glick, and J. Xu. 2002. A Model of (Often Mixed) Stereotype Content: Competence and Warmth Respectively Follow from Perceived Status and Competence. Journal of Personality and Social Psychology 82: 878–902. Ridgeway, C.L., and S.J. Correl. 2004. Unpacking the Gender System: A Theoretical Perspective on Gen- der Beliefs and Social Relations. Gender and Society 18 (4): 510–31. Lueptow, L.B., L. Garovich-Szabo, and M.B. Lueptow. 2001. Social Change and the Persistence of Sex Typ- ing: 1974–1997. Social Forces 80: 1–36. Spence, J.T., and C.E. Buckner. 2000. Instrumental and Expressive Traits, Trait Stereotypes, and Sexist Atti- tudes: What Do They Signify? Psychology of Women On Norms and Agency Quarterly 24: 44–62. Wall, J. 2012. Can Democracy Represent Children? Toward a Politics of Difference. Childhood 19 (1): 86–100. 26 This is in line with Wall’s (2012) emphasis on the need for societies to strengthen their capacities to recognize and address children’s needs and interests in the context of advancing deeper and more effective democratization processes. 35 CHAPTER 1 The rules we live by: Gender norms and ideal images “ Frankly speaking, women here are very miserable. They suffer from a lot of pressures. Pigs scream, kids cry, and husbands ask for sex. — Village man, Hung Yeng District, Vietnam � D rawing on the nearly 4,000 voices of the people husbands, were very consistent across countries and who participated in the qualitative assessment, communities. Chapter 1 shows how their views on the chapter 1 explores the prevalent gender norms behaviors expected of both sexes in a household con- surrounding women’s and men’s lives in the communities text are translated into and reinforced by community- where the focus groups were held.27 As key components level (collective) expectations of behavior as much as by of a society’s culture, norms and roles allow people to collective behaviors (as Bicchieri 2006 notes). organize their lives in consistent, predictable ways. But sometimes normative role behavior becomes rigidly de- fined and curtails freedom of action and agency. 1. Normative frameworks for household gender inequalities To capture the ideal views of gender roles in a household, we look first at how the focus group participants defined Gender norms and roles are reproduced in the private a “good wife� and a “good husband� in their communities. and public spheres and all other areas of life. Under- Masculinity and femininity are more than sex-appropriate standing how this framework operates at the household behaviors. They are also defined by a power relationship, level is important because it provides a sort of mental and it is in the domestic sphere where subordination and map of the acceptable roles, responsibilities, and behav- domination are more clearly revealed. Chapter 1 shows iors for each household member. The codes that govern On Norms and Agency how little the core practices that define the identities of men and women’s relations in the household are strict wives and husbands have changed.28 and gender-specific. They have varied a little, but when changes have occurred, they have not always remained The focus groups of adolescents (conducted in nine stable over time. countries) held conversations about what traits charac- terized a “good girl� and a “good boy,� and a “bad boy� or How much is the traditional intrahousehold normative “bad girl.� Like the adults, the adolescents’ views on what framework changing today? Are norms relaxing and makes the girls good and turns them into good wives, changing as gender equality increases around the world? 36 and what makes the boys good so they become good To answer these questions, all the adult focus groups explored stereotyped notions of a good wife and good Figure 1.1 lists the four most-mentioned topics in the fo- husband. The depictions below of a good wife and good cus group discussions of a good wife and good husband. husband reflect the participants’ most idealized views of The frequencies in the figure show the number of times gender roles and norms. They do not necessarily reflect a topic was brought up in the 194 adult focus group dis- the composition of their households, the realities of their cussions. The figure does not specifically assess favor- daily lives, or their aspirations for their lives (see box PI.1 able or unfavorable perceptions associated with these above). They do, however, describe the normative frame- attributes.29 work that binds both women and men. As observed in figure 1.1, domestic responsibilities re- We find that the normative frameworks governing the ceive, by quite a remarkable amount, the greatest em- roles within the household have remained largely un- phasis in discussions about the definition of a good wife. changed. Consistently across both men’s and women’s Opinions of this overriding role for women outnumber focus groups, and across the urban and rural contexts, all other descriptions, whether we asked urban men and and across different economic, political, and social cir- women or rural men and women. For a good husband, his cumstances of the 20 countries, men and women hold economic role is mentioned the most, but in comparison similar views of the wife’s and husband’s roles. Strict with the domestic role of the good wife, this received gender norms may be relaxing some, but they still retain less stress. a tight grip over women’s and men’s idealized roles and behaviors as husbands and wives. Almost every partici- a. The good wife pant described a good husband as the highest house- hold authority and responsible for being a benevolent The strong emphasis on women’s domestic work and decision-maker and a good provider for the household. care is all the more striking because focus groups were The focus group accounts of a good wife depicted her first and foremost as an obedient, caring, and respectful mate to the good husband. She is responsible for all the housework and the care of all members of the house- FIGURE 1.1: CHARACTERISTICS OF A hold, and is held strictly accountable for her domestic GOOD WIFE AND A GOOD HUSBAND responsibilities day in and day out. Domestic 408 responsabilities 233 The wife and husband roles are quite stable across the Attitudes or 296 focus groups, but we also see signs of flexibility around behaviors 245 these norms. In some places, the norms are relaxing, and Economic roles 242 some of the factors that are driving this relaxation seem 272 to be associated with increased education levels, wom- Merital relations 164 en’s participation in the labor force, and urbanization. 160 But we cannot affirm that there is a direct relationship with these drivers. Men, however, appear to have more Frequency of mentions leniency within their prescribed norms. In the aggregate, Good Wife Good Husband urban communities are ahead of rural communities in Note: Data from 194 adult focus groups (men and women). On Norms and Agency norm relaxation and negotiation. 27 The title of chapter 1 is adapted from Bicchieri (2006). 28 In fact, the exploration of what makes a good or bad wife, husband, girl, or boy was conducted within the context of the current community, the previous generation, and the future (of the adolescents). Focus group facilitators launched discussions in all focus groups with this question: “For a woman to be seen as a good wife in your community, what is she like? What does she do? Why?� All adult groups were asked how things had changed compared with the previous generation. Adolescents were asked how they saw themselves in the future at age 25 and what they thought of their parent’s lives. 29 It is important to note that the frequencies (or numbers) of mention are not statistically representative data. The samples were not randomly selected and the frequencies show only how often certain themes appeared in focus group conversations. We present the coded frequencies at different junctures to help convey the pattern of findings that emerged from systematic work with the narrative data in the field reports. 37 “ A good wife stays home, takes care of the house and children, cooks, feeds livestock. The important In Peru, wives must have a “good character, love their husband, help their husband, and be a homemaker.� In thing is that the woman should do this by goodwill. — Village men’s group, Floresti District, Moldova � Levuka, Fiji, a good wife is “a good listener and obedi- ent to the husband ... [and] a good advisor.� In Rafah, “ West Bank and Gaza, a good wife is “obedient, polite, [A good wife] looks after the children, does all behaves well.� housework, keeps her husband happy by doing everything, contributes to household income somehow, In most contexts in the communities sampled, women and thinks about husband and children first. — Village woman, Naitasiri Province, Fiji � who have children and husbands who provide well for the family generally do not work. In Bhubaneswar (Od- “ isha), India, the women said that a good wife does not If the husband is not good, the impact on the family have to take a job, but “men whose wives contribute are is lighter because the mother is the foundation of the happy because they feel a little relieved from their eco- family. If she is not good, it affects the family more. — Village woman, Dirbas, West Bank and Gaza � nomic responsibility.� Urban men were generally more likely than rural men to voice appreciation for wives who earn income and contribute to a household’s prosper- ity and happiness. Nevertheless, the overall picture from specifically prompted about the income-earning roles of the focus groups of a good wife’s economic role is quite both the good wife and the good husband, and how they mixed. Simple urban and rural differences in whether each balance work and family life. Domestic responsibili- women work for pay or do not work cannot capture the ties are clearly the dominant tasks assigned to women. complex realities of women’s lives. Often their quite- active economic participation may go unrecognized or Domestic responsibilities for women cover a broad range even be hidden because of the status their communities of activities, mostly associated with home care. In a vil- attach to being “just a housewife.� lage in the Sumadija District, Serbia, a woman described a good wife as a “housewife, obedient, loyal, good mother, Still, in many communities, a good wife may mean she good cook, cleans the house.� To a women’s focus group earns income. The urban focus groups, more often than in Nellore (Andhra Pradesh), India, a good wife “looks af- rural groups, mentioned the economic participation ter the family well, gives all her time to the family, and un- of good wives. (Chapter 5 looks specifically at working derstands the family problems.� Similarly, in Hato Mayor, mothers and women’s economic participation.) Their dis- Dominican Republic, men portrayed a good wife as “the course about working women, however, may just be glib, one who cares for the house, the children, and the hus- reinforcing expectations that women’s traditional domes- band.� And in a semi-rural community of the Ngonyameni tic role remains the more important one. This synthesis area outside Durban, South Africa, men said a good wife of a good wife from men in a neighborhood of Hoang Mai “makes sure that everything runs smoothly in her house- district in Hanoi, Vietnam, is typical: hold, she takes care of her husband and children.� In- deed, almost every focus group elaborated on a good A good wife should make her husband proud of her. wife’s paramount role of caring for her family. A good wife is not necessarily a high income earner, but she has to have a stable and decent work. She has On Norms and Agency Focus groups also attached the highest ethical attitudes to be a good daughter in-law. Most important, she must and behaviors to a good wife (see table 1.1 below). Wom- be a good mother who knows how to raise her children en from urban National Capital District, Papua New to be healthy and smart. Guinea, imagined her to be “honest, friendly, smart, sharing, caring, helpful, submissive, loving, understand- In a similar vein, the men’s group in Nsenene village, Tan- ing, faithful, [have] the heart of a servant, hardworking, zania, highlighted how their town’s expectations of an respectful, responsible, and wise [with budgeting].� urban good wife have become more relaxed and now Many focus groups stressed that a good wife respects include a provider role and activities beyond the house- 38 her husband and is faithful, supportive, and submissive. hold—in addition to traditional care duties: Table 1.1: Characteristics of a good wife and good husband described by adult men and women in Ba Dinh District, Vietnam A good wife A good husband Women Men Women Men Takes good care Is faithful Is responsible for Is faithful of her house Educates children family, kinship and Is a good earner Takes good care well society. Has sympathy and of her family and Takes good care of Contributes to helps wife and children the family family income. children Is a good cook Should be a good Has social status Cares for kinship Earns money breadwinner Is hard working /relatives Has social status Takes care of his Does not drink or Contribute to children gamble heavily the family income Does not come (not necessary) home late Is not adulterous; does not associate with sex workers She does all the cleaning. She prepares breakfast. where it is not customary for women to work for pay, She works on the plantation in the morning. the women declared that “income is not our responsibil- She prepares lunch. She goes to work on the plantation ity.� In a semi-rural community of Ngonyameni, South Af- in the afternoon. She attends association meetings rica, the women indicated that good wives can work, but in the late afternoon. She comes back to make sure there was disagreement in the men’s group. One man supper is ready. She serves supper. She goes to bed and from this community suggested that a working woman should have sex with her husband. can even contribute more income than her spouse, while another insisted that “good wives do not work; they stay Despite the economic role of the urban good wife, her at home and care for the children.� Many rural women principal priorities remain domestic and her authority is work on family plots and tend their own gardens (plus clearly subordinate to her husband’s. A good wife today, raising small livestock), but these farm or household ac- noted by men in Balti, Moldova, is likely to work for pay; tivities are often perceived as extensions of their house- she “may contribute to the family budget, but if the hus- hold roles. However, focus groups noted that a good wife band is a good provider, then she should not. Her role earning outside income also announces that the house- is to create appropriate conditions for her husband to hold is experiencing undesirable circumstances. For in- earn money.� In urban Mongar District, Bhutan, a good stance, poor widows may have little choice but to seek wife “stays home, looks after the children, listens to her jobs outside their households that provide some cash husband, and does not roam around. … During her free income (see box 1.1). time, she works to earn extra income for the family. [She can] weave, raise vegetables or poultry.� An urban good When asked how a good wife balances her many re- On Norms and Agency wife’s provider role is also second to her reproductive sponsibilities, from work to family life, a common re- roles. She will likely not work (earn income) if she has sponse from both urban and rural groups was that a many children or her children are very young. good wife can do everything skillfully and with ease. Whether she works for pay or not seems to be second- When we compare our urban and rural communities, we ary to household obligations. In rural Velugodu (Andhra find that rural settings more often stick close to the tra- Pradesh), India, a good wife, according to the women’s ditional prescribed norms. Particularly with the produc- group, “always chooses to work from home.� In Olsz- tive role, we see some dissent among focus groups par- tyn, a large city in Poland where women have been out ticipants. In rural Afghanistan and Yemen, for instance, in the workforce for decades, a good wife “copes per- 39 Box 1.1: Nontraditional households Focus groups most often discussed good wives and good husbands in relation to a monogamous couple in a nuclear family. On the rare occasions when they mentioned other types of households, women living in such arrangements were often portrayed as more vulnerable and powerless than when attached to a mate. If a woman in a village in (Odisha) India separated from her partner, she had to forfeit custody of her children; she could not expect alimony or a share of household property; she would “feel helpless even in her parents’ home�; and she might be forced to remarry an elderly man. In communities in the sample where polygamy is practiced, monogamy was named by both women and men as a quality of a good husband. Sharing husbands and resources with multiple wives or living with in-laws, parents, or other relatives can be disempowering for women. In Tangerang, Indonesia, a 38-year-old widow lamented, “At present I live with my kids and parents, and still rent a house. I used to be happy because I had a hus- band.� Sudanese focus groups talked about the great stigma and gossip that widows face because villagers assume they are “having relations with men.� In a few communities, family laws are making separations somewhat easier for women and men. A single woman, whether a mother or not, may in some contexts enjoy greater freedom of action, status, and control of assets than a married woman. A women’s focus group in Liberia explained that better-off widows can get on with their lives, “but if the deceased husband was poor and her kids are still young, then she would suffer a lot unless relatives stand beside her.� Similarly, in rural Papua New Guinea, better-off widows are the only women who can own land in the village. And in Afghanistan, focus groups explained that elderly widows enjoy a lot of independence and can travel in public because they are too old to shame family honor and are perceived to be the “mothers of society.� fectly with her obligations.� However, another woman In a semi-rural community of Ngonyameni, South Africa, retorted, “She lives 26 hours a day.� In urban Bukoba, men voiced nostalgia for earlier times when wives were Tanzania, the men’s attributes of a good wife included more obedient: “They respected their husbands. They that “she must do business as well,� but when asked did not argue with them. What is happening today is just how a good wife finds time for all this, one of them of- a shame.� Similarly, a woman from University Quarter, fered, “I think that, if she is employed, it is her fault. Let West Bank and Gaza, recalled, “[The good wife from my her do all her work also.� mother’s generation] used to remain quiet and not argue with the man.� Discussion groups also compared today’s and the previ- ous generation’s good wife. Many recalled that, in their Most focus groups of both sexes concurred that today’s mother’s time, a good wife was more submissive, patient, good wives are less obedient, less respectful, and less quiet, and tolerant of being ill-treated. And a good wife patient, and more likely to talk back and argue with their in the past typically did not earn income, which was often husbands. While there are exceptions, most women viewed as undesirable. In rural Chiclayo, Peru, men said viewed these changes in a good wife—and in gender rela- On Norms and Agency that a good wife in the previous generation was “dedi- tions generally—favorably and described their families as cated to the home … scared of her husband, and hard now closer and friendlier. According to a village woman working.� Women in Umlazi township A (near Durban), in Velugodu (Andrha Pradesh), India, “a good wife then South Africa, maintained that a good wife “would have was more accommodating and patient, and today’s good stayed in the marriage even if the husband was beating wife is smart and ambitious.� her.� In Olsztyn, Poland, one women’s group did not men- tion problems of violence, but they felt that a good wife This perception of change is crucial. For women to be- of their mother’s generation was treated like a servant or come empowered in the domestic sphere, they must use 40 “kind of slave.� their agency to negotiate the nature of gender relations “ in the household, which in turn may influence the de- A man who just stays home and has little responsibility cisions made within it. Women’s public roles may have around the house is good enough, because most men � changed in recent decades, but the limited changes in drink and hardly stay home. gender relations within the private sphere allow unequal — Village women’s focus group, Samtse, Bhutan gender relations to persist.30 b. The good husband Set against the many ideal qualities of a good wife, focus “ There is a difference. In the past, the men didn’t want to help in the kitchen. Now, men are more willing to help in the kitchen. In the past, boys weren’t even allowed to go into the kitchen; now boys are told to help in the kitchen. � groups depicted a good husband as the “real head of — Village women’s focus group, household,� “a worker,� “employed,� and “always work- Nagari Bukik Batabuah, Indonesia ing hard for his family.� In addition, women in rural Su- madija District, Serbia, said he should be “handsome, open to compromise, capable, responsible, reasonable, who brings in steady income and provides for the whole mature, smart, permissive, and realistic.� In Chiclayo, family.� The women also indicated that a good husband Peru, the women believed that a good husband should does not necessarily have to make a good income, but work with his wife “as a team� to make decisions and “he can contribute in a lot of other ways, such as spend- raise their children. He is also described as a loving and ing time with his family and doing jobs that require physi- engaged father. cal strength.� In this village, conditions were difficult, jobs scarce, and domestic violence common. Yet, relative to a good wife, both women’s and men’s focus groups were much more likely to qualify a good Compared to rural men, urban good husbands shoulder husband by what he should not do. They often listed un- more parenting and housework obligations. In Aden, Ye- desirable behaviors that a husband needs to avoid in- men, women said that a good husband “helps the mother stead of affirming positive characteristics or mentioning raise the children... [and is] loving and attentive to his the prescriptions of the husband role. A good husband children.� Likewise, in Hyderabad (Andhra Pradesh), “does not cheat,� “drink [alcohol],� “gamble,� “scold and India, a woman noted that a good husband “nowadays beat his wife or children,� or “stay out late.� According even takes care of the children if it is necessary.� In rural to a men’s group from Ba Dinh District, Vietnam, their Tanzania, where gender norms remain quite traditional, view of a good wife is a woman seemingly “perfect in all a good husband is involved in few household tasks and aspects,� and her partner is expected “to be faithful,� is appreciated if he only visits with friends “until early and to cease the drinking, gambling, adultery, and late evening and then comes home for supper.� nights (table 3). The focus groups also reflected on the good husband of Focus groups across all countries agreed that men’s role the previous generation. They generally reported him as as main provider and responsible for the economic secu- being a good provider and hardworking, but more strict rity of the household is central. Reflecting on the strong and less affectionate toward his wife and children than hold that this specific aspect of the male role has, par- today’s good husband. Groups disagreed about whether ticipants voiced concerns about a good husband’s ad- a good husband of the past was more violent toward his On Norms and Agency equacy with breadwinning. In communities with a weak wife or more likely to have affairs than today; however, local economy, it was often enough if the good husband most conveyed that good husbands used to drink less. earned something and, ideally, stayed out of trouble. In The women’s focus groups were much less likely than a village outside Paro, Bhutan, women viewed a good men’s to look favorably upon the good husband of the husband as “one who is very understanding, supportive, past who “hardly listened to his wife.� In contrast to the hard working, who does not have extramarital affairs, women’s views of favorable trends, the men often ex- who does not resort to physical (domestic) violence, and pressed nostalgia for the days when husbands enjoyed 30 Rowlands (1997) shows this in Honduras, and Das (2008) in Bangladesh. 41 more respect and knew better “how to keep his family that the physical changes of puberty appear, gender under control.� A man from University Quarter, West role defi¬nition intensifies. Girls move from childhood to Bank and Gaza, mentioned a time when a good wife adult roles, such as wife, mother, and worker; and boys “helped her husband and did not make him feel that he become workers, providers, and fathers.31 is lacking anything.� When imagining a good girl, the focus groups of ado- In sum, men almost everywhere continue to be the lescent girls and boys provided remarkably consistent dominant household authority figures and breadwinners. responses (figure 1.2).32 Most often, they mentioned For their part, women continue to be held to strict ac- that a good girl models expected gender behaviors for count for the large majority of household work and care, a woman inside the household: she helps around the no matter what other roles they may play beyond the house and is well-behaved, obedient, and respectful. A household. Depending on the neighborhood or village, good girl also goes to school. Likewise, a good boy goes the relaxation of gender norms was portrayed as either to school, but like a good man, he is also described also incremental or a somewhat faster blurring of women’s by certain things that he is expected not to do (figure 1.3). and men’s roles and responsibilities. But these changes Girls see a good boy as helping around the house—much are generally happening more quickly in urban contexts. like a good girl—but boys tend to think this is less impor- tant than being respectful or doing other things. These Finally, the focus groups conveyed idealized notions patterns suggest that girls perceived fewer gender dif- about good wives and good husbands who enjoy harmo- ferences between the sexes in household roles and iden- nious and cooperative relations, but their ensuing dis- tify with more modern gender norms than did boys. cussions of how marital relations have changed over the generations suggested a more stressful picture. The men In Thimphu, Bhutan, girls depicted their good girl as “a expressed a particular concern that the good husband is very reliable daughter; she can take care of the house and losing control over his life and his family, and their testi- at the same time behave well outside too.� These girls re- monies seem to question their present roles in society. quired that both good girls and good boys “take care of Their frustrations cannot be taken lightly. In general, the parents.� In rural Velugodu (Andrha Pradesh), India, boys discussions about a good wife and good husband display said, “A good girl is religious, soft spoken, and obedient; the same patterns that we find elsewhere in the dataset: she does all the domestic chores under the supervision women perceive that they are gaining more indepen- of her mother. A good boy is religious, respectful to el- dence and freedom, while men often expressed a grow- ders, and helpful.� In many contexts, boys do help around ing sense of powerlessness. the house, but more often with discrete tasks and not the time-consuming, constant daily chores expected of girls. In Yemen and the Dominican Republic, none of the boys’ 2. The good girl, the good boy focus groups associated their good boy with being helpful to their family. In Suva, Fiji, as the opening quote to this “ chapter attests, the girls imagined a good boy as useful [The traits of a good boy are the] same as girls, around the house, while the boys offered no such indica- but [the good boy] also does house-work like raking ... � tion in their list of traits for a good boy and imagined that plus his own laundry, like washing his school uniform. a good girl “stays home, cooks, and washes dishes.� — Urban girls’ focus group, Suva, Fiji On Norms and Agency Other markers of a good girl and boy, as with a good wife Girls and boys learn at an early age how they are ex- and husband, include many desirable behavioral traits, pected to behave according to their sex. Gender norms such as honesty, good morals, and respectful and de- are passed on by parents, school teachers, and peers. cent treatment of others. And good children (boys and Furthermore, when they reach adolescence, any flexibil- girls) are expected to be obedient, deferential to adult ity that the girls and boys may have had while growing up authority, and good students. Yet, a good girl also should tends to disappear, and compliance with gender norms dress decently, act politely, and not date, which were not 42 is tightly enforced. In adolescence, at the same time included in a good boy’s attributes. (If anything, a good boy should not date too many girls.) And, unlike good In the descriptions of a bad girl and a bad boy, the fo- girls, good boys were cautioned by many focus groups cus groups specified stronger gender differences in their not to smoke, drink, use drugs, or steal. Focus groups defining traits (figures 1.4 and 1.5). Nearly one-quarter were also more permissive of a good boy’s behaviors and of all the comments by boys about a bad girl identified interactions with the opposite sex than with a good girl. her as promiscuous (e.g., many boyfriends, a prostitute); In Ouagadougou, Burkina Faso, the boys said that a good however, less than 10 percent of girls’ comments men- boy does not “run often with girls�, while girls indicated tioned promiscuity for a bad girl. Next in frequency were that a good boy “does not force the girls�; but in Lautoka, references, by both girls’ and boys’ focus groups, to bad Fiji, the good boy “might have a girlfriend.� girls’ bad behavior: gossipy, dishonest, low morals, dis- FIGURE 1.2: CHARACTERISTICS OF A GOOD GIRL 25% Share of total mentions 20% 15% 10% 5% 0 Helps at Good Respectful Obedient Studies Decently Good Polite Does not Religious home behavior dressed friends date Accordings to girls According to boys Note: Data from 82 focus groups. FIGURE 1.3: CHARACTERISTICS OF A GOOD BOY 25% Share of total mentions 20% 15% 10% 5% 0 On Norms and Agency Good Respectful Obedient Religious Helps at Studies Not bad Good behavoir home habits friends Accordings to boys According to girls Note: Data from 82 focus groups. 31 Bruce, Mensch, and Greene (1998) note that, during this period, health and social behaviors are established that have a lifetime of consequences. Puberty triggers a marked divergence in gender-based trajectories. 32 The field work with adolescents was conducted in eight countries: Burkina Faso, Dominican Republic, India, Fiji, Bhutan, North Sudan, Yemen, and West Bank and Gaza. Facilitators asked the children to imagine: “what is a good girl like?; what makes her a good girl?; what is a bad girl like?; and how would you describe them?� The same questions were asked about a good and bad boy. 43 obedient, and disrespectful (talks back, does not obey adult years and need to manage relationships with the parents, swears). Boys were more likely than girls to opposite sex. Girls and boys are both under pressure to think that a bad girl shirks her domestic responsibilities. conform to similar desirable expectations. Yet, for girls, In rural Koudipally Mandal (Andhra Pradesh), India, the expectations of gender-defining household roles are boys said that a bad girl “won’t help her mother with the changing. In their testimonies, the girls themselves re- household chores.� defined housework as a practice that ideally should be normative for both boys and girls, and their good boy re- If boys expressed more concern about a bad girl’s pro- flects that ideal. Boys are not as eager to include domes- miscuity, the girls’ focus groups significantly stressed tic responsibilities in their concept of a good boy. Also, a bad boy’s vices—smoking, drinking, and drugs. both adolescent groups felt that good and bad boys are Boys saw this as a problem, too, but not nearly as im- at risk for the same risk and violent behaviors that bur- portant. Next in frequency for a bad boy are deeply den their fathers. aggressive behaviors, such as killing, gang fighting, and rape. Both boys’ and girls’ focus groups attributed vio- lence to a bad boy, but only rarely was it mentioned 3. Community-level views about a bad girl. In urban Lautoka, Fiji, for instance, the of gender norms boys described a bad girl as “wearing a mini skirt, being too outgoing, going to nightclubs, and being a discipline What constitutes ideal gender-appropriate behavior in problem,� but these characteristics were much less se- the household does not stop when individuals leave the vere than those portraying a bad boy: “smoking, drinking domestic sphere. As the girls’ and boys’ focus groups dis- alcohol, sniffing glue, roaming around, having a future in cussed, their behavior outside their home marks them as jail or a future of poverty and crime.� being good as much as their domestic responsibilities do. Prevailing views on gender-appropriate behavior, when Norms of good and bad femininity and masculinity take reproduced at the community level, can open or close root at very young ages and help explain the persistence opportunities for women. of gender differences in attitudes and behaviors from one generation to the next. The different normative In this section, we explore two areas where community yardsticks that are applied to girls and boys set the stage sanctioning is more likely to have an impact on women’s for stressful gender relations as adolescents head into empowerment, capacity to take advantage of new op- FIGURE 1.4: CHARACTERISTICS OF A BAD GIRL FIGURE 1.5: CHARACTERISTICS OF A BAD BOY Does not help Promiscuous at home Promiscuous Stays out a lot Disobedient Does not study Not religious Stays out a lot Does not study Indecently Steals On Norms and Agency dressed Bad behavior Disobedient Disrespectful Disrespectful Aggresive behavior Bad behavior Bad habits 0% 5% 10% 15% 20% 25% 0% 5% 10% 15% 20% 25% 30% Share of total mentions Share of total mentions According to boys According to girls According to girls According to boys Note: Data from 82 focus groups. Note: Data from 82 focus groups. 44 portunities, and decision-making: 1) the ability to com- men,34 concerns for safety that prevent movement, and bine their productive and reproductive tasks, and 2) concerns over women’s honor being tarnished in public. freedom of movement. As shown in World Development Other traditions or established practices regarding wom- Report 2012, women’s ability to use their time to work is en’s economic participation are equally powerful. For central to their economic empowerment, while freedom example, in Hung Yen District, Vietnam, where women’s of movement is one of the dimensions of their agency.33 participation in the labor force is almost 70 percent, the range of women’s productive activities is limited because a. Views on working mothers “according to long-standing social norms and prejudices, women are supposed to take care of the family and breed Given how norms affect women’s roles, we examine the chicken or pigs at home. They can only do small things.� qualities that young adults and adults find desirable and As a consequence, women in these contexts usually opt undesirable in a mother, and their opinions of mothers to work from home or in jobs that do not contravene the who balance family and work (if any). In many communi- severe norms of their households and communities. ties, the desirable qualities of a mother are the tradition- al ones of care, affection, and household management. Where no such restrictions are present, acceptable pro- But the younger groups in our study overwhelmingly fessional jobs for women are often just extensions of highlighted as noteworthy women’s (mothers’) desire and their caretaker roles at home: teacher, nurse, and cook. ability to earn and income and contribute to the house- Moreover, women’s household (gender) duties are as hold wellbeing together with caring for the family life. powerful as any religious belief in limiting women’s move- ments and ability to work for pay. Women negotiate work While opinions split over defining the primary role for choices from the options that meet their communities’ women, the participants clearly preferred that women requirements for appropriate work and accountability accomplish both work and care tasks (figure 1.6). For for household duties. adult men, however, the overwhelming majority (regard- less location) considered care to be women’s major and When women do find employment, they face a new chal- only role. Little was said about the conciliation of roles lenge. Can they perform well in their dual roles of workers between men and women. and mothers? Women sometimes fail in their attempts to do both equally well. Whether or not women fulfill their The choices that women and men make regarding work role as mother and wife often is the deciding factor for are informed as much by their definitions of masculin- accepting working mothers as a positive development. ity, femininity, and their views of care responsibilities at Taking on outside and household work simultaneously home as by market opportunities. Leaving the economic can exact a high cost from women. A young woman from context in the background for now—and how it constrains the University Quarter, West Bank and Gaza, saw no or facilitates work choices for women—we probe deeper problem with a mother who decides to start working, but into the justifications men and women give for choosing she noted that there are trade-offs one way or another: to work or stay home, and the normative (traditional) “No woman is a superwoman [and] no woman who works gender limitations imposed upon women in job searches will be able to balance that with another role. One has and actual employment. to happen at the expense of the other. For example, my mom’s work as a teacher happened at the expense of On Norms and Agency In many communities, women approach the prospect of her health.� work through the lens of norms that restrict their initia- tive, mobility, and autonomy. Examples of these restric- The weight of a double workload, hard as it is, has even tions appeared throughout focus group conversations more burdens, including what it means for women to step in all communities and included taboos on commercial out into the public space. Working mothers must keep or social exchanges between women and unrelated their (and their family’s) reputations intact. Even in urban 33 See World Development Report 2012 chapter 4 for further discussion of women’s freedom of movement and chapter 5 for time use and work. 34 The consequences of such “inappropriate� exchanges range from refusing permission for women to work outside the home to insisting on segregated jobs. 45 FIGURE 1.6: PERCEPTIONS ON WHAT WOMEN’S ROLE SHOULD BE Rural young men Rural young women Rural adult men Rural adult women Urban young men Urban young women Urban adult men Urban adult women -1 -0.5 0 0.5 1 Share of total mentions Women’s sole role is care Women should combine work and care Note: Data from a total of 370 focus groups. contexts, working women and mothers must behave with norms, and between generations. For young women, in decorum and politeness, pass up after-work recreational particular, the decisions women make to work are not in activities, and return home at an appropriate hour; oth- sync with their community’s normative framework. But erwise, they face being stigmatized as negligent, materi- it does not deter them, especially when they need an alistic, hedonistic, and even sexually promiscuous. income or are determined to challenge the norms and affirm their independence. Young women see them- The contents and the consequences of the stigma at- selves as not needing or wanting to depend on men; tached to working mothers vary between communities, they believe a woman should provide for herself first. but it tends to center on the effects on children. In many And although they know older generations oppose this cases, the widespread view holds that a working mother way of thinking and believe that women should only be has abandoned her children. In Jaipur (Odisha), India, housewives, many young women feel it is important to young women described the attitudes of their community be able to provide everything needed for a home. More as “friends and neighbors who don’t have high opinions and more, younger women support the right of women of [working mothers] who think money is more important to get a job and earn a salary; younger generations do than children and don’t have the maternal feelings for not condemn such behavior as vehemently as older gen- children that mothers had 20 or 30 years ago.� Some sam- erations do. ple communities expressed concern that the children of working mothers will be negatively affected by their ab- Young urban women’s views reveal that they are expe- sence. They turned working mothers’ own worries about riencing—by themselves—the changes in the normative the well-being of their children against them in the form frameworks of their societies. Their male counterparts of social criticism: their children may end up spending are increasingly aware of notions of equality and are On Norms and Agency too much time in the street, engage in criminal behavior, slowly showing willingness to share in the responsibili- and in general deviate from the right path because their ties of household and child care. Not only do some men mother is not present. These negative perceptions and (especially younger men) agree that women should work, pressures have serious impacts on women’s self-esteem but they also think that giving women access to areas and behavior, as well as their agency. where they have been excluded (such as certain types of jobs) benefits everyone. Even among young rural men, In our sample communities where normative change the extent of approval for working mothers is higher than around gender seems to be unfolding, accounts show a with older generations, and they are willing to voice their 46 mismatch between personal, household, and community support even in the face of community opposition or “ criticism. This generational change signals a trend in the As long as boys are everywhere, we can’t move direction of greater gender equality. While partly driven freely. They say that you are walking around and � by a shift in norms, a large part of the changes in opin- showing yourself off. ions about working mothers is stimulated or facilitated — Young urban woman, Aden, Yemen by economic conditions. From Poland to Vietnam, there is evidence that young women expect men to take on as much housework as “ Women go out less. They go out only if we take them. — Village man, Comendador, Dominican Republic � “ women do if they both are working: She [wife] stays at home and does the house chores. – [It is a] partnership. He cooks dinner for me; he washes, clears up, dusts. I would not choose to marry him, if we did not share these responsibilities. She does not move around the community gossiping. — Village man in a semi-rural community of Ngonyameni, South Africa � (Adult woman, Martynice, Poland) reputations and family honor also curtail their freedoms. – [Husbands and wives] should do everything together. These barriers are easing, however. From comments Everyone is tired when they get home after work, so heard in the focus groups, women and girls are able to they share the work. They can help each other with leave the house more than previous generations. Yet, the many things. (Adult woman, Hung Yen District, Vietnam) difficult reality is that they have much less freedom than men and boys to go out in public. Younger generations are embracing change. When asked what they think about working mothers, the re- Many of the focus groups with younger and older adult sponses of young adults in the focus groups were un- village women reported that they face heavy pressures equivocally positive. From Bhutan to Fiji, from North not to move around their communities independently. Sudan to Yemen, young women are seeing and demand- “The men can move freely in and out of the village and ing change. While it is still common practice for women even travel far away from their homes, too. Women can- to stay home and take care of household chores, this is not go anywhere, much less alone, because of tradition changing as more and more women question the strict and culture,� said a young village woman in Boyina Bagh, gender divisions in responsibility for the household, be- Afghanistan. In Koudipally Mandal (Andhra Pradesh), In- come more educated, and work in more diverse occu- dia, young women explained that the school is too far pations—especially those considered inappropriate for away from the village for girls to attend and “in our Reddy women in the past. community,35 even if we are poor, they don’t allow us to go out and work. Men care [too] much about the social b. The good woman (not) in public status.� As a result, women have developed alternative strategies: “Some of the women like me have started A bad girl goes out a lot and a bad wife spends more to learn tailoring and are doing the work at home.� In time away from home than inside the house; men and a semi-rural community of Ngonyameni, South Africa, boys, on the other hand, routinely have enjoyed more women were asked what it means to have little freedom, freedom in their ability to move. Women’s limited mobil- and one replied, “It means she is controlled by her hus- On Norms and Agency ity has long affected their social connections, enjoyment band and cannot visit her friends.� of free time, economic options, and possibilities for civic engagement. Diverse, interlocking factors tie women to In urban areas, women are also pressured to stay home their homes, including local traditions governing women’s or nearby in their neighborhoods. Young women in the and girl’s roles, religious practices, limited public safety, focus group in Jaipur (Odisha), India, said they are free lack of time, and poor transportation infrastructure. Gen- to move about their neighborhood during the day, but der norms surrounding concerns for women’s and girl’s “if [a woman] comes back late at night, then neighbors 35 A higher caste predominantly found in Andhra Pradesh. 47 talk bad about her, and parents and other family mem- Problems of safety also disproportionately affect wom- bers get pushed [to make her behave].� Young women in en. “By six o’clock in the evening, you should be indoors Dirbas, West Bank and Gaza, agreed that women were if you are female,� warned a woman from Umlazi Town- restricted as if they were small children by “social cir- ship B(near Durban), South Africa. And in Lautoka, Fiji, cumstances … customs, traditions, and public opinion.� “A man can travel alone at night, but not a woman be- In this community and many others in the sample, seclu- cause of safety issues.� Almost nowhere is it safe for sion practices require that women be accompanied by a women to move about at night, and even during the male guardian when in public, although educated women day large sections of their neighborhoods may also be with good jobs and poor or widowed women who need off limits because they are unsafe. Given the incidence to work enjoy more mobility. Still, young men in Dirbas of mugging, groping, and verbal abuse, public transpor- warned that it is better for a woman to “stay at home� tation can be especially risky for women and girls. Risk instead of “tiring herself� and bringing “problems to the of assault is reported even in communities considered family from her work.� safe by the focus groups, indicating how vulnerable women are to harassment and violence, which are of- In diverse contexts, women indicated that they may face ten acceptable. harsh discipline for leaving housework and children for short periods. They may be openly shamed and mocked Although a minority, focus groups with young adults for speaking with strangers, wearing insufficiently mod- sometimes indicated that young women are flouting the est attire, or making the slightest missteps that may be norms that limit their freedom. Although not as much as deemed sexually provocative or a sign of loose charac- men, more women are driving cars in Moldova, Poland, ter. “I may have freedom as a woman, but if I want to go and Serbia, for example, and more villages are providing and dance, my husband will give me a black eye,� said a electricity, water, and transportation services that free woman from the National Capital District, Papua New women from the worst drudgeries and give them more Guinea. Aside from household obligations, concerns for time to venture beyond their households. women’s reputations and family honor give rise to strong normative pressures on women to refrain from working Yet, in most places sampled in this study, getting out the or socializing outside the home. front door remains a daily struggle for women. And men, as well as women, acknowledged that women’s restrict- Even in communities in the sample where large num- ed physical mobility in public is a manifestation of much bers of women have been part of the workforce for de- wider gender power struggles. According to a 39-year- cades, the extent of some women’s mobility is still very old man from Dirbas, West Bank and Gaza, constrained. “Some women cannot leave the house or even talk to or hang out with neighbors. We call them a man can go anywhere to work … [yet he] gets jealous ‘kuper’ (lacking a social network),� explained a woman of his sister if she goes to a nearby place to work. from East Jakarta, Indonesia. In this same women’s So how do you think it will happen that she can go to focus group, three women volunteered that they are a faraway place? … But the man’s responsibilities are largely secluded in their homes, but for three different different. He has to work; otherwise, he will never build reasons. One woman saw herself as having no freedom his home or get married. But for the girl it is different. because of her husband’s views: “I can’t attend gath- On Norms and Agency erings at the mosque or go sightseeing. My husband Similarly, a man from another urban neighborhood in the doesn’t let me. I obey him because he is my husband.� West Bank and Gaza argued that the lack of public safety The second woman’s limited mobility was due to time for women is a troubling consequence of the transition constraints of meeting her responsibilities at home: underway in gender norms: “Women have just started “I don’t really have freedom because I always have to entering society, so the man is still trying to maintain his do the laundry.� The third participant was mostly con- control.� cerned about community perceptions: “If I go out by myself, the neighbors will talk. So now I sell stuff or Whether in their homes or out in public, deeply embed- 48 wash clothes [from home].� ded gender norms continue to guide the preferences and behaviors of girls and boys, women and men, in the com- munities visited by the research teams. But around these norms, we also see diverse manifestations of change along with tension over these changes. Young women, in particular, harbor ideals for a world where they too can enjoy whatever opportunities present themselves. Men, on the other hand—across generations and especially in rural areas—appear less willing to accept women’s chang- ing roles and aspirations. References Bicchieri, C. 2006. The Grammar of Society. Cambridge University Press. Bruce, J., B.S. Mensch, and M.E. Greene. 1998. The Un- charted Passage: Girls’ Adolescence in the Develop- ing World. New York: The Population Council. Das, M. 2008. Whispers to Voices: Gender and So- cial Transformation in Bangladesh. Bangladesh De- velopment Series Paper, no. 22. Washington, DC: World Bank, South Asia Sustainable Development Department. Rowlands, J., 1997. Questioning Empowerment: Working with Women in Honduras. Oxford: Oxfam. World Bank. 2012. World Development Report 2012: Gender Equality and Development. Washington, DC: World Bank. On Norms and Agency 49 CHAPTER 2 Negotiating the norms that bind: A winding road “ Men in the past were very strict about tradition and that everything they do be done the traditional way. … Nowadays, parents— including fathers— let their daughters have more � freedom, [for example] if they want to dress a certain way. … Because life in general is changing, it’s hard to keep the traditional way of raising children. — Adult woman, Fiji N egotiations—explicit or not—have been happen- Yet, despite encouraging signs of norms bending toward ing over gender norms over the last generation greater gender equality, this chapter also addresses in our sample communities, as this chapter de- countervailing forces that impede faster change. The tails. Significant shifts are taking place in these norms comments from the focus groups make clear that new and, more importantly, are not generally the result of practices emerge and exist with ease alongside tradi- conflict between couples. Instead of abrupt and conflict- tional inequitable practices in the same households ridden changes, we find that norms are relaxing slowly and communities. Within most every group, participants and sometimes inadvertently, and are creeping into the raised discordant perspectives and opinions about how everyday lives of the people throughout our sample. But much change is perceived or desired. To the progressive people do not always notice the subtly shifting forces, so voices driving change, normative ideals and aspirations often the process goes unrecognized.36 may be well ahead of actual behaviors. Focus groups from diverse contexts also elaborated on how slowly- The evidence collected from the many focus group con- changing norms that still sanction male dominance and versations on intrahousehold and gender relations re- aggression were significant triggers for marital conflict, veals a widespread tendency toward greater acceptance particularly under conditions of economic stress. of more equitable gender relations and modification of gender-ascribed responsibilities. Members of almost To explore these dynamics, we first discuss how we dis- every group testified to the presence of these changes tinguish a relaxation from a change in gender norms in On Norms and Agency when asked to reflect upon generational changes in the the dataset, and then present focus group accounts of lo- definitions of gender roles and people’s aspirations. The cal trends in daily time use, household roles, marital rela- large majority of the adults and young adult participants tions, and parental aspirations for their children. We then expressly wanted more cooperation and discussion be- explore the participants’ perceptions of norm change and tween men and women. Moreover, as discussed in chap- how uneven they are. We close by synthesizing accounts ter 1, men’s involvement in domestic work and childrear- about the persistence of domestic violence against wom- ing is increasingly valued, as is the women’s right to work en (in nearly one-third of the study communities), which 50 and pursue productive activities. is perhaps the starkest evidence of gender inequality and “ lack of agency in the case of women. The chapter high- If she [the wife] is hardworking, goes to work early, lights forces at play that press both for and against norm comes back home late, and cares for her family, surely � change in the lives of the study communities. people would understand and sympathize with her. — 23-year-old urban woman, Ba Dinh district, Hanoi, Vietnam 1. The quiet relaxing and changing of norms challenge the norms: as the dominant gender, they can more easily flout or champion norms to suit their inter- We asked focus groups to consider gender norms from ests as long as that challenge does not aim at breaking many angles and how they are changing, as discussed in the basis of their own male identity.37 By contrast, gender the introduction to Part I. By comparing responses by the power relations put women at a decided disadvantage if younger and older generations, we can assess how these their interests require breaking a gender norm or holding norms may be evolving in the study communities. In this their partner to account.38 report, we also look at the focus groups’ responses to questions about differences between current and past In addition to the differences between norm compliance generations and about their aspirations for themselves and resistance, it is important to distinguish relaxation and their children. (The recall period for these questions from change of a gender norm. By relaxation, heard in ranges from a decade to one generation, depending on observations by the focus groups, we mean that wom- the question) The sampling frame in each country in- en and men are challenging and perhaps crossing the cludes 1–2 communities chosen by the researchers for boundaries of traditional gender roles or conduct, al- specific attributes (e.g., girls’ school enrollment was high though their actions are not recognized as a legitimate in the country or the local economy was dynamic). We and acceptable norm. They are assuming new roles or hypothesized that these select communities may likely responsibilities, but are not setting a new standard. For provide more incentives or opportunities for individu- instance, in Zabibu village, Tanzania, the key informant als (of both sexes) to challenge the boundaries of stifling (who was interviewed to provide information about his norms. For comparative purposes, the country-level village) reported, “Rarely do women work for pay here. sampling frames also contain urban and rural communi- They have to stay at home and take care of the house- ties that have the opposite experience, i.e. with fewer hold and family. But what work can these women do? educational and economic opportunities. They have not even gone to school.� Later in his inter- view, however, he added, “Women are mostly involved When investigating norms, it is important to differentiate in digging holes to plant grapes, tending vegetable gar- acts of compliance with the norm from resistance be- dens, and making ‘vyungu’ [clay pots].� Women are work- cause they imply different levels of agency. For example, ing, which is a relaxation of the norm that restricts them a woman who limits her fertility in a community where from breadwinning tasks—and they may even be prepar- having many children is the norm necessarily pushes ing and selling goods in the market. Nevertheless, these against the expected choices and behaviors for local activities have not yet been recognized as income gen- women. She is demonstrating a different level of agency eration, but rather as an extension of their domestic du- or of empowerment (if successful) than a woman who ties. Hence, it is acceptable within the traditional norms. limits her fertility in an environment where small families On Norms and Agency are the norm (Kabeer 1999). It is important to recognize Across the community sample, we find signs of gender that men have an advantage in initiatives that uphold or norms in transition and uncertainty. Women’s pursuit of 36 For discussion and other cases of how normative change can ensue over time through processes that imply both more and less hidden forms of resistance, see Rao and Walton (2005, 23–26) and Scott (1985). 37 For a more detailed analysis see, for example, Coston and Kimmel (2012) and Connell and Messerschmidt (2005). 38 Gender relations are power relations, even if they are not recognized as such and are willingly entered into. For example, Rich (1976, 57–58) defines patriarchy as “the power of the fathers: a familial-social, ideological, political system in which men— by force, direct pressure, or through ritual, tradition, law, and language, customs, etiquette, education, and the division of labor—determine what part women shall or shall not play, and in which the female is everywhere subsumed under the male.� Men often do not realize that they benefit from patriarchy and women do not ques- tion such male privilege. 51 an economic role or a larger say in managing their house- The contexts where we see these rarer changes in norms holds can have unpredictable consequences where these are important. Notably, Poland and Serbia have a longer practices are not widely accepted. On one hand, such history of gender-sensitive legislation and higher levels initiatives may be interpreted as a gender-norm violation of girls’ and boys’ education. There, men and women can and thus subject to sanctions that can damage a woman’s draw from a wide range of public and private discourses reputation and status. On the other hand, if a new role and institutions that favor gender equity. We also see or conduct becomes widely accepted, it may confer on important movements in norms in the other countries women new status, respect, and influence (or power) in the sample—especially in rapidly urbanizing cities in their relations with others. The day-to-day activities and where higher education rates are nibbling at the that quietly resist or flout norms for women’s roles and edges of traditional cultures. In several contexts, the behaviors may not trigger a complete change or cause data reveal the presence of other factors that can drive an unfair practice to be discarded. These initiatives are change, such as more egalitarian traditional practices, nevertheless important because over time they can po- campaigns to introduce new norms, or decisive policy tentially affect power relations within a household. In action by a government.39 many cases in our dataset, where a relaxation of gender norms is evident, men’s and women’s unequal authority in the household remains largely intact, even though the 2. Timing is all: Negotiating gender differences in their roles, responsibilities, and be- opportunities and gender-specific haviors are narrowing and disappearing. responsibilities Although it may be a subtle difference, we are referring The allocation of resources—particularly scarce ones, to a change in gender norm in contexts where both wom- such as time—reflects prior normative considerations in en’s and men’s focus groups reported a significant shift in communities and households. One area where gender a gender role or in gender power relations. For example, differences are most conspicuous is individuals’ use of in some communities, husbands shop for groceries when their time. Time use responds to the “gender contract� a wife does not feel like cooking and are deeply involved obligations, reflecting dominant norms about the divi- in raising and educating their children. But reports of sion of labor among family members. These obligations such engagement were relatively rare in our dataset. A affect individual work choices, physical mobility, access man who admits that he prepares his own dinner may be to opportunities, and the possibility of free time. In our teased rather than encouraged in many focus groups. analysis of time use by the young adult and adolescent focus groups, we find that they have shifted their daily The processes of relaxing and changing norms of behav- schedules in ways that offer more opportunities for edu- ior, which are so central to our identities and status with cation, work, and free time, despite the persistence of others, seem to be slow and hidden. As Ridgeway and gender disparities and urban-rural differences. Correll (2004, 528) explain, We expected the adolescents and young adults in our to the extent that people continue to deeply hold sample to more easily articulate views about progress beliefs that men and women are essentially different, and change in norms because their life experiences re- separate categories of people, they will likely resist flect gains from previous generations. Indeed, the ado- On Norms and Agency beliefs that there are absolutely no instrumental lescents’ accounts of how they spend their days reveal differences between men and women in the “things that that they constantly negotiate between taking advan- count,� even in the face of clear displays of competence tage of “new� opportunities and tending to “old� duties by women. This tension between the belief that men in their households. Yet, almost everywhere, girls and and women are fundamentally different and the young women still must devote more time to household displays of similar levels of competence between men and care-giving tasks than boys and young men. This is and women may facilitate some cultural redefinition in the case whether the young women attend school or not; what counts at a given period of history. however, those in school who must study spend fewer 52 hours on household tasks, particularly in urban areas. The main message from the evidence regarding time In both rural and urban areas, young and adult women use is simple: men have more free time because they have no or very little free time. They switch from one do little or no housework. Even when men work long activity to another and, with few exceptions, are con- days, they usually end it with some form of distraction stantly engaged. In a less dynamic rural setting, such or entertainment. Women very rarely pause to enjoy as Floresti District, Moldova, “women don’t have much some leisure time. When they are not working for pay or spare time. They work and take care of children, their studying, they are tending the house and children. The husband, and the household. Free time appears late very definition of free time is adapted to this gendered in the evening or during the winter when there is less arrangement of the use of time. Unlike men, women use farm work.� Women in the more dynamic city of Bhu- their free or spare time to work; they simply shift activi- baneswar (Odisha), India, reported being just as busy ties. Women are the losers in the time distribution game. during their free time hours. They work for the family Men and women readily acknowledge this fact and, while business, teach their children, clean and organize the there is evidence that urban men particularly are doing house, go shopping for clothes and food, and the list more to help around the house, inequalities in free time, continues. Only three activities the women mentioned however, remain. can be associated with a traditional idea of leisure: eating snacks outside, reading magazines, and buying Young women have more opportunities now for educa- cosmetics. A woman from urban Dobrowice, Poland (a tion and work than previous generations, and they are middling community in terms of economic opportuni- aware of it. Young men likewise find themselves seeking ties and normative change), noted that while men “play more education than in the past, but they also realize ball, go fishing, sit in the front of the television, or drink more free time now. To some extent, their gain in free a beer, we are ironing as a form of relaxation.� time is due to entering the labor market at an older age and decreasing demands on young men to supplement The different uses of time by men and women under- their household’s income by working for pay or helping line the prevailing gender order in the community. If in family businesses and farms. Along with increased men are expected to enjoy free time after their working opportunities for schooling, young women in both rural hours, norms that insist women not forget their repro- and urban contexts also seem to have rid themselves of ductive duties, even when engaging in productive paid some of the burdens of being an unpaid family worker. jobs, will also persist. This disparity was overwhelm- But only urban women see themselves doing less work ingly affirmed throughout the sample. For men in many inside the home than the previous generation. Their communities, women’s work at home sustains the right circumstances have changed the most, due to living in order of things and is inviolate. The normative frame- cities, and so have their expectations. Young women work and power relations that impose a double burden in urban communities want more time to work for pay, on women (in and out of the house) do not disappear in line with their rising aspirations and opportunities. or change when economic conditions do. Unemployed Young women in rural communities long for paid work, young and adult men have to find ways to employ their but mostly they aspire to do less work inside the home, extra time when they have less income-generating which primarily occupies their days. However all wom- work, but housework is not a viable option. When job en—urban and rural—strive for free time. and business opportunities for men diminish or unem- ployment strikes, and women must take on the role of a. Incessant activity On Norms and Agency provider and working to support their households, the norms shaping the allocation of time and men’s duties “It seems like there is no end to what we women must do. — Village woman, Sungai Puar District, Indonesia � do not change. Men have more alternatives to “kill time,� a valuable re- source that is under their control. Women never speak 39 As Jensen and Oster (2009) document in the case of the introduction of cable television in some communities in India. Also Fogli and Veld- kamp (2008) report how in the United States the expansion of women’s participation in the labor force responded to access to information and peer examples. 53 in terms of killing time, most likely because their job b. The rise of the Swedish husband: description as homemaker or caregiver, for example, An emerging male double-burden? does not end after eight hours or at twilight. The pre- “ scription means that if women wish or need to work I am sensitive to my wife’s affairs. I don’t wake her up in outside the home, they may not abandon their “base the morning just so she can prepare my morning coffee. line� household duties and must factor in the time I wake up to make tea and breakfast, and bring it to her to accomplish them. This female responsibility is en- because I know that she likes to sleep in the morning and grained to the point that it has become “naturalized�; it is the best time to sleep. I take care of all the house women’s incessant activity is not seen as an unequal affairs in the morning because she has a right to expect burden but as part of the female nature. “Women are me to do so. When I get home in the evening, she does � different. Women like working. Women often keep everything to make me happy, so I try to do the same. themselves busy because they feel the need to. They — Urban man, University Quarter, West Bank and Gaza do not care about themselves. They may just go out to have a chat in their free time. In rural areas like ours, women often keep themselves busy all the time, If, in previous generations, the primary defining charac- which makes them different from men,� noted a wom- teristic of a good husband was that he provided for his an from rural Vietnam. It is beyond nature—“God has family, today’s younger generations of men and women given women the art of multi-tasking, so we won’t find demand more of the husband, along the lines of the ex- anything difficult�—was the view of women in Nellore ample of the Swedish husband.40 Although still essential, (Andhra Pradesh), India. bringing home the money is no longer enough. Good hus- bands must not only secure their families’ economic well- The naturalization or customariness of household being but in addition must be sensitive to their wives’ work as women-only work also arises from the norma- and children’s emotional needs, spend time with them tive refusal by men to engage in home production. This at home (instead of out with friends, drinking, gambling, belief (and practice) has been reproduced and solidi- or cheating), share domestic chores, and devote time fied over time, producing a pattern that is transmit- to help the children with their homework. If and when ted to the next generations with little change. Younger men actually do all these things, we may see the emer- women are accustomed early to household work and gence of a masculine double burden. The difference with stepping into the older woman’s place: “When mother women’s double burden is a matter of emphasis and time is absent, I am there to take care of everything. Wom- allocation. While women may or may not work, they must en take care of everything. The man is the household in any circumstance look after the household; men have head, but the woman takes care of everything� (young little flexibility in the imperative that they must provide woman, Serbia). for the family, but they have more scope in how much time they spend nurturing the family. The novelty is the While education and work opportunities can bring posi- expectation that a man spends any time or resources at tive chances for female autonomy, at the same time, all in the domestic sphere and that he shares in its man- however, they imply a trap that increases women’s time agement with the wife. burden. The current generation of girls can study and have been freed from part of the burden of housework, Notice that we are deliberately moving this discussion On Norms and Agency but not of its totality. Indeed, looking at the perceptions away from the prevalent patterns of time use and house- of what it means to be a good girl (chapter 1), helping at hold roles in order to learn from contexts that feature sig- home is still one of the main attributes of a good girl. The nificant relaxation or incipient changes in gender norms. appeal of the world beyond the confines of the house- Adult women and men in Poland and Serbia, especially in hold beckons women, but such aspirations have to take urban locations, provided eloquent accounts of changes into account the need to earn income to support them- in the norms that define gender roles in their communi- selves and their families, form a family, and run a house- ties. Compared with their fathers, whose role as provid- hold, plus manage the expectations of time associated ers was sufficient to qualify them as good husbands, the 54 with each of them. partners of the new generation of women are expected to “understand the woman’s needs,� “sympathize that I munity where the good husband of the past was “the have a stomach ache due to my period,� “recognize that chief income earner� and “did not help his wife in the we [wife and children] need his help,� “realize that I do household.� A strong push for normative change has not feel like cooking and prefer to go to a restaurant,� risen due to the high levels of male unemployment in “remember the children’s birthdays,� and “remember the community and the emergence of opportunities for our anniversary [celebrate it] and know how to make the women. Men are no longer the sole breadwinners and pleasure� (adult women, Dobrowice, Poland). they have had to adapt. In addition to being sensitive, the good husband engages In Belgrade, Serbia, adult men claimed that a good hus- in essential house chores. “He cleans the flat, does the band “does not divide male and female chores. [A hus- basic shopping, cooks� and participates actively in the band and wife] should complement each other and find education of the children. Not only “is [he] willing to mutually beneficial solutions.� “He cannot just sit and spend the time with the child,� and “participates in fam- watch television if the child needs to change clothes ily life and helps with the children’s homework,� but he and the wife is cooking dinner.� Their good husband is has to be good at it. He needs to be “creative while play- a “good parent, caring, thoughtful.� In Pomoravlje Dis- ing with children, like a football game,� added the Dobro- trict, Serbia, the good husband of the past “used to be wice group of Polish women. Women in urban Olsztyn, undisputed in decision-making: his word was final. The Poland, reported that “now he distinguishes the cabbage wife was to obey, love, even wash his feet.� Now, how- from the lettuce.� The same definitions of the good hus- ever, these men insisted that the good husband today band appear in Belgrade, Serbia, where women claimed “washes, irons, helps his wife, even vacuums, and usually that, apart from having a “secure income … he divides does the hard physical labor.� the household chores with his wife, goes shopping, and is resourceful (can take care of himself) and tolerant. He In a radical departure from tradition, one man from Po- must not be an idiot who is not capable of cooking a meal moravlje District argued that “if your wife is at home, you or ironing his clothes.� have to contribute something. You can’t expect her to do everything and you do nothing. You can’t spend your Crucially, these notions are shared by men and rein- time in a pub, you have to help her.� The good husband force the assumption that the bending of norms occurs also “has to have time for the children. And this does in a relatively non-conflictive manner in everyday inter- not mean simply hanging out with them every once in a action. A man from Justynowo, Poland, explained that while.� “He has to know what to do to get the children sometimes “[when] my wife comes back from work, her prepared for school, even cooking lunch if necessary, if female friends come to visit and she tells me to cook he wants to have children who love and respect him. He dinner by myself.� He sees this as fair and does not has to be dedicated to them, not just let them see him think his wife is behaving as a bad wife. The fact that once in five days.� Other men in the group agreed that women and men expressed similar views testifies to a household cooperation is the ideal. shift in what is consensually considered desirable, prob- ably as a result of the interactions between education, While men do not question these additional responsibili- economic opportunities, and messages in the media and ties as diminishing their masculinity, they have some con- other public spaces, where discourses about gender cir- cern that the balance of power may shift in women’s favor On Norms and Agency culate. As a quite forward-looking adult male in urban in the future: “Men are slightly more dominant now, but Sjenica, Serbia, remarked, “[the good husband must be women may become more dominant soon� (adult man, willing] to serve his wife, to be obedient, to make pies, Pomoravlje District, Serbia). For the moment, the extra and to wash the dishes.� Simply making this statement duties taken on by men do not alter the core of their role constitutes a massive shift in the status quo of a com- as breadwinners. As another man in the same group ex- 40 Sweden has consistently been ranked as one of the most gender-equal societies in the world. It has consistently appear as one of the top five countries for gender equality in the World Economic Forum’s Global Gender Gap Report (2012); and UNDP’s Gender Inequality Index (2011). Gen- der equality in the household and domestic responsibilities sharing between men and women has been attributed to family supportive legislation and gender-sensitive social policies (for more details see Nyberg 2011). 55 plained, although men are expected to do more at home, spend quality time with wife and children.� Similarly, in the “husband still has to provide most of the household urban Jaipur (Odisha), India, men agreed that the good income. His income is the biggest and the most impor- husband must “take out some time from the day to day tant for the functioning of the household. It gives him activity to help the children in studies.� By comparison, self-respect.� When men do not feel threatened in their in the rural groups from India, we find more modest signs main role as providers, they are more willing to accept of change. But even though small, women perceived a changes in their secondary, domestic roles. difference at the margins of the strict norms that regu- late their actions: In nearby Moldova, we also find evidence of norm re- laxation, but it is more incipient and appears only in ur- There is a difference [from the previous generation]. ban contexts. Women in Balti see their good husband as In those days, the husband was treated like a god. someone who “participates actively in his children’s edu- The wife in my mother’s generation would wait for her cation� and “balances his work and family in the same husband to come home and she would eat her food way as a woman does. After work he goes home and only after her husband had his dinner. These days, the spends his leisure time with the family.� Men in the same wife doesn’t wait for her husband, but after he wakes city agreed that a good husband “manages to balance her up, she serves dinner to him. (Village women, work with home responsibilities and support his wife�; Koudipally Mandal (Andhra Pradesh), India) however, he does so only as long as the traditional asso- ciation between income and self-respect that are central Sisum’s story, which opens chapter 1, makes clear the dif- to his idea of masculinity remains intact. ferences between urban Thimphu and rural Samtse. In the countryside, the central characteristic of the husband But again in Sjenica, Serbia, increased domestic work by as the provider of the family remains, but men now seem men is a direct result of their unemployment and men more open-minded about the restrictions that some tra- expressed their frustration. In contrast to husbands in ditional norms impose on women. Radical change in gen- the past, “today we are not working and we are unable to der roles across all countries and communities is still far provide for and make our families happy.� These circum- away, but the emergence of more progressive views of stances may trigger violence, as explored more below, men’s contribution to housework mean, at least, greater but may also produce peaceful accommodation. The ap- visibility of the double demands upon women. Men are parent ease with which these new changes fit into male also challenging the dictate of staying removed from re- identity in some households and communities indicates productive tasks demanded by traditional masculinity. the potential for future behavior change. The more main- Becoming more involved in the lives of their children stream these ideas become, the easier it is for young arises, one assumes, not only from a sense of duty but men and boys to follow their fathers as role models and from the pleasures and rewards associated with it. Still, be socialized in more egalitarian gender norms. men are “helping,� rather than taking responsibility for household chores or children’s education. The relaxation and change of norms that shape gender roles is also evident in the focus groups in communi- c. Open dialogue and emergence ties where women’s education took off later and where of household power sharing cultural views of women are more traditional. In some On Norms and Agency sample communities in India, for example, changes in the The notion of cooperation in household discussions— economy and educational opportunities have spurred open dialogue—is requisite to the success of a couple, new awareness of the need to share responsibilities, and central in the view of a Vietnamese man from Hung particularly in urban areas. In urban Bhubaneswar (Odis- Yen District, who said, “Husband and wife should dis- ha), India, adult women’s changed expectations for their cuss [an issue] and come to the same decision. The partners are evident in that they define the good hus- spousal relationship should be equal. For instance, they band as someone who “takes responsibility of house,� have to discuss and agree on buying a television or not. “takes responsibility of children,� “does household work A decision made without their [mutual] agreement is 56 when wife is sick,� and “should work, but also should seldom successful.� In Kim Dong District, Vietnam, an adult woman described her belief about domestic bal- tions as a result of more open dialogue between men ance of power: and women: “For me, I think that if you want to remain powerful, you must collaborate with people. You must In my opinion, balance of power means each partner also listen a lot, especially to your wife, because some- is free to share their ideas in the couple’s discussion times women have good ideas.� His further comment is before a decision is made. It is similar to a couple even more revealing: “Look at ‘H.’ He was educated and [reaching] consensus. If, for example, the husband takes had money, but he quarreled with his wife. And what the lead of the family and his wife and children have happened? When he abandoned his wife, he married an- to ask him for money to buy everything, his wife and other one who one simply ate his money, so where is he children have no share of the power. now? At the bottom [of the power ladder].� Although not abundant, evidence in this study of peace- Dialogue has an important normative component to ful routine negotiations, daily conversations, and trans- it—meaning that both men and women adhere to it as actions between spouses or partners offers glimpses of a matter of principle. But it is illuminating if one also con- ground-breaking changes in household cooperation, open siders that, when women have their own income,42 their dialogue, and even power sharing. Dialogue and harmoni- capacity to engage men in negotiation greatly increases. ous relations have an instrumental, beneficial value, not Income gives women bargaining power, as the younger only for family life but also for economic decision-making women in the study particularly are discovering and ex- within the household.41 From Fiji to Moldova to India, the ploring. In addition, due to their gains in education and possibility of partners being able to express different public engagement, the young women have also become opinions without conflict—open dialogue—is significant more aware of their rights and are developing different progress as gender norms bend and change. Its presence expectations of their relationships with men. Informed, is mnemonic for contemporary visions of a “good� couple connected, and aspiring women with some income of and the emerging equality between women and men. their own are likely to demand more open dialogue in their households, but they also have more resources to Notably, the task of initiating more open dialogue is prevail in some decisions or reach agreement with their placed on men. It is both seemingly contradictory and ex- husbands or partners. However, as we discuss later in pected, given that men are the main power holders and chapter 5, norms do not always change with women’s traditional leaders of the household. For example, wom- economic independence. en in rural Kalahandi District (Andhra Pradesh), India, felt that it is a good husband’s duty to “consult his wife for decisions made in the family and discuss household is- 3. Intergenerational transmission sues with women.� Numerous examples in the data show of the possibility of change that women observed positive changes in their husbands and new spaces opened up for negotiation and dia- logue—where husbands consulted their wives, permitted “ [A woman] knows the traditions and picks what is important and tries to transmit conversations to take place around family decisions, did not make decisions alone, and were open to dissenting views. Given that men are the main power holders and what she thinks is good for kids now. — Village woman, Malangachilima village, Tanzania � On Norms and Agency expected household leaders, the creation of more space Perhaps the final quiet mechanism that ensures move- for dialogue is in their hands. ment toward more equal gender norms is the infusion of this aspiration into the next generations’ ideals. The A comment from a man in rural Zabibu Village, Tanzania, “capacity to aspire� is the first step toward change (Ap- underlines this shift—by men—in domestic power rela- padurai 2004). When prompted about their hopes for 41 See for example, World Development Report 2012, Pronyk et al. (2006), Panda and Agarwal (2005), Swaminathan, Walker, and Rugadya (2008), and Aizer (2010). 42 Changes in bargaining power within the household have been widely documented by Lundberg and Pollack (1993), Stevenson (2008), and Aizer (2010), among others. 57 their children’s futures, the parents in the sample agreed Peru, as in many other countries, have defined masculin- that they wished for them a good education and work ity for a long time. opportunities. Both are tools that increase their agency, fuel their drive to seek a better life, and teach them to In Thimphu, Bhutan, mothers’ aspirations for their make sound choices. Education and economic well-being sons include a more direct reference to gender equal- dominated the discussions on aspirations for sons and ity: “Equality for both would make the biggest differ- daughters alike. ence in the lives of our daughters and would imbue our sons with the understanding that all humans of The parents’ narratives display a diverse set of aspira- opposite gender are same.� These women “would love tions, ranging from specific changes in gender norms to see their girl and boy children take equal stand in (daughters do not have to obey their husbands unques- all sectors, where girls will not be the underprivileged tioningly or will not be abused), gender relationships gender.� Just as with their daughters, mothers social- (open dialogue and more equality), and the traditional ize their sons and can also be key agents for change practices that are detrimental to women, to positive in their son’s attitudes and behaviors. The willingness character traits (be more decisive and stronger). We find of mothers and fathers to embrace gender equality in these in urban and rural communities, but significantly, their children’s education may bring massive change they appear mostly in adult women’s aspirations for girls and make gender relations in the next generation more and younger women. This is crucial, given the central role equitable and harmonious. of women in the socialization of children in family norms and their potential roles as agents of change. As gender norms loosen, today’s young women are less content to recreate the family dynamics of their mothers. The hope of women that their daughters can develop When the young adult women in the sample were asked stronger personalities, learn from their [mothers’] mis- if they wished to lead lives similar to their mothers, they takes, and take better charge of their lives appears responded consistently that they wanted to be more across the sample. Notably, in countries like Burkina proactive, less tolerant of abuse, and more informed. For Faso, the very suggestion that young girls be more alert example, in an urban area of Hato Mayor Province, Do- and seek different life paths from their mothers consti- minican Republic, one young woman asserted that she tutes a significant departure from tradition and a shift in did not want to be “passive in her life� like her mother. the norms that link good life choices today with those of Another urban young woman, in Nsenene village, Tanza- the previous generations. Mothers in Burkina Faso de- nia, disapproved of her mother’s tolerance of her father’s scribed the development of more courageous, less pas- violent behavior: “She does not say anything to my fa- sive personalities and a reflective or critical attitude as ther who beats her up.� In Bhubaneswar (Odisha), India, a desire for their daughters: “They must fight more for young women rejected their mother’s “innocence� and ig- themselves and be more daring.� In East Sepik Province, norance: “[The women] never came out of their houses, Papua New Guinea, women wanted to teach their daugh- so they did not even know what was happening outside.� ters from their own experiences and help their daughters avoid making the same mistakes that limited their own Further signs of change are appearing in the norms that life choices, “[like] those girls who got pregnant early and surround marriage. A group of mothers, also in Bhubane- missed out on opportunities.� swar, had revolutionary hopes for their daughters: “[We On Norms and Agency wish] them to find a good life partner. If they marry by Mothers’ hopes for positive character traits in their their own choice, then the boy can be of a caste lower daughters applied to their sons as well, although they than theirs.� This aspiration that their daughters have typically spoke of sons’ avoiding bad behavior. For in- a choice in marriage not only significantly challenges a stance, an adult woman in Chiclayo, Peru, wished her son powerful traditional gender norm but also breaks a sta- would “continue working, be as responsible at home as tus-related norm. he is with his two sons, and not fall into alcoholism or a life of vice.� If her aspirations for her son are successful, Mothers also pass along traditions to their daughters 58 this mother will have bucked a set of attitudes that in in order to safeguard their reputations, ensure proper “ marriages, and preserve family honor in compliance with In the past, everyone knew which roles belonged to gender norms. When asked about her desires for her men and which roles belonged to women. � child’s future, a mother from Sumadija District, Serbia, Today you do not know which role belongs to whom. declared, “She should be able to go out during the day, — Urban woman, Belgrade, Serbia but not during the night. And she must get an education, otherwise she will have no chance of finding a good job.� Like any mother, she wanted what is best for her chil- and more equitable gender roles and relations is not lin- dren, even if some of her wishes still limit her daughter’s ear: old and new co-exist. Advances toward more eq- freedoms and reinforce gender inequalities.43 uitable norms on some fronts may not be matched by progress on others. Gains may sometimes be reversed. Mothers in our study also wanted to spare their daugh- Actual behaviors may deviate from changing ideals and ters from physically harmful traditional practices. For aspirations. And, perhaps to be expected, focus group example, rural women in Burkina Faso and Tanzania op- members often disagreed among themselves about the posed having their daughters undergo the traditional rit- nature of normative changes occurring in their house- ual of female genital cutting.44 These mothers expressly holds and communities. This muddled story on norm talked about the harm to their daughters and that they changes, moreover, emerges in almost every context we did not want them “circumcised, so they can give birth sampled. On balance, change is surely happening, but at more easily.� One mother in Malangachilima village, Tan- the local level it is often patchy, gradual, and difficult to zania, was particularly eloquent about the negative im- discern. (Also see box 2.1 below.) pact of the practice and the need for change: For example, an earlier quotation by the women of Koudi- I hope one thing happens to this community. We pally Mandal45 (Andhra Pradesh), India, described how a used to have a reality of circumcising girls. Many new dinner rule has taken root, which no longer requires organizations came and sensitized the society about them to wait to eat until after the men have finished. This how bad it was and the situation seemed to improve. is also a community, however, where female seclusion is I have one request for my fellow women: let us be practiced and violence against women remains high and honest with ourselves and our daughters if we love generally acceptable. Women there “should not go out, them. There are still some women who are still doing even if the husband scolds or beats her. She should ad- this to their daughters when they are very young and it just.� Over and over again, the data show that, although cannot be noticed easily. We take our children to visit some norms do relax, other significantly inequitable norms their aunties, but that [genital cutting] is what we go to persist. Also, it is not only in more traditional contexts that do. Let us try to convince each other to completely stop overtly coercive forces uphold gender hierarchies. Urban it. We are not telling anybody lies, except ourselves; our Dobrowice, Poland, is another community (quoted ear- daughter will blame us and we shall feel ashamed at lier) where husbands are engaging meaningfully in house- some point. hold work and family care.46 Yet, the Koudipally Mandal and Dobrowice focus groups reported that domestic violence occurs there regularly. (Koudipally Mandal and 4. Gender norms in transition Dobrowice are outliers on the continuum of domestic vi- olence of all focus groups, with greater levels than most.) On Norms and Agency In order to learn more about processes of norm change, The point here is simply that, under diverse conditions, we cast a spotlight above on evidence in the dataset new and old norms can exist together in the same house- where this is more pronounced. The transition to new holds and communities. (Also see box PI.1 in chapter 1.) 43 There is a large literature on intergenerational transmission of norms. For thoughtful discussions, see Farré (2011), Quisumbing (1994), Inglehart and Norris (2003), Bisin and Verdier (2001), and Farré and Vella (2007). 44 Despite active campaigns in these countries to raise awareness about health and other risks of female genital mutilation, norms in some quar- ters persist in shunning “uncircumcised� girls, which leaves them severely disadvantaged in marriage markets (see Mackie 1996 and 2000). Note that the World Bank does not use the term “female circumcision,� preferring the more accurate term, female genital cutting. 45 In chapter 2, subsection 2b, “The Rise of the Swedish Husband.� 46 Also in chapter 2, subsection 2b. 59 In many rural men’s focus groups, particularly, it was not emancipated have family problems because the hus- uncommon for members to express open dissatisfaction band and the wife will always be in endless discussions that women are gaining a stronger and more indepen- and quarrels.� One way for a wife to avoid these quar- dent voice. The men of Hung Yen District, Vietnam, as rels, according to women in Kalahandi District (Andhra noted earlier, mostly lauded consultative decision-mak- Pradesh), India, is when “consulted by her husband … she ing processes between couples as key to making bet- agrees to what he says.� ter choices. Still, others in this group cautioned against women gaining too much power and freedom because Consideration for the husband’s feelings or reputation “you can kill yourself� or “living that way, you are consid- continues to be a prominent factor that prevents women ered self-indulgent.� One man added, in the study from pursuing their ideas or projects. As dis- cussed in the next chapter on key decision-making pro- Local women are supposed to meet four attributes cesses, women do not always succeed in negotiations or of an ideal woman: industry, appearance, respectful in bringing about dialogue. This situation may be chang- speech, and proper behavior. Therefore, it is not good ing in the new generation for whom the norms are no for wives or mothers to have much freedom. The image longer sacred. An intergenerational dialogue between of a man who gambles and drinks alcohol has been two women in Levuka, Fiji, made this evident. The older around for ages. A woman should enjoy freedom only woman (in her 50s) asserted that “a good wife will not try within a certain scope. to outdo her husband. Even if she is working, and he is not, she will still treat him as if he were the head of the The women’s group from Hung Yen District disagreed household. She will not belittle him.� The younger wom- among themselves whether it is good for women to have an (in her 20s) argued that “maybe in the past it was like extensive freedom, although they too valued more open that, but my husband helps me wash clothes and cook dialogue among couples, now generally accepted. Gen- food, so I think it is all right to be a good wife and ask der differences in status and acceptable behaviors are your husband to help in the house.� narrowing, but views can vary greatly, even among peers of the same sex and same generation, over how much Even one of the most progressive communities in the normative change is taking hold and is even desirable. sample, urban Olztsyn, Poland, produced diverging views on the qualities of a good wife. One woman opined that Moreover, although we often heard reports of better a good wife should be very attentive to her husband’s communication among couples, the discourse described needs: “She doesn’t sleep at night� if required to manage as beneficial by both men and women did not necessar- her work load, and “doesn’t talk back.� But other women ily involve a full and fair consideration of one another’s in her group disagreed, arguing for a more give-and-take views. As is evident from the data, women frequently relationship and that a woman needs to “motivate her tended to concede or compromise in order to avoid dis- husband to help her.� Mixed signs of progress were also rupting family harmony or to protect the husband’s feel- evident in other exchanges in this group. As one woman ings, or out of concern for the welfare of the children and indicated, her son-in-law “creates partnership in his mar- their own physical well-being. So, in many cases, there is riage: he and his wife go shopping together, he always dialogue, but often women pull out of the conversation makes the bed, and he cleans up after himself. My hus- before it even starts. band doesn’t.� Another woman countered, “My son-in- On Norms and Agency law just does the opposite: he does not help my daugh- For example, one woman from Sungai Puar District, In- ter, like my husband. He claims that he works very hard donesia, explained that “generally, the wife is more pa- and deserves to rest—he has been brought up this way.� tient due to family considerations. Maybe, because the children are still little, she will let things be for a while.� If some women in the study seemingly gave up in the pro- In Zabibu Village, Tanzania, where there is plenty of evi- cess of negotiation, others adopted strategic responses dence of increases in women’s autonomy, one adult wom- to unequal power distributions and pretended to comply an thought that “good couples simply have wives who are with the norm. They either included the husband perfunc- 60 submissive to the husbands. Couples where women are torily: “You see, if you are not careful with men, you can Box 2.1: Co-existence of norms and support for women’s work outside the home in Islamic communities No causal connection has been shown between low numbers of women in the labor force and Islam’s religious beliefs and ideals. Islam is, arguably, no more gender biaseda than Judaism and Christianity. Other factors, such as geog- raphy, culture and history—more so than religion—are more responsible for fewer women working, for example, in the Middle East and North Africa (World Bank 2012, Rauch and Kostyshak 2009). In fact, female labor varies widely across the Arab world, from lower levels in the West Bank and Gaza to much higher ones in Indonesia. Indeed, in Offenhauer’s (2005) extensive review, these different labor rates for women reflect variations in national economic structures and strategies or in local pre-existing cultural values. The evidence from the countries in our sample is similar, which shows no single standard or norm for women’s domestic and breadwinner roles or for working mothers.b For example, normative restrictions are more intense in Afghanistan, North Sudan, and Yemen, although there are signs of normative change in West Bank and Gaza and open support for women’s work autonomy in Indonesia, especially among young adults. While the communities in these five countries adhere to women’s “natural� role as the guardian of domestic order (or “original duty,�c as pointed out by a Sudanese Muslim man), these views do not differ from other sample countries where different religions are practiced. The norms, which underlie whether women can work, frame women’s decisions about jobs in terms of compatibility with household duties and supplemental income to what the man earns. “They [men] deal with women’s work as a necessity, but if their living conditions improve, then women should stay home and not do any kind of work� (young man, North Sudan). In Yemen, Muslim men generally felt the same way, given that “most of the available jobs for women interfere with their family care obligations. Women’s work for pay gets in the way of their original (traditional) duties� (young man, Aden, Yemen). But the economic struggles faced by Yemeni households are forcing gender norms to slacken to the point that men now expect their wives to contribute to the family income. Notably, these views co-exist with other attitudes that see no problem with women working. These positive views may develop from exposure to other realities, more education, or economic hardship. In a rural village in Afghanistan, adult men explained that “before, mothers could not work out of home, but now they can because they have more education and are exposed to the opinions of other women—who immigrated when war erupted—that women can work.� Urban young women in Afghanistan also sought to work as a logical consequence of their education: “It is better for women to go out to work, but if her family does not agree, it is better to stay at home. …But if the family sends the girls to school, then they must accept that the girls will work and earn an income.� In Indonesia, although the adult women were more compliant with traditional norms that dictate that they must consult their husbands before working outside the home, educated young Muslim women living in cities expressed strong opinions about their right to work and to be independent. This does not mean, however, that they do not face the same normative limitations as their mothers, but that they are willing to fight them. “In the past, the girls were not allowed to go out or even go to school, but the current situation has changed. I have a sister who is studying and will work for sure,� pointed out a young man from Dirbas, West Bank and Gaza. On Norms and Agency Overall, the situation in Islamic countries is not substantially different from other countries. Changes in views and practices are slower and uneven, but the normative frameworks in Muslim communities are clearly adapting to the new aspirations and realities of the women and men living there. a. The positions on women held by very conservative or extreme proponents of Islam (whose aggressive actions have dominated media headlines in recent years) do not reflect what Islam actually says about women and their roles and activities. b. Similar variation in norms is reported for Bangladesh in World Bank (2008). This study finds, for example, more support for gender equality in education, but it does not translate into views that husbands and wives should have equal education. c. “Original duty� is the view in Islam. 61 lose it all. So you need to make sure that you involve them, ceived to be a principal trigger, but focus group narra- so they feel involved, and then you can continue with the tives revealed that slowly changing norms for acceptable business� (woman in Tanzania)—or they simply acted as if roles and conduct also contribute to violence and the they followed the rules, but then did what they wanted forms that it takes. The focus groups’ narratives consis- when their husbands were not present or did not notice. tently reported that men who are unable to fulfill their provider role often act out their frustrations with vio- In a few cases, focus groups reported trends that sug- lence, and that it remains acceptable in many communi- gested backsliding on normative gender conduct. In ties to sanction women harshly for minor infractions that some sample communities in Liberia, women perceived are perceived as challenging male authority or norms of they had made significant strides in gender equality. They feminine conduct. felt that they are clearly gaining more powerful public roles, but they also expressed frustration that their part- In general, private interviews rather than group discus- ners are becoming less cooperative. “The men refuse to sions are preferable when investigating questions of in- cut palm nuts and brush on the farm, so we now have timate partner stress and violence. (We included both to do it,� complained a woman from Greenville District. in the study). Nevertheless, the focus group discussions Throughout Liberia, men are simultaneously struggling still painted a rough picture of what is deemed “normal� to adapt to women’s new roles and dealing with the or perhaps “acceptable� in conversation about this sensi- slow recovery of the economy after the end of the civil tive topic among friends and neighbors. A large majority conflict in 2003. Unfortunately, in some cases, women’s of the adult focus groups spoke in some detail on this economic participation can fuel violence by men against difficult subject.47 them. Women’s empowerment and gender norm change do not always move amicably together. The gender literature varies in its emphasis on more- versus less-coercive mechanisms that make gender norms difficult to dislodge. Actual or credible threats of 5. “A woman should be beaten if she violence are the most extreme, costly, and risky of the deserves punishment� mechanisms that sustain gender inequalities and enforce gender norm compliance. Much more potent and effec- “ tive, however, are the everyday routine interactions and — Does a good husband have to be internalized psychological processes that sustain gender good provider? (Facilitator) hierarchies, as well as other social inequalities, with less — Yes, that is the main reason why he is the head disruption.48 As discussed in the introduction, norms of the family. You know, if he doesn’t do that � carry with them a set of socially acceptable sanctions people will make jokes about him. that may be invoked in instances of deviation from the — Village man, Blue Nile State, Sudan norm. These acts of sanction are part and parcel of the “ common, acceptable conduct of household members. In If she is nagging me and I tell her to stop, addition, norms are held in place by deeply internalized and she continues nagging, then it is her fault � beliefs about men’s greater authority and competency and she deserves to be beaten. (Ridgeway and Correll 2004). Foucault (1995), for exam- — Urban man, Belgrade, Serbia ple, argues that social control most often works through On Norms and Agency internalization, self-discipline, and vigilance rather than In this section, we present men’s and women’s accounts external coercive mechanisms that inflict pain. of domestic conflict and violence in their communities. Facilitators introduced the topic by asking the focus In rethinking of the notion of “hegemonic masculinity,� groups to reflect on what typically happens in their com- Connell and Messerschmidt (2005, 842) stress the less- munities when a wife is not a good wife or a husband passive ways in which men use language to meet their is not a good husband. Their responses make evident “interactional needs� and favorably “position themselves that domestic violence is all too common, albeit at vary- through discursive practices.� 49 They also reflect on ways 62 ing levels across the sample. Economic factors are per- that “boys and men choose those discursive positions that help them ward off anxiety and avoid feelings of indicated that a wife may be harshly scolded or even powerlessness.� These discursive acts to reaffirm men’s beaten, should she “not be pleasant,� talk about “small dominant position are widely evident in the focus group matters,� “gossip,� or serve a meal that is “not tasty� or accounts about marital strife. Their narratives pointed to “late.� Men in diverse communities mentioned chiding or causes of domestic violence as men’s affirmation of their sternly reprimanding women for triggering conflict with dominant role and response when it appears to be chal- their “useless talking and interference.� In Pomoravlje lenged. Other studies also suggest that men’s peers play District, Serbia, a man referred to women as “sharp- important roles in upholding gender norms of masculin- tongued; they are masters of mental abuse, they nag, ity, dominance, and aggressive behavior: peer pressure they harass, they badger.� In Caedir Lunga, Moldova, an- on men pushes them to earn respect and demonstrate other man warned, their competency by subordinating woman (see, for in- stance, Holland, Ramazanoglu, and Sharpe 1998). These for example, if a friend of mine comes to visit me—not demands on men to secure and display control greatly the family—then my wife should pour the wine into our complicate women’s agency and pursuit of goals that glasses. She may also drink a glass, but immediately require resisting or relaxing the gender norms that gov- after she has to leave the room. God save her if she ern their roles and responsibilities. They also complicate starts chatting, making comments, or judging someone women’s power to compel their partners to uphold their or something. prescribed roles and responsibilities—for example, to insist that their husbands’ behave respectfully, be good Yet, in the study, for women to punish or belittle a man decision-makers, and provide for their families. for these same “misbehaviors� is generally unthinkable and unacceptable. In addition, a woman is expected to a. Extent and forms of domestic violence discipline herself and if she fails—for example, by not holding her tongue—then she may be harmed or abused Sisum’s questioning, in the opening of Part I, of a dinner as a consequence. Focus groups reported that a woman rule that sparked a wild tirade by her father, was not at all who is a victim of abuse may be advised by her parents, unusual in our focus groups’ accounts of the normative friends, in-laws, and neighbors of her duty to accept the behaviors that surround marital stress. While for some shame and mistreatment for provoking her husband. In participants—mostly women—the outburst itself could Ba Dinh District, Vietnam, for instance, the men’s group qualify as an act of violence, this view was not shared by noted, “Neighbors will give advice to the wife, something most men. The men’s focus group from Samtse, Bhutan like ‘your husband has a hot temper, so you should find (Sisum’s village), reported domestic violence as a rare a better time to talk with him. Try not to complain too event in their community, although the women rated it as much.’� Such are the stark realities of the unfair sanc- occurring regularly. This discrepancy may arise because tioning practices that govern daily life and the unlevel these men do not perceive displays of rage or overt playing field of marital conflict. threats of violence toward women as domestic violence, while the women register the physical intimidation or the Figure 2.1 reveals that on average focus groups perceived knowledge that such rage often precedes violence. domestic violence to be occasional events in their commu- nities, although significantly 31 percent of women’s groups Both women and men in the study often pointed out, thought domestic violence was a regular or frequent oc- On Norms and Agency in very different contexts, that seemingly minor actions currence in their communities versus 19 percent of men’s by the female could spark explosive reactions from the groups. Rural men acknowledged the least amount of male authority figure, which were wholly disproportion- domestic violence. Focus groups on balance indicated, ate to the immediate circumstances. Many focus groups however, that such violence is ebbing, compared to a de- 47 A small number of groups dismissed domestic violence as a problem for their community and had little to say on the topic. 48 For an economic perspective on these processes, Hoff and Stiglitz (2010) show how power historically has constructed social categories and identities that can reproduce status differences long after opportunities across groups are made equal. 49 The concept of hegemonic masculinity was developed as a reference framework encompassing the different traits of the normative ideal of male behavior. It refers to the dominant way to be and act like a man. 63 cade ago. While these perceptions of how much domes- The averages in figure 2.1 hide significant variations. At tic violence occurs are not derived from representative one extreme are the sample communities that both samples, they do signal that violence remains a serious women and men deem to be very safe and, at the other, threat and challenge.50 And given the shame and stigma are those where the respondents’ reports echoed this associated with family conflict, the discourse among fo- man from Koudipally Mandal (Andhra Pradesh), India: cus group members likely understates the extent and se- “The women are not physically abused as frequently as verity of marital stress and domestic violence. they used to be. Of course, around 40 percent are still being abused physically. The reasons are mainly related to earnings.� The large variation in local levels of violence is consistent with the findings of randomized surveys of FIGURE 2.1: PERCEPTIONS OF THE domestic violence (Heise 2012). PREVALENCE OF DOMESTIC VIOLENCE AGAINST WOMEN INTHE STUDY COMMUNITIES Figure 2.2 presents a snapshot of the most common Rural abuses mentioned in the focus groups’ descriptions of Women marital conflict. Most often, and by quite a large margin, their comments centered on men’s acts of physical vio- Urban Women lence against women, which was mentioned more than other types of abuse. This pattern emerged despite the Urban Men facilitators’ explanations to the groups of our interest in learning about all forms of violence common among Rural Men couples in the community, whether verbal, emotional, economic (deprivation), or physical. Still, male-on-fe- 1 1.5 2 2.5 3 Average of ratings male physical abuse dominated the discussions, usu- Average 10 years ago Average now ally reported with single-word descriptions of violence, Happens: 1 = almost never; 2 = occassionally; 3 = regulary; 4 = frequently such as slapping, punching, hitting, bashing, or raping. Note: Data from individual ratings by members of 194 adult focus Sometimes the participants were more specific: hitting groups. The �gure shows the average of ratings by focus group with a broom, pulling hair, breaking arms and legs, be- members in 194 adult focus groups. The ratings were done privately by individuals during the focus groups, and then the general ing denied food, forcing sex, using knives, or being left pattern of response was discussed. outside at night. Facilitators had the option of aiding the discussions about marital conflict by drawing a cause–impact dia- FIGURE 2.2: REPORTS OF FORMS OF gram and jotting down the main comments from the DOMESTIC ABUSE AGAINST WOMEN group about the different causes, forms, and impacts of abuse. Figure 2.3 displays the diagram from a wom- Rural Women en’s focus group in a densely populated neighborhood of Ba Dinh district in Hanoi. The diagram shows more Urban Women detail in the forms of abuse, but otherwise is typical of other diagrams generated by the focus groups. Both Urban On Norms and Agency Men men’s and women’s groups from the Hanoi neighbor- hood reported declining levels of violence, although Rural Men the women’s diagram indicates quite diverse forms of violence. The diagram also highlights how expected 0 20 40 60 80 100 behaviors and gender roles are at the center of the Frequency of mentions causes, as well as in the types, of violence. For exam- Physical abuse Verbal abuse Economic abuse Emotional abuse ple, such practices as making all the decisions or not Note: Data from 194 adult focus groups. allowing a wife to go out are nothing but a tightening 64 of some norms already in existence. The Hanoi women also listed many good reasons why a woman would hide stronger and in a dispute, when a man finds it difficult being abused, including neighborhood gossip, public to control his anger, he may use his hand to punch or shame, the stigma attached herself and her family, and slap his woman. For the woman, because of the pressure high risk of further beatings. to care for the family, it is easy for her to get angry. Women tend to have this habit of complaining to the Women from this neighborhood of Hanoi conceded that husband when he comes back home late. It is meant to they “scolded their husbands,� and many focus groups of terrorize men psychologically. both sexes reported that women also can be emotionally and verbally abusive. Yet, the greatest preponderance of Well beyond Hanoi, other men’s groups viewed a wife’s violence described was men’s physical abuse of women. “banter,� “sulking,� “grumbling,� or “complaining� as This passage from a 37-year-old man from the neighbor- “mental torture� that merited the harshest discipline. hood captured this pattern: In Hato Mayor, Dominican Republic, the men listed one reason for violence as “women at times put on a lot of In our district, there are cases where men beat women pressure on them.� In Balti, Moldova, urban men said and women mentally terrorize men. Because they have that violence can happen “because the wife couldn’t different points of view on some problem, men and stop from making caustic comments about some mis- women argue with each other. Men do not talk about takes her husband made.� Clearly, deeper stresses are rights because men and women are equal. Men are driving these processes. FIGURE 2.3: CAUSES AND CONSEQUENCES OF VIOLENCE, WOMEN'S FOCUS GROUP IN BA DINH DISTRICT OF HANOI, VIETNAM CAUSES FORMS OF ABUSE CONSEQUENCES OF ABUSE Husband drunk; wife Divorce or suspects separated adultery Husband scolds, humiliates wife; wife Family scolds husband; husband �nancesfall; prohibits his wife from unemployment; going out; husband Family wife does not give controls wife’s time and unhappy husband money money. No respect. for drink Husband dominates all decision-making in the house. Husband beats, slaps, punches, abuses On Norms and Agency wife. Sexual violence. Dispute between wife Children are and mother affected in-law psychologically 50 Focus group participants were asked individually in private to estimate the incidence of domestic violence in their communities before any detailed discussions took place on this subject with the group. 65 b. Causes and consequences deremployment. Some women also reported that violent of domestic violence abuse by husbands may stem from the women taking out loans, being unable to repay debts, or refusing to get ad- Perhaps to be expected, economic factors, such as ditional loans on behalf of their husbands. Men’s acute poverty, joblessness, hunger, and financial problems frustration in their inability to provide is well recognized and mismanagement, emerged most often as causes of as fuelling anti-social behaviors. Both women’s and men’s domestic violence (figure 2.4). The narratives revealed, groups gave numerous accounts of men suffering emo- moreover, that many times these fights are not just tionally (when under- or unemployed) and either with- about economic hardship. Men may also become bellig- drawing or lashing out in violence, while at the same time erent because they feel they need to re-assert or main- they may be squandering scarce assets on drinking, gam- tain their dominance, because they have lost the pro- bling, or other women. Yet, it is the woman—the tradi- vider status or ability that underpins their power in the tional household and budget manager—who may be ter- home, or because their partners are gaining economic ribly scolded or even beaten for raising questions about independence.51 running the household “on air.� A woman from Velugodu (Andhra Pradesh), India, a woman explained, This interchange from a men’s group in Martynice, Po- land, displays the tight links between money and power: if the husband comes home drunk, then the wife bursts out at him because she is suffering equally, or — Women are more often smart alecks [about money]. more than he is, due to their financial problems. Then — The money issue is not settled. Most of the the husband becomes angry and beats her. It is very disagreements [in a marriage] are connected to money. common for him to beat the wife. The wife and the — The wife asks, “Why are you earning so and so? children watching start crying and finally after some Couldn’t you find a better job? How am I supposed to time they sleep. live on this money?� And if she’s making more, that’s even worse. She says, “What do I need you for?� Rather than poverty alone, which certainly complicates — If the wife is making more than the husband, it’s over. a good wife’s household role, men’s insecurity in their She asks, “How come I can make such and such amount ability to lead or provide, for whatever reasons, underlie and you can’t?� many domestic clashes. In many of the contexts of economic stress on couples in Marital conflict (understood as disagreements in the way this sample, men are struggling with joblessness and un- the marital relationship is being conducted) is another FIGURE 2.4: PERCEPTIONS OF REASONS FOR DOMESTIC VIOLENCE Economic hardship 94 72 Marital conflict 81 74 Alcohol, drugs, or gambling problems 73 74 On Norms and Agency Challenge male authority 49 39 Unfaithfulness 36 43 Household responsibilities 30 27 Frequency of mentions Men Women Note: Data from 194 adult focus groups. 66 major trigger in figure 2.4 and is perceived variously to The Hyderabad men’s accounts of violence were surpris- be a cause, form, and effect of violence. Focus groups ingly similar to the women’s. They concurred that “lack of described marital conflict with terms such as disrespect, money� was a principal factor, that “men abuse women jealousy, lack of trust, miscommunication, misunderstand- at the slightest pretext,� and that “when abuses fly thick ing, personal differences, and too much stress. Among and fast, the man storms out to have a drink.� Men ex- the four adult groups, urban men had by far the most to pressed awareness of women’s rights, but they clearly say about the triggers of violence in general, mentioning felt that women invited the violence and, hence, it was poor marital relations the most. implicitly justified: “The fights carry on for a longer time and sometimes the women succumb to pressure and Current theories of domestic violence attribute it to agree that they were wrong.� an “interplay among personal, situational, and socio- cultural factors� (Heise 2011, 7), rather than just one Although norms for gender roles are clearly relaxing in factor. Men resort to violent behavior as a tactic for this Hyderabad neighborhood, the process seems to be asserting and demonstrating their manhood and their fraught with stress for many households. Domestic life power and control over women when they feel their po- is changing and many women now work for pay, but the sition is threatened (Connell and Messerschmidt 2005). community continues to value and award prestige to In the urban focus groups, in particular, both men and couples who play their traditional roles. Earlier in the women marked the stressful conditions for households focus group discussion, before the topic of marital con- and communities brought on by higher costs of city liv- flict was introduced, a man lamented about the change ing and more unemployed or underemployed men. In in local women’s attitudes and behaviors: “They want response, women are stepping up their economic role power. They want jewelry. They always want more and and, as a result, may be expecting more authority and they are highly demanding. ... They always want more.� cooperation. The men also expressed exasperation with women who get together “to compare jewelry and tales of beatings.� Focus groups from a bustling peri-urban neighborhood The men favorably recalled the wives of their father’s of Hyderabad (Andhra Pradesh), India, perceived that generations who were “good� and “sacrificing,� unlike to- domestic violence was intensifying due to these forces. day’s wives who “[c]ould not give time to children and The women from Hyderabad reported that, although family,� made “no sacrifices,� and “asked for rights.� For tension between couples often has economic roots, any their part, women were happier with the changes in gen- trifling reason—real or imagined—that men perceived as der roles and reported that men are better now at show- defying their authority (or altering women’s prescribed ing their love for their children. The women cautioned, conduct) may ignite violence: however, that a good wife should not pick fights with her husband for “silly reasons.� Even if the husband is not drunk, there may be quarrels at home for different reasons. When the household Indeed, women in many communities conceded they also faces financial crisis, he thinks that any talk about played a part in marital conflict. Women in a neighbor- finances points to his failure. The wife cannot avoid hood of Rafa Governorate in Gaza said they were less talking about the lack of money because basic needs, submissive than in the past and that violence sometimes like food, school fees, rent, or medicines cannot be happened because “often she just doesn’t obey him as a On Norms and Agency ignored. She becomes an “irritant� when she lists what way of abusing him.� Scott (1985) usefully classifies tac- they must have, so he starts fighting. He may take tics that involve resisting submission as the “weapons of anything as a pretext to show his anger and then he the weak,� and draws on the wider literature to show how beats her. Sometimes he starts a fight over whether the acts of insubordination and evasion of responsibilities food is not tasty or is too salty. are common strategies among powerless groups strug- 51 Generally, survey research is mixed on the association between domestic violence and women’s economic independence (Heise 2012). Two recent studies finding a significant link, however, include Hjort and Villanger (2011) and Heath (2012) on Bangladesh. See Chin (2011) who shows that reduction in domestic violence to rural women is associated with less exposure to their abusive husbands (reacting to their wives working for pay outside the household) rather than with their (wives’) increased negotiating power. 67 gling for recognition and influence, and how powerful wives. They do so because they are stronger and they groups will frequently go to great lengths to keep these are patriarchal and they want to satisfy their ego.� Or behaviors in check in order to prevent them escalating in Milne Bay Province, Papua New Guinea, the men’s to more costly cycles of resistance and reaction. Agarwal group warned what underpins ongoing conflict: “When (1997, 18) similarly argues that “persistent complaining, husbands and wives don’t trust each other, then there is pleading ill-health ... withdrawing into silence, and with- always misunderstanding between them.� holding sex are all means by which women are known to bargain within the family.� The women’s group in Ba’adan In short, findings on marital strife reveal a world where center, Yemen, for instance, agreed that “lack of money intensely coercive behaviors against women are widely and the husband’s income are some of the most impor- acknowledged. The violence is perceived to be easing, tant causes of fights between a couple, as much as when but still persists at varying levels. “Now women are be- the husband is jobless and sitting at home. ... [In this case, coming [more] powerful than men and men get furious,� when] his wife and children make lots of requests [for observed a woman from National Capital District, Papua money], the wife does not respect him.� New Guinea. Because the abusive tactics that character- ize marital strife have strong roots in everyday initiatives When a marital relationship becomes locked into tense that uphold and resist gender norms, violence against and abusive interactions, the focus groups consistently women is likely to have a dampening effect on gender described how deep stress and mistrust can then take norm change. Given the risks of overt challenges to gen- hold and become impossible to reverse. In a neighbor- der norms, quieter everyday negotiations and a gradual hood of Moldova, the men stated bluntly: “Besides, no relaxation of norms may provide a safer route for many, matter if the wife is good or bad, the husband will cheat but certainly not all. on her anyway.� In the focus groups, women sometimes argued that the specific reasons for marital discord are Empowering women and increasing their agency, as not only arbitrary, but generally boring or unimportant. much as finding exit options for women who are in a Men more often associated ongoing stress and discipline threatening situation, remain central challenges. Initia- (of the woman) with somehow failing to demonstrate ad- tives that address the stressful and costly consequences equately their competence and authority as household of male gender issues require stronger policy attention head. Some testimonies from men suggested they felt and research. A young woman in urban Emputa village, they need to redouble their efforts and regain control, Tanzania, urged, “There is huge need for education that while others just give up on the relationship, as women addresses men and their problems as men. I think drink- do. This quote from an urban men’s group in Al Fashir, ing too much and womanizing is their disease and needs Sudan, is disquieting: “Disputes are something normal. a cure in order to have a better community. Men should Spouses may disagree over any of the daily life details stop beating their wives. There is too much of that here and this may lead to conflict. In such cases, men have the in this community.� right to beat their wives. This is normal.� The testimonies make clear that domestic violence is nei- References ther hidden nor mysterious. Men and women displayed plenty of awareness and understanding. A village man Agarwal, A. 2003. Gender and Land Rights Revisited: Ex- On Norms and Agency of Sumadija District, Serbia, commented, “It’s a known ploring New Prospects via the State, Family, and Mar- fact in the village who behaves like this, but nobody in- ket. Journal of Agrarian Change 3 (1 and 2): 184–224. terferes in other peoples’ business.� Similarly, an urban Aizer, A. 2010. The Gender Wage Gap and Domestic Vio- women’s group in Pomoravlje District, Serbia remarked, lence. American Economic Review 100 (4): 1847–59. “There are women who pretend. They go out. They laugh. Appadurai, A. 2004. The Capacity to Aspire: Culture And they were just beaten.� Focus groups also provided and Terms of Recognition. In Culture and Public Ac- eloquent explanations of its causes and the logic of the tion, ed. V. Rao and M. Walton, 59–84. Stanford, CA, cruelty. In rural Ba Dinh District, Vietnam, the men ex- USA: Stanford University Press. 68 plained that “men often use their strength to abuse their Bisin, A., and T. Verdier. 2001. The Economics of Cultural Transmission and the Dynamics of Preferences. Jour- Indicators. OPHI Working Paper 04. Oxford: Oxford nal of Economic Theory 97 (2): 298–319. Poverty and Human Development Initiative. Chin, Y. 2011. Male Backlash, Bargaining, or Exposure Inglehart, R., and P. Norris. 2003. Rising Tide: Gender Reduction? Women’s Working Status and Physical Equality and Cultural Change around the World. Spousal Violence in India. Journal of Population Eco- Cambridge University Press. nomics 25 (1): 175–200. Kabeer, N. 1999. Resources, Agency, Achievements: Re- Connell, R.W., and J.W. Messerschmidt. 2005. Hegemon- flections on the Measurement of Women’s Empower- ic Masculinity: Rethinking the Concept. Gender and ment. Development and Change 30: 435–64. Society 19 (6): 829–59. ———. 2001. Reflections on the Measurement of Women’s Coston, B.M., and M. Kimmel. 2012. Seeing Privilege Empowerment. In Discussing Women’s Empowerment: Where It Isn’t: Marginalized Masculinities and the Theory and Practice. Sidastudies No. 3. Stockholm, Intersectionality of Privilege. Journal of Social Issues Sweden: Novum Grafiska AB and Sida. 68: 97–111. Hausmann, R., L. D. Tyson, and S. Zahidi, eds. 2012. The Das, M. 2008. Whispers to Voices: Gender and Social Global Gender Gap Report 2012. World Economic Fo- Transformation in Bangladesh. Bangladesh Develop- rum, Geneva, Switzerland. ment Series Paper, no. 22. Washington, DC: South Lundberg, S., and R.A. Pollack. 1993. Separate Spheres: Asia Sustainable Development Department, World Bargaining and the Marriage Market. Journal of Politi- Bank Group. cal Economy 101 (6): 988–1010. Farré, L. 2011. Background report prepared for World Mackie, G. 1996. Ending Footbinding and Infibulation: A Development Report 2012, World Bank, Washington, Convention Account. American Sociological Review DC. 61: 999–1017. Farré, L., and F. Vella. 2007. The Intergenerational Trans- ———. 2000. Female Genital Cutting: The Beginning of mission of Gender Role Attitudes and Its Implications the End. In Female Circumcision: Multidisciplinary for Female Labor Force Participation. IZA Discussion Perspectives, ed. B. Shell-Duncan and Y. Hernlund, Paper, no. 2802. Bonn, Germany: IZA. 245–82. Boulder, CO, USA: Lynne Reinner Publishers. Fogli, A., and L. Veldkamp. 2011. Nature or Nurture? Nyberg, A. 2011. Sweden Country Case Study. Back- Learning and the Geography of Female Labor Force ground Paper for World Development Report 2012: Participation. Econometrica 79 (4): 1103–1138. Gender Equality and Development. The World Bank Foucault, M. 1995. Discipline and Punish: The Birth of the Group. Washington D.C. Prison. 2nd ed. New York: Vintage Books. Offenhauer, P. 2005. Women in Islamic Societies: A Se- Hjort, J., and E. Villanger. 2011. Backlash: Female Employ- lected Review of Social Scientific Literature. Wash- ment and Domestic Violence. Department of Eco- ington, DC: Library of Congress. nomics, University of California, Berkeley. Panda, P., and B. Agarwal. 2005. Marital Violence, Human Heath, R. 2012. Women’s Access to Labor Market Oppor- Development and Women’s Property Status in India. tunities, Control of Household Resources, and Do- World Development 33 (5): 823–50. mestic Violence. World Bank Policy Research Work- Pronyk, P.M., J.R. Hargreaves, J.C. Kim, L.A. Morison, G. ing Paper, no. 6149. Washington, DC: World Bank. Phetla, C. Watts, J., and J.D.H. Porter. 2006. Effect Heise, L. 2012. Draft. Background report prepared for of a Structural Intervention for the Prevention of World Development Report 2012, World Bank, Wash- Intimate-partner Violence and HIV in Rural South Af- On Norms and Agency ington, DC. rica: A Cluster Randomized Trial. Lancet 368 (9551): Hoff, K., and J.E. Stiglitz. 2010. Equilibrium Fictions: A 1973–83. Cognitive Approach to Societal Rigidity. American Quisumbing, A.R. 1994. Intergenerational Transfers in Economic Review 100 (2): 141–46. Philippine Rice Villages: Gender Differences in Tradi- Holland, J., C. Ramazanoglu, and S. Sharpe. 1998. The tional Inheritance Customs. Journal of Development Male in the Head: Young People, Heterosexuality and Economics 43 (2): 167–95. Power. London: Tufnell Press. Rauch, J., and S. Kostyshak. 2009. The Three Arab Worlds. Ibrahim, S., and S. Alkire. 2007. Agency and Empower- Journal of Economic Perspectives 23: 165–88. ment: A Proposal for Internationally Comparable Rao, V., and M. Walton. 2004. Culture and Public Action: 69 Relationality, Equality of Agency, and Development. In Culture and Public Action, ed. V. Rao and M. Wal- ton, 3–36. Stanford, CA, USA: Stanford University Press. Reich, A. 1976 (1995). Of Woman Born: Motherhood as Experience and Institution. New York: W.W. Norton. Ridgeway, C.L., and S.J. Correl. 2004. Unpacking the Gender System: A Theoretical Perspective on Gen- der Beliefs and Social Relations. Gender and Society 18 (4): 510–31. Samman, E., and Santos, M.E. 2009. Agency and Empow- erment: A Review of Concepts, Indicators, and Em- pirical Evidence. Research Paper 10a. Oxford: Ox- ford Poverty and Human Development Initiative. Scott, J.C. 1985. Weapons of the Weak: Everyday Forms of Resistance. New Haven, CT, USA: Yale University Press. Stevenson, B. 2008. Divorce Law and Women’s Labor Supply. Journal of Empirical Legal Studies 5 (4): 853– 73. DOI: 10.1111/j.1740-1461.2008.00143.x. Swaminathan, H., C. Walker, and M.A. Rugadya, eds. 2008. Women’s Property Rights, HIV and AIDS, and Domestic Violence: Research Findings from Two Ru- ral Districts in South Africa and Uganda. Cape Town, South Africa: HSRC Press. United Nations Development Program UNDP. 2011. Hu- man development report 2011: sustainability and eq- uity: a better future for all. United Nations Develop- ment Program. New York World Bank. 2008. Whispers to Voices: Gender and So- cial Transformation in Bangladesh Washington, DC: World Bank, South Asia Sustainable Development Department. ———. 2012. World Development Report 2012: Gender Equality and Development. Washington, DC: World Bank Group. ———. Forthcoming. Opening Doors: Gender Equality in the Middle East and North Africa. Washington, DC: World Bank Group. On Norms and Agency 70 The power and ability to have choices and make de- cisions, and especially to have a say at crucial junc- tures of one’s life are arguably the elements most fre- PART II: quently associated with agency. Together with having control over assets, including income from earnings, they are also the measures of agency most often used On Norms and Agency Having and in the literature.52 As Ibrahim and Alkire (2007) note, while asset and income control are more precondi- making choices tions to agency than actual reflections of it, together with decision-making these measures allow for interna- tional comparisons, which is central to a study like this. 52 For a review of agency, see Jejeebhoy (2000), Ibrahim and Alkire (2007), Kabeer (2001), Samman and Santos (2009), and World Bank (2012). 71 The power to choose largely arises in household deci- The choices in crucial household decisions are subjec- sion-making because there “individuals confront basic tive, reflecting individual preferences and interests, but livelihood concerns, norms, values, power, and privilege� are also influenced by specific contexts and constraints. (Narayan et al. 2000, 219), including gender roles, as dis- Partly dependent on the opportunity structure around a cussed in chapter 1. Women’s participation in decision- decision—for example, the presence of schools, the state making at the household level is essential to their well- of the local market economy, or availability reproductive being and sense of self-efficacy. But this is not true for health services—choices are also subject to norms and all domestic decisions. In this section, we look at intra- how they shape women’s and men’s preferences. Here, household decisions linked to agency, such as those re- “culturally produced dispositions, beliefs, and behaviors� lated to family formation and education of children. are likely to operate as “constraining preferences� (Rao and Walton 2004, 15). They reflect both the internaliza- Women’s life trajectories are dependent on certain choic- tion of the possibility of success or failure, given the dom- es that they may (or may not) be able to make. While inant norms regulating a man’s or woman’s position in the women and men make decisions every day, not all carry broader social structure, and the ability to take advan- the same weight. Certain decisions have a greater impact tage of the structure of opportunities. The slow pace of on women’s and men’s lives’ paths, particularly whether change in the “terms of recognition,� as Appadurai (2004, and when to work, when and who to marry, and how many 64) notes, emphasizes the conditions and constraints un- children to have and when—which Kabeer (1999, 436) calls der which women (and men) negotiate the gender norms strategic life choices, or those choices “which are critical that frame their lives. These terms of recognition are for people to live the lives they want.� Women participate present in different forms, from rituals to cultural prac- in many routine decisions appropriate to their traditional tices to public discourses and internalized beliefs –includ- role and the gender division of labor in a household. But ing whether a woman or man is recognized as entitled to these decisions are unlikely to be “strategic� and trans- be the decision-maker. They are central to having a voice late into empowerment or improve gender equality as a and they affect the outcomes of decisions at all levels. consequence. Women’s gains in power to decide or nego- tiate more significant issues are more evident in strategic What is evident, according to many of the focus group life choices, where they perceptibly affect the course of participants, is that a window to aspire to a different women’s lives or influence desired outcomes. life, to more education, or to have a choice in impor- tant matters (such as family formation) has opened up. Strategic decisions include how much to invest—or Whether or not these aspirations materialize depends whether to invest at all—in the human capital of children on the structure of opportunities and available resourc- (girls’ and boy’s education) and how to allocate different es, but the existence of such aspirations drive women (or responsibilities, assets, duties, and rights inside and out- men) to achieve them. The focus groups made clear that side the household (e.g., who works, who does the care- changing aspirations regarding children’s education lead taking, and who makes major spending decisions and on to greater-than-average achievement by adolescents and which expenditures). They are influenced by prevailing youth in the community and encourages girls and boys to gender roles and norms, and by the relative voice and dream more ambitiously. What focus groups recounted bargaining power of the adult members of the house- about educational achievements in their communities hold. In a continuous feedback loop, as described in tends to fall midway between norms and desires, due On Norms and Agency World Development Report 2012, the endowments that both to the presence of more schools and the change individuals have accumulated, the opportunities avail- of view regarding the value of education for children in able to them, and their control over resources give in- general and girls in particular. A similar situation can be dividuals increased decision-making power and greater seen in other cases such as age of marriage or childbear- agency. Jointly with the relaxation or change of gender ing, young women’s and men’s aspirations are not yet norms, such agency allows women to take advantage of achieved but they see the most likely materialization in opportunities to accumulate assets, challenge disadvan- their lives to be somewhere in between what has been tages arising from gender inequalities, and gain more the prevalent norm in their communities and what they 72 control over their lives. aspire for their own lives. References Ibrahim, S., and Alkire, S, 2007. Agency and Empower- ment: A Proposal for Internationally Comparable In- dicators. OPHI Working Paper 04. Oxford: Oxford Poverty and Human Development Initiative, Appadurai, A. 2004. The Capacity to Aspire: Culture and Terms of Recognition. In Culture and Public Ac- tion, ed. V. Rao and M. Walton, 59-84. Stanford, CA, USA: Stanford University Press. Rao, V., and M. Walton, eds. 2004. Culture and Public Ac- tion. Stanford, CA, USA: Stanford University Press. Kabeer, N. 1999. ‘Resources, Agency, Achievements: Reflections on the Measurement of Women’s Empow- erment.’ Development and Change 30 (3): 435–64. ———. 2001. Reflections on the Measurement of Women’s Empowerment. In Discussing Women’s Empowerment: Theory and Practice. Sidastudies No. 3. Stockholm, Sweden: Novum Grafiska AB and Sida. Jejeebhoy, S.J., H.B. Presser, and G. Sen. 2000. Women’s Autonomy in Rural India: Its Dimensions, Determi- nants, and the Influence of Context. In Women’s Em- powerment and Demographic Processes: Moving Be- yond Cairo, ed. H.B. Presser and G. Sen. New York: Oxford University Press. Samman, E., and Santos, M.E. 2009. Agency and Empow- erment: A Review of Concepts, Indicators, and Em- pirical Evidence. OPHI Research Paper 10A. Oxford: Oxford Poverty and Human Development Initiative. On Norms and Agency 73 CHAPTER 3 Strategic life decisions: Who has the final say? T he women—and men—participating in this study prevailed over traditional gender norms and roles. The make or influence their life choices via a process second set of decisions the chapter looks at pertains set within a non-egalitarian gender system that to the first job. This decision sometimes remains in the constrains their agency. In this chapter, we look at sev- hands of the parents, who may push children into the eral specific strategic life choices (Kabeer 1999), which labor market to get an early return on the investment in can be critical in determining the life men and women their education or to acquire extra income to help deal actually live and the life they want to live.53 We do not with economic need. In other cases, the adolescents ignore the relevance of making small decisions or “em- or young adults themselves make the decision to start powerment in small matters� (Schuler and Rottach 2010, working for pay. 381) or its potential link to the ability to make large deci- sions. As Malhotra et al. (2002) note, being able to decide The most visible of all strategic decisions for women what to cook—while not equivalent to having the power center on family formation. Women’s control over their to make decisions about children’s schooling, health, or own bodies is a strong marker of their agency, although marriage—when aggregated with other small decisions it has been—and remains—highly contested. In traditional may provide useful insights on intrahousehold decision- settings, where early marriage and childbearing for girls making processes. We asked the focus groups how much is common, parents usually decide when (what age) and freedom women and men, young and old, have to make whom a girl will marry. Also, mothers-in-law and other decisions about their own lives. Could they identify con- family members often have significant influence over a straints? Did different social and gender norms affect young wife’s child bearing, specifically when she should their decision-making processes? start having children, how many, and what desired sex. Sen (cited in Bardham, Dattachandhri, and Krishnan The chapter begins by looking at education decisions. 1999, 458) notes that reproductive choice can be about Here we assume that parents make the decisions rather agency, but it also may be a trade-off for other sources than children. Deciding whether a girl or a boy goes to of power: “Bearing the approved number of children will On Norms and Agency school, continues in school, and completes school has grant a woman the rights and privileges accorded to a more to do with the parents’ decision-making authority fertile woman, but does not necessarily give her greater and their views on education. As revealed by the “good autonomy in decision-making.� student� attribute of both good girls and good boys (see figures 1.2 and 1.3 in chapter 1) and parents’ general aspi- Similarly, decisions to wed may bring both new freedoms ration of education for their children (as well as by young and new constraints for women. A woman may decide women and men for themselves), we can assume that, for herself when, who, and under what conditions to across the sample countries, focus groups so valued marry, but she may also “choose� under family and social 74 education as an investment in future well-being that it pressures to comply with expected norms. Marital prac- tices usually reflect cultural norms and are a sure sign of to education in Rafah, West Bank and Gaza, has perse- how gender relations and social relations are generally vered and risen, like most of the sites visited for this re- organized within a society (Malhotra 1997). As Kabeer port: both women and men are graduating from second- (2005) observes, little agency exists when there is little ary school and university or vocational school (tertiary choice or just a passive form of it. Agency that conforms level) at higher rates than ten years ago. According to the to traditional gender norms may help women be effec- community’s focus group in the study, these changes oc- tive in their gender-assigned tasks and roles, but it does curred due to constant negotiation and interaction with not challenge the gender system. prevalent gender norms. The chapter ends with the matter of intrahousehold de- The adolescents’ focus groups from Rafah pointed out cisions on use (expenditure) of assets and household that, in line with traditional gender roles, their fathers’ purchases. Women’s ability to get a job or start a busi- voice prevails in household decisions, including who stays ness—in other words, to earn independent income—is a in school and who must drop out. While some acknowl- strategic means of increasing their bargaining power and edged that both parents decide about their schooling, participation in household decisions. Independent in- the decision is largely out of their (girls’ or boys’) hands, come also gives women something to fall back on if they regardless whether the parents consider their prefer- need to leave a difficult or violent domestic situation. Be- ences or not. ing able to accumulate some assets, and to control them, also increases women’s agency and voice.54 The parents participating in the Rafah focus groups gave similar reasons for pulling their children out of school Across the 20 countries in our study, some similarities that we find in other sample countries. In the case of in the major barriers to the exercise of agency emerged. boys, household financial problems often dictate break- It is not surprising that strategic life decisions are not ing off their education: “The boy and his parents decide necessarily affected by economic development. In some he should leave school in order to find a job and help urban communities in the sample, we see more opportu- provide for the family.� For girls, marriage trumps edu- nities for women to actively exercise their agency and a cation: “If the girl is pretty, then her parents stop her greater universal value of education for both boys and schooling to get her married.� (This may also apply if the girls. But in the private sphere, behind the household’s girl is a bad student.) If a suitable man asks for a girl’s front door—regardless of location—conformity with tra- hand, she no longer needs to be educated because her ditional gender norms and practices remains persistently future is guaranteed. intact. While the aspirations of the younger generation are changing, many are unable to realize their goals, but In decisions about education, gender norms are in full signals show that they are on the way to do so. play—the father’s authority, the good boy who works and provides for his family like a good man, the good girl who becomes a good wife and manages the household. When 1. Investing in education: a girl leaves her parents’ house, though, education is then Why should girls and boys negotiated with the new man in her life. One 20-year-old go to school? woman in Rafah lamented, “I was studying to be a vet- “ erinarian, but because I had to go out to the farms with � On Norms and Agency For those who have education, every single door is open. men as part of the practical study, neither my husband — Urban man, Kragujevac, Serbia nor his family would accept it. My only solution was to change my major.� These normative constraints, however, Despite ongoing conflict and deterioration of local eco- are now pushing up against people’s growing recognition nomic conditions and employment opportunities, access of the value of an education, both as an investment for 53 This is akin to Amartya Sen’s (1985) notion of “functionings,� which range from basic concerns (e.g., being healthy, having a good job, and being safe) to more complex states (e.g., being happy, having self-respect, and being calm). The freedom to achieve functionings obviously has an instru- mental value, but it also has intrinsic value to a person’s quality of life. 54 Chapter 4 in World Development Report 2012 analyzes available evidence from around the world on this relationship. 75 future well-being (e.g., getting a good job) and as a trans- Boys, girls, and their parents invest in education be- formative power that opens up previously unattainable cause they expect to see a return on it, whether it is possibilities and expectations, or the capacity to aspire functional literacy or an advanced graduate degree. As (Appadurai 2003). Educating children—including girls— Patrinos and Psacharopoulos (2004) document, educa- has become a new norm and deemed necessary to en- tion returns—expected earnings—have risen in all world sure their future, as almost all our focus groups agreed. regions. The increase in returns is higher for girls than Yet, the results of educating girls and boys are not as boys when they go beyond primary school.56 “Education straightforward as they seem: the new aspirations and op- takes us to good places; it is our road to employment and portunities for those with more schooling are not always a path out of poverty,� exulted an adolescent boy from enough to overturn longstanding social and gender norms. Fiji. A young man from Sudan agreed, “Education lets us join the modern world and offers us better jobs now. In The impact of education on access to future opportuni- the past, it was not important because our people were ties for boys and girls is undeniable. We know that the par- farmers and did not pay attention to the future or look to ents in our focus groups place high value on their children change the present.� Boys from urban Mongar District, being good students and getting an education because Bhutan, explained that the gains are not only for them they told us. But did the adolescent boys and girls agree? but for their entire families: “We can go to college, earn We asked the 670 adolescents in our study, who were a salary, and help our struggling parents. It is payback 12–17 years old and lived in 41 urban or rural communities time. We want to help our parents who got us educated in 8 countries.55 According to their average school enroll- with their hard-earned money.� ment, almost all of them go to school, but the girls aspire to higher levels of education than the boys (figure 3.1). But when people perceive that returns from education are low, investment in it falls off. Boys from urban West More than 60 percent of the girls in both urban and rural Bank and Gaza commented that staying in school lon- areas hoped to obtain advanced degrees, but only 40 ger than necessary was a waste of time and had little percent of the boys sought this level of education. Girls effect on their future economic participation: “[Higher and boys both expected that getting an education would education] is not important to us because we can learn give them a better future than their parents and would a craft and work without a university degree.� In rural permit them to fulfill their parents’—and their own—aspi- Sudan (Blue Nile State), boys also did not believe that rations for them. education could really help them in the future. They FIGURE 3.1: IDEAL LEVEL OF EDUCATION REPORTED BY ADOLESCENT BOYS AND GIRLS IN THE STUDY 70% 70% 60% 60% Share of total mentions Share of total mentions 50% 50% 40% 40% 30% 30% On Norms and Agency 20% 20% 10% 10% 0% 0% Finish Finish Finish Get an Other Finish Finish Finish Get an Other primary secondary college advanced... primary secondary college advanced... Ideal level of education Ideal level of education Rural boys Rural girls Urban boys Urban girls Note: Data from 670 students between 12-17 years in 8 countries. 76 mentioned knowing college graduates in the area who nities to earn their own income—all elements of indepen- were unemployed. Worse, when prompted to describe dence. A girl from Hyderabad (Andhra Pradesh), India, themselves 10 years in the future, a group of boys from explained that, as part of achieving her goal to graduate urban Sudan (Khartoum) answered, “jobless.� from college, she tries to imagine what it takes to be self- sufficient and learns how to manage her money. Another From education’s instrumental relation with income girl from urban Lautoka, Fiji, believed that education not generation we can see how gender norms and expected only provides her with knowledge but also gives her the behaviors pressure boys to contribute to the household courage to “become someone in life, to even become the in advance of their future role as expected provider— head of the household.� so much so that the opportunity to earn income takes precedence over any education opportunity: “If there “Education is a girl’s best weapon to face the world,� sum- is money, there is no need to learn� (young adult man, marized a girl from neighborhood of Rafah, West Bank Olsztyn, Poland). The poor economic conditions of and Gaza. And she is right. Education has started reshap- many of the communities in the research sample also ing local norms that define women and changing per- contribute to the lack of confidence in the value of edu- ceptions and expectations. It is often the mothers who cation: “People don’t have any interest in education be- envision a different life for their daughters and strongly cause they are too busy making a living� (adult man, Naw advocate for them to stay in school. In Fiji and Bhutan, Da, Afghanistan). However, as discussed earlier in part both rural and urban girls told us that their mothers en- I, boys and men both are starting to resist and contest courage them to study hard, so they are able to become the expectations laid on them, and are contesting the independent and to look after themselves. masculine ideal of being the main provider and deciding to stay in school or to combine education with working Moving up half a generation to the 18–24-year-old women in view of future aspirations. in the sample, who may have at least finished secondary school, we see that they share a similar view of educa- More education also imbues people with increased self- tion. They highly value education because they expect efficacy and confidence. For adolescent girls specifically, it to help them better their own and their children’s the added value of an education—beyond mere econom- economic well-being and, more importantly, advance ic returns—is its positive impact on their agency, which their personal development. Education leads to better is less important to boys as a reason for valuing educa- decision-making and strategic decision-making requires tion. Similarly, World Development Report 2012 shows information and education. “If you are not educated, you that higher levels of education reduce the grip of social cannot think. When other people tell you things, you norms that restricts women’s autonomy. For example, in take too long to understand,� said a woman from Emputa South Asia and the Middle East and North Africa, wom- village, Tanzania. “Education brings awareness of more en with more education are not as likely to have to ask things so that I make better informed choices,� noted a their husbands or family for permission to seek medical young woman from Bhubaneswar (Odisha), India. care. Education gives them more freedom than earnings (World Bank 2012).57 The literature shows that women with more education tend to have more control over other life decisions, and The adolescent girls in the focus groups agreed, reporting they tend to marry later and have fewer children,58 which On Norms and Agency that education helps them gain more control over their was confirmed in many comments by the girls in the study lives, bolsters their self-esteem, and opens up opportu- about the lives they see their mothers leading: “My mum 55 Fiji, Dominican Republic, Bhutan, India, Burkina Faso, Sudan, West Bank and Gaza, and Yemen. Togo adolescents were also interviewed but the data was not included in the figures in this section. 56 The highest returns on education are recorded in low-income and middle-income countries when looking at income levels. When looking at regions, Latin America and Caribbean region and sub-Saharan Africa region see the highest returns. The lowest returns are in non-OECD European countries and the Middle East and North African countries. 57 See World Development Report 2012 (World Bank 2012, 169) for estimates on the probability of women who need permission to get medical care based on demographic and health data. 58 See Field and Ambrus (2008), Goldin and Katz (2002), and Pezzini (2005), for example. 77 had to help look after her brothers and sisters [when she was young], so she could not complete her schooling or have the opportunity to work. She then bore six children “ My parents wanted me to keep learning up to 8th grade. I wanted to finish high school, but my father forbade me. He said, “What do you need school for? You and had to stay home and look after them. She spends all will get nothing out of going to school. her time doing housework and looking after the family� You are just wasting your days. So I secretly went to (adolescent girl, rural Fiji). “My mother only finished 10th school, but he found out and yelled at me, “I warned grade. Her life was very simple in order to raise us. That is you about school!� I feel remorse that I listened to why I certainly don’t want to be like her (urban adolescent him. I could have continued, but you need books and girl, Dibras, West Bank and Gaza). Social norms and the other things for school, and you need a lot of money for fulfillment of traditional gender roles, again, are the main education. My mother only does what my father tells reason that the girls’ mothers left school and remain as her to do. My brother only finished 5th grade because barriers to future opportunities for the new generation. he did not like it. I liked school, but my father would not allow it. He was afraid that I’d meet somebody there � When we asked parents in all 20 sample countries di- who would take advantage of me because I was a girl. rectly about their expectations for their daughters and — Urban woman, Kragujevac, Serbia sons, they tended to mention education as one of the primary options that can offer their children a better future. They felt that having a school nearby or in the sion was made jointly with an adult. For young women, community was essential for their daughters, even in 15 percent acknowledged that they did not have a say difficult locales, such as the West Bank and Gaza, rural in decisions about ending their schooling, compared to Afghanistan, or traditional communities in Burkina Faso, about 7 percent of boys (figure 3.2). Young men were and among such minority communities as the Roma in more likely to say that the decision to leave school was Serbia and indigenous populations in Peru. Fathers and their own than were girls. mothers alike noted that their daughters gain more equal standing and have more independence from their future Over half the young women and men in our sample husbands when they can earn their own income. Young dropped out of school early, which will have a significant girls will “not be so easily confined� (Burkina Faso) when impact on their future. Unfortunately, it means the level they secure the freedom to look after themselves. “For of education that many of them hoped to reach will re- my daughter, I want her to have power. I want to give main merely an aspiration. For example, from Yemen to her an education, so she has more opportunities, and Bhutan and from Burkina Faso to India, the adolescent even a degree, so she can be independent. I want my girls wanted to get at least a college degree, and the ru- daughter to be better than me,� asserted a rural woman ral girls were aware that their education will likely only from Peru. Neither mothers nor girls want to replicate go through secondary school, if they are lucky. When we the lives women in the past have endured and they rec- asked girls from India how far they thought they were ognize education as their main outlet to change. truly likely to go in school, their answers varied. Like other girls from rural communities, some guessed they would just make it to secondary school and not fulfill their 2. Why should I leave school? dreams of becoming doctors, lawyers, or engineers. Not my choice! On Norms and Agency The boys, too, aspired to get college and professional Who decides when it is time for a child to leave school? degrees, but realized that, like those in Burkina Faso and What factors influence this decision? Looking at our data, Fiji, primary school might be the end of their education. one-third of the young adults reported leaving school Other boys in Burkina Faso did not want to go beyond because they completed their education. Depending on primary levels, and boys in Dominican Republic put a low the context, completion meant primary school to college value on the promise of education to provide a better level.59 A little less than one-third of the young men said future. Indeed, parents of boys in four of the six Afri- that the decision to leave school early was their own, can countries in our sample felt that their sons—not their 78 while an additional 22 percent indicated that the deci- daughters—did not take education seriously or value its FIGURE 3.2: WHO MAKES THE DECISION FOR CHILDREN TO LEAVE SCHOOL? 40% 35% Share of total mentions 30% 25% 20% 15% 10% 5% 0 Completed education Own decision Decided with adult Adult decided Other (bad marks) Young women Young men Note: Data from young women and men (18-25 years old) in 113 focus groups. potential benefits enough.60 Young men more often re- pants always described the decision to pull children out ported not liking school over other reasons for dropping of school as difficult. The same two fundamental reasons out, including the need to work and the absence of a that drive families to educate their children—a better life school in their community. and more income—are also the first ones parents men- tioned for stopping their children’s education. On one Often, though, the adolescents’ and young adults’ focus hand lies lack of economic resources for school fees, groups mentioned lack of money and their parents’ deci- transportation, and school materials, and on the other is sion as the two main reasons they left school. The lack of the desire or obligation for young people to generate in- power to have the final say over their education cannot come or relieve their families from the economic burden be disentangled from the reasons behind a decision to that education represents for the household budget. leave school. Stated preferences and individual decisions do not take place in a void, but are constrained by op- Foregoing the opportunity to invest in future returns portunities, power imbalances in the household, gender from education may, at first glance, point to a lack of systems, lack of information, and more. Distance and lack agency by a young woman. Both the Bhutanese girl, who of infrastructure—from no roads to no nearby school— opposed her father taking out a loan to pay for her edu- affect both boys and girls, but girls have more difficul- cation, and the young woman from India, who decided to ties if there is no school in their village, if they have to leave school because her family cannot afford her edu- travel long distances to school, and if their friends drop cation, are making a deliberate choice. Financial issues out (they have no peers to walk with them to school) be- and family loyalties are so intertwined, however, that cause their mobility is more restricted than boys’. Both it is not that clear that this is an empowered or strate- also leave school to help support their families (boys by gic choice. We can mark their decision as either a sign earning an income, girls by working at home), to prepare of self-determination or a passive choice due to pres- On Norms and Agency for and marry early, or—as girls from Bhutan, Burkina sure and lack of alternatives. When we judge a boy who Faso, Yemen, India, and Sudan added—so they will be leaves school because of financial constraints, the same safe from violence and unapproved pregnancies. duality appears. Given the high valuation of education held by the com- Especially with entrenched poverty, young people seem munities in the study, it is not surprising that the partici- to have little bargaining power to state their prefer- 59 Our survey did not collect specific data about the education level attained by the participants, just their comments on, and their perceptions of, levels reached in their communities. 60 Participants in urban and rural communities in South Africa, Tanzania, Burkina Faso, and Togo also echoed this perception. 79 ences. A young man especially faces strong gender man- es both lack of agency and her belief in what a good wife dates calling him to be a family provide and act like a should do. While it may seem contradictory, because man. Young men’s accounts of leaving education due to more girls than boys currently attend school, social- and family financial straits, however, show a proactive com- gender-normative restrictions are more evident in re- ponent. “Our family condition (financially) was not so gions where education for women still lags behind men’s strong. We thought of supporting our families� say young or where general enrollment in school remains low, such men from Jaipur (Odisha), India, and like them, other as sub-Saharan Africa and some regions of India:62 young men leave school not only to reduce stress on the family budget but also to make a positive contribu- – The pressure of endless domestic tasks: “You see, tion. “I had to leave school because my father separat- if we went to school, who would do the housework?� ed from my mother and I had to support her,� noted a (Rural young women, Malangachilima village, Tanzania) young man, from Lambayeque Province, Peru. The same imperative to generate an income was heard from boys – Early childbearing: “My boyfriend got in the way in Moldova, West Bank and Gaza, and other countries. of my education. When I was in the 6th grade, I got But when some boys and young adult men in the sample pregnant. I had the child and my boyfriend left me.� left school to gain economic independence, they surpris- (Urban young woman, Greenville District, Liberia) ingly reported a sense of empowerment that other ac- counts did not have: “My parents weren’t happy when I – Traditional mobility restrictions on women in order to decided not to go to high school. But I already had a job protect their virtue, propriety, family honor, and safety: and was earning my living. I couldn’t focus on my studies “How can we walk on deserted and lonely jungle paths anymore� (young man, East Jakarta, Indonesia). to reach school?� (Rural young women, Kalahandi District (Andhra Pradesh), India) Looking further into young men’s accounts, we find that, even when they reported making the decision to leave – Household preference for educating boys: “As a girl, I school independently, they regretted it. They sometimes had to agree with my parents to support the education expressed as strong a wish to go back to school as the of my brother. Time passed and now I am married young men who had no say in the process. This signals with kids, and I have not been able to complete my that the decision was influenced more by restrictive cir- education.� (Rural young woman, East Sepik Province, cumstances, whether need or norm, and less by prefer- Papua New Guinea) ence.It may be that the young men adapted their pref- erences to practical conditions and needs.61 Again, social We cannot claim, however, that it is more altruistic for norms associated with a sense of duty appear to be a pri- boys to drop out of school to financially support their mary factor. One young man in Lautoka, Fiji, put it matter- household than for girls to leave school to do house- of-factly, “Father got sick, so I made my own decision [to work or to let a male sibling attend school. Agarwal (1997) leave school].� An Indonesian young man from Tangerang reaches a similar conclusion: girls and boys equally have presented another constrained choice: “I wanted to con- no choice but to agree to these “voluntary� concessions. tinue my studies, but we didn’t have the money. So I de- cided to stop after I finished junior high school.� A frequent justification for preferring to educate sons rather than daughters involves inheritance laws and tra- On Norms and Agency Overall, for young women, the traditionally-instilled ditional roles associated with family financial support, es- norms of inequality inside and outside the home raise pecially in rural areas. Because young women will join an- more barriers to finishing their education. How are they other household when they marry turns their education able to exercise agency from their unequal bargaining into a bad investment for their family. Their acquired ca- position in the face of the traditional authority of fathers pacity from education will not serve their family: “There and husbands? A young woman in Tewor District, Liberia, is a common belief that when we [women] get married, justified leaving school: “[It was] my own decision be- our education benefits the husband’s family, not our fami- cause I had a husband and I had to follow his way. He said lies. So a father feels that his family loses if he educates 80 no, that I could not go to school.� Her situation showcas- you� (young woman, Malangachilima village, Tanzania). Educating young men, on the other hand, has direct re- turns to the household. And while male power affects FIGURE 3.3. WHO DECIDES WHEN YOUNG both girls and boys when it comes to education, only girls ADULTS OR ADOLESCENTS FIRST GO TO WORK? are bound to transition from the school to the household. 80% 70% Share of total mentions Despite diverse barriers, both boys and girls are stay- 60% ing in school longer than previous generations, and the 50% overwhelming embrace of education is causing deeply 40% engrained norms to slowly relax and bend. Signs of 30% change are visible in narratives from the adolescents’ 20% focus groups, which related accounts of traditional, re- 10% strictive fathers who push their daughters to study and 0% of mothers whose gains in voice may be counteracting Young Young men Young Young men women women the fathers’ resistance to schooling their daughters. “I am Urban Rural studying because my mother insisted that girls should at Own decision Decided with adult Adult decided least complete 10th grade, even though my father doesn’t Note: Data from 194 male and female focus groups. want me to study in a regular school,� announced an ado- lescent girl from rural India. For a Bhutanese girl, it is her educated brother who is pushing for change—“My broth- stances, gender roles, entrenched norms, and market er forced my parents to put me in school. My parents opportunities all play into their decisions. never felt that I needed to go to school�—even though her parents views did not change in the long term. “They be- In most of the communities in our survey, women have lieved that I have to stay home and take care of the land participated in the labor market for more than 10 years, and the main house of the family.� This girl eventually was but, like the 500 million women who have joined the glob- pulled out of school when her brother left the house, but al labor force since 1980, they have worked under dis- her knowledge and aspirations changed in the process. advantaged conditions, with limited access to assets and services, and coped with the unequal gender distribution of household responsibilities.63 Getting a job requires 3. From school to work: that women, like men, have the skills that fit the work and Getting the first job access to information about labor opportunities; but un- like men, they also need an enabling environment that Strategic choices do not arise frequently in a person’s includes options for childcare, redistribution of domestic lifetime, which makes their impact less visible in the tasks, access to transportation, and mobility. short term, but their impacts are significant over a longer time frame. Starting a productive activity, such as finding In many developing countries in our survey, contexts a first job, is one example. As Malhotra et al. (2002) rec- of scarce opportunities drive the timeline for starting ognize, getting a job can be a manifestation of women’s work. By age 13—some in rural areas were as young as agency as a decision-making exercise, as well as driver 10 years old—the majority of the adolescent participants to promote greater agency. Women’s economic partici- were already economically active (much earlier than we On Norms and Agency pation can be an enabling factor to predict women’s in- expected), even if not continuously; they also frequently creased control over other important decisions in the worked while going to school. Not surprisingly, some felt household and their lives. Not all women, however, are that they were thrust into the job market despite their free to make (or capable of making) the decision to leave desire to learn a skill or to complete their education. In the domestic space to start working. Household circum- addition, although more than half the young adults in the 61 “Adaptive preferences,� as described by Sen, respond to the material conditions of individuals’ lives. Nussbaum argues that these choices are not real expressions of agency, but a simulacrum of choice. 62 World Development Report 2012 refers to these groups as severely disadvantaged populations or regions. 63 World Development Report 2012. 81 focus groups described making an independent decision evant under changing economic circumstances that are to work (figure 3.3), in most rural settings and for women not under their control: (regardless of location), it was less likely to be their deci- sion than an adult’s. Urban young men appeared to have You work at what you can, at whatever is available. the most freedom to decide to work, as well as a larger, I know someone who finished three years of vocational more diverse pool of opportunities. school for plumbers and got a job in “Goša� paying 35,000 dinars. Now, because the company was sold, Like leaving education, the decision to start working is he works as a security guard for 20,000 dinars where also usually made within constraints: the need to sup- his education credentials are not recognized. He can’t port the family, the lack of resources to afford education, find any work in his profession.� (Village young man, the desire to be economically self-sufficient. “It was my Sumadija District, Serbia) own decision to help my parents� is something that we heard from urban young men and women in Indonesia In a few communities, however, where job prospects and in Yemen, where almost all adolescents interviewed appear to be improving, the narratives of boys and girls stopped their education, due to lack of resources, to were more positive in terms of decision-making power. start working or start helping at home. It was rare in the A group of young girls in Hyderabad (Andhra Pradesh), focus groups to find examples where adolescents or India, said, “mostly the boys decide where they want to young adults managed to combine education and work work ... [but] now the girls can too because there are successfully in the long run. Going to school and holding many job opportunities.� And with more work oppor- down a job, when household finances were tight, often tunities, education becomes more valuable. “For un- made continuing their education impossible, but also educated people, there are limited choices for work, so gave them a sense of independence from their parents’ they have to take the first one. But in the many malls and designs—a first step toward self-efficacy and the capac- shops, there are more jobs available if you have school- ity to act. According to adolescents in Umlazi township ing,� the girls from Hyderabad explained further. A young A (near Durban), South Africa, working and studying man of Hoang Mai district of Hanoi, Vietnam noted that helped them learn to be responsible and to manage “in Hanoi … those who have little education can work as their own money, but it is not easy, and more often than motorbike drivers or masons. It doesn’t matter as long not students ended up choosing immediate returns over as the work brings an income.� Unfortunately, these ex- long-term investments. “Work is very demanding; it takes amples are the exception. Adolescents and young adults a lot of energy. If you are a student, it interferes with predominantly pointed out that acute economic need in studying because your mind is always tired,� said a young the family prompted their job-seeking decisions regard- girl from Umlazi township B. less of the availability of employment choices. For about 20 percent of the adolescent focus groups, a. You say, I say: The weight of adult voices male and female, their first job was the first available on job decisions job they could find. Education does not always guaran- “ tee more choices for economic engagement or meet all [After leaving school] I started working as an � expectations—for young men or young women (Jeffrey et al. 2005). One young man in Khartoum, Sudan, noted, apprentice in a store with the help of my grandmother. — Urban young woman, Assoli Préfecture, Togo On Norms and Agency “It is difficult to find a job, so we grab any that are avail- able. You are lucky if you find a job in your profession. I studied computer science, but am working in a bank. It is A young woman from Floresti District, Moldova, ex- good job, but I want to be a computer programmer; I just plained that “it is hard to say who pushes you to start didn’t find the chance.� “I think that even after completing making a living. When the economic situation of your their studies, some girls still have to stand behind a shop family is poor, you have to make difficult decisions and, counter and some guys have to work at construction or in this case, you have to ask for advice from your fami- something,� said a young girl from Justynowo, Poland. In ly.� Her comment refers to three elements that impinge 82 some cases, their education qualifications become irrel- on the decision to start working and run throughout the narratives we heard in different communities: pov- ple in the sample, especially when training and local mar- erty or lack of resources, a sense of responsibility to- ket opportunities combine. They take ownership of their ward family, and the requirement to consult the family. decision and sense they can succeed. As a young woman Relatives and parents not only enforce behaviors— from Hung Yen District, Vietnam, said, “I am highly in- what a young man or woman can and cannot do—but dependent and make all my decisions by myself. Even if also (particularly in contexts of limited resources and my parents give their opinions, I will still choose the job information) act as facilitators and guides in the quest I like.� Changes in her community, including a new road to find a job. connecting to the large city nearby, are catalysts of such behavior. Even the norms for parent-child relations give Parents and relatives are heavily involved in the search way in face of the autonomy of the decision to work by for the first job. On one hand, as the authorities in the young men. As one young man in urban Santiago de los family, parents must sometimes grant permission for Caballeros, Dominican Republic, explained, “It is difficult their children to work (when they are not actively push- in Santiago de los Caballeros for a father to tell his son ing them into the job market). On the other hand, young what job he should get.� people can tap into adults’ experience and knowledge of the labor markets, as well as their connections or in- On the other side, norms that restrict women’s roles to formation about job opportunities. For young women the household can overwhelm their ability to make de- particularly, “it is important to get approval [from the cisions to work in the public sphere and may increase adults and heads of their households]. Without parents’ parental involvement to the point that there is no in- approval, you cannot work� (young woman, Sungai Puar dividual choice. Young women in rural communities in District, Indonesia). India in the sample stressed over and over their com- plete lack of individual agency and that their only choice Parents fiercely protect young women’s and girls’ safe- was total compliance with parental decisions. Young ty and take measures to compensate for the girls’ lack women “will do whatever work the parents assign� and of experience and information when it comes to deci- if “parents tell [them] to stop going to school and sup- sions such us to work outside the home. For example, port the family,� they submit without a word. In Kou- in Hung Yen District, Vietnam, young women think that dipally Mandal(Andhra Pradesh), girls recognized that “it is better to make a decision with an adult. Because “mostly parents decide what we should do�; in Velugodu you are inexperienced, you need your parents’ direction (Andhra Pradesh), the entire focus group recounted and support. After finishing school, I did not know much how adults determined their first jobs; and in Kalahandi about the outside world, so my parents guided me.� And District (Andhra Pradesh), a girl’s parents “refused to let even young men, who rely less on adults (see figure 3.3 me study further and insisted I look after my younger above), admitted that they seek advice from adults in siblings. They also made me go to the jungle and collect their families. “I would ask my parents what they think produce, and hired me out for wage work.� Nowhere did about a particular offer or if they know the company. young men in the sample relate such stories. Even when However, I would mostly ask my father, not my mother,� pushed to take certain jobs to comply with masculine said a young man from Dobrowice, Poland. But gender expectations or to augment the family’s income, they bias, as always, is never far away. Due to men’s privileged believed they contributed to the decision and could engagement with the labor market, as household pro- give their opinion. On Norms and Agency vider, they not only have better knowledge about the labor market but also better contacts and networks to break entrance barriers. Mothers and other women may 4. “First comes love, then comes have similar experiences and information, but they are marriage, then comes baby in a baby not seen as the best option, or sought out first, for coun- carriage� sel about getting a job. From early in life, we face constant reminders of the Sufficient labor opportunities and exposure to them of- relations expected between women and men. Even ten drive individual expressions of agency by young peo- playground games and songs—such as the title of this 83 section,64 which is a song used by children to taunt boys Marriage may free young women from their father’s con- and girls seen as getting too close or too romantic with trol, but it often is simply a transition into different situa- each other—charge our lives with gender signification. tions of disadvantage with another male (their husband) Thorne (1993) refers to this process as gender play and and of decreased agency as a junior female among the children, as well as adults, recreate gender in everything women in the husband’s extended family (Kabeer 2001). they do, such as creating a couple when they see a boy Arranged marriages are still customary in some of the and girl together. For most children, when they grow up, study communities, as are financial payments, such as a this becomes a reality. Starting a new household and bride price (lobola in southern Africa) or a dowry. In oth- having children are the most visible and significant life er sample communities, women have to yield to strong decisions, and are both the norm and the aspiration of pressure from husbands and in-laws over the number most girls and boys. of children they bear. Retrospective accounts of young women who regretted getting married and having chil- This section discusses how norms and prevailing prac- dren too early make evident the importance they attach tices around marriage and childbearing are bending, to starting a family at the right moment. The accumula- although they are infrequently challenged. As women’s tion of experiences, awareness of women’s rights, gains empowerment grows, they gain more control over their in bargaining power, and accumulation of resource are bodies and fertility choices, such as contraceptive use, all curtailed by early family formation, as are women’s family size, spacing between births, and the sex com- achievement of their individual aspirations and expecta- position of their children (Jejeebhoy 1995; Malhotra tions for the future. 2002). In turn, these new reproductive behaviors and changes in family formation influence women’s major There are two main decisions that lead to family forma- life decisions. tion: childbearing and marriage (both arranged and con- sensual unions). What signs tell women and men when The position of women in the household is central to is the right time wed and start a family, whether to have their ability, or lack of it, to exercise their agency, and children (or not) and how many, and whether they have this position varies with age, the bearing of children, eco- enough autonomy to make these decisions? Agency de- nomic participation, and more. Marriage and reproduc- velops throughout life on a continuum of small gains in tion have a different effect on men’s lives and agency. For empowerment: gains early in life emerge later as an im- many men, family formation moves them from a subordi- proved capacity to decide. Factors, such as marriage cir- nated position—as sons under the authority of an older cumstances and family formation, may at times be at odds male—to the position of power in their own households. with gains in agency, especially from education or a job, But with that power come responsibilities, such as the and have to be negotiated with existing social norms. economic support of the new family and the pressures to comply with associated norms. For example, one of the most contentious areas of au- tonomous decision described by the focus groups has One of the messages that emerged from the discussions to do with reproduction. The arguably central position with young adult women and men in the study is a de- of sexuality in shaping gender relations and reproduc- sire to delay starting a family until they have greater con- ing inequality between women and men constrain sexu- trol over their lives. They consider having an education ality and reproduction both socially and politically.65 In On Norms and Agency and a job with a steady income, as well as physical and fact most gender issues, whether in work, family life, psychological maturity, to be preconditions for a secure or divisions of labor, revolve around reproduction and adult family life. These yearnings for control, however, sexuality.66 Reproduction encompasses more than the constantly interact with social norms and expected be- single event of having a child. It triggers a set of other haviors about how and when family formation should be- future choices that push women toward specific paths in gin. Different views on the appropriate age for marriage line with norms and social and cultural ideas, and prac- may have an impact on the accumulation of endowments tices surrounding childcare and motherhood. The cross- (e.g., education) and the capacity to take advantage of cultural character of this study reveals how the beliefs, 84 economic opportunities. norms, and values surrounding women’s reproductive behavior significantly affect all realities, in both rural to Most young women and some men in the sample, rural urban locations, in all countries. These norms are so en- and urban, wished to marry when they were older, even trenched, according to the focus groups, that a woman those young women who were already married at the who does not opt for reproduction is seen as denying time of the interviews (the majority). Some communities, what she really wants. But a man, however, who is not however, did not approve of couples marrying or having particularly involved in childrearing, does so because kids over a certain age and, although the appropriate fatherhood is not based on the same natural drive as a window of time varied, most agreed that the ideal age woman. Among the study communities, it is still assumed was 18–20 years and not before. that motherhood is a core marker of adult femininity and that the normal outcome of marriage is the production Reasons for delaying marriage were similar to those for of children. And more often than not, motherhood exac- delaying reproduction: maturity and social and financial erbates gender inequalities derived from gender roles. stability. One young woman from Jahran District, Yemen, was married at 15 and explained that “getting married at a a. So he proposed... young age was a disadvantage and hard on me because a girl cannot handle all the household responsibilities and Data dating back to the 1950s, from a set of non-OECD and have children at that age.� Marrying too early also inter- OECD countries, show that women’s age when they first feres with education: “I got married at 16 years old, but marry has increased, although they are still marrying at a it was supposed to be when I was to be 24 years old and younger age than men, and that men’s average age at mar- had completed my education,� lamented a young woman riage has remained stable.67 The young adults in our study from River Nile State, Sudan. Marrying later leads to a bet- mirror these global averages along with rural and urban ter outcome for both partners: “Both of us are employed, differences with marriage. Almost half the rural women’s so that we started life together on a solid basis� (young groups said that women in their communities marry by woman, Sumadija District, Serbia). Indeed, research age 17. In urban settings, only 30 percent of young women shows that women’s education is a stronger determinant marry this young (table 3.1); most wed between the ages than men’s for higher age at marriage and first child.68 of 18 and 25, around the age of majority in many coun- tries; however there is a difference between the rural and Rural young women and girls in our sample were five urban young women is more than 10 percentage points. times more likely than rural young men and boys to drop On the other hand, rural young men, like rural young out of school to marry or have children. For urban young women, marry earlier than their urban counterparts. women and girls, pregnancy or marriage had no differ- Table 3.1: Age of marriage for women and men in focus group communities 15 years or less 16-17 years 18-25 years 26+ years Men Rural 15% 23% 54% 8% Urban 14% 17% 57% 12% Women Rural 26% 22% 44% 7% Urban 13% 18% 48% 10% On Norms and Agency Note: Average age of marriage in each community as reported by 194 young adults focus groups. 64 World Development Report 2012 (World Bank 2012, 217–219) has a section with this same title, which deals with the differences in time use be- tween men and women and their impact on women’s labor market outcomes. It notes that gender differences in time use patterns stem from the gender division of roles and responsibilities inside a household that exist and develop individuals decide to start a family. 65 Brickell (2006) and Ginsburg and Rapp (1991) provide a good overview and summary of these issues in sociology and anthropology. 66 In feminist work, and in sociology more generally, sexuality usually refers to both individual’s practices and identities, as well as to the ideologies, discourses, and social arrangements around it (Walby 1990; Holland et al. 1998). 67 Diez Minguela (2010), based on United Nations’ data on world marriage. 68 For example, see research by Breierova and Duflo (2004) in Indonesia, Kirdar et al. (2009) in Turkey, Osili and Long (2008) in Nigeria, and Perelli-Harris (2005) in Ukraine. 85 ence on school dropout rates, although girls were more pressures to marry in case of pregnancy, while they have likely than boys to interrupt their education to marry. not disappeared entirely, are fading away, even in rural Like other studies, couples where the women had more or more traditional communities like hers. On the other education tended to have fewer children (Iyigun and hand, arranged marriages are still customary in West Bank Walsh 2005), both because the women may have more and Gaza, Yemen, and India. In these countries, according agency within the household and because they have to our study, the bride and groom may not necessarily op- better opportunities and are more likely to participate pose the union, but have no say in selecting their partner. in the labor market. And educated mothers were more This is particularly true for young women and girls, whose likely to invest in education and better nutrition for their roles are mostly passive. Men may indicate their wish children’s well-being (Thomas et al 1990). to wed a particular girl without convening the norm, al- though it may or may not be approved by their families. In The reasons young women in the study married younger India and West Bank and Gaza, traditional arranged mar- than they wanted highlight the imposition of strong so- riages are more common than freely chosen ones, even cial and cultural norms, namely, marriages arranged by though some love marriages do occur (between 3 and 20 their families or forced unions for financial reasons and percent, according to focus groups participants). pregnancy. For example, in Firestone District, Liberia, young women and girls explained that “pregnancy can Frustration with current marriage norms is most evident force people here to live together in a Congo [makeshift in sample countries where traditional rituals involve structures attached to the main house to accommodate economic costs for the bride’s or groom’s family, such the new family]. As soon as a boy impregnates a girl, he as bride price (e.g., South Africa, Tanzania, Papua New can bring the girl to his family’s home.� Similarly in Peru, Guinea, and Afghanistan) or dowry payments (e.g., Bhu- co-habitation was reported as the obligatory step af- tan, Sudan, Liberia, India, and Serbia). While the young ter an unintended pregnancy. In Samtse, Bhutan, young people did not always disagree with these cultural prac- women and girls felt strongly that arranged marriages tices, they spoke of the difficulty in complying with them, should cease and that men and women should have the when facing economic hardship, and the power differen- freedom to choose their partners. tials that they can create in the bride’s future home.69 For example, girls who do not have an adequate dowry have The younger adult women’s groups expressed discon- trouble getting married or are mistreated and abused by tent with current marriage practices, especially young the groom’s family. “If girl brings a large dowry, she will urban women, and tended to question them, although be treated well�; “girls who don’t have a proper dowry their opinions were divided. Young men’s views were are not treated well�; and “they will be treated well only also split, but urban young men were slightly less in- for a few years,� said young women from India. Another clined to reject current marriage practices, probably be- young woman in Jaipur (Odisha), India, related how her cause they have more freedom to choose. Young women sister had been deserted by her husband and in-laws be- in towns and cities seem more dissatisfied (or more able cause of what was considered insufficient dowry: “She to express dissatisfaction) with marriage practices than was physically tortured for a period of time to get INR70 all other groups. In some cases, the expenses associated 50,000 more from our parents, which they did not have with a formal wedding were a concern; in others, tradi- the resources to give.� Young Indian women in Bhubane- tional practices, such as dowries and bride prices, were swar (Odisha) strongly felt the dowry system should be On Norms and Agency questioned as being unnecessary and costly. stopped entirely: “The bride’s family should decide how much they want to give to a daughter on her wedding.� In Most of the women and men in the study felt that they cases where the bride’s family receives the assets, girls in freely decided their marriages, and reported seeing Malangachilima and Zabibu village, Tanzania, complained greater autonomy in some communities in selecting their that they were being married off for economic gain: partners, compared with their parents’ generation. For example, young people in Poland generally felt free to be Some parents force their children to get married in with whom they wanted and to decide whether to live order to get income from the bride price. Girls are 86 together or get married. One young woman noted that married off at an early age; but they are young and behave like children, so they never last in marriages. to be able to freely choose who to marry, but they did not Most times, these girls are poor because they have had believe that women should have the same right. no jobs, except farming, and their husbands mistreat them because they are dependent on them. Yes, there Most focus groups indicated that formalizing a union via needs to be change. Our parents should not see us wedding or civil ceremony is customary, but they called as an income-generating asset. We should be able less forcefully for expanding that practice. Interestingly, to choose the right time to marry and which men we in Peru where informal unions are more frequent, some want to marry. (Rural young woman, Zabibu village, young women wanted legal unions because they felt Tanzania) that a marriage contract brought them more benefits and rights, such as financial support for childrearing and Bride prices paid to a household represent a valuation social status and respectability. They also believed that of the woman’s productive and reproductive capacities. they gain voz y voto (voice and say) in their household. Dowry, on the other hand, speaks to the groom’s capacity For them, a formal marriage license, by securing their to earn an income, as well as a valuation of his status in status as wives, is a means of getting more equal footing the social hierarchy.71 Generally, young women’s decision- with their partners. making power in marriage choice or timing is completely nonexistent when financial gains are at stake. Young But norms and practices are hard to change. The desires women in the study also rejected the consequences of of young women to change certain practices restricting payments in terms of male “ownership� of the wife. For their freedom to choose do not go unchallenged. Young example, for women in National Capital District, Papua men in Afghanistan, Yemen, Tanzania, Sudan, Vietnam, New Guinea, a bride price renders them more vulnerable South Africa, and Fiji, and some adult women in South to domestic abuse since “customary marriages mean once Africa, Yemen and Vietnam, pushed back strongly. They the husband buys the lady, he can do whatever he wants argued for the protection of traditional marriage cus- to do with her. Her parents and brothers are not able to toms to preserve the norm—the way things have been do anything to help her. A wife cannot go back to her fam- done in their culture as passed down by the ancestors— ily if problems arise in her marriage.� Young women have for the future. little leeway to choose their partner and have little voice once they enter into a domestic partnership. b. How can a child take care of another child? While, in theory, bride price can be interpreted as ex- The average age of a girl or young woman when they plicit recognition and valuation of women’s potential con- bear their first child varied among the communities stud- tribution to marriage, in practice, it often limits women’s ied. As seen in table 3.2, rural focus groups reported that control over their own lives. Similarly, in theory, dowry women start having children much earlier than their ur- may endow daughters with property (or an inheritance) ban peers, and earlier than the average age that young early in life to protect them (or give them some agency), men become fathers for the first time. Nearly 50 percent but in practice, it transfers “their� property rights to the of rural groups said most girls were mothers by age 17, husband. It is worth noting that it is not just women who compared with 30 percent for urban women, who started want to change traditional marital practices. A young man having children at the same age they married—between in Koudipally Mandal(Andhra Pradesh), India, explained 18 and 25 years—much like urban men and a significant On Norms and Agency that, although a dowry is traditional and common practice, share of rural men. although he took a dowry for his wife, and although he (like other men) has control over the assets taken under Whether urban or rural, male or female, the majority of dowry, he still felt that “more love marriages should take focus groups agreed that the current average age that place.� Young men in Sudan (Red Sea State) also wanted women had their first child or pregnancy was not appro- 69 Anderson (2007) notes that asset transfers per marriage can be as high as six times the annual household income in South Asia (Rao 1993), and four times in sub-Saharan Africa (Dekker and Hoogeveen 2002). 70 Indian rupees. 71 Rao (1993) shows how dowries in India increase in amount for higher-ranked castes and when the groom is more educated. 87 Table 3.2: Age of men and women at birth of first child 15 years or less 16-17 years 18-25 years 26 years + Men Rural 15% 23% 54% 8% Urban 14% 17% 57% 12% Women Rural 26% 22% 44% 7% Urban 13% 18% 58% 10% Note: Average age of marriage in each community as reported by 194 young adults focus groups. priate and should change, and that women were having Both boys and girls in their focus groups mentioned children too early in life. But like the decision to marry, pregnancy as one of the reasons girls and boys left urban men were the only group that split between ques- school early. A young woman in Firestone District, Li- tioning current practices and trends, and not changing beria, warned that, at early ages, girls are “not ready for them; but again, they have more flexibility about when to children. When you have a child, you will not be able to wed and more options in choice of potential partners, as go far in school. You will suffer because the boys will well as a larger range of opportunities to work, study, and disown the pregnancy and will not support you when access contraception. you get pregnant.� Even young men agreed that wom- en should prioritize education over childbearing. For a The gap between the current and desired age to be- young South African man, a woman should have children come a parent, described by the focus groups, is signif- “at the age of 23 because she has completed her tertiary icant. Most groups preferred 20 years of age or more, education, maybe has a decent job, and is able to sup- and sometimes even older than 30. Their main reasons port her children.� A young woman from Cusco Prov- for delaying childbearing were similar: mother’s physi- ince, Peru, who had her daughter at 18, wished that if cal health, parents’ maturity, and parents’ social situa- she “could do everything all over again, [she] would have tion (marital status, financial situation, and education had her daughter at 25 so [she] could continue study- level). A young man from Monrovia, Liberia, summed it ing.� In Velugodu (Andhra Pradesh), India, a young wom- up, saying “parenthood is not for young people.� an who was married when she was 10 and had her first child at 15, asserted, “This is not the right age. ... A girl Indeed, the discussions about delaying childbirth of- should have a child when she is able to understand what ten raised the issue of the health (and even the life) is right and what is wrong.� Overall, the focus groups felt of the mother when she is too young, but more as a that having children too early means that the parents, general concern than a challenge to the norm that especially the mothers, must give up their dreams for marked young age as appropriate for starting repro- the future. duction. Still, many communities consider teen preg- nancies problematic. According to a young woman c. How many children is too many? from Tchien District, Liberia, “[At 14] the girls’ bodies are too small. They suffer too much... . They get sick. Once reproduction has started, couples should be able On Norms and Agency Some of them have to go to the hospital because their to negotiate the number of children they want to have. As bodies are so small.� Another Liberian young woman shown in figure 3.4, the discussions in most urban focus from Zorzor worried that “sometimes you can die when groups, and by rural men, indicated that couples jointly you have a baby this early [16 years old].� One young decide on the number of children to have. Only 20 per- Yemeni woman’s 14-year-old sister died due to early cent, however, of the young adult women in both rural and pregnancy, as did the school friend of a young woman urban areas said that the decision on how many children from Dominican Republic. Even a man from Emputa vil- to have was in their hands. Rural women (40 percent), lage, Tanzania, noted, “In our community, women have though, described it differently: what men consider a 88 children at the age of 12.� joint decision in rural communities is basically the man’s decision. The husband decides the number and spacing of children because it is inherent in his role as household FIGURE 3.4: WHO DECIDES ON NUMBER head; the wife agrees with his authority and accepts the OF CHILDREN? outcome of the decision.72 Some young women justified a 60% man’s right to decide because he pays the bride price or 50% Share of total mentions has the power to impose his will through violence. Other 40% women indicated that, when faced with disagreement about having children, men compelled their decisions ei- 30% ther through forced sex or by threats to leave the wife or 20% take an additional wife. 10% In some cases, the decision to keep having more children 0% is also imposed by men as a means of keeping women Young Young men Young Young men women women under control: “The decision of how many children to Rural Urban have comes from the man. The man tells the woman the Women’s decision Joint decision Men’s decision Others’ decision number of children he wants. If she says she’s tired, the Note: Data from 194 young adult focus groups. man will beat her,� explained a young woman from Zor- zor, Liberia. And young men are also aware of their deci- sion power: “Mostly the man says he will marry another More groups than we expected described childbearing wife if she doesn’t have [a] baby� (Old City, West Bank as a non-decision. In Peru, the Dominican Republic, and and Gaza). In the Roma community of Kragujevac, Serbia, many African communities, the adults and even younger comments extended to the extreme of valuing women women spok e of childbearing as accidents or as events only for their reproductive capacity. “A woman exists that just occur. In the experience of young women in Chi- only to give birth to a child. What does she have to do clayo, Peru, and Tchien District, Liberia, having children “is with deciding how many children to have?� not something you decide. It just comes and you have to look after [the baby].� In these cases, while social norms In the sample countries, Yemen, Afghanistan, Sudan, may not be binding, their lack of agency combines with India, West Bank and Gaza, Burkina Faso, Fiji, Liberia, lack of information about their own reproductive process. and Vietnam, and among the Roma population in Ser- bia, young women felt that the decision to have children There are indications that the grip of some social norms was forced on them, not just by their partners, but by on family formation has started to relax (Malhotra 1991; family members (his and hers). Sometimes it is an older Jejeebhoy and Halli 2002). Better information and access woman who imposes her will on a young woman. The role to family planning (mainly contraception) enable women of mothers-in-law in determining the number of children to claim agency and greater control over their bodies, young wives have appears unchanged in Fiji, India, and even though progress may still be limited by the opinion of West Bank and Gaza, especially the pressure for male others—family members, community, religious groups, and children. Furthermore, in many of these communities, others. Women’s lifetime earnings and education are neg- women attributed their future reproduction to forces atively associated with the number of children they have, beyond their control, such as God’s or Allah’s will, fate, particularly those who begin childbearing early, so acting On Norms and Agency and more. While the exception, a few focus groups re- to take control of their situation is an important step.74 jected any public discussion about reproduction as inap- propriate, showing how little control over reproduction In the rural areas that make up almost half of the commu- is talked about.73 nities in the research, we expected to find restrictions on 72 These opinions came from Naw Da, Afghanistan; Labasa and Sigatoka, Fiji; Koudipally Mandal and Velugodu, India; Sungai Puar District, Indone- sia; Briceni District, Moldova; and Milne Bay Province, Papua New Guinea. 73 In the Boyina Bagh, Afghanistan, focus group, the young women claimed “not knowing about this question because it is a private issue between husband and wife.� Participants from the Red Sea area of Sudan reacted to the question with surprise and anger, and refused to continue the con- versation until the subject was changed: “No almighty but Allah, you disbelievers! This comes from God, and we have nothing to say about that.� 74 A review of this evidence can be found in Buvinic, Das Gupta, and Casabonne (2009). 89 Table 3.3: Number of children of mother compared with desired fertility for self 3 or more children 2–3 children 1 child Location 81% 19% – Rural Mother’s average number of children 76% 24% – Urban 44% 47% 9% Rural Women’s expectations for number of children 39% 49% 12% Urban 61% 31% 8% Rural Men’s expectations for number of children 49% 41% 9% Urban Note: Data from 194 young adults focus groups. contraception availability and use. Both rural and urban tions. Urban women—probably due to their exposure to women in the study, however, can get contraception and more relaxed norms, more certain supply of contracep- use it frequently. Control over one’s own reproduction tion, and better economic opportunities—were more and fertility, and use of fertility-control mechanisms are likely than rural women to have less than three children. clear signs of agency, control and ownership, and self- (These same factors hold for younger women as well.) determination by women. Rural women, however, wanted more children than ur- ban women. The picture changes when we look at young We see a crucial change when we compare the genera- men, who aimed for a larger family size than women, par- tions of adult and young adult women in the sample in ticularly in rural contexts. their access, knowledge, and acceptability of family plan- ning services. Young adult women have far more control Like marriage, childbearing changes the status of women over their reproduction, due to the life-altering changes in some localities, which influences their ability to control made possible by the availability of family planning ser- when to have children (and the number). When the same vices. For example, in Samtse, Bhutan, the young wom- women in Tchien District, Liberia—who asserted that chil- en pointed out that “before there was none ... but now dren just happen—get married, they face heavy pressures around 90 percent of women use injections to limit the to have a large number of children, which has become number of children in the family.� In Comendador, Do- a competition among families in the community. It is no minican Republic, a young woman felt that “one has con- surprise that the man’s status is at play: “The men decide. trol because there are so many methods,� and in Tewor They tell me that they want 7, 10, many children. They make District, Liberia, another young woman noted that “now sure that their women bear that number,� related a young women can decide. Now there is family planning. Before, Liberian. Having many children reinforces the husband’s the man used to tell the woman how many children to conformity with the prevalent norms of masculinity. “You have.� Either because of contraception’s positive effect cannot tell men to use birth control; they want children. on women’s empowerment or because its use is driven The more they have, the more manly they appear to be� by the soaring costs of raising children (as mentioned by (young woman, Zabibu village, Tanzania). On Norms and Agency the Vietnam and Papua New Guinea participants), wom- en are controlling their fertility more as they desire. In line with perceived ideal masculine behaviors, young men did not oppose the use of contraceptive methods This increased control and attitude change is striking by women, but they generally refused to use them, even when we compare the average number of children their where HIV/AIDS is a known risk. (Tanzania is one excep- mothers’ had with the number of children the younger tion.) Young men from Umlazi Township A and B, South generation of women desire to have (table 3.3). An over- Africa, rejected condoms for their impact on their sexual whelming majority of the older mothers in the study had enjoyment, preferring “skin to skin� relations because 90 a minimum of three children or more, with some varia- “you cannot eat a sweet with the wrapping.� In Santiago de los Caballeros, Dominican Republic, young men wear as family property and business; having boys guarantees condoms only when having sex with women other than that assets or patrimony remain in the family: their wives to avoid bearing bastard children. In the spe- cific case of vasectomies (available in public reproduc- In our community, life is difficult for the man [and his tive health services in India and Bhutan), men and wom- immediate family] who does not have a son. Most of en feared that it affects men’s physical strength, capacity his relatives expect to get his property when he dies to work, and sexual drive. Because it might alter their and his property is distributed by his relatives and not manhood, men preferred that women undergo steriliza- [given to his] wife or daughter. If he has a son, all his tion, if that is the couple’s decision. property will belong to his son. (Rural young woman, Naw Da, Afghanistan) d. Pink or blue? Girl or boy? Males carry the family or clan name and guarantee the The preference for sons is associated with cultural cus- continuity of the lineage; the family’s status increases tom and norms, local community characteristics, the if the son does well: “I prefer a boy. I will be social- ability of women to have a say about their reproductive ly accepted if he succeeds� (young woman, River Nile preferences, and household characteristics (Astone and State, Sudan). Boys are expected to help support the Pande 2007). Social norms in this context probably play household with their earnings and provide for elderly the largest role, in terms of the value of women in soci- parents: “We have to depend on our sons to take care ety, the association of sons with transmission of bloodline of us. Daughters have to care for their husband’s family� (Das Gupta 2009), inheritance practices (Carranza 2012), (young man, Hung Yen District, Vietnam). Many partici- and other traditions. In some countries, the quest for a pants considered sons easier to care for because they son is reflected in a larger number of children when the represent less risk to the family’s honor: “It is more diffi- first born is not a male, while in others—most noticeably cult to provide a good upbringing for female children. To China and India—it has led to skewed preferences and go out to a café and drink alcohol is all right for men, but unbalanced sex ratios.75 When asked about their prefer- not for women. If a girl sleeps with 115 guys, everyone ence for the sex of their children, the participants in our labels her in one way; if a man does the same, everyone study indicated that the traditional preference for sons sees it differently [as not so bad]� (young woman, Su- wanes in cities, especially among young women. Along madija District, Serbia). Boys who have more freedom with conventional arguments on the benefits of having boys emerged narratives about the advantages of raising girls. Figure 3.5 shows that all the men clearly preferred FIGURE 3.5: SEX PREFERENCE FOR CHILDREN male children, particularly rural men where 46 percent of preferences are for a son. Urban men and women 100% show similar levels of indifference regarding their chil- 90% 23% Share of total mentions 80% 36% 39% dren’s sex, but have opposite views when it comes to 70% 46% preferring a boy or a girl. More women overall, though, 60% 39% actually prefer girl children than having no preference 50% 24% 34% 19% for the sex of a child, although rural women are equally 40% 30% split between preferring male and female children. 20% On Norms and Agency 30% 35% 38% 37% 10% A deeply engrained set of normative views about gen- 0% der distribution of responsibilities and perceptions by Rural young Rural young Urban young Urban young women men women men communities permeated the reasons for preferring sons, echoed by the participants. It is based on the conse- Prefer male Prefer female No preference quences of inheritance laws and customs that determine Note: Data from young adults focus groups (258 mentions). what the male members of the family are entitled to, such 75 A detailed analysis of the phenomenon of “missing girls� can be found in chapter 3 of World Development Report 2012 (World Bank 2012, 120–27). 91 carry positive externalities for the father: “A father Control over resources—measured by women’s ability can have fun with a son or drink with him� (young man, to earn income, control their income, and own assets— Justynowo, Poland). And in conflict areas, such as Af- is instrumental to women’s agency and may increase ghanistan and West Bank and Gaza, sons help protect their bargaining power, their position, and their ability the household better than daughters. to make decisions or gain voice in a household, as noted clearly in World Development 2012. It gives them power The reasons for preferring girls also reflect the nor- over their lives by allowing them to leave an unequal mative and institutional systems underlying gender in- family situation and raises their value in society and to equalities. When daughters were preferred, the focus their families.76 Lack of assets severely limits women’s groups desired them for their feminine qualities. Daugh- choices by rendering them powerless to negotiate bet- ters take care of the parents, especially in old age: “It’s ter terms for themselves in their households or with a different when you grow old. A daughter will more like- range of formal and informal institutions where control ly take care of the mother than a son (young woman, over income or assets becomes important for them to Justynowo, Poland); “[I want] my first child to be a girl have a voice (Narayan 2002)77. Not having control over so she can help me with house work� (young woman, one’s own assets basically means not having those as- East Sepik Province, Papua New Guinea). Girls have a sets at all, which diminishes one’s capacity to make stra- better understanding of household management and its tegic life choices. difficulties. Girls are easier to bring up because “a girl is more docile and she keeps you company� (young wom- Women are not strangers to asset management. They an, Chiclayo, Peru). Families look at the future economic tend to manage the everyday expenses required to run reward that girls represent when they marry: “I prefer a household and care for dependants—the small choic- daughters because I know that at some stage I will ben- es we mentioned at the beginning of this chapter. Men, efit from her [bride price]� (young man, Umlazi Township however, remain in charge of major expenditures, such B, Durban, South Africa). as expensive animals, real estate or land, and (generally) education fees. The differences in asset or financial con- These views plainly reflect the expectations attached trol are exacerbated in poor households where women to different life paths for male and female children. The have much less control or management of their income preference for girls has increased in our sample com- and assets. The rational explanations for women giving munities, especially in urban areas, which may be due to up decision-making power to men or for men taking such women’s greater access to education and economic op- power from women are a matter of strong social norms portunities, but may also point to a change in the over- that govern specialization of duties in the household. all position of women in society—as experienced and witnessed by the young women interviewed. The fact The narratives from the communities in our research that preference for girls is still very much based on their predominantly put men in charge of the family income, roles as homemakers and caregivers, however, shows including any wages earned by other members of the the persistence of traditional gender roles and power, household, especially women. Women never control and asset distribution in the hands of men. men’s money (or at least not all of it), and shared control (by men and women) of a man’s money is rare, but com- mon for a woman’s money. 5. What is mine is yours: On Norms and Agency Asset control and decision-making Our focus groups discussed the situation of Judith and James, a fictional couple living in their communities, “ When the money is mine, I spend it on my family because we share our life. … Money is for all the family. It is mine, who negotiate Judith’s decision to start a business and the use of its profits.78 Starting when Judith manages to get start-up capital for a small business, the focus group � but my husband and I decide how to spend it. discussions tracked her ability to actually go ahead with — (Rural woman, North Darfur State, Sudan) her idea, how much support James would give her, her 92 chances for success if James opposed her foray into business, and her authority to decide how to use the where the business was owned by James, women and returns of the business. The discussions proved quite men both were more likely to share control over Judith’s telling. Figure 3.6 shows the different opinions in the fo- money and less likely to share control over James’ money. cus groups about control of income earned by Judith. Shared control of Judith’s money or Judith gets to con- When it comes to major expenses, the process is similar: trol her income were the most common answers, across women in the study were almost never solely in charge of all locations and age groups. these decisions. Even when women reported that they had autonomy to decide how to use their income, they More than 40 percent of the groups acknowledged that still followed the directions of a man. More often than Judith should have a say about her money, except rural not, women’s accounts of their independent decision cit- adult men—only 27.6 percent agreed that Judith should ed compromises reached with their husband or partner. control her money (compared with 45.8 percent among On the whole, there is overwhelming evidence, reported rural young adult men). A significant number of partici- by both women and by men in a number of communities pants chose shared control over Judith’s business prof- (showing no specific regional or country pattern), of how its. A small number of groups discussed whether the little autonomy women actually exercise when it comes couple should also make joint decisions about James’ to their own assets and income. money, if he started the business. The predominant view preferred that James control his money over sharing Men and women participants justified male control control with Judith. over all earnings and assets belonging to the family or belonging to the women in the family by referring to The first impression of fairness and women’s control traditional customs (some of which are religious) and in figure 3.6 is deceiving. A little probing by the discus- equally traditional entitlements attached to the head of sion facilitators into how the shared decision-making the household. For example, in three of the communi- process operates revealed that men’s opinions carried ties in Papua New Guinea, men claimed to control as- greater weight in most decisions about Judith’s income, sets on the basis of tradition or custom: “It is the man even when reported as shared control. And among the who dominates, who owns everything.� One of these 64 groups that account discussing the reverse situation, communities has a matriarchal tradition, but even there FIGURE 3.6: WHO CONTROLS JUDITH’S MONEY? Share of total mentions Share of total mentions 34% 42% 35% 36% 35% 53% 46% 45% 22% 16% 22% 18% 13% 15% 6% 26% 44% 46% 49% 41% 42% 40% 47% 28% On Norms and Agency Women Men Women Men Women Men Women Men Rural Young Adults Urban Young Adults Rural Adults Urban Adults Judith Controls her money James controls Judith’s money Both control Judith’s money Note: Data from 388 focus groups. 76 See chapter 4 in World Development Report 2012 (World Bank 2012) for a detailed review of the evidence. 77 A detailed review of the evidence is on Chapter 4 of WDR. 78 The exercise was conducted with young adult and adult men and women. 93 men are perceived as the natural heads of households, noted, “She can decide how to use her money, but she although “women are financial controllers� and “the needs to consult her husband. Otherwise, her rela- bosses.� When it comes to making certain decisions, tionship with her husband will be negatively affected.� the women’s tradition-bestowed power does not hold. Decisions about their money are, for women (unlike Women “don’t sell anything because women’s place is men), part of the primary considerations for keeping in the house.� Men from this Papuan matrilineal com- family harmony and balance. “The money can be mine, munity asserted that the husband has a large say in the but the moment I need to do something for the fam- use of money earned by his wife. To them, James as ily, I need to have him also decide on [how to use] my head of household was endowed with the entitlements money. Sometimes we women do this, not because we and responsibilities associated with that role, and they think it is right, but simply because you need the family overrode even the traditionally recognized lead role of relations to keep going well,� said a woman from Zabi- women in their community. bu village, Tanzania. There is little distinction between being a man and hav- Some women in the study referred to traditional norms ing authority. A good husband has the biggest voice and to rationalize their lack of power: “According to our tra- the final say over any decisions for managing women’s dition, the biggest things are not controlled by women. income and assets. The Judith and James scenario Also, women cannot buy and sell anything in our village�; prompted a group of women in from a village of Sumadi- and “according to the culture of our community, women ja District, Serbia, to present these situations: do not do business, so of course women alone cannot do business in our village� (women in Naw Da, Afghani- – There would be bloodshed if she [the wife] spent stan). In a semi-rural community of Ngonyameni, South the money by herself. When I received some money, Africa, women described a tradition mixed with patriar- 250,000 dinars, we put the money in one place chy: “This is a very patriarchal community. We believe because my husband said so. If my husband had that men are the heads of the households and they make received the money, he would have made the decision all the decisions. This is also a traditional community, so himself. He is my husband, he is the boss, he is the man. the man has a final word in any decision regarding the household. It does not matter if Zodwa (Judith) worked – [This is] an example that is applicable to the majority for that money; she cannot decide alone how to use it.� of us here. Some men are cunning and manipulate References to decision-making power and control over their women in a nice, slick way and others pound their assets, as one of the rights of the head of household, fists on the table to get their way, but the end result is were heard repeatedly in communities in Africa, India, always the same. and the Middle East. But this automatic authority may also be detrimental In a rural community of Ngonyameni, South Africa, a mix to men, who are expected to provide all necessities, of cultural tradition, men’s reputations and respect in buy big things for the house, and always pay because the eyes of the community, and a sense of marital duty it makes them the man of the house. This expectation combined to justify male control over assets and income. remains in effect, even if lost employment, poor local Its tradition supporting unequal control of assets was economic situations, or lack of skills prevent men from first based in magical-religious elements. The villagers On Norms and Agency fulfilling this role. defended the impossibility of women sharing ownership of the house they live in with their husbands with their Sometimes it is less about explicit dominance, despite belief that “his ancestors have lived in that house, so the men always having a say. A careful look at what the land belongs to them.� A woman cannot own a house be- women in all 20 countries said reveals that many men cause “she left her ancestors when she got married, so it genuinely welcome cooperation and shared opinions, is her husband’s ancestors [in his house] that watch over but only if they do not affect household balances. In the family. If she claims the house is hers, it might bring Hung Yen District, Vietnam, when asked if Judith can her misfortune.� Only after the death of the husband is 94 decide alone how to spend her money, one woman a woman allowed to take control of the property. In Ngonyameni everything belongs to men. Women take men’s authority. Women are often accused of bothering their husband’s surname as a sign of becoming his prop- their husbands if they ask him about minor needs. erty and their assets also get a new owner, whether they are brought into the marriage or generated after. “You At this level of decision-making, norms are slightly more may buy livestock on your own, but the purchase is in flexible and women find ways to exercise some degree your husband’s name, and that gives him respect in the of autonomy. More than a few women in the study were community.� Women are not only barred from owner- able to set aside money occasionally and did not tell ship but “some women have to tell their husband about their husbands or refused to surrender control over it. every penny they earn.� Indeed, women were recognized as wiser money manag- ers than men by a men’s focus group from the Domini- By and large, women are economically dependent for ev- can Republic in several different discussions, echoed erything, including necessities, even if they bring assets by a (Tchien District) Liberian man who admitted that to the marriage. Furthermore, in some sample communi- “some men are good managers, but women are better. A ties, the institution of marriage also meant that women man can decide to do whatever he wants with his money are transferred to the husband’s family as property. A if he is a rich man; if he is less well off, he should consult man in Liberia challenged whether a woman can own his wife on how to spend his money because she is more property since she is part of the household assets: likely to prioritize household expenditures.� These men saw themselves at high risk of spending the money on She can get property, but she can’t own property. For themselves and falling into the bad husband category. example, if her brother gives her a cow, it is hers. But before she sells it to someone else, she must consult the Men’s tendency to squander their income outside the man (brother or husband) and the man must agree. Even home was a recurring problem and may underlie the the woman herself is your property. The only reason she perception by the focus groups that women are better is respected is because she is a human being. managers. Both men and women in widely diverse coun- tries in the study recognized that men spend too much Finally, community pressure to conform to expected of their earnings on alcohol, women, and gambling, while behaviors can be the most effective method to keep a women used their (earned) money to sustain the house- woman (or put her back) in her rightful place. In a com- hold. Evidence of differences in expenditures has been munity in Vietnam, if Hoa (Judith) makes her own deci- widely documented, proving that women favor invest- sions about spending her money, the local women’s as- ing in children. This preference partly relates to social sociation will visit and explain to her that women are norms, which imbues women with greater knowledge supposed to ask their husband’s opinion before deciding of children’s needs due to women’s traditional gender on the spending, even though it is her money. They will roles. Women’s control of income and assets is also remind her of the Vietnamese saying: “Couples living in important for their children’s well-being.79 The motives harmony have strength to displace the Eastern Sea.� behind men’s spending on alcohol have been less re- searched, but they appear to be related to demands to a. Pocket money assert their masculinity in the eyes of the community, as much as an escape from living up to the hegemonic Not everything acts as a constraint, although it may seem masculinity model or gender role prescriptions.80 On Norms and Agency like it. Even under restricted conditions described by the focus groups, women are usually allowed to handle ev- The focus groups again brought up men’s private con- eryday expenses independently, as long as they purchase sumption of income in relation to control over household inexpensive household items for household consump- finances. “Most men will spend money on mistresses and tion—nothing that will increase the women’s bargaining gambling, expenditures that will not be discussed with the power or their voice in the household and supersede wife� (adult man, Paro, Bhutan). “A man does not provide 79 In studies of countries as varied as Brazil, Côte d’Ivoire, Ghana, Mexico, South Africa, and Turkey, women’s spending on goods that benefit chil- dren has prompted the introduction of social policies, such as conditional cash transfers. See Schady and Rosero (2007) and Doss (2006). 80 See Lemle (1989), Holland et al. (1998), and Barker (2005). 95 any money to his wife when he wastes money on drinks. Once women accumulate wealth—or their family does— He spends a lot of money on alcohol and then starts ask- they also seem both to accumulate decision-making ing the wife to get money from her father� (adult woman, power and gain access to larger assets. Women in less Velugodu (Andhra Pradesh), India). Even the fictional dire economic circumstances, and certainly women with James was matter-of-factly described as spending the more independent economic means, have more control profits of his (or Judith’s) entrepreneurship on partying over their assets and earnings than poorer women, even by a man from Comendador, Dominican Republic. if women’s autonomy is tightly restricted in other areas of their lives. Poorer women more frequently require On rare occasions, women actually manage the entire permission from their husbands to use their own money, family income with their partner’s consent, much like as attested to by women in Boyina Bagh, Afghanistan. In this woman in Serbia: “My husband receives his salary Sigatoka, Fiji, as family finances ease, women often gain and gives it to me. He does not carry money with him, more control over income from agricultural products and I pay all the bills. He never asks me how I spend and small livestock and poultry, for example, and even- the money.� Other women reported some awareness tually ownership, control, and decision over land. of their rights that led to some level of control or that associated autonomy with effort. Women in Fiji, India, Women’s right to self-determination is starting to be Sudan, and Dominican Republic felt that, if Judith is the recognized by the women and men in the focus groups, one who makes her business succeed without James’ even when autonomy and control of money they earn help, she has the right to decide alone how to use her remains elusive. “I respect her and her right to start her money. The idea of bestowing the right to decide on own business,� said a man from East Sepik Province, women is a sign of attitudinal change challenging the Papua New Guinea, “but apart from money for clothes, norm of sole male control of financial decisions in the for everything else, her husband will have access to and household. Chapters 4 and 5 present more evidence of control of her money.� A Vietnamese woman from Hung women actually acting upon these views of their own Yen District saw a more positive future: “Women’s roles entitlements. have changed a lot. Women’s social relations are the same as men’s. Both daughters and sons get the inheri- b. Moving up and forward tance. However, the sons are given larger inheritance than daughters, even though it is stipulated by the law Even when women’s right to decide is recognized, wom- and the court that daughters and sons should have the en are still not always able to decide alone or may not same inheritance.� opt for selfish consumption. When women have access to earnings or assets, they tend to include others in their decision-making more often than men. Even the women 6. When does choice mean agency? in the study cannot easily detach themselves from their household roles and responsibilities as mothers and In the strategic decisions covered in this chapter, women home-makers.81 have gained some autonomy to decide about their edu- cation, jobs, marriage (who and when), and reproduction, Justifications for how responsibilities and entitlements although they still are permanently challenged not to ne- in households are distributed, however, appear to be glect their domestic duties. Men in the study are showing On Norms and Agency changing in some of the sample communities. While more willingness to consider sharing power (if not actu- some women and most men accepted traditional cus- ally share it) and to release some control over household toms and norms for asset control (the rights of heads of decisions to women. Shared decision-making means men households and husbands, and rights handed down by have to bend constraining norms, but it introduces a bet- ancestors), other men and women saw a need to alter ter decision-making process into their households. And them and referred to changes that have already taken as these men and women change, they transform the tra- place, such as inheritance laws that include women or ditional playing field in their communities. In the domes- the benefits of women generating and managing their tic sphere, the women are stealthily altering traditional 96 own income. definitions of duties and responsibilities associated with Box 3.1: Choosing not to change things Agency does not always entail making progressive choices. Conservative choices are made in at least two scenarios: 1) where options are limited or where restrictions and conservative choices increase people’s status or improve their general well-being, and 2) where true commitment to the norm is not strategic. An example of a conservative choice that has been interpreted multiple ways is the use of the veil among Muslim women. It has been argued that religious and social ideals and norms are not always (or even usually) rejected by women as oppressive, but are sometimes appropriated and creatively used to increase their agency and achieve positive goals (Mahmood 2001). Adolescent girls and young women in West Bank and Gaza refer to restrictions on their mobility in ways indicating that it constrains rather than enables their agency. But wearing the veil in public is also an apparent strategic choice to gain a degree of autonomy, even though it appears constraining. Although the focus groups did not discuss it extensively, enough comments gave evidence that the veil is accepted both as a matter of course and as a passport to greater mobility and autonomy. Adolescent boys and girls in commu- nities near Rafah (West Bank and Gaza) both agreed that it is easier for “veiled girls to go to school� or move around in public. In a neighborhood of Rafah Governorate, the adolescents also reported that girls work while going to school to help support their families, but it is more difficult if they do not wear the veil. Making a progressive choice (in this case, not to wear a veil) affects their ability to earn income and reduce their agency. The evidence for a strategic use of the veil, however, does not rule out adherence to the norm from true commit- ment to religious or moral values. The statement by an adolescent girl from Rafah that “we have to be veiled, it is our religion� should be taken seriously and not explained away by reference to social conditions. Truly conservative choices that are not a response to restrictions were rare in our study, but this may be the result of the emphasis in the research on the conditions that enable or limit to agency, rather than on values and religious ideals. their expected roles, which may induce change in the tional power roles and identities. However—perceptions norms or make them more flexible. and voices are changing and they mark real movement in women’s power and freedom. Whether women make The evidence from the different communities in our more traditional or modern choices (box 3.1), the goal is study shows that it is young women in urban areas to make their own decisions with fewer constraints. who are more overtly acquiring a greater sense of self- determination and agency in the decisions discussed Kabeer (1999) sees the ability to exercise choice in terms here. Within persistent constraints, they are beginning of three inter-related dimensions: resources, agency, and to envision a future similar to young men: education, in- achievements or outcomes. Resources, the contexts, and dependence, greater financial autonomy, and shared re- conditions that the men and women in the study have sponsibility for their family. These younger generation’s to make choices, vary with each decision, as they have On Norms and Agency mothers, as much as other adults in rural areas, have a clearly described. Women’s acquisition of education, in- less positive view of the changes occurring in their so- come, and jobs, and access to services—resources—act cieties, both given the difficulties they have faced or as preconditions to exercising their agency. The actual because of uncertainty from challenges to their tradi- ability to make strategic decisions, however, does not al- 81 These and other accounts should be complemented by economic research on women’s expenditure preferences, as well as anthropological research. Miller’s (1998) study on shopping shows that housework and homemaking are strong sources of self-identity and solidify family bonds in working- and middle-class English homes. Miller argues that everyday shopping and management carry meaning, among other reasons, because they maintain crucial relationships and allow women to positively mold family members. The assumption is that male and female members of the household adhere to and follow a moral scheme of what is good to instill in others, using provisioning and housekeeping to do it. This is not normally recognized because action in the domestic sphere is considered, by default, as inferior to action in the public sphere. 97 ways follow immediately, but as women’s horizons and ing Paper, no. 5148. Washington, DC: World Bank. options broaden, so does their ability to envision a dif- Buvinic, M., M. Das Gupta, and U. Casabonne. 2009. ferent life and act upon it—and achieve it. Gender, Poverty, and Demography: An Overview. World Bank Economic Review 23 (3): 347–69. But aspirations that lead to action do not take place in Dekker, M. and H. Hoogeveen. 2002. Bridewealth and a void. As Ray (2003) notes clearly, individuals still fall Household Security in Rural Zimbabwe. Journal of back on their local reference frameworks as guidance. African Economics 11 (1): 114–45. They make comparisons with their peers and qualify Diez Minguela, A. 2011. Mating (Marriage) Patterns and their own situations relative to community references, Economic Development. History of the Family 16 (4): which shape their aspirations. From their communities 213–30. flows information and how to interpret it; communities Doss, C.R. 2006. The Effects of Intrahousehold Property identify role models and produce “others like me.� Aspi- Ownership on Expenditure Patterns in Ghana. Jour- rations for economic mobility and power are contextual nal of African Economies 15 (1): 149–80. and depend on how much mobility one perceives in a Field, E. and A. Ambrus. 2008. Early Marriage and Fe- local community and in society. male Schooling in Bangladesh. Journal of Political Economy 116: 881–930. Ginsburg, F., and R. Rapp. 1991. The Politics of Reproduc- References tion. Annual Review of Anthropology 20: 311–43. Goldin, C., and L.F. Katz. 2002. The Power of the Pill: Agarwal, B. 1997. “Bargaining� and Gender Relations: Oral Contraceptives and Women’s Career and Mar- Within and Beyond the Household. Feminist Eco- riage Decisions. Journal of Political Economy 110 (4): nomics 3 (1): 1–51. 730–70. Anderson, S. 2007. The Economics of Dowry and Bride- Holland, J., C. Ramazanoglu, S. Sharpe, and R. Thompson. price. Journal of Economic Perspectives 21 (4): 151–74. 1998. The Male in the Head: Young People, Hetero- Appadurai, A. 2004. The Capacity to Aspire: Culture sexuality, and Power. London: Tufnell Press. and Terms of Recognition. In Culture and Public Ac- Iyigun, M., and R.P. Walsh. 2007. Building the Family Nest: tion, ed. V. Rao and M. Walton, 59–84. Stanford, CA, Premarital Investments, Marriage Markets, and Spou- USA: Stanford University Press. sal Allocations. Review of Economic Studies 74 (2): Astone, N.M., R. Pande. 2007. Explaining Son Preference 507–535. in Rural India: The Independent Role of Structural Jejeebhoy, S. 1995. Women’s Education, Autonomy, and versus Individual Factors. Population Research and Reproductive Behaviour: Experience from Develop- Policy Review 26 (1): 1–29. ing Countries. Oxford: Clarendon Press. Barker, G. 2005. Dying to be Men: Youth, Masculinity, Kabeer, N. 1999. Resources, Agency, Achievements: Re- and Social Exclusion. London: Routledge. flections on the Measurement of Women’s Empower- Breierova, L. and E. Duflo. 2004. The Impact of Educa- ment. Development and Change 30: 435–64. tion on Fertility and Child Mortality: Do Fathers Re- ———. 2000. The Power to Choose: Bangladeshi Women ally Matter Less than Mothers? NBER Working Paper, and Labour Market Decisions in London and Dhaka. no. 10513. Cambridge, MA, USA: National Bureau of London: Verso. Economic Research. Jejeebhoy, S. J. 1995. Addressing Women’s Reproduc- On Norms and Agency Brickell, C. 2006. The Sociological Construction of Gen- tive Health Needs: Priorities for the Family Welfare der and Sexuality. Sociological Review 54: 87–113. Program. Asia and Near East OR/TA Project Special Carranza, E. 2012. Islamic Inheritance Law, Son Prefer- Report. New York: Population Council. ence and Fertility Behavior of Muslim Couples in Jejeebhoy, S.J., S.S. Halli, C.B. Lloyd, J.R. Behrman, N.P. Indonesia. Policy Research Working Paper, no. 5972. Stromquist, and B. Cohen. 2005. Marriage Patterns Washington, DC: World Bank. in Rural India: Influence of Sociocultural Context. In Das Gupta, M. 2009. Family Systems, Political Systems, The Changing Transitions to Adulthood in Developing and Asia’s “Missing Girls�: The Construction of Son Countries: Selected Studies, ed. C.B. Lloyd, J.R. Beh- 98 Preference and Its Unraveling. Policy Research Work- rman, N.P. Stromquist, and B. Cohen. Washington, DC: National Academies Press. Ray, D. 2003. Aspirations, Poverty, and Economic Change. Kabeer, Naila. 1999. Resources, Agency, Achievements: Photocopy. Department of Economics, New York Uni- Reflections on the Measurement of Women’s Empow- versity. erment.’ Development and Change 30 (3): 435–64. Rao, V. 1993. The Rising Price of Husbands: A Hedonic ———. 2001. Reflections on the Measurement of Women’s Analysis of Dowry Increases in Rural India. Journal of Empowerment. In Discussing Women’s Empowerment: Political Economy 101 (4): 666–77. Theory and Practice. Sidastudies No. 3. Stockholm, Schady, N., and J. Rosero. 2008. Are Cash Transfers Made Sweden: Novum Grafiska AB and Sida. to Women Spent Like Other Sources of Income? Eco- ———. 2005. Gender Equality and Women’s Empower- nomics Letters 101 (3): 246–48. ment: A Critical Analysis of the Third Millennium Schuler, S.R., and E. Rottach. 2010. Women’s Empower- Development Goal. Gender and Development 13 (1): ment across Generations in Bangladesh. Journal of 13–24. Development Studies 46 (3): 379–96. Dayioglu, M., M.G. Kirdar, and A. Tansel. 2009. Impact of Sen, A. 1985. Well-Being, Agency, and Freedom: The Dew- Sibship, Size, Birth Order, and Sex Composition on ey Lectures. Journal of Philosophy 82 (4): 169–221. School Enrolment in Urban Turkey. Oxford Bulletin of Sen, G. 1993. Path to Fertility Decline: A Cross-Country Economics and Statistics 71 (3): 399–426. Analysis. In Development and Change: Essays in Hon- Lemle, R., M.E. Mishkind. 1989. Alcohol and Masculinity. or of K.N. Raj, ed. P. Bardhan, M. Dattachandhri, and Journal of Substance Abuse Treatment 6 (4): 213–22. T.N. Krishnan. New Delhi: Oxford University Press. Malhotra, A. 1991. Gender and Changing Generational Thomas, D., J. Strauss, and M.-H. Henriques. 1990. Child Relations: Spouse Choice in Indonesia. Demography Survival, Height for Age, and Household Characteris- 28 (4): 549–70. doi: 10.2307/2061422. tics in Brazil. Journal of Development Economics 33 ———. 1997. Gender and the Timing of Marriage: Rural- (2): 197–234. Urban Differences in Java. Journal of Marriage and Thorne, B. 1993. Gender Play: Girls and Boys in School. Family 59 (2): 434–50. New Brunswick, NJ, USA: Rutgers University Press. Malhotra, A., S.R. Schuler, and C. Boender. 2002. Mea- Walby, S. 1990. Theorizing Patriarchy. Oxford: Black- suring Women’s Empowerment as a Variable in Inter- well. national Development. Background paper prepared World Bank. 2012. World Development Report 2012: for the World Bank Workshop on Poverty and Gen- Gender Equality and Development. Washington, DC: der: New Perspectives, Washington, DC, June 2002. World Bank. Miller, D. 1998). Theory of Shopping. 1st ed. Ithaca, NY, USA: Cornell University. Narayan, D., R. Chambers, M. K. Shah, and P. Petesch. 2000. Voices of the Poor: Crying Out for Change. New York: Oxford University Press for the World Bank. Long, B., and U. Osili. 2008. Does Female Schooling Re- duce Fertility: A Nigerian Experiment. Journal of De- velopment Economics 87 (1): 57–5. Patrinos, H.A., and G. Psacharopoulos. 2004. Returns to On Norms and Agency Investment in Education: A Further Update. Educa- tion Economics 12 (2): 111–34. Perelli-Harris, B. 2006. The Influence of Informal Work and Subjective Well-Being on Childbearing in Post- Soviet Russia. Population and Development Review 32 (4): 729–53. Pezzini, S. 2005. The Effect of Women’s Rights on Wom- en’s Welfare: Evidence from a Natural Experiment. Economic Journal 115 (502): C208–C227. 99 In our 97 sample communities, we invited men and wom- en to reflect on the role of power and freedom in their lives. What does it mean to be a powerful woman or a powerful man in their community? How can a woman or man become more powerful and free? How can they PART III: lose power and freedom? On Norms and Agency Empowerment Our aim is to systematically record the factors that women and men in the study saw as helping increase their feelings of empowerment. As outlined in the in- troduction, agency and empowerment are contested concepts with different definitions and perspectives of their importance to processes of social change. 100 Kabeer (2001, 19) defines empowerment as the “expan- sion in people’s ability to make strategic life choices in “ I am free and I have some power; my partner has the same: sovereign decisions are freedom and power. � a context where this ability was previously denied to — Urban man, neighborhood of Olzstyn, Poland them.� We do not hold the focus groups to this bench- mark. In fact, we set aside the academic terms of agency and empowerment and instead explore their dynamics Overall, in our sample, women more often than men re- with the focus groups by turning to local understandings ported becoming more empowered and spoke of gain- of the more common terms of power and freedom. ing more influence and freedom of action in either their domestic or public roles. Men’s changes in their sense The capacity of women and men (or groups) to pursue of power and freedom, however, were far more tightly goals, their agency, may or may not achieve the desired tied to their role as providers and to the health of the outcomes. Local context also significantly influences local economy. their pursuits. “Clearly a process of empowerment is in- complete,� explain Ibrahim and Alkire (2007, 9), “unless it Our assessments of the focus groups particularly reveal attends to people’s abilities to act, the institutional struc- that women’s sense of empowerment and the factors ture, and the various non-institutional changes that are shaping it can be very different from men’s, even in the instrumental to increased agency.� In our exploration of same community. This discrepancy in large measure has the gender dimensions of exercising agency, we explic- to do with gender norms. Although we observe many itly look at the interplay of three dynamics that may lead new norms slowly taking hold, when reviewing all the re- to a sense of greater power and freedom, or empower- sponses in aggregate, important gender differences still ment: 1) the behaviors or actions that men and women persist in aspirations and in perceptions of, and access associate with exercising agency, 2) the conditions and to, opportunities. These differences were reflected in trends in their local structure of opportunities, and 3) the men’s and women’s evaluations of their capacities to the change in gender norms as part of that opportuni- act and empowerment trends for their own gender. As ties’ structure. Kabeer (1999) and Mahmood (2001) point out in differ- ent ways, empowerment does not always or necessarily Social norms reproduced across institutions feed into a result from deviations from the norms. While it is true gender system that demands that men and women act in that complying with norms may lead to increased status, certain ways, although it also provides a sense of iden- and that this may be antithetical to autonomy, the ac- tity and a position in the community structure. (Whether tual outcome depends on the context of women’s social this is equal or favorable is a different matter.) As such, relations and individual histories. Alkire (2009, 4), in a this system, and how it is changing, determines what is discussion of approaches to assessing agency, explains possible to imagine and to achieve. We also look at the that, “people who enjoy high levels of agency are en- enabling local conditions that make it possible to pursue gaged in actions that are congruent with their values.� A and realize aspirations. strong desire for harmony between values and behaviors is common in local understandings of what brings power We use two tracks to assess factors and processes and freedom, and the values often embody strong gen- gleaned from the focus groups’ understanding of empow- der dimensions. erment. We first parse out women’s and men’s different On Norms and Agency perceptions of the factors that increase their power and In chapter 5, we connect the men’s and women’s assess- freedom, and the role that gender norms play in their ments of trends in empowerment for their own gender interpretations of what creates empowerment (Chapter to data gathered about local markets and state and civic 4). In chapter 5, we explore the local context in more de- institutions. From this, we note that the perceived em- tail and how it influences the processes and outcomes of powerment outcomes, trends in local norms, and access agency, especially those elements over which individuals to opportunity structures do not necessarily move to- may have no control, such as market forces, local gover- gether. Changes in one domain may not be matched by nance, and civic action, and the norms for women’s inclu- changes in others, with gender norms often lagging. In sion and leadership in these public spaces. general, we often observed that communities with more 101 dynamic economies stimulated a greater sense of em- expansions in female empowerment hold no meaning powerment in men and women and greater relaxation when applied to men. Male pathways are narrower and of norms. But even in the most supportive of contexts, dominated by local and national economic conditions. women and men pursued local opportunities in ways that did not conflict with their gender-prescribed roles, The identified pathways do not present a comprehen- responsibilities, and conducts. sive picture, but are a starting point for chapters 4 and 5, which delve deeply into perceptions of agency Given the great diversity of individual perceptions and lo- and freedom, and how inequalities inherent in gender cal conditions, we aim first to identify the main pathways, norms and gender power relations create different sets or combinations of factors, that lead to increased senses of opportunities and challenges for women and men. A of power and freedom in the sample communities.82 In a woman who seeks greater power and freedom of ac- nutshell, we find that a clearer path emerges for urban tion in her life may constantly have to negotiate norms women than rural women. Urban women perceived more that discourage her from taking initiative and modifying extensive gains in their power when they control major expected gender-typed behaviors. For example, norms assets, are free (or freer) from domestic violence,83 ac- of femininity prescribe submissive behaviors; however, quire greater social capital, and have a supportive local women speak of the need for acting boldly and taking opportunity structure. For instance, the agency of women risks in order to pursue aspirations, such as earning an in the study benefited from the presence of active wom- income outside the home. And if success in these initia- en’s organizations that tap into partners and resources tives means a woman faces greater possibilities of do- outside their localities. Urban women also benefitted mestic violence, then it renders her gains in economic from residing in neighborhoods where gender norms are independence and assets useless. By contrast, factors more relaxed, markets are stronger, and public services that propel men upward are clearly in line with accept- more accessible. ed definitions of masculinity and mandates for what it means to be a good man. Men’s access to opportuni- For rural women in the sample communities, identifying ties is less dependent on their own efforts than women’s causal factors associated with empowerment was more because being a good provider and being powerful and difficult. This likely reflects the diverse barriers that vil- free are widely accepted traits of men. Fundamentally, lage women confront in order to attain more autonomy women’s rising empowerment in a community—and their for themselves. Like women in the cities, however, the eventual transformation of expected feminine attributes model suggests that rural women do somewhat better of domesticity, docility, and obedience—is a process that when they have supportive local opportunity structures, ushers in significant change in gender norms, power re- enjoy more mobility, and face less domestic violence. lations, and institutional inclusion. These factors are more present in the sample communi- ties marked by deeper poverty and little human develop- ment.84 A fourth factor affecting rural women’s capacity to increase their agency is scarce labor opportunities for men in the formal economy. Urban and rural men’s gains in power are largely depen- On Norms and Agency dent on economic growth and the existence of and access to jobs. In fact, the explanatory forces behind perceived 82 In order to learn more from our dataset about women’s empowerment, we invited sociologist Charles Ragin to collaborate with us in investigat- ing causal factors. We used qualitative comparative analysis (QCA) to track each community’s changes in empowerment on a “Ladder of Power and Freedom�(explained more fully below) as a configuration of traits or aspects, rather than representing aspects as separate, independent variables, which neutralizes and thus sacrifices context. QCA ensures that it is not just single variables acting independently that drive outcomes but combi- nations of causal conditions (Ragin 2000, 2008). Appendix 1 on the methodology we used has more details on this exercise. 83 These findings are consistent with broader surveys, such as Agarwal and Panda (2007), which show that women’s ownership of major assets can be a protective factor. 102 84 As measured by the Human Development Index of each country (UNDP 2011). CHAPTER 4 What drives agency? What crushes it? “ The moment that you know that you can do things by yourself and not have to depend on a man is the moment you begin moving up. I see some women being beaten by their husbands every � day. When you talk to them, they say they are married and cannot leave their husband. These [women] will never climb out of their situation. They will stay at the bottom. — Urban woman, Emputa village, Tanzania M en and women in the study reported that generally recognized to understanding the capability quite similar factors fuel their empowerment. of local level institutions to serve the public good, and They spoke most often about how their own hence the pace at which communities are able to shift to economic initiatives help them gain more power, fol- more inclusive and prosperous development. lowed by positive attitudes and behaviors, then educa- tion and skills. When we compare discussions by women and men from the same locality about what triggers em- 1. Step by step: Climbing the powerment, however, significant gender differences of- “Ladder of Freedom and Power� ten emerge. Understanding how local gender norms are evolving locally is often critical for making sense of why To guide our analysis of the complex factors and pro- women and men may perceive sharply different causes cesses that underpin how and why individuals are able to of and trends in empowerment. gain power and freedom—become more empowered—we draw on a concept of agency as “the ability to define one’s Indeed, one of the findings from our study is that urban goals and act on them�(Kabeer 1999, 438). If successful in women perceive significantly more gains in their power their pursuits, individuals may both increase their agency, and freedom over the past decade than any other group or capacity to negotiate and make decisions, as well as sampled. This is consistent with the rural and urban dif- their power and freedom to control resources and shape On Norms and Agency ferences, noted in previous chapters, in relaxation and institutions that affect their lives (Narayan 2002). Yet, not change in gender norms. The reports by urban women, all women’s and men’s initiatives for agency are success- moreover, starkly contrast with urban men’s sense of ful. Empowerment is conceived as a product of the in- loss of power and freedom over the same time period teraction between, on one hand, individuals and groups and the challenge that this presents to their compliance seeking to exercise agency and advance their interests, with expected models of masculinity. The findings also and on the other, changes to their local opportunities suggest that gender norms may be more important than structures.85 In our dataset, for instance, we observe how 85 See Petesch, Walton, and Smulovitz (2005) for a macro-level framework that inspires this community-level analysis. 103 “ After each group built their ladder, the discussion shift- According to the tradition of our village, ed to how someone can climb up each step to the next, women cannot move about freely. But old women and to what factors may push someone down the lad- who are on step 2 or step 3 (the top step) der. Finally, participants were requested to identify how can go out and about in the community � they would distribute 100 women (or men) on the differ- to the homes of relatives, friends, and neighbors. ent steps to represent the current distribution of power — Village woman, Naw Da (Parwan), Afghanistan and freedom among community members of their own “ gender. The same sorting exercise was repeated, but At the top of the ladder are people who feel this time the focus groups were asked to imagine where confident about their lives and their future. They are � these same 100 individuals would have been found on well placed to realize their life’s goals and ambitions. the ladder a decade ago. A completed ladder, with steps, — Urban man, Hyderabad (Andhra Pradesh), India dynamics, and distributions of power for two points in time (2000 and 2010) produced a rough roadmap of a changes in local opportunities seem to exercise strong community’s social structure for that gender, and wheth- effects on aspirations and agency, but sometimes these er and how this structure is perceived to be changing. effects vary markedly by gender. Each ladder is highly influenced by local characteristics; Before moving into the analysis of the different factors however, commonalities can be discerned among the dif- and dynamics associated with power described by the ferent ladders in the traits of the different steps and in focus groups, we briefly review the data collection meth- the reasons for movement up and down. Also, the gen- od we used, which is important for making sense of the eral mobility trends reported by the focus groups can be findings that follow. The instrument we used in the field compared. For this purpose, we synthesized the numeri- to explore agency and empowerment is the “Ladder of cal data from the ladder sorting exercise into a “mobil- Power and Freedom,� which builds on similar instruments ity index� that equals the difference between a ladder’s in other studies of poverty and economic mobility.86 The mean step now and the mean step 10 years ago. We will ladder exercise was conducted only with the 194 adult fo- return to this index after highlighting common character- cus groups (one group for each sex in the 50 urban and 47 istics of the ladders. rural communities). The focus groups initially spent some time building their ladder to establish a common under- a. Defining the steps on the ladder standing, or framework, for assessing the dynamics of gaining and losing power in their communities (see box 4.1). To illustrate the exercise of creating a Ladder of Power and Freedom, we turn to one created by the women’s Facilitators introduced the topic by asking participants focus group from a traditional mountain town of 6,000 to identify the characteristics of the most powerful and in Ba’adan center, Yemen. According to the women from freest women or men of their neighborhood or village. this community, “enjoying a lot of freedom means that (Men described men and women described women.) women can express their views and move about freely, Similarly, they discussed qualities of the least powerful but only within certain limits and under the authority and least free women (or men) of the community. With of men and the customs, traditions, and social norms.� this information, they defined the top and bottom step of In this town, for instance, seclusion practices—such as On Norms and Agency a ladder, and the facilitator annotated the key traits for the requirement that male guardians accompany them each rung on a large piece of paper in front of the focus if they travel any distance from their neighborhood—re- group. Intermediate steps were then determined by the strict women’s movements. participants.87 While complex and multidimensional, the ladders do not completely describe all the power struc- The traits associated with the most powerful women tures for all the different types of women (or men) who included being married to powerful or wealthy men in reside in a study community. Some of the information the town, who may be members of the local council or provided is more stylized than based on actual women sheikhs. A powerful woman in Ba’adan center may or 104 and their characteristics. may not be well educated, but her sons and daughters go to school, a few all the way to university. These women their spouses, family name, or their adept management have money and can own jewelry, houses, land, and cars, of household affairs. although “men have a say in how women handle [their money] and often its disposal is for the benefit of the At the bottom step of the Ba’adan center ladder, women family.� A small number of these powerful women work cannot express their opinions “and are totally dependent outside the home, but “only in a government job as a on men for everything.� They are very poor, illiterate, teacher or nurse,� where they do not have to interact and may have to work both “in and outside the house� with the opposite sex and potentially risk their reputa- to make ends meet. If they work outside the home, it is tion, safety, and family honor. The majority of Ba’adan a sign of great economic stress in their households and center’s most powerful women do not have economic they take low-status jobs as domestic workers or ven- independence and their influence is mainly derived from dors. Some on the bottom step may earn an income at Box 4.1: Challenges with measuring social change from below Analyzing and comparing complex processes of social change, which necessarily transpire over time and across di- verse contexts, is inherently a great research challenge. We approached this by building on qualitative research tradi- tions of learning inductively from local people’s own interpretations and understandings of what power and freedom mean and how they lose or gain them in their lives. Issues of recall and context-specificity, however, are two key con- cerns that must be addressed in such analysis. First, asking individuals to recall situations always introduces the risk of getting partial information or an interpretation of events that the individual has developed to make sense of their current condition (Dempsey 2010). Some of our questions about agency required study participants to identify factors and recall conditions affecting agency for their own gender a decade ago. Yet, the natural course for individuals is to remember most clearly those actions that they made happen themselves and that best explain their circumstances now. Such processes mean that our focus groups at times might tend to downplay the relevance of seemingly unimportant events (in the course of pursuing goals) or wider environmental factors (which also influenced their choices) over which they often feel they have little control, such as weather, birth caste, presence of roads, access to services, etc. Second, this study applied a rapid and relatively standardized method of constructing a “ladder� to facilitate compara- tive analysis of men’s and women’s understandings of agency across their diverse contexts. Yet, we recognize that our method just provides a general picture; much of the nexus of agency and changing gender norms is deeply contextual to each specific location and is more clearly delineated with techniques that feature small samples, revisits, extended observation, and detailed life-story tools. Without question, dynamics of change are better captured through longi- tudinal techniques that involve tracing social change across generations in specific localities. Examples of insightful multigenerational investigations that provide a valuable perspective on how gender and poverty dynamics operate in specific localities or population groups include Perlman’s work in the favelas of Rio de Janeiro, Brazil (2010); Moser’s study of a neighborhood in Quayaquil, Ecuador (2009); Epstein’s two villages in India (1998); Fishburne Collier’s study On Norms and Agency of a Spanish village (1997); and Lomnitz’s (1977) study of a Mexican shantytown. 86 Our tool builds on the “Ladder of Life� from the World Bank’s global Moving Out of Poverty methodology (Narayan and Petesch 2005), which was designed to assess socio-economic mobility from the perspectives of men and women who had escaped poverty or remained trapped in pov- erty. While the Ladder of Life focuses on conditions that affect the poverty transitions of households, the “Ladder of Power and Freedom� assesses factors that affect the “transitions� in the agency of individuals’ of the same gender. 87 The number of steps on each ladder varies according to what each group deemed was necessary to represent their community’s reality. On average, 3 to 5 steps are defined, with rural communities typically adding fewer rungs than urban ones. 105 home from embroidery, hairdressing, sewing, and other percent to 50 percent of the community’s men over the activities. These women may have some more freedom past decade. The men spoke about no longer being able of movement, but their husband’s economic situation is to work in other Gulf States or the United States, and the not good and it is reflected in the entire family. lack of stable jobs for workers in their town. And unlike the women, these men saw a large decline in the share of The Ba’adan center women’s focus group identified four men on the top step over the last decade.88 steps on their ladder. For the women on the middle steps, their assets and influence are more limited than those b. A representative ladder on the top step, but they enjoy some decision-making inside their households. They may also have some as- When we examine the general characteristics of the lad- sets, thanks to inheritances, or work for pay within the der steps, we see important consistencies across gen- limits of what is acceptable for women, for instance, sell- ders and contexts. Focus groups mentioned traits as- ing goods to women door-to-door, teaching, nursing, or sociated with economic assets and occupational status holding government jobs. far more often than any other traits. Figure 4.1 provides highlights of the major characteristics that emerged in Much like the general traits of the ladder steps in other the ladder discussions. communities, women in Ba’adan center attached great importance to certain behaviors and attitudes as signs To the men’s focus groups, occupation and position of of power. Women on the highest step have “good morals authority in that occupation mattered intensely to their and good reputations, have the experience and ability to perceptions of the stature a man commands in their com- solve problems, and have a lot of money and authority to munity. The men on the top step of the ladders are at express their opinions and [give] advice.� Ba’adan center the pinnacle of local farming, business, politics, religious women on the lowest step, by contrast, have “weak per- institutions, and sometimes civic groups. In addition to sonalities� and “are very tired. Their lives are full suffer- authority roles, they frequently described the most pow- ing, deprivation, and daily misery.� erful men as very wealthy, commanding great respect, able to do whatever they want, and possessing excellent In Ba’adan center, as in most of the communities in the social skills and networks. Women attached an economic study, the top rung of the men’s ladder sits “higher� than role to power on their ladders almost as much as men the women’s, and men’s status is more tightly linked to do. Even in traditional rural communities, a woman on their public roles and economic might. In Ba’adan center, the top step may receive rental income from inherited these men are traders, elected officials, sheikhs, court land or have a nursing job; in urban settings, however, the employees, and “they get everything they want.� Men most powerful women are doctors or lawyers. Women on the bottom step are described as “tired. They work and men with little power do not work for pay or work only for one or two days a month and are responsible for in low-status jobs, for instance, as a domestic servant if a an entire family.� In most communities, the men with the woman or a daily-wage hauling job if a man. least power are jobless or rely on insecure daily wages. For both men’s and women’s ladders, education levels To measure how things have changed in Ba’adan center, associated with power can be quite mixed, and this likely women put twice as many women on the top step (about reflects the limited schooling opportunities that were On Norms and Agency 40 percent), compared to a decade ago. They indicated available to most adults in our sample. Focus groups that women are moving up the ladder as they become bet- largely attributed a stable and agreeable family life to all ter educated and find jobs; their husbands and children but those on the bottom step. In addition, they attached also have better jobs, which brings them status. Although great importance to conduct and attitude in their discus- they perceive they have greater power and freedom, these sions of power and freedom. For example, men and wom- women’s lives remain primarily in the domestic sphere. en on the top step were frequently seen as driven with strong leadership and social skills, while those with little The men of Ba’adan center reported the opposite trend power were more often described as unhappy, lacking 106 about their agency: their bottom step doubled from 25 confidence, withdrawn, or voiceless. Such assessments FIGURE 4.1: REPRESENTATIVE LADDER OF POWER AND FREEDOM (BOTH WOMEN’S AND MEN’S) • High status occupation • Harmonious marital and family life • Strong leadership, social skills, and networks • Great freedom of action Top step • Great self-efficacy and independence • May or may not be well educated • Some economic assets • Stable occupation • Harmonious marital and family life • Some freedom of action Middle step (s) • Generally self-con�dent • Has social networks • May or may not have some education • Scarce economic assets Bottom step • Jobless or insecure occupation • Stressful marital and family life • Uneducated • Little freedom of action • Excluded, voiceless, oppressed • Suffering, hopeless, no self-con�dence reflect what each community considered desirable val- Because women derive status from both domestic and ues and behaviors to be admired and respected. productive roles, they arguably enjoy more sources of power on their ladders than men. In an urban com- Where women’s and men’s ladders differ most is the munity near Zorzor, Liberia, a woman on the top step important status that women derive from their gender- may achieve that rung because she has 9 or 10 children ascribed household roles; however, as gender norms (and “many more grandchildren and relatives�), because change, these markers of status are also changing. In she is a community leader, or because she runs a big more traditional communities, focus groups often po- farm. Still, in the highly gendered playing fields of power, sitioned women with many children on the top step women’s status remains subordinate to men’s. The posi- and considered them influential due to their reproduc- tions of authority, assets, and occupations that women tive success; other focus groups, however, considered on the top step command are almost always of less sig- large families an attribute of the bottom step. Women nificance than the men’s (on the top step) in the same On Norms and Agency may gain the top step if they have a powerful spouse community. or a strong voice in their household, even though, as in Ba’adan center, their physical mobility may be restricted. A large proportion of the participants assessed them- But in contexts where gender norms are relaxing more selves as being on the middle steps, where their capacity quickly, women on the top step may have important civic to negotiate their interests is more in flux. Some women or political positions and enjoy great freedom of action. in Tanzania noted that: 88 In Ba’adan center, the women’s mobility index is a substantial 0.40, while the men’s plunges to -0.85. 107 Most women are on the middle or second step. They and men’s views are not always in sync. In 36 percent can provide some income for their families, but depend of the urban communities and 17 percent of rural com- on their husbands and their children. Their houses munities, women are climbing the ladder, but men reg- are always clean, but they do all the housework ister zero or negative mobility. Declines in power and by themselves, with assistance from the children. freedom by both sexes—�twin falling�—and men outper- Although they are the link between their families and forming women are more prevalent in rural communi- other families, they have little time to concentrate on ties than in urban ones. friendship because they are too busy trying to manage their household and family. They go to civic meetings, Importantly, the notion of “mobility� here and in chapter but they rarely speak out. They work hard, but a lot of 5 refers to gaining or losing power and freedom—in other their work is on the household farm or plantation, or at words, movements up and down the ladder of power small tables where they tend small retail businesses in and freedom—and not to the more traditional applica- front of their houses. They are hardly employed. They tions for assessing economic mobility. While there is sig- get little cooperation or help from their husbands. nificant overlap and emphasis on economic well-being, focus groups in this study did not completely associate Before turning to the reasons why individuals can rise accumulation of wealth or high-status occupations with and fall on the ladder, it is useful to show the com- great power. For example, elders, scholars, and religious parative findings on the change in positions of the 100 or civic leaders of either sex may be poor, but also are representative women and men on the ladders. Figure awarded great authority and respect by virtue of their 4.2 breaks apart the dataset by urban and rural focus benevolent deeds, wise counsel, or compassion for oth- groups, with the first set of bars on the left displaying ers. Alternatively, downward mobility may ensue from the share of communities where both women and men perceived misbehaviors and misconduct, some of which reported at least some upward movement over the have a strong gender component. In rural Kim Dong past 10 years—or gains in power and freedom for their District, Vietnam, women said loss of power can occur gender.89 With 55 percent of rural communities and 42 if women spend money “without discussing it with the percent of urban showing men and women advancing husband� and make “the husband think he is unneces- up the ladder, this is the most common pattern in our sary.� We explore these and other triggers for climbing dataset, which we called “twin climbing.� But women’s and falling in the next section. FIGURE 4.2: OUTCOMES OF ALL MEN'S AND WOMEN'S LADDERS IN URBAN AND RURAL COMMUNITIES 0.60 0.50 Share of total mentions 0.40 0.30 0.20 0.10 On Norms and Agency 0.00 Twin climbing Women climbing, men falling Twin falling Men climbing, women falling Rural (n=43) Urban (n=49) Note: Shares display outcomes from men's and women's ladders in 49 urban and 43 rural communities (or 184 ladders in all). Four women’s and eight men’s ladders did not register any movement, so were classi�ed with the falling set. The 5 Sudanese communities are not included in this analysis because the women's focus group did not conduct the sorting exercise of 100 women at the end of their ladder discussions. 89 Figure 4.1 does not include the 5 men’s and 5 women’s ladders from Sudan because the women’s focus group did not conduct the sorting exercise 108 at the end of their ladder discussions. Figure 4.2, however, includes the 5 Sudanese men’s ladders in the average ratings for the men’s mobility. Figure 4.3 also breaks out urban and rural samples, but instead presents the average rates of mobility on the lad- FIGURE 4.3. AVERAGE MOBILITY INDEX ders of the men’s and women’s focus groups. The “mo- OF MEN’S AND WOMEN’S LADDERS IN bility index� captures the difference between the mean RURAL AND URBAN COMMUNITIES step now and mean step 10 years ago, and portrays the general direction and extent of change in power and freedom perceived by a focus group over the past de- cade. We find a striking contrast between urban women Mobility index value 0.32 and men, with women seeing significant change in their 0.16 0.15 power and freedom and men, on balance, reporting loss of control. The same dynamic does not translate to rural communities, where women and men show a more “twin -0.09 rising� trend on average.90 Women Men Women Men Urban Rural From the accounts by the women’s focus groups of what drives their sense of empowerment, urban women (and Note: Results from 189 focus groups reflect the mean step today versus the mean step 10 years ago. The 5 Sudanese women's to a lesser extent, rural women) perceived that they have ladders are missing from this analysis because they did not conduct the sorting exercise. more voice and choices in their lives, and more space to negotiate and pursue goals. We expect urban women to feel more empowerment than rural women because cities offer more anonymity and freedom from confining 2. Perceptions of factors norms, as well as more institutional outlets for exercising shaping agency “ agency. Stronger trends can also be felt in cities from uni- versal education, promulgation and awareness of gender But you cannot climb up from the very bottom. There is laws, investments in public services and infrastructure, no way that you can even learn how to use money, if not the spread of communication technologies, and other only to drink. You cannot go to school anymore because forces. Women in the focus groups talked about how there is no way you can understand what they teach their lives are changing for the better, although exercis- ing agency does not always require them to challenge you. So, once down, you are doomed to stay there. — Urban man, Nsenene village, Tanzania � “ and change gender norms. The weak woman can work as a warden serving coffee, Men did not register as much empowerment as women, at a sewing factory, or as a secretary, nurse, or teacher. which is to be expected, given that they are more ac- The economic situation that forces her customed than women to being in positions of power as a traditional norm of masculinity. Yet, many urban men reported feeling a loss of power and freedom to shape to work can make her strong. — Village woman, Dirbas, West Bank and Gaza � their lives. As shown throughout this study, urban and This section presents the leading factors mentioned by rural men commonly expressed difficulties with adapting focus groups for gaining or losing power and freedom in to changes in their local structures of opportunities—and their communities. As with the ladder traits, the gender On Norms and Agency gender norms play an important role in constraining their dimensions of the mobility factors are more remarkable agency and flexibility to adapt. From men’s accounts, it is for their similarities than differences. Nevertheless, the clear that, in their eyes, opportunities and other factors strong role that gender norms play in mobility process- shaping aspirations and capacities to act are not equally es will become more evident in the sections to follow, distributed and few available opportunities are consid- where we probe more deeply into and compare women’s ered suitable for them. and men’s ladder discussions in specific contexts. 90 If we used calculations based on the median mobility indexes for each sample group, urban men perform somewhat better, but the rural ladders be- come a more moderate version of the urban ladders, with rural women’s average mobility doubling the rural men’s. The median, as opposed to average, mo- bility indexes for the urban ladders are 0.35 for women, but 0.0 for men. In the rural sample, the median mobility index is 0.20 for women and 0.10 for men. 109 Both men and women explained that they gain more ban women in Zorzor, Liberia, warned that “money can’t power and freedom in their lives through occupational carry you up here [to higher steps], only how you talk and economic initiatives that mainly include getting a to people and help pull them together.� Focus groups better job and sound financial management. Among the mainly associated desirable behaviors and attitudes with four groups (men and women, urban and rural), economic moving up the ladder, and undesirable or antisocial be- activities account for about one-third of the factors men- haviors and attitudes with falling down the ladder. But tioned (see figures 4.4 and 4.591). In Ba Dinh District, Viet- these divides sometimes blurred: focus groups, on occa- nam (a neighborhood of Hanoi), one way to move up the sion, assigned movements both up and down the ladder ladder is if “a man dares to borrow a big loan from dif- to acts of selfishness or corruption. ferent sources to open a business.� Women in the same neighborhood said that what helps them climb the ladder As with the ladder traits, the role of education and train- is “getting promoted,� “using money saved from a govern- ing in gaining power and freedom is not as straightfor- ment job to buy land,� or “being successful in the stock ward as conveyed in the figure because the frequencies market.� Across the focus groups, women and men not of mention in the focus groups include contradictory only stressed earning income, but also careful budgeting, reports that it is both necessary and unnecessary for borrowing, and management of their finances to build sav- upward climbing. While it is central to aspirations that ings. In rural Nagari Bukik Batabuah, Indonesia, women younger generations described for their own lives, more can climb, “but it doesn’t happen quickly. We first have to than half of the adult focus groups neglected to mention save the wages that we receive from working in someone education as a factor in upward mobility, while others of- else’s field before we are able to move up to a higher level.� ten qualified its importance: Individual behavioral and psychological factors were the Women on the highest step are supposed to be second most often mentioned factors that stimulate up- [educated], but the reality is that any creative, diligent, ward climbs, except by rural women (who mentioned ed- and aggressive woman is capable of influencing others. ucational factors slightly more). Both women’s and men’s They can be community leaders. There are educated explanations routinely indicated that ladder ascents re- women in the lowest class, so education alone is not quire courage, sincerity, humility, service, temperance, enough. Woman should have other skills. (Village spiritual guidance and growth, good reputation, hard women’s focus group, River Nile State, Sudan) work, consistency, positive mind set, honesty, willing- ness to change, goal-setting, avoidance of alcohol abuse, In Suakoko District, Liberia, the men advised that “edu- utilization of individual talent, rejection of corruption or cation is needed, but in our area, people can do with- bribery, and optimism. In a village of Morobe Province, out�; and in Jakarta, Indonesia, the men explained that Papua New Guinea, some of the factors that allow men “some uneducated people have become big bosses. The to climb the ladder included “the drive to start a busi- important things are capital and ambition.� Nevertheless, ness,� a good character, and a “dream of being success- a strong cross section of groups highly valued education ful.� A sense of self-efficacy and purpose —or capacity to and, where available, adult literacy and vocational train- aspire (Appardurai 2004)—were deemed to be vital: “A ing opportunities. woman has to believe in herself, break through her fear, gain more self-confidence. Then everything becomes Ladder descents most often ensued from business fail- On Norms and Agency easier� (a woman, Justynowo, Poland). ures, jobs losses, bad investments, and poor financial management (figures 4.6 and 4.7). Urban men and women More generally, the focus group narratives about behav- stressed these risks more than their rural counterparts. ioral and psychological factors revealed that both wom- The second cause of descents, behavioral and psycho- en and men fervently believed that their own positive logical factors, can be triggered by dishonesty, prejudice, mindsets and self-confidence, as well as how they con- lying, conceit, wasteful spending, extravagance, corrup- duct themselves and treat others, matter greatly when it tion, cheating, jealousy, selfishness, laziness, infidelity, comes to gaining power and freedom. For many, power is loss of trust, fights, disregard for the rules, and excess al- 110 associated with being respected by the community. Ur- cohol consumption (by men). These concerns were more FIGURE 4.4: URBAN UPWARD MOBILITY FACTORS 0.45 0.40 Share of total mentions 0.35 0.30 0.25 0.20 0.15 0.10 0.05 0.00 Occupational Behavioral and Education and Marital and Social networks and economic psychological training familial Urban men Urban Women Note: Data from all 100 men’s and women’s urban focus groups. FIGURE 4.5: RURAL UPWARD MOBILITY FACTORS 0.40 0.35 Share of total mentions 0.30 0.25 0.20 0.15 0.10 0.05 0.00 Occupational Behavioral and Education and Marital and Social networks and economic psychological training familial Rural men Rural women Note: Data from all 94 men’s and women’s rural focus groups. often raised in the countryside and, again, it may be that losses of power more than men. “Marrying into a wealthy behavioral factors are less prominent as a cause of losing family� and cooperative couples and harmonious house- power in urban groups because of the greater anonymity holds bring clear advantages. In Zabibu village, Tanzania, in urban environments. for instance, a woman explained that to move up in her village a woman “needs to get a good and understanding On Norms and Agency Overall, factors related to marital or familial relationships husband, who allows her to go into business and engage were less frequently mentioned as triggers for climbing in educational activities.� Women from a more urban area and falling. As expected, women mentioned the role of (Nsenene village) in Tanzania agreed: “You can divorce a their spouse or other family members in their gains and hopeless husband and marry a good one who can put you 91 Figures 4.4–4.7 present mobility factors mentioned by least 5 percent of the groups in the coding exercise. We do not, however, include the many factors coded under “other.� Due to time constraints with informing the World Development Report 2012, it was not always possible to refine and clarify how the coders understood some of the factors coded as “other.� A review of these factors reveals that a good share could have been coded under the already defined categories. We do not feel, however, that the overall pattern of findings would be much affected by a re-coding. Some of the additional factors under the “other� category that we did not code for include religion, sorcery, aging, and luck (winning the lottery). 111 FIGURE 4.6: URBAN DOWNWARD MOBILITY FACTORS 0.6 0.5 Share of total mentions 0.4 0.3 0.2 0.1 0 Occupational Behavioral and Marital or Ilegal activities Social networks and economic psychological familial Urban men Urban Women Note: Data from all 102 urban focus groups. FIGURE 4.7: RURAL DOWNWARD MOBILITY FACTORS 0.45 0.40 Share of total mentions 0.35 0.30 0.25 0.20 0.15 0.10 0.05 0.00 Occupational Behavioral and Marital or Ilegal activities Social networks and economic psychological familial Rural men Rural women Note: Data from all 92 rural focus groups. higher on the ladder.� Dialogue and agreement between on “is worse than his death.� But it is urban women who the couple were key for women from Saylla District, Peru, displayed the greatest relative concern for marital and when it comes to leaving the bottom step. familial triggers. In urban Lambayeque Province, Peru, a woman explained that “when you separate from your Meanwhile, falling down the ladder can be triggered by husband, assets must be divided and this weakens you.� On Norms and Agency any number of household problems, such as marital or family discord; widowhood; family dissolution or divorce; Non-family social networks are the last of main factors too many wives where polygamy is practiced; marrying in movements on the ladder (see figures 4.4 and 4.5), “down�; sons who are unemployed, leave home, or en- accounting for about one-tenth of the upward factors gage in vices; and inability to bear children or bearing mentioned and less so for falling. In some cases, these too many children. Rural men mentioned these hazards connections can be instrumental in securing other deter- as often as rural women, and urban women much more minants, such as economic gains, as suggested by a man so than urban men. In rural Martynice, Poland, and else- from Tewor District, Liberia: “Make friends with people at 112 where, women reported that being divorced or cheated the top step and ask for land.� In other cases—mentioned by women—connecting with other women (informally or 1 “have nothing to invest,� and so some of the only ways in formal organizations) can open information flows or for these women to move up the ladder were to “win the build self-confidence and aspirations. In Old City of He- lottery,� a “wise wedding,� or “organized crime.� Scaling bron, West Bank and Gaza, “a weak woman can go up subsequent steps, by comparison, was seen to be more [the ladder] by hanging out with a strong woman�; in Um- doable because each step potentially brings more earn- lazi Township A of Durban, South Africa, a woman “must ing power, assets, self-confidence, knowledge and skills, involve herself with other women� in order to climb. In reputation for good work and honesty, and so forth. the refugee camp visited in Al Fashir, Sudan, women said they can become more powerful through literacy and In addition, the five upward mobility factors described “training courses, workshops, and seminars, and inter- above are seen to be interlinked, and it takes purposeful acting with active women leaders in society.� In National combining or sequencing of them to produce real gains Capital District, Papua New Guinea, getting involved in in power and freedom. The introduction to part III men- church activities helps women increase status and move tions how their interconnections help open up pathways up from lower rungs. Earning a good reputation through of empowerment and how some factors are necessary charitable works or taking a leadership role in self-help conditions for others to have an actual impact in trigger- groups are sometimes mentioned as stepping stones for ing upward movement on the ladders. These pathways women and men alike. By contrast, women indicated that are not the same for men and women. isolation from their community, self-help groups that lack cohesion, and “bad friends� who expect or owe loans can A man from Umlazi township B of Durban, South Africa, cause ladder descents. married three factors—related to better management and generation of economic resources and self-discipline—as necessary for leaving step 1: “If people work hard, save 3. Combinations of mobility factors their money, and do not waste it on alcohol, they can manage to move up to the next level.� In Floresti, Moldo- In surveying the narratives about movements up and va, the men similarly noted that a man on step 1 can get down, two processes stand out. First, the bottom step to step 2 if he “finds a job or some possibility for earning of powerlessness is widely seen as the hardest of all to money,� but he also needs to “stop drinking and start tak- leave. Second, rather than describing one factor as more ing care of his family,� and to get “a good wife.� In Paro, important than another, most focus groups conveyed Bhutan, one man said that in order to climb, “a man has that moving up the ladder requires mobilizing combina- to be hardworking, reliable, and good at heart; be able to tions of factors. learn from others and help others; avoid indulging in al- cohol; take good care of his family; not have extramarital Those on the bottom step are widely perceived to face affairs; and be positive in life.� Numerous accounts about the toughest climb of all. Not only is great effort and sac- men needing to stop indulging in vices and focus on their rifice required, but some assistance from others or great family obligations to be a good provider revealed the ex- luck are often seen to be needed as well. And these tensive problems of jobless and deeply frustrated men may well be less in one’s control. In Umlazi Township of in this dataset. In the Umlazi township B focus group, the Durban, South Africa, men say that “It is very difficult men lamented how their lives have changed, compared for a person to move out of step 1 because the major- to a decade ago: “Men had work before, now they have On Norms and Agency ity of them are not educated all. Maybe unless someone no job opportunities.� wins the lottery jackpot.� Or, in University Quarter of Hebron in the West Bank, escaping step 1 takes “cred- Women’s combinations differed from men’s for many ibility, luck, and people’s support.� In a neighborhood of reasons, including different gender norms and different Monrovia, Liberia, women explained that “those that are positions and status in the communities and households. down on are not able to send children to school� and so While economic factors featured broadly for women as their children, when they become adults, are less likely well as men, women were less likely to mention prob- to be able to help them move up. In Olsztyn, Poland, lems with antisocial vices and more likely to speak about the women’s group thought that because those on step a need to become less passive and submissive, although 113 these traits are often valued traits in a woman. In Lam- by “being more responsible, emulating other people who bayeque Province, Peru, a woman can increase her pow- work hard, and consistently working hard themselves; er and freedom but “she needs to take more initiative, having a positive attitude and a desire to learn; being take risks, begin a business and commit to it, take out a ambitious; and not indulging in bad habits.� Yet, like else- loan, be responsible and careful, and set goals.� where, a Paro woman’s household role and domestic re- lationships may also be ingredients in her efforts to gain In urban Ba Dinh District, Vietnam, women said that ris- power, such as aspiring to be a “husband and wife [who] ing above step 1 may be possible if, among other mea- encourage each other and give moral support,� or having sures, their husbands stopped beating their wives and “educated and independent children.� Across diverse if the women are able to get a loan to open a business contexts, women reported climbing their ladders as they “and improve the family economy.� Domestic violence is gain more voice and autonomy in their domestic roles; a strong sign that women have little power and freedom. although, overall, women stress their own economic ini- Economic initiatives are generally seen as ways for those tiatives more than any other single factor as a reason for on the bottom step facing difficult household circum- gaining power. stances to forge more voice and autonomy. Downward mobility is also linked to intersecting causal Climbing up the ladder, moreover, still means women factors that can send individuals spiraling out of control have to tend to their household responsibilities even of their lives. The nexus of hardships mentioned often fea- as they mobilize diverse assets and capabilities to fuel ture job loss or financial mismanagement that may then their ascent. In Firestone District, Liberia, women can trigger undesirable or antisocial behaviors and marital or become more powerful if they are extremely industrious familial problems. According to a man from urban Emputa and patch together different initiatives to build assets, village, Tanzania, “you only have to drink a lot, hang out but they cannot neglect their domestic duties: with women a lot, and sleep in bars and guest houses, and that will be the end of your power. There were rich men If you have a small farm, you make the farm bigger. here who once had boats with engines, but now they are If the crops come up, you sell the produce and make at the bottom begging.� In Chiclayo, Peru, women par- whatever business you want. But if you wash for one ticularly mentioned how a woman on the top step can fall person, you can just as well wash for two or three down the ladder if she mismanages her business, makes a people. In doing these things, you should tie your bad investment, or has problems with workers; although, stomach [eat less] and ask your family to help you take women can also lose status in the village due to difficul- care of some of your children. ties with their marriages or children. Similarly, escaping the bottom step for women from ur- In the next two sections, we look more closely at the gen- ban Shirabad Ulya (Kabul), Afghanistan, also involved a der dimensions of agency processes in an urban (Jaipur) flurry of initiatives that implicitly require carefully navi- and a rural community (Malangachilima), both with espe- gating old and new norms for women’s conducts: cially strong twin climbing. In each locality, women’s and men’s groups reported that many of them are gaining Women must work hard and follow precisely the power. Unfortunately, twin climbing to the extent report- decisions made by the men household members. They ed in these two communities is rare in our sample, but On Norms and Agency must not quarrel with the men. Meanwhile they should these contexts are nevertheless useful to examine more consult their friends and get information and assistance closely because they provide valuable insights into what from them. They must send their children to school, but strong agency looks and feels like for both sexes. They be must be careful about the expenses. They should also clearly reveal the gender differences in the factors avoid any extra, unnecessary spending. Whenever they that underpin agency. A third section then assesses have free time, they can go meet their friends. an urban community with a polarized mobility pattern, where the women perceived significant empowerment, Or, in the less traditional setting of urban Paro, Bhutan, a but men reported falling down their ladders. Taken to- 114 woman can climb up by being successful in business and gether, these localities display how men’s and women’s mobility on their ladders is interdependent, and the sig- posure by “meeting with other women frequently,� and nificant (albeit different) influence of gender norms and trying to have more say in their household. These are dif- local institutions in their agency processes. We take up ficult hurdles for women on step 1, where gender norms the latter themes more systematically in chapter 5. for women’s physical mobility and expectations of obedi- ence and submission are at their strictest. a. Moving up the ladder in the city At the top of their ladder, they placed 25 percent of the Jaipur (Odisha), India, a growing town of 7,000, used to women in their community on step 5, up from the re- be mostly farms and pasturelands a generation ago.92 spectable 15 percent of a decade ago. Women said that With rapid urbanization, the town’s level of poverty has they can climb to the top step if they gain experience plummeted over the past decade from 60 percent to 20 “dealing with community elders,� learn more about mo- percent. The women of Jaipur have good jobs as teach- bilizing resources inside and outside the community, and ers or administrative staff in private schools, as workers are “ready to provide a helping hand� to other women. in tailoring and other industries, as owners of big and These very same triggers are also on men’s ladders, but small shops. The town’s men work in daily-wage agricul- these women’s interactions with elites and local activism tural jobs, but there are also good jobs available in con- are more ground-breaking for their town. struction, stores of all sizes, travel agencies, computer shops, and the government. The men’s focus group in Jaipur paid most attention to assigning substantial economic and political roles to the The women’s focus group from Jaipur described wom- men on the top step. Men on bottom step, in contrast, en on the top step of their ladder as having significant have little earning power and education: “They are daily visibility in civic roles in their community and enjoying laborers and they have to fight every day to earn their greater authority in their households. These empow- livelihood.� Unemployed men are also commonly found ered women participate in local meetings and interact on bottom steps, but Jaipur’s economy is unlike many with important community leaders. They also make in our sample because men can find sufficient jobs that substantive decisions related to their children (such they deem appropriate for them. as sending them to school, paying tuition) and do not have to seek permission from their husbands for ev- The men’s ladder also features five steps, but most of erything: “She can go out of the house at will.� At the the movement happens in the middle. The majority of other extreme, Jaipur women on the bottom step are men, on step 2 at the beginning of the past decade, depicted as having no education or economic indepen- rose to step 3, and the share on step 3 changed from 40 dence, little voice, and no aspirations for themselves, percent to 60 percent of the community’s men. Due to and are secluded and isolated in their homes. (The full education, better savings, and the assistance of self-help list of characteristics of women’s top and bottom lad- groups, many men in Jaipur on step 3 have been able to der steps from the Jaipur women’s focus group are in find better livelihoods over the past decade, working in table 4.1 below.) the government or the private sector, or even running their own profitable businesses. The men’s bottom step With five steps on their ladder, the women in Jaipur grew slightly, but stayed close to one-third of the com- provided a fine-grained analysis of the mobility factors munity, similar to the women’s bottom step. On Norms and Agency and extent of changes shaping their lives. The group re- ported that 10 years ago nearly half the women of their Men attributed their bottom step and reasons for falling community were on step 1, the bottom step, with little to troublesome behaviors that include “losing the trust of power and freedom. For women to escape the bottom or not being accountable to� their self-help groups, not step, the focus group suggested that they must aspire saving, drinking too much, and “not letting their women to get ahead and develop an attitude for “self-growth,� work.� This last statement, implying that women’s econom- improve their networks, and gain information and ex- ic participation affects men’s mobility processes, is a quite 92 This Jaipur is not the large city with the same name in Rajasthan State in India. 115 Table 4.1: Top and bottom steps of women’s and men’s Ladders of Power and Freedom in Jaipur (Odisha), India Women's top step Women’s bottom step • Has good understanding of others’ needs] • Does not interact or mingle in • Is the community’s face [represents the community] community; stays at home • Has more exposure and is mobile • Is not very expressive • Is soft spoken and approachable • Lets husband make all decisions, • Is patient and tolerant economic, future of children, marriage • Can adjust and accommodate to everyone of children, education of children • Is determined and courageous • Is not very responsive General attitudes • Takes initiative and responsibility for community concerns • Feels (internalizes) that they cannot and behaviors • Is trusted and looked up to by other women understand anything because they are • Has awareness of community development issues, but is not illiterate well educated • Has no awareness of or interest in • Has vast experience and seen as a guardian of the community participating in community issues • Able to speak in front of 10 people or to outsiders • Does not have decision-making power • Has good relationship with others and does not have vested for self interest • Has high aspirations for children’s • Turns cold shoulder to the women in community education • Has an education • Is illiterate, did not complete primary • She completed 10th grade with honors education Education • She completed a nursing course in Cuttack • Is less literate than husband • Worked eight years before marriage • Has husband with university degree (BSc or MPhil) • Involved in community groups and is responsible for looking • Does not have a job Jobs and after financial management of those groups • Mostly does house work specific • Sometimes acts as midwife for poorer communities free of community cost responsibilities • Has time to devote to community causes and concerns • Has relationship with spouse where they trust and respect • Sees husband as the ultimate authority each other in making economic decisions • Husband doesn’t stop wife from being active in civic • Husband is also in business Family and organizations or holding positions in the community • Submits completely to husband’s household • Aspires to raise daughters to be independent. decisions and demands • Does not need permission from husband to go anywhere • Reveres her husband] • Has equal voice in household’s economic decision-making for small and large things • Saves through a self-help savings group or a chit fund • Has very little savings Savings and • Has received a loan from an MFI and repaid it • Has only one member of family earning borrowing habits • Has a bank account in her name income Men's top step Men’s bottom step • Has money • Has low income General attitudes • Has political power • Has no bargaining power and behaviors • Donates money to the needy Education High school or university graduate Minimal or no education On Norms and Agency • Is helpful to the community Is a daily wage worker Jobs and specific • Lets community members consult him and ask advice for free community • Donates land for community development works, temple, responsibilities school • Has politically-oriented, supportive family • Has a large family Family and • Motivates and grooms wife to take open seat reserved for • Has many friends in the same economic household women on the ward council situation Is politically active and effective; can get work done for the • Has very little savings Political power community and individuals • Has only one member of family earning 116 income Financial power Has their own business Has no savings strong signal that Jaipur is undergoing real change in gen- women. The village also hosts a busy bimonthly market der norms, not just relaxation of compliance with gender that draws buyers and sellers from across the area. Some norms, as more women participate in the local economy. of Malangachilima’s smaller farmers have been through difficult times, however, due to swings in the area’s econ- Both the women and men of Jaipur, like many other omy, drought, and Rift Valley fever. And the village still communities in the study, spoke of the strong mutual has important development challenges. Most school-age support that defines the marital relationship on the top children attend the local primary school, but only half go step. Women appreciated the presence of trust among on to secondary school. There is no electricity, and most these couples and the husbands’ support for women’s residents obtain their water from local stand-pipes. civic activities; the men’s ladder specifically mentioned, for example, how powerful men groomed their wives to When building their ladder of power and freedom, the take advantage of the gender quotas for local council women’s focus group in Malangachilima described their seats. No such harmony and cooperation are mentioned most powerful women with these attributes: on the bottom step. The women warned that they could fall back down to step 1 if they “completely surrender to • She works very hard. [their] husbands’ authority.� • She harvests a lot. • Most of her children go to good schools; some go The women in Jaipur are visibly gaining power and free- even to schools in town or farther away. dom, and in this process are changing some of the prev- • She completed primary and even secondary school. alent gender norms in the community. The most pow- • She is a leader; she gets involved in politics beyond erful women enjoy authority positions both inside and the village. outside the home and have gained a significant pres- • She has a good husband, but does not always have to ence in the formal economic, political, and civic institu- have a husband. tions of their town. The men are moving up, too, and • She can be rich and own cattle and land. see their education, financial management, and self- help groups playing important roles. This simultaneous In stark contrast, they label the woman on step 1 “a slut,� movement likely helps ease the way for local women who “hardly has a husband; she simply gives birth to chil- to exercise agency and for men to accept the women’s dren who have different fathers. She is poor: she hardly new conduct and roles. owns a cow, possibly has a chicken. She has no land, un- less it is inherited land.� She also is “confined at home ... b. Moving up the ladder in the countryside [and is too] poor to buy anything useful for herself. She never speaks in public.� Polygamy is common in the vil- “ lage, which is especially disadvantageous for women on Powerful women take good care of their families, the bottom steps because their husbands are stretching their husbands, and children. They are also very resources across multiple households. hard working. A powerful woman will never say that there is nothing in her sock [store for food]. — Village woman, Malangachilima, Tanzania � What is surprising about Malangachilima is that the wom- en felt that they have almost entirely escaped the agency- crushing bottom step. While they assigned an astonishing On Norms and Agency Malangachilima, Tanzania, is a large village of nearly 95 percent of women on the bottom step just a decade 7,000. The village chairperson, a woman, estimates that ago, now only 5 percent are on this rung. On step 2, where poverty fell in the 2000–2010 decade from perhaps 70 most of the women now reside, women have their own percent to 30 percent. She attributes most of the vil- farms and other ventures, most of their children are in lage’s rising prosperity to the recent introduction of school, and they worry much less about hunger. modern farming methods. Farmers are raising cattle and growing sunflowers, ground nuts, and grapes. Donated Women said they can climb their ladders through hard equipment for making sunflower oil stimulated the rise of work, leadership in local groups, saving money, buy- a successful community cooperative for village men and ing cattle, and visiting their children in town and com- 117 ing back “with good ideas.� One woman explained that are finding ways to gain power, the men identified sev- children send money from town, which has helped some eral factors: gaining confidence from their own edu- villagers “become very powerful because nothing is dis- cation and educating their children, attending training turbing them anymore and they have become busy in courses and developing entrepreneurial skills, obtain- the community thinking about development issues.� Al- ing financial assistance, and improving their networks to though “nothing disturbing them� might be interpreted get advice from other knowledgeable people who are in different ways, women and men reported decreasing considered powerful. One man elaborated, “Ten years incidences of domestic violence. Also, once their chil- ago, we rarely left the community; mingling with other dren grow up, no longer require care and school fees, people was very minimal. That is why our community and start to give back, women generally have more time was backward.� In later discussions about community and resources to focus on themselves and become ac- organizations, a man reported that now they have vari- tive in their community. To move up to the top of their ous outlets to obtain loans to expand their farming and three-step ladder, women need to “make their mind big� other ventures, and that local men gather together in and “think how to change for the better.� In reflecting agricultural groups, carpentry groups, and beekeeping on their growing strength, a woman declared, “I think activities. that we have very few very hopeless women because we know what work means now, not like 10 years back when In fact, external groups seem to have played a valuable we were just following men. And in fact, we have many role in the village. Focus groups of younger women re- women now who are very powerful, manage their own lated that programs offered by non-governmental or- lives, and are not at all dependent on men.� ganizations encouraged them to study and raised their awareness of their rights. And both women and men By comparison, Malangachilima’s men have not been as took advantage of NGO and governmental programs successful at escaping the bottom step, but a good share to mobilize their own farmer, producer, and finance is nevertheless on the move. The men estimated that groups and have reached out to external partners to their bottom step has shrunk from 85 percent a decade strengthen their livelihood activities. For instance, a vil- ago to about 50 percent still trapped there. The men on lage woman noted that, “we can borrow money from step 1 are described as letting their wives do everything our associations; we also have SACCOS [savings and for their families’ needs and, when harvest comes, “he credit cooperatives] from which we can get money. ... steals the crops. He is a lazy man.� The leading reasons We had World Vision here, which has been a good re- keeping men on the bottom rung (or descending to it) are source for us.� World Vision spent a decade working in too much drinking and too few assets “because the few the village and departed several years ago. Women and cows they have are shared among the many wives they men have also rotated in and out of the local executive married.� The village men also mentioned that these men officer position. on step 1 are the only decision-makers in their house- holds and they may be “cheating a lot.� And if you give In sum, gender norms seemed to have transformed very a man on the bottom step a loan, “forget about getting rapidly in this village. The large majority of women and it back.� These men are “never trusted� and “will never a good share of the men of Malangachilima are rising have good ideas.� off their very harsh bottom steps. They both are being helped by diverse factors, including their own economic On Norms and Agency With half of the men in the village seen to be power- initiatives, greater exposure to new ideas, educational less and still on step 1, the Malangachilima community opportunities, and collective action. Jaipur and Malan- should be facing extensive hardship. Yet, the men’s nar- gachilima are more hopeful communities than most in ratives also spoke of progress for the many men mov- our study. Very often we find women climbing, but men ing up. They described the men on the top step (which are mostly stuck or falling on their ladders. Very often has grown from 5 percent to 20 percent of the men) the men’s focus groups are deeply discouraged by the as commanding great respect in the community, help- poor condition of their local economy; problems of job- ing others in need, having very strong family lives, and lessness, frustration, and vices send more men tumbling 118 being good providers. For the share of village men who than climbing. c. Women climbing, men tumbling in the city Ceadir-Lunga, Moldova, illustrates the polarizing dynam- ics of economic stress on men’s and women’s agency. A As noted above, the general mobility trends on the city of about 23,000, Ceadir-Lunga’s economy has been urban ladders display a pattern of women perceiving weakening over the last decade and worsened during significant empowerment and men seeing disempow- the global economic crisis of 2008 and 2009. Focus erment. The forces driving this pattern mainly seem groups conducted in mid-2010 indicated that many men to be a consequence of communities in the study that and smaller numbers of women have become economic have been hit hard by various shocks, which had heavi- migrants. Five of the eight men in the adult focus group er impacts on agency processes in our urban sample, in this town were unemployed at the time of our inter- compared to the rural. In a few cases where the role of views. Unlike the men, most of the local women in the shocks is less evident, men may be feeling emasculated adult focus group had jobs. And in spite of the economic by their exposure to new lifestyles or ways to earn a situation, this group of women mainly saw the past de- living that have raised their aspirations, but they do not cade as favorable for their agency. have the means or local structure of opportunities to pursue such goals. When asked about the best ways to make a living in Ceadir-Lunga, the men identified diverse opportuni- A quarter of the sample countries had been stressed ties for both sexes, from working in the government by violent political conflict during the 10-year period and the private sector, running small- and medium- of the ladders; other urban neighborhoods in peace- sized businesses, to engaging in wholesale and wine- ful countries reported significantly harmful effects making ventures. Openings for such good jobs, they from the global economic crisis of 2008 or other major explained, are rare and few of them have the means economic shocks. Gender differences in responses to to start their own business or to access the requisite periods of turmoil are important for interpreting the finance. Some immigrated to find work in other Euro- patterns on the urban ladders. With the deterioration pean cities, but they do not speak highly of their expe- of economic opportunities, men struggle with identity riences and reported that the men who are still work- issues as they strive to make ends meets and provide ing abroad are lonely and unhappy away from their for their homes, often having to undertake economic families and friends. activities that may damage their status or erode self- esteem (Schrock and Schwalbe 2009). Some men The men also felt that women have better opportuni- seem to become passive and opt for unemployment ties to get jobs in the city than men; local gender dif- until better times. It is in these contexts where men’s ferences attached to the status of jobs clearly shape emotional struggles—and coping strategies that involve their perceptions. In response to questions about the drinking, gambling, drugs, affairs with other women, worst ways to make a living in the city, for instance, the and marital conflict—are often reported by women’s men identified digging graves and cleaning streets or and men’s groups (although these are common prob- houses; however, one of the men countered, “There is lems for men on bottom steps even where shocks have no such thing as a bad job. These are just insufficiently not been reported). paid jobs, like watchmen, sweepers, cleaners.� Yet, an- other interjected that men would never take a cleaning Women in struggling economies, meanwhile, try their job because it is so poorly paid and demeaning: “A man On Norms and Agency best to pick up the slack for their families. They begin who respects himself will never accept such a job.� They new economic initiatives or intensify their existing ones, also said that local men would be unlikely candidates and carry their households through the difficult times. for administrative positions in firms. “They choose the For some of these women, the grip of various gender woman [for a secretarial position] because she is more norms relaxes due to the exigencies of these stressful responsible and conscientious about the work she has periods. This relaxation, however, does not necessar- to do,� one man stated. These remarks about the gen- ily lead to a significant change in the overall climate for der appropriateness of one job or the other are not women’s economic, political, and civic participation, as uncommon in other communities. The poor returns and occurred in Jaipur and Malangachilima. potentially strong reputational harm attached to bad 119 jobs or women’s jobs (which may often be the same) Beyond Ceadir-Lunga, economic factors also emerge seem to sap men’s resilience in the face of deteriorating as the triggers most often named that move individuals economic opportunities. down their ladders. Falls can be precipitated by losing a business, being retrenched, losing property, losing a Women in Ceadir-Lunga also mentioned struggling with job abroad, going bankrupt, making bad investments, de- the difficult economy, but they said this has pushed many faulting on loans, mismanaging money, dealing with na- of them into new jobs and expanded their income-earn- tional economic crises, facing war, and so forth. In rural ing role. They described the advent of more powerful areas, droughts, floods, pests, and diseases add further women in their neighborhood, who did not exist a de- severe risks to livelihoods. cade ago. The women on their top step are better edu- cated, have fewer children, and work at professional jobs In our sample, urban men seemed to struggle the most in the government or run their own enterprises. Some with adversity. In many urban communities, where men’s women have been able to launch businesses with funds “good� jobs have grown scarce, they often perceived that they earned themselves while working abroad or that women have easier access to work in certain sec- that their migrant husbands sent home. “These women tors, such as service industries, because of the premium are financially secure, smart, self-confident, good in on “soft skills� and personal appearance. Whether this business, but sometimes unreliable and unkind,� they perception is accurate or not, it is clear that many urban explained. A few of them climbed up the ladder by mar- men feel deep insecurity about their role as a provider rying rich husbands or taking advantage of market op- and this is a source of great frustration for them. portunities left by men who went abroad in search of better jobs. The focus group estimated that nearly 20 The dataset also contains four economies affected by percent of the women in their neighborhood had pulled violent political conflict during the 2000–2010 reference themselves up to the two top steps, which did not exist decade for assessing changes in power and freedom. for them 10 years ago. The gender and conflict literature is bringing to light women’s deep vulnerability to sexual and other violence The much larger share of women still on steps 1 and 2 are in the case of war, but it also documents how periods of also educated, may have jobs, and are raising families. conflict force gender norms to relax as the institutional Their husbands often are unemployed, so their house- structures that control them are dismantled.93 Women holds rely on the women’s income. The Ceadir-Lunga enlarge their public roles, but men go into retreat. These focus group estimated that 80 percent of the women in processes were especially on display in some of the sam- their neighborhood used to be on step 1 a decade ago, ple communities in post-conflict Liberia. Women there but this has now dropped substantially to 20 percent. widely saw their economic, civic, and political leadership Women on step 2, however, may well have been com- strengthen in their communities, while many men report- pelled to take a job due to the economic crisis, but this ed feeling emasculated as they tried to recover their live- has strengthened their voice. lihoods since the war ended in 2003. Some of the men’s narratives openly identified women’s gains in power as Men’s and women’s different coping strategies in the face challenges to their authority, if not causes for their loss of adversity affected their sense of agency in competing of power. Women, on the other hand, voiced frustration ways in Ceadir-Lunga. When we asked the men about at men’s inability to adapt to women’s new roles or to On Norms and Agency new gender laws taking effect in Moldova, one of them the changing economic and institutional circumstances responded, “Mainly nothing has changed, especially for in their communities. Gender relations in these mainly the good.� Another elaborated, “It didn’t change any- urban Liberian localities seemed very tense. thing because the financial situation in Moldova is very bad. I think that women should stay at home and take In Greenville, Liberia, for instance, the women mentioned care of the family, but when women go into politics, busi- that they have a new marketing association (established ness, and so on, it is not a good thing.� Unlike in Jaipur by their elected chairwoman), but the men are not help- and Malangachilima, the men seemed loath to recognize ing make it a success. Rather, the men are leaving women 120 women’s expanded and important provider roles. alone to do the arduous work required to “cut palm nuts “ and brush on the farm.� Women are gaining power, but It is good for women to be strong, but the most men in Greenville are falling in droves down their ladder. important opinion is the man’s opinion. He is the person Just half of the men were on the bottom step 10 years ago, who controls everything and has the power to make � but the men’s group placed 90 percent of them there now. decisions in the family. And this is how the men describe themselves on step 1: — Village women’s focus group, Kharef District, Yemen “ They are not working, there is no business. They cut A person has to look for opportunity, take risks, � palms and give them to their wives to sell so they can push themselves, and fight to get ahead. get food. They do the weeding and brush contracts; — Urban man, Lambayeque Province, Peru they collect kiss-me (tiny snails) to sell; they cut wood and make charcoal to sell. Any day they don’t work, they have no food. They live in thatched houses and also play roles in climbing the ladders, as do family re- have a junior-high school level of education. They have lations and social networks. Yet, gender norms interact a fine and happy family that goes to church together with local opportunity structures and individual initia- and sits together. ... tives to make processes of exercising agency quite dif- ferent in reality for men and women. Women seem able [Some couples have a] fighting relationship; both to climb their ladders in good times or bad, and claim women and men grumble and fuss every day. Even more empowerment whether they gain more voice in today, when we have gathered for the focus group, the their domestic or public roles. Men’s sense of agency, by wife is asking, “Why didn’t [the man] go to the farm? comparison, appears to be much less multidimensional Will sitting in that group give us our daily bread for and more contingent on their status as providers and lo- today?� And that becomes another source of conflict cal economic trends. between the man and woman today. The frequent uneven movement—where women gain a And in another Liberian town, Harper, where men also sense of power and freedom and men feel stagnant or reported extensive disempowerment, they talked about see minimal movement—may have detrimental effects on how, since the war, women have taken over leadership the community as a whole. When men backslide and feel positions at the local university and in the local market, insecure, or perhaps even when they remain content not while jobs that used to provide many local men with to climb, this may impede the functioning of local institu- good and reliable income, such as at the port or with tions and slow relaxation of and change in gender norms. logging companies, have not been recovered. Like many Alternatively, when men see desirable outlets through urban communities in this sample, the men of Harper which to exercise agency, it may be easier for them to and Greenville feel voiceless and trapped, but women recognize and welcome women’s changing roles. And say they are finding their voice and building better and it is in this sense that men’s and women’s capacities to freer lives for themselves. exercise agency in a community interact and together shape the prospects for the relaxation of gender norms. 4. Men’s and women’s interdependent In most communities, we observed a gradual process agency and gender norm change of norm contestation and negotiation. Women’s aspira- On Norms and Agency tions are changing, but they are not necessarily finding The ladders highlight the fact that men and women re- promising openings for acting on them. In Ba’adan cen- counted few differences in what drives their ability to ter women felt empowered, but their stronger agency gain more power and freedom in their lives. They most continues to be heavily bound by traditional roles for often stressed economic initiatives, although they ac- their gender. Moreover, as explored in chapter 5, wom- knowledged that attitudes, behaviors, and education en who take on new public roles in very constrained en- 93 See, for instance, Petesch (2011), Menon and Rodgers (2011), Bouta, Frerks, and Bannon (2005); Bop (2001), Meintjes (2001), El-Bushra (2000), and Sørensen (1998). Petesch (2012) elaborates on the four conflict-affected countries in this sample in more detail than was possible in this study. 121 vironments may not necessarily derive greater agency strong enough to approach a man with great power. like the women in Caedir-Lunga. Power is circumscribed by the norms of social interac- tion, as much as by physical, economic, or political might. In the eight communities94 similar to Jaipur and Malan- Jaipur’s most powerful and free woman “takes initia- gachilima, however, where men are climbing the ladder tive and responsibility for community concerns,� and is in large numbers along with the women, we see gender “trusted and looked up to by other women.� Women on norms relaxing quickly and local economic, political, and Jaipur’s top step are also able to interact with the com- civic institutions becoming more accepting of women’s munity elite and access public resources. In other words, participation and leadership. These sample communities as women find ways to climb their ladders, more and with extensive twin climbing most often featured both more gain entry into and benefit from the formal spaces sexes finding better livelihoods, becoming more active of their communities. And in doing so, the attributes of in local economic organizing, and women entering lo- greater power, identified by the focus groups, suggest cal politics. These changes may send hopeful signals to that the presence of these women may help shift infor- other men and women about the potential for them to mal governance capacities toward more idealized nor- climb, and hence these contexts more reliably demon- mative codes of conduct. strate rapid and significant institutional transformations in gender equality on the ground. Another crucial point is that a woman on the top rung can slip down if she “loses the trust of the people in the In his work on systemic shifts in inequality at the nation- community,� or “stops communicating with people lower state level, Tilly describes how the rhythm of social on the ladder� or becomes arrogant. Leaders who dis- inclusion and democracy building can shift, almost un- play less than best behaviors are likely to be sanctioned predictably, from slower events of individuals breaking by their communities. But this fundamental mechanism down barriers to exclusion (and only limited numbers of for institutional accountability locally is far weaker in people climbing the ladder) to a much more rapid “cate- excluding environments. Where women can gain entry gorical phenomena� where whole social groups suddenly and influence, local institutions need to become more find pathways to full citizenship in their societies (2007a, responsive to serving the entire public good and not just 64–70; 2007b, 35–50). Similar processes seem to be at the male half, or a small group of elites. play at the micro level. It may be helpful to conceive of extensive twin climbing seen in the focus groups as “in- We note that women’s and men’s self-help groups and novative� moments in the lives of the communities, mo- economic collective action are also present in the com- ments when more democratic markets, politics, and civic munities with extensive twin climbing. The power of action become reinforcing—with greater gender equality civic networks in forging more inclusive societies cannot as a core driver of this shift in institutional functioning. If be underestimated, especially their role in increasing this is the case, there may be potential for more holistic women’s empowerment. Appadurai’s (2004) research policy designs that can buttress both men’s and women’s on organizational efforts among slum-dwelling women in agency and support their communities to make a shift to Mumbai, India, shows dramatic shifts in women’s mind- the more inclusive institutional equilibrium reported by sets due to the solidarity and awareness-raising of their study participants in Jaipur and Malangachilima. organizations. When oppressed social groups see that the structures of their subordination are not necessarily On Norms and Agency Women’s and men’s accounts of the processes that in- immovable, they seem able to take bolder actions to help crease their agency hinted at why their community insti- themselves—finding jobs and earning income, becoming tutions may function better with more gender-inclusive active in civic organizations or politics, claiming a voice participation and leadership. The power holder perched in their households—in ways that before were inconceiv- on the top step of the men’s ladder in Malangachilima able. The women can then help one another to escape “loves people� and does not discriminate; whereas the the bottom steps and marshal ways to mobilize and ac- women on the top step remarked that they are now cumulate assets and capabilities. 122 94 The communities with extensive twin climbing reside in Bhutan, India (three), Liberia, Peru (two), and Tanzania. Half are urban, half rural. References ———. 1966. La Vida: A Puerto Rican Family in the Culture of Poverty—San Juan and New York. New York: Ran- Agarwal, B., and P. Panda. 2007. Toward Freedom from dom House. Domestic Violence: The Neglected Obvious. Jour- Lomnitz, L.A. 1977. Networks and Marginality: Life in a nal of Human Development and Capabilities 8 (3): Mexican Shantytown. New York: Academic Press. 359–88. Mahmood, S. 2001. Feminist Theory, Embodiment, and Alkire, S. 2009. Concepts and Measures of Agency. In the Docile Agent: Some Reflections on the Egyptian Ethics, Welfare, and Measurement, Vol. 1 of Argu- Islamic Revival. Cultural Anthropology 16 (2): 202–36. ments for a Better World: Essays in Honor of Ama- Meintjes, S. 2001. War and Post-War Shifts in Gender rtya Sen, ed. K. Basu and R. Kanbur. Oxford Univer- Relations. In The Aftermath: Women in Post-Conflict sity Press. Transformation, ed. S. Meintjes, A. Pillay, and M. Tur- Appadurai, A. 2004. The Capacity to Aspire: Culture shen, 63–77. London: Zed Books. and Terms of Recognition. In Culture and Public Ac- Menon, N., and Y. van der Meulen Rodgers. 2011. War and tion, ed. V. Rao and M. Walton, 59-84. Stanford, CA, Women’s Work: Evidence from the Conflict in Nepal. USA: Stanford University Press. HiCN Working Paper 104. Brighton, UK: University Bop, C. 2001. Women in Conflicts: Their Gains and Their of Sussex, Institute of Development Studies, House- Losses. In The Aftermath: Women in Post-Conflict holds in Conflict Network. Transformation, ed. S. Meintjes, A. Pillay, and M. Tur- Moser, C. 2009. Ordinary Families, Extraordinary Lives: shen, 19–34. London: Zed Books. Assets and Poverty Reduction in Guayaquil, 1978– Bouta, T., G. Frerks, and I. Bannon. 2005. Gender, Conflict 2004. Washington, DC: The Brookings Institution. and Development. Washington, DC: World Bank. Narayan, D. 2002. Empowerment and Poverty Reduction: Dempsey, N. 2010. Stimulated Recall Interviews in Eth- A Sourcebook. Washington, DC: World Bank. nography. Qualitative Sociology 33 (3): 349–67. Narayan, D., and P. Petesch. 2005. Moving Out of Pover- El-Bushra, J. 2000. Transforming Conflict: Some ty: Understanding Freedom, Democracy, and Growth Thoughts on a Gendered Understanding of Conflict from the Bottom Up—Methodology Guide. Washing- Processes. In States of Conflict: Gender, Violence, ton, DC: World Bank, Poverty Reduction Group, and Resistance, ed. S. Jacobs, R. Jacobson, and J. Poverty Reduction and Economic Management Net- Marchbank, 66–86. London: Zed Books. work. Epstein, T.S., A.P. Suryanarayana, and T. Thimmegowda. Newman, K.S. 1999. Falling From Grace: Downward Mo- 1998. Village Voices: Forty Years of Rural Transforma- bility in the Age of Affluence. Berkeley, CA, USA: Uni- tion in South India. New Delhi: Sage. versity of California Press. Fishburn Collier, F. 1997. From Duty to Desire: Remaking Nolan, B., and R. Erikson. 2007. Intragenerational Income Families in a Spanish Village. Princeton, NJ, USA: Mobility: Poverty Dynamics in Industrial Societies. In Princeton University Press. Moving Out of Poverty: Cross-Disciplinary Perspec- Ibrahim, S., and S. Alkire. 2007. Agency and Empower- tives on Mobility, ed. D. Narayan and P. Petesch. Vol. ment: A Proposal for Internationally Comparable 1. New York and Washington, DC: Palgrave Macmillan Indicators. OPHI Working Paper 04. Oxford: Univer- and World Bank. sity of Oxford. Rogers, E. 1995. Diffusion of Innovations. New York: Free Kabeer, N. 1999. Resources, Agency, Achievements: Re- Press. On Norms and Agency flections on the Measurement of Women’s Empower- Perlman, J. 2010. Favela: Four Decades of Living on the ment. Development and Change 30: 435–64. Edge of Rio de Janeiro. New York: Oxford University Kabeer, N. 2001. Reflections on the Measurement of Press. Women’s Empowerment. In Discussing Women’s Petesch, P. 2012. The Clash of Violent Conflict, Good Empowerment: Theory and Practice. Sidastudies 3. Jobs, and Gender Norms in Four Economies. Back- Stockholm, Sweden: Novum Grafiska AB and Sida, ground paper for World Development Report 2013, 17–57. World Bank, Washington, DC. Lewis, O. 1961 The Children of Sánchez: Autobiography ———. 2011. Women’s Empowerment Arising from Violent of a Mexican family. New York: Random House. Conflict and Recovery: Life Stories from Four Mid- 123 dle-Income Countries. Washington, DC: USAID. Petesch, P., C. Smulovitz, and M. Walton. 2005. Evalu- ating Empowerment: A Framework with Cases from Latin America. In Measuring Empowerment: Cross- Disciplinary Perspectives, ed. D. Narayan, 39–67. Washington, DC: World Bank. Ragin, C. 2001. Fuzzy Set Social Sciences. Chicago: Chi- cago University Press. ———. 2008. Redesigning Social Inquiry: Fuzzy Sets and Beyond. Chicago: Chicago University Press. Small, M.L. 2004. Villa Victoria: The Transformation of Social Capital in a Boston Barrio. Chicago: University of Chicago Press. Sen, A. 2002. Development as Freedom. New York: Anchor. Sørensen, B. 1998. Women and Post-Conflict Reconstruc- tion: Issues and Sources. WSP Occasional Paper, no. 3. Geneva: United Nations Research Institute for So- cial Development. Schrock, D., and M. Schwelbe. 2009. Men, Masculinity, and Manhood. Annual Review of Sociology 35: 277–95. Tilly, C. 2007a. Poverty and the Politics of Exclusion. Cit- ed in D. Narayan and P. Petesch. 2007. Moving Out of Poverty: Cross-Disciplinary Perspectives on Mobility. Vol. 1. New York and Washington, DC: Palgrave Mac- millan and World Bank. ———. 2007b. Democracy. New York and Cambridge, UK: Cambridge University Press. On Norms and Agency 124 CHAPTER 5 Structures of Opportunity and Structures of Constraint “ Education and development of the country also matter. If the country is rich, � the people on the bottom can rise to a higher level. If [not], people are depressed, unemployed, homeless, and hopeless. — Urban man, Dobrowice, Poland B oth the women and men in our study recognized gender norms, however, remain central because they so that certain factors and processes under their greatly shape aspirations and access to opportunities. control can enhance their capacity to shape their lives. This chapter ponders factors that are also The traits associated by the focus groups with the lad- vitally important for exercising agency, but over which der steps reveal that, as more and more women move they typically have far less control: local level institu- up, a share of them are penetrating—perhaps for the first tions in the public sphere of their communities, and the time in their communities—formal institutions and labor normative climate for inclusion and accountability in markets. For instance, in urban Kart-e-Bakheter (Par- these arenas. It is in the creation of structures of op- wan), Afghanistan, the women pointed out that on step portunities that are open and equal for men and women 1 of their ladder of power and freedom, a woman is not where policies have largely focused. And while changes working for pay (or cannot), but on step 2 she may be have been implemented and more opportunities cre- involved in “tailoring, embroidery, weaving carpets, tend- ated –that have translated to changes in endowments, ing home poultry.� On step 3, a woman may be employed economic participation, and aspirations, much hard in a more remunerative job outside the home (e.g., earn- work remains. ing income as a nursing aide or from livestock). And a woman on their top step is usually well educated, can be “Progress toward gender equality,� in World Develop- “nominated even for election,� and “may be a doctor or ment Report 2012 (World Bank 2012, 330), “entails shifts teacher, and has a good economic position. They have On Norms and Agency towards a new equilibrium where women have access to freedom and power.� Even in more traditional societies, more endowments, more economic opportunities, and such as Kart-e-Bakheter, greater participation of women more ways to exercise their agency—and where this new in public spheres is displayed on the ladders. The basic arrangement becomes the dominant order.� Chapter 5 question for this chapter is what role do institutional and draws on our dataset to probe the role that community normative factors play in shaping perceptions of agency. characteristics play in constraining and enabling agency. Do these three pieces move together? Or separately? Our initial focus is on local labor markets and then public and civic institutions, and especially how gender equal- Gender norms influence not only women’s (uneven) ity is reflected in legal frameworks. The constraints of capacities for exercising agency—explored in previous 125 chapters—but also the extent to which local institutions tions of whether and under what conditions women may welcome women’s participation and leadership.95 This seek a job, as well as the types of work that they con- social context, which Kabeer (2001, 47) argues can be sider desirable and available to them. more aptly understood as “structures of constraint,� means that individual women acting alone are unlikely In the first part of this chapter, we show how normative to challenge and change the conservative elements of change is impeded by numerous factors, including the their local institutions: “The project of women’s empow- interplay between preferences and the local opportuni- erment is dependent on collective action in the public ties that combine to funnel women into less productive arena as well as individual assertiveness in the private.� areas of the economy. For other reasons (explored later), While we observe many signs of individual women cross- women’s economic participation in even dynamic mar- ing gender boundaries throughout our dataset, signs of kets may not be enough to challenge and change gen- effective collective action (e.g., the women of Jaipur and der structures meaningfully. In the final sections of the Malangachilima in chapter 4), and institutional change chapter, we look at how better laws, political leadership, in the gender order are relatively rare. This chapter and community organizing provide other potential out- looks at more typical communities, where local opportu- lets for women to gain more power in their communities. nity structures are less open to women’s initiatives and Yet, our findings reveal that norms act as important con- chiefly appear to reinforce, rather than alleviate, gender straints on these routes to empowerment as well. inequalities. Gender norms in these environments prove more resistant to change. 1. Community factors that fuel agency The operations of markets and other institutions reflect local gender norms. The interaction between beliefs Certain community-level conditions and trends, gath- and attitudes, on one hand, and women’s participation ered from information from local key informants,97 seem in the labor market, on the other, is tamed partly by the to be more conducive to men’s and women’s empower- potential benefits women can obtain for themselves ment. To frame the comparative analysis, we sorted the and their families, as well as by the norms within their focus groups’ ladders into three categories, according to households.96 Escriche, Olcina, and Sánchez (2004) note the extent of the mobility they reported: high, moder- the obvious changes in women’s ability to work (outside the home) and in gender roles in the last 30 years—also seen in the decade changes study participants reflect on during the focus groups—including changes in atti- FIGURE 5.1: AVERAGE MOBILITY INDEX tudes toward women in the labor force. But we cannot really say whether these changes are due to adjust- Low mobility ladders ments in preferences and norms governing gender roles or due to a different socialization process for younger women and the transfer of preferences and aspirations Medium mobility ladders from parents to daughters. It is not only the household and the market, but the overall structure of opportuni- ties and normative climate in a community that helps or High mobility ladders On Norms and Agency hinders women’s and men’s ability to negotiate gender- allocated roles. -40% -20% 0% 20% 40% 60% 80% Average mobility index Economic participation can be a strategic move to give Women Men women voice. But our dataset makes plain that, while Note: The mobility index is calculated based on the difference communities may be growing or shedding jobs, the im- between the mean step now and 10 years ago on each ladder. The distribution was divided into three groups according to the pact of economic trends on women’s access to income average distribution within each country set of communities. We opportunities and their sense of empowerment is far used a tercile distribution function that classi�ed the ladders based on the distributions of the mobility index for each country. 126 from linear. Social and cultural factors affect expecta- ate, and downward. Figure 5.1 shows the average mobil- presented in chapter 4. One important explanation for ity for the three sets of men’s and women’s ladders, and why so many urban men feel disempowered or in limbo table 5.1 presents the ladder distributions in each set by is that difficult local economic conditions provide few job sex and location. opportunities for them. This—combined with their views of what constitutes an appropriate “male� job (which al- At first glance, the average mobility indexes are quite lows them to fulfill their provider role)—prevents men similar for both men’s and women’s ladders in all three from adapting more easily to a changing environment. categories. However, when looking at the specific num- ber of ladders located in each mobility set, more wom- Figure 5.2 compares local economic data with the lad- en’s ladders are in the high mobility group than men’s, der mobility categories. The focus groups’ ladders that particularly urban women’s; and more men’s ladders, show high mobility (more empowerment) correspond also heavily urban, are in the downward mobility group. to communities that were more prosperous on average These patterns underpin the polarized urban ladders over the past decade and have lower rates of poverty Table 5.1: Number of ladders in each mobility category by sex and location High mobility ladders Moderate mobility ladders Downward mobility ladders Men’s Women’s Men’s Women’s Men’s Women’s Urban 4 27 21 14 26 9 Rural 12 14 17 14 17 14 Note: Results from 189 adults focus groups. The 5 Sudanese women’s ladders are missing from this analysis because they did not conduct the sorting exercise. FIGURE 5.2: MOBILITY ON LADDERS IN MORE PROSPEROUS AND POORER COMMUNITIES 2.8 60 1 = more prosperous; 3 = less. 2.7 50 2.6 Share of poverty 2.5 40 2.4 30 2.3 20 2.2 2.1 10 2 0 Women’s ladders Men’s ladders Women’s ladders Men’s ladders High mobility Moderate mobility Downward mobility On Norms and Agency Note: Data from 189 adult focus groups. The 5 Sudanese women's ladders are missing from this analysis because they did not conduct the sorting exercise. 95 See Feher and Hoff (2011) for a thoughtful discussion of the literature on circumstances under which norms and preferences may be more mal- leable than often recognized. 96 Also noted by Duflo and Udry (2003), Guiso et al. (2006), and others when looking at household bargaining and female labor force participa- tion. Both studies highlight the impact that beliefs and preferences have on economic behavior and thus economic outcomes. Also see Munshi and Rosenzweig (2006) for an example of how traditions surrounding men’s caste-based occupational networks in Mumbai limit young men’s returns from education more than young women’s. 97 Most of the key informants interviewed for each sample community were local officials, elected leaders, or civic leaders, although business leaders and teachers sometimes worked with our teams and completed the surveys, too. 127 In economic sociology, markets are as cultural as any FIGURE 5.3: RATES OF WOMEN WORKING other aspect of social life, and norms and values are a FOR PAY WITH TWIN CLIMBING AND FALLING central part of their constitution and functioning (Zel- izer 2010; Wherry 2012; Spillman 2012). Markets are not 2.4 1 = many work for pay, 3 = few do gender neutral; they are embedded in societies and take 2.2 up (and reflect) their specific gender norms. Thus, when 2.0 assessing women’s agency, it is important to bear in mind 1.8 that women’s knowledge and evaluations of their actual 1.6 chances in local labor markets are intimately shaped by 1.4 the hierarchy of values to which they adhere, as well as the values of their community, family, local leaders, and 1.2 employers. In other words, whether jobs are plentiful or 1.0 not in local markets, they may or may not be open to Twin Women Twin Men climbing, climbing climbing, falling women falling women or women may not see existing opportunities men falling Now 10 years ago as appropriate for them. Norms and values act here not Note: Data from 184 adult focus groups. The 5 men's and 5 women's only as external secondary factors brought in as part of ladders from Sudan are missing from this analysis because the the black box of individual preferences, but are funda- women's groups did not conduct the sorting exercise. mental to, for example, how employers define potential or ideal employees. (as reported by key informants). We anticipated these findings for the men’s ladders, but were surprised that World Development Report 2012 documents the rapid favorable economic conditions and lower poverty levels expansion of women participating in the labor force in proved equally important for women’s perceptions of recent decades together with a more slowly declining empowerment as well. We also find that men’s and wom- occupational segregation by gender across the world. en’s mobility is higher in communities where women cur- Women continue to be over-represented in less produc- rently serve as local elected leaders. The significance of tive sectors of the economy and in positions of lower the presence of civic groups and empowerment trends authority (see Tzannatos 1998). To explore how gender is more ambiguous. norms may be affecting economic agency and the de- sirability of particular local jobs on offer, we asked the The narratives in our dataset reflect the global trend of focus groups to reflect on differences in men’s and wom- women’s increasing participation in the labor force in the en’s economic roles and capacities and key challenges in past decade. The rate of women working for pay is also their labor markets. higher in the communities where the focus groups of both sexes reported greater gains in power and freedom Discrimination in general is a problem for both sexes in (figure 5.3). the study, although women see gender discrimination against them as a stronger barrier than men do. It is a fac- tor that influences which jobs women and men can get. 2. Whose jobs? Men reported discrimination against men and against women, but both women and men reported more dis- “ On Norms and Agency crimination against women. In Kalahandi District (Andhra More women are making business now and can do � Pradesh), India, a woman explained simply, “We are fe- anything for themselves. male. That is why we get lower wages.� It is relevant that — Rural women’s focus group, Suakoko District, Liberia young men and young women have the strongest views “ regarding discrimination. Women study to be teachers or pedagogues because 128 they love children. Men often inherit a private business. — Urban women’s focus group, Belgrade, Serbia � Opportunities for jobs vary significantly by gender, in fact, when asked to sort the “best� and “worst� ways to earn a living for workers their localities the focus groups, across countries and communities, identified about 50 find jobs easier, while young women had the opposite percent of the jobs they mentioned as gender-specific opinion.98 “I browse through the newspaper ads in Bel- ones (either men’s or women’s jobs). Individual accounts grade looking for work, and I see more demand for girls of the factors that determine their ability to get a job re- to work in cafés and pizzerias, boutiques, counters, ev- flect these gender differences. Normative perceptions, erywhere—for them it is easier,� remarked a young man as well as predominant gender roles within a society, de- from Sumadija District, Serbia. fine a job as male, female, or gender neutral. And not only are jobs gender specific, but the skills jobs require Figure 5.5 shows how women and men recognize that are as well. Men in the study, for example, believe that there is more discrimination in the market against wom- having good connections is important, and they mention en than against men. But a significant number of views this more than women (figure 5.4)—as well as give more also points to some discrimination against men. Women relevance to information sources about job opportuni- more often perceive that in line with this discrimination, ties. Women depend more on the demand for gender it is easier for men to find a job. Men are more evenly specific abilities such as soft skills for “female� jobs (such divided over which gender has better job prospects. as education, nursing or retail). The ability to balance work and family life and having previous job experience “Women excel in health, education, and housekeeping. matters more for women than for men. But both equally But men excel at everything else, like engineering and po- recognize that education is a central factor at the time lice,� suggested a woman in Rafah, West Bank and Gaza. of finding a job. In Karta-e-Bakheter, Afghanistan, a woman said that “tai- loring, embroidery, and carpet weaving are for women, Gender discrimination –the most relevant factor for and construction, metal work, carpentry, and jobs in the women to find a job- plays a role in the type of jobs they government are for men.� In Monrovia, Liberia, where can find, because it influences perceptions of women’s gender norms are more relaxed than in West Bank and abilities and the opportunities that are open to them. Gaza and Afghanistan, women are servers at “drinking Again, jobs requiring “feminine� skills in social relations— spots,� while men purchase supplies and handle the cash. better at conversation, more attentive—are deemed bet- In Lambayeque Province, Peru, where the tourist indus- ter for women, but jobs involving authority, technical try is booming, men are the chefs in the city’s world-class knowledge, strength, or public safety often remain solely restaurants that cater to the tourists, while women, who for men. Young men in the sample felt that women can have less access to training and opportunities, are some- FIGURE 5.4: MOST MENTIONED FACTORS AFFECTING ACCESS TO JOBS IN THE LOCAL LABOR MARKET, ACCORDING TO THE MEN’S AND WOMEN’S FOCUS GROUPS Information sources of job opportunities (family) Work schedule to balance household needs Previous work experience Information sources of job opportunities (not family) Physical strenght Connections On Norms and Agency Soft skills (attitudes and behaviors) Hard skills Transport and distance Education and other credentials 0 0.05 0.10 0.15 0.20 0.25 0.30 Share of total mentions Men Women Note: Data from 388 adult and young adult focus groups. 98 This difference is not as stark among older adults in the study. 129 FIGURE 5.5: PERCEPTION OF DISCRIMINATION BY SEX IN THE LABOR MARKETS (MENTIONS) 140 120 Frequency of mentions 100 80 60 40 20 0 Discrimination Discrimination Easier for men Easier for women against women against men to �nd work to �nd work Note: Data from 388 adult and young adult focus groups. Women Men times cooks in smaller establishments and catering com- because it required women to “work long hours on the panies. In Tandjouaré Préfecture, Togo, a young woman street and deal with different types of people who may explained that jobs requiring “physical force [are] for treat them indecently.� They mentioned that tea sellers men, and courage and patience are for women.� under age 40 can be treated like prostitutes and even older women are harassed as well. Unmarried women Social norms, and the institutions that support them, de- doing domestic work for other families risked sexual termine to a large extent the types of jobs, and their par- advances that “deprived a girl of modesty.� Women’s ticular working conditions, that women and men can ap- jobs cleaning schools and selling in telecom centers ply for, as much as the views the communities hold about and shops are also deemed undesirable in Sudan, but specific jobs (Jutting and Morrison 2005). These views are considered acceptable, good-paying jobs for wom- take several forms. A detailed look at the data suggests en in many sample communities in Sub-Saharan Africa three specific ways that norms and jobs are in constant and elsewhere. Similarly, call-center jobs in Hyderabad dialogue for women. (Andhra Pradesh), India, are not well considered, in spite of the potential economic gains: “Actually, there is good First, we see that the women must negotiate their work money in call-center jobs, but society doesn’t consider choices around the different constraints and norms gov- this a decent job. Women engaged in such jobs are not erning a woman to avoid female-inappropriate work and considered respectable because it has night shifts and potential risks or difficulties—verbal, physical, and sexual the workplace is full of young men who have fat sala- harassment, physical injury, and more. (Most of these ries.� In these communities, women approach the labor norms follow the attributes of a good woman, advanced market with already constrained choices and miss out on in chapter 1.) In the process of searching for work, wom- good opportunities for employment. en in the study not only considered salary and status (which everyone takes into account) but also had to fac- In more conservative communities, such as those in West tor in the potential threats to their moral integrity and Bank and Gaza, Yemen, and Afghanistan, for example, their family’s honor. In communities that still tightly hold where women’s physical mobility is restricted, it is not On Norms and Agency to more traditional norms for women and gender roles,99 surprising that the women are less attractive as potential some jobs considered inappropriate or immoral for a employees. Similarly, women who have the qualifications woman are perfectly acceptable for men. For example, for a job, but who need their husbands’ or father’s con- any job in trade, which involves dealing with strangers, sent to work, are considered less reliable, even if more or in services that require interaction with men, or a job qualified than male colleagues. Sometimes, women from with night hours, often are off limits to women. the sample communities where no childcare services available—for example, in the Dominican Republic—self- In all the Sudanese focus groups, one of the worst jobs censored their work choices—as much as employers do— 130 for women was hawking tea, coffee, and homemade food when faced with the possibility that household demands may compete with their work hours. The whole set of so- to work to feed their children, norms relax and adapt. cial relations that makes up the labor market is affected In Serbia, women do construction work under the same by the gender norms of a community, and not only the conditions as men, rain and snow. And in Liberia, oppor- part pertaining to gender relations in the household. tunities opened up for women during and after the war, and they began running businesses, farming, and selling Second, a good job for a woman has to be flexible. How cash crops. women internalized or appropriated the norms of their communities shows in the way they described what con- Clearly the choices women make when it comes to work stitutes a desirable job for women. These jobs tend to vary, depending on several factors, and so does the way mesh well with domestic work and childcare in terms of they negotiate social norms. For some women in the location and time flexibility, and do not involve physi- study, taking a particular job is a response to certain cal strength. “Safe and easy jobs are best for women� restrictions in their choices; for others, it is an expan- pointed out a woman from Jaipur (Odisha), India, mean- sion of their ability to choose. In rural areas, as Agarwal ing the best jobs are not physically demanding, are close (2003) reports, nearby farm work, household gardening, to the woman’s house, and allow them to devote time to or small livestock tending can help women increase their housework. agency because the jobs are accessible, manageable with their domestic work burden, within their mobility Flexibility is one of the main reasons why being a small restrictions, require less education, and do not need as- business owner is frequently mentioned by the focus sets or investment. Some women feel motivated to work groups as a suitable job for a woman. It provides some to reach specific aspirations or goals. All of them are act- freedom in the allocation of time and does not require ing as economic agents and expressing their preferences traveling long distances away from home. Women’s to earn income, perform their domestic (gender) roles, search for flexibility solves both self-imposed limits (to safeguard their reputation, and more (Kabeer 2000). accomplish the many household tasks they have more easily) and ways to adhere to social norms. While some women find jobs that conveniently complement their 3. It takes a village: Local economic domestic duties, others face demanding physical work dynamism and empowerment in fields or factories in addition to their housework and “ are aware of the price they pay. “Selling coconuts is the I worked in Zastava [a factory] for 24 years and then worst because women have to climb coconut trees, cut I was fired. I automatically lost everything. I lost any � the nuts, husk them, and then carry them to the market freedom and power I had. Everything was lost. to sell. Women do all the hard work and men come home — Urban man, Kragujevac, Serbia and find dinner ready and served to them each evening,� “ pointed out a young woman from East Sepik Province, Unfortunately, the economic situation has a big effect Papua New Guinea. on an individual’s personality. … If [the man] is able to Third, economic need is stronger than traditional norms, so the norms have to adjust. In East Jakarta, Indonesia, selling carpet used to be only a man’s job, particularly provide for his family, his personality will be stronger. — 45-year-old urban man, Old City, Hebron, West Bank and Gaza � On Norms and Agency because it requires travel around the city. But today, be- ing a female itinerant carpet seller is considered a good Having a job that offers the possibility of generating an in- job, although “it wasn’t a few years ago. But because of come and accumulating assets is at the core of both men’s poor living conditions [and economic stress], it must be and women’s accounts of what helps them gain power done—by men and women,� noted a young woman. It is and freedom. Economic well-being is one of the main obviously better if a woman can sell from home and take attributes they associated with being on the top step of care of the house, but when both wife and husband have the ladder. For men, having a job so they can be a good 99 Specifically West Bank and Gaza, Yemen, and Sudan, but also Umlazi township A in South Africa, Hyderabad in India, and East Jakarta, Indonesia. 131 provider is the identity of their gender and meets the ex- on men’s opportunities? The answers vary, but it is clear pectation that they will be a good husband. For women, a that normative views are shaping women’s sense of their job and its income are means to increase their bargaining choices, along with the stressful market conditions. power inside the household. But men and women alike are largely dependent on the structural conditions of Among communities that have seen less growth, econom- their local economies in the pursuit of employment. ic need is forcing a relaxation of gender norms. In the least dynamic communities where everyone is struggling, The links between empowerment, norms, and economic especially men who cannot continue providing for the dynamism are not clear in this study. As women become household, women spoke of being thrust into the mar- more economically active, they gain a sense of empow- ket to earn an income. They end up competing for and erment. The exceptions are contexts where norms are working in whatever jobs are available. The market reg- deeply discouraging or desirable job choices extremely isters fewer differences between genders when needed scarce. Here we provide additional evidence from our workers are unskilled and eager to work. There is little study that women perceive empowerment from their self-selection out of a job, except for some clear gender economic initiatives, regardless whether their economies differences, such as physical strength (from which norms are growing rapidly or moderately, or shrinking. Looking were derived), but even that does not keep women from at market trends described by the focus groups, how- working in construction or agriculture. In agriculture, men ever, gender norms do not keep up with women’s chang- do the jobs that require plowing or heavy lifting, while ing roles. This variability means that perceived changes women seed, weed, harvest, and pack. Choices of jobs, in empowerment and gender norms cannot be assumed according to the focus groups, are more often constrained to be complementary forces or direct and universal con- by availability, and job selection may not be strategic or sequences of economic dynamism. based on a sense of self-efficacy or empowerment. To explore these interactions more fully, we examine In Serbia, Poland, and Moldova (as discussed in other some weaker local economies in the study against a chapters), women’s presence in the labor force has a lon- more dynamic one to see what happens to local norms ger history; nevertheless, it too segregates jobs by gen- around women’s paid work. In these contexts, the varia- der, similar to contexts where women’s economic partici- tion in the stringency of norms does not correspond with pation is more recent. In the East European communities, the state of the economy in the community or with how as elsewhere in the sample, women opt for careers in women are able to handle both norms and employment the humanities and teaching, while men go into sciences, opportunities. engineering, and business. “Women are less capable in business than men,� and “women study to be teachers a. When choices disappear or stay poor or pedagogues because they love children more� is the view among Serbian women in Belgrade. Yet, men and “ women from this region voiced growing disenchantment There is no work. Nobody is doing anything. about which jobs can be found and kept in their weak If you find a job, you cannot hold it. � local economy. The type of job may vary, from domestic And the jobs which you do find are the worst ones. work for young women to factory work for young men, — Urban young woman, Kragujevac, Serbia but the feeling seems pervasive that opportunities are On Norms and Agency limited at best. The traditional gender division between a household’s productive and reproductive tasks assumes that there In some communities, like rural Floresti District, Moldo- are sufficient economic opportunities for men to find va, opportunities are deteriorating for both genders. The paid employment or other productive activities to sup- community had good stable jobs for both women and port the household. But reality is not always that accom- men that disappeared when the greenhouses and dryers modating. What happens to women’s agency in weak closed, the state-owned enterprises were liquidated, and economic situations where there are not many jobs avail- some private companies went out of business. Floresti 132 able, and where lack of economic dynamism takes a toll District is poorer than it used to be, and many people grow vegetables and raise small livestock to survive. Per- choice [with a distant job], he will not forbid it.� These haps not surprising, women reported disempowerment women are not abiding by traditional norms, but are they on their ladder. Men, by contrast, said they have moved increasing their agency in the process? Not always, as off their bottom step in large numbers by finding stable they made clear when discussing their ladder of power work locally or with temporary migration, which has al- and freedom. To move up, they needed either to start a lowed them to “find a good wife� and “stop drinking and small business or “find a more prestigious job.� start taking care of the family.� In line with their gender role, these factors allowed men to escape the bottom Even though these women are well educated, the lim- step in this village (box 5.1 presents the particularly dif- ited desirable opportunities seem to be holding back ficult case of Roma men in Serbia struggling to overcome more rapid change in gender norms. Some young wom- a harsh economic and social climate). en expressed progressive views in terms of aspirations of freedom and productive use of their acquired skills: For Floresti District’s women, by comparison, the local “I studied and if I find a job opportunity to use my edu- jobs still on offer strain their gender roles, identities, and cation, then I don’t want to stay home just taking care of expectations. Their discouragement is high, although they my children.� But as so commonly found in our dataset, face little gender discrimination in job opportunities or in this view sits side by side with other women’s opinions equal income. For example, a recent renovation of a lo- that highly value the mother’s role. Nor have more flex- cal school employed about 30 people, of whom 12 were ible norms changed young men’s mindsets: “I consider it women; all received equal pay regardless of the specific better for the entire family if the woman stays home and tasks performed. Moldovan men and women work equal- takes care of the children until they are 3 to 8 years old.� ly in good and bad jobs and get similar earnings. Their dis- Although some local women have been working in jobs cussion about local jobs portrayed men as truck-drivers, outside the home for decades, norms remain parked metalworkers, and locksmiths, while women clean and between community disapproval and endorsement: cook, as well as work in construction and agriculture. “There are some people who say that a working mother has abandoned her children�; and “it may be, though, And while some gender specialization is evident in the that when a mother leaves her child and gets a job, then jobs the focus groups listed, women are competing for she is in a difficult economic situation and that is what and taking men’s jobs that require strength, despite al- she needs to do.� most universal preference that women do less arduous work. For example, men clearly saw strength as a factor Sometimes the need for cash in a household and the de- that should favor men in construction work: “Jobs that mand for unskilled workers in nearby markets together imply physical work are harder and men are better at unlock doors to women’s economic participation. They those kinds of jobs, for example, construction. Moreover, may even cross, and change, the boundaries of gender- men are braver and therefore take more dangerous jobs.� appropriate work, and open up job possibilities for other While women see construction as “one of the worst ways women. But in contexts where choices are limited or de- of making a living� because “it is very physically taxing� teriorating, the process of norm relaxation is slower than and pays poorly, it is nonetheless viable if it is the only in more economically dynamic communities (see box job available. 5.1). In sum, in Floresti District, where previous economic opportunities and education levels have raised expecta- On Norms and Agency Women in Floresti District defied gender differences and tions, where desirable economic options are now scarce, restrictions outright. They challenged the concern for and where old and new norms coexist, not just any job their safety and risked travelling at night if they needed will lift women off the bottom ladder step. “I didn’t suc- the job. Safety became a secondary issue when it came ceed in finding a job I wanted,� lamented a 19-year-old to getting a job, as some young women noted: “In our woman in her focus group, “so I decided to stay home situation now, it is more important to find a good job with and take care of my children and household.� a wage that covers commuting expenses. I prefer to work closer to my home in order to save time and money, and Hato Mayor, a secondary city in the Dominican Re- my husband prefers this for me too. But if I don’t have a public, is another local economy that is losing jobs. 133 Box 5.1: The Roma of Kragujevac: Where disadvantages and strict norms overlap and trap Kragujevac is one of the oldest Roma settlements in Serbia. Small in size, its population is dense with a high unem- ployment rate. Multigeneration families have at least three children and 95 percent of the inhabitants are poor. Many people have been laid off in the last decade in the economic downturn. Men find it hard to reach employers and get a job. They mainly engage in manual labor and agricultural and seasonal work; collect secondary raw materials, or work for the local garbage disposal company. Some find work through the National Employment Service and personal contacts. Women work to a lesser extent than men: a small number of them clean other people’s houses, work for the municipal gardening company, or pick fruit in season. Most women do not receive fixed salaries. The few girls who finish school actively seek jobs, but have difficulties finding regular work. Kragujevac does not present a very uplifting panorama. “Men and women have a lot of free time because, among other reasons, they cannot find anything to do for pay,� related a young woman of Kragujevac. Even when they find a job, in their free time men tend to socialize with other men, gamble, or drink. The younger men, who have lost their motivation and aspirations, do the same. “Most young men in the community have a lot of free time. There are a lot of idle guys who do nothing; they look for work, but hope not to find it,� explained a young man. Women, as traditional, are in charge of the housework, whether they are working for pay or not. When men don’t work, they sleep late and spend their time with their friends, away from their homes. “He only comes home when he is hungry. He brings no money. How could he bring money when he does nothing? We receive child support, which is not enough, but what can we do when there are no better opportunities? Our families help us a little, as much as they can afford to,� commented one woman. Women notice that men, unlike them, find it harder to accept the loss of work, given the pressure they feel to provide as breadwinners. Even though they would like to have more work and better business opportunities, however, when faced with unemployment, men would rather be idle than do “female tasks� and contribute to the care of the household. Unemployment and poverty in the sample communi- Yet local opinions about women’s dual roles appear ties jumped when the free-trade commercial zone to be even more contested in Hato Mayor than in Flo- they depended on was hit by a down turn. Today, it resti District. The women’s focus groups support work- is difficult for workers to find stable jobs and the lo- ing mothers and their attempts to reconcile productive cal market is less active. But compared to rural Flo- and reproductive demands, but they were also aware resti District, there are enough work outlets for many of the opinion this generates in the community. When informal workers and entrepreneurs in Hato Mayor to a working mother leaves her children, people think and make ends meet. Perhaps surprising, both women and say that she is going “to prostitute herself.� Moreover, men perceived upward movement on their ladders and women reported discrimination and physical risk at their On Norms and Agency singled out their initiatives with their own businesses jobs: “Women get paid less for more work and they get and better financial management as central to these abused.� And they do not dare work at night due to un- gains. Women, for instance, said they can move up the safe streets. In their focus groups, the men were clearly ladder by opening their “own small businesses selling more conservative than the women. Despite the women’s food, making and selling sweets, selling second-hand changing roles, men made plain that gender differences clothes, and cleaning streets.� These women see them- should remain intact and women should not go after or selves as empowered and “willing to push forward� to take male jobs, for instance, working construction and contribute to their families’ well-being during the dif- driving motorcycle taxis. Younger men were as conserva- 134 ficult times. tive as older men and echoed these negative views. b. When choices grow to them, and not only available to men. They asserted that they can be police officers, lawyers, and doctors, “ Now women can go out to work and hold a high ranking job, even in the army and the police. just the same as men. They also remarked that they want to be like their fathers and “have prestige like him, and � This is a great change since our parent’s time. get [public] exposure like him.� For these young women, — Urban young man, Khartoum, Sudan domestic work and care giving are not attractive jobs. But while their hopes for the future include professional Chapter 4 looked at Jaipur (Odisha), India, a rapidly ur- careers, they also want marriage, recognizing that pursu- banizing town, and Malangachilima, Tanzania, a village ing both has a cost: “If you are a married woman, it is moving into commercial agriculture. These two commu- even more difficult [to work far from home] and it can nities were in the midst of particularly fast change, and destroy the marriage. Men cannot wait for a woman. If both women’s and men’s focus groups presented clear you are gone too long, by the time you come back, he evidence of a strong relaxation of gender norms for may have moved out to live with someone else.� These women’s public roles. Jaipur and Malangachilima give us young women are forced to hang on to this dual role of a valuable look at how quickly women’s roles can some- professional worker and proper wife to accommodate times change, when local opportunity structures support male peers who may or may not welcome their income- their economic initiatives. In most other sample commu- earning role. “Yes, it is acceptable that they have the nities with dynamic local economies, however, gender right to work,� noted a young man, “It is good because norms are more resistant and do not shift as quickly. they can assist their husbands in meeting household de- mands. In some households, you find that the man is un- In Umlazi township B, a community of 3,000 outside employed and only the woman works and supports the Durban, South Africa, a local official interviewed for the family.� Yet, another young man disagreed strongly, “It is study estimated that perhaps 80 percent of the women not acceptable because a woman needs to be at home currently work outside the home, mainly as teachers and caring for the children. Most of the time, working wom- nurses, and some in offices, retail, or the police. Just 10 en are promiscuous and don’t respect their husbands.� years ago, few women earned any income at all. “Wom- Unlike the communities that are struggling with limited en are no longer regarded as just housewives,� a young jobs and high levels of poverty, these young women at woman remarked, pointing to a shift in the predominant least can take heart in the ready availability of desirable association of women with their domestic role. In this jobs and the prospect that some men are broadminded suburb of Durban, women generally have more educa- enough to welcome this development. tion than men, with young women attaining the highest levels. Women in Umlazi township B feel empowered; As quickly as women’s public roles are changing in Um- they placed 75 percent of the women in their community lazi township B, traditional gender identities continue at the top of their ladder and characterized them as be- to frame desirable jobs. Other young women consider ing “powerful women who can afford anything. They own construction jobs as more suitable for men due to their vehicles and houses. They are single parents and are in- greater strength. And when men are employed in health- dependent. They do not wait for men to do things for care, “male nurses are discriminated against and people them. They send their children to university. They have call them homosexuals.� Gender stereotyping allocates On Norms and Agency lots of money. They have everything they need and can such jobs as nursing and office administrative work more eat whatever they feel like eating.� often to women and authority positions to men: “If the school principal is a man, the school is highly respected Young women here do not see themselves as bound by because men are known for enforcing discipline.� any restrictions when it comes to finding a job: “Nothing [prevents us from getting a good job]. In today’s world, In Ouagadougou, Burkina Faso, the market has become women fit anywhere as long as you have the right quali- more dynamic, but it is still not easy for men or women fications.� They listed highly skilled work—some conven- to find jobs. Education remains a distant objective for tionally masculine jobs—as suitable for women, desirable everyone. Some young people aspire to technical and 135 Table 5.2: Desirable and undesirable jobs in three urban communities Dynamic labor market: Middling market: Slow labor market: Umlazi township B, South Africa Ouagadougou, Burkina Faso Hato Mayor, Dominican Republic Nurse* Trader Moto taxi driver Teacher Tailor (women) Construction worker (men) Police officer Gardener (men) Retail and shop clerk Desirable jobs Lawyer Mason (men) Security guard Plumber (men) Mechanic (men) Concierge Carpenter (men) Drug dealer Pickpocket Waste picker Taxi driver Drug dealer Domestic service worker (women) Security guard Prostitute (women) Undesirable Caregiver (women) jobs Day worker in agriculture Street cleaner Drug dealer * All professions not marked were considered gender neutral by the focus groups. professional jobs, but they are not attainable by lo- ing, professional jobs as part of their future (table 5.2). In cal workers. Indeed most of the good jobs are clearly Ouagadougou, Burkina Faso, jobs also remain scarce and manual skilled labor that is highly segregated by gen- traditional norms continue to segment the labor market der: mechanic’s work, carpentry, and construction for and dampen aspirations. men; and sewing and housekeeping for women. A fe- male mechanic will be hired “only if she has real experi- In his review of the literature on youth aspirations in Af- ence and because women mechanics are very scarce,� rica, Leavy (2010) finds that higher poverty rates and the noted a man, but women “don’t have the strength to lower socio-economic status of rural communities limit carry bricks,� so they should not work in construction. the hopes of young people. In our rural sample, howev- The Ouagadougou focus groups also mentioned that er, there are village economies with enough vitality and they avoided illegal or risky jobs because they threaten growth that young people can imagine better futures for people’s honor, which appears to be very important to themselves than what is immediately available. In other men and women alike. communities, norms are relaxing and reinforcing wom- en’s economic initiatives. A case in point is Firestone The qualities of desirable and unappealing jobs, as well District, Liberia. as the determinants of access to them—norms of femi- ninity, flexibility, and need—vary when considered in the Firestone District, a rural town described by a key in- On Norms and Agency context of specific communities. The Umlazi township formant as poorer than a decade ago, is still recover- B community has a more dynamic economy, so young ing from the recent downsizing of its Firestone rubber women are reaching for and expecting better oppor- plantation and processing plant. This event sent men tunities than their mothers had, which are less defined looking for work in other communities or in alternative by traditional gender norms than in urban communities livelihoods, such as starting small businesses, and pro- experiencing tougher economies. Young women in Hato pelled women into the labor market to assume the role Mayor, Dominican Republic, are completing high school, of provider or augment income that used to be gener- for instance, but their local economies are stressed and ated solely by their husbands. Rather the traditional 136 discourage women from conceiving of, much less pursu- unskilled jobs of fixing hair and dealing in second-hand clothes, the women discovered better options for paid ents, partners, siblings, and peers for advice and saw in work in the private sector, such as selling fish or char- their experiences notions of what is possible for them. coal, making and selling soap, refining palm oil, running a And they make choices weighing similar factors as men— taxi service, and renting properties. opportunity, economic need, returns from their labor, and the best use of their talents and skills. In these more The need for women in Firestone District to support their inviting contexts, young men and women alike expressed families—and their new autonomy—has changed women’s their desire to be independent, but are aware of respon- views of their own capabilities. As a result, younger wom- sibilities toward their families. “Nowadays both parents en are completely disregarding the gender assignment of [mother and father] are working [and] ... women also jobs, do not automatically assume that unattractive jobs spend most of their time at work,� explained a young are only for women, and do not feel forced to take what- woman from Umlazi Township B in Durban. While anoth- ever work is available, like their mothers did. Inspired, er from this focus group elaborated, “Nowadays women too, by the successful and educated women on the upper no longer have to care of children by themselves. Their rungs of the Firestone District ladder, the young women husbands and boyfriends are also accepting the respon- are finishing their education and envisioning good jobs sibilities of caring for the children.� not currently available in their community. This growing empowerment of the women, vis-à-vis the 4. Impact of laws and local civic men’s (which is more moderate), is permeating other ar- action on empowerment eas of their lives and influencing how they envision the “ future. The women are not only working and saving but We understand that there are laws establishing the are engaging in what other communities term exclusively rights of women, but most of us do not take them male activities, such as going out in the evening or with seriously. As men, we are the heads of the family. In friends, having affairs, and financially supporting their the past, no one knew about these laws, and women households. Even though many in Firestone District dis- respected their husbands. Now, because of these laws, � approve of women stepping out of their traditional space women try to control their husbands, which is not good. and even though these activities can indeed cause them — Village man, East Sepik Province, Papua New Guinea to fall down their ladders, the younger women’s attitudes denote freedom from normative constraints. Markets are not the only structures that influence what Sometimes changes in market signals induced changes is possible on the ladders of power and freedom. Na- in individual behaviors of the focus group participants, tional policies, local governing bodies, state agencies, which may slowly modify social norms around jobs. But and community-based organizations also play a role in the relationship between market trends, agency, and shaping local opportunities and the climate for women’s norm change is mediated by a complex set of delibera- inclusion and influence in the public sphere. We un- tions that include valuations of material benefits, time fortunately had limited time with the focus groups to costs, and reputational risks for different individuals in explore these pathways for exercising agency and how a household.100 In harsher contexts, the gains of women’s they may interact with gender norms. What was shared, economic agency are not at all clear, and the norms that nevertheless, indicates a good deal of unrealized poten- On Norms and Agency surround these gains are more resistant to change. tial for policy action. In countries and communities in the study, where con- We begin by reviewing focus groups’ accounts of their text allows for more economic choices, young women understanding and implementation of gender laws. As were as likely as young men to feel both constrained and part of institutional attempts to alter practices by direct- empowered to find a job. Women looked to their par- ly sanctioning them, laws represent one of the resources 100 Gary Becker (1981) conceives of economic models that account for household specialization of human capital and division of labor. For wider reviews of this literature, see Doss (2011), Alderman et al. (1995), and Dasgupta (2000). 137 for states to challenge prevalent norms. Next we move Overall, the urban adult women’s focus groups demon- to discussions about the (limited) resources that are strated the most specific knowledge: “Yes, we are aware available to couples facing disagreement and conflict in of the laws. There is the act against dowry and there is their marriage or union. Viable exit strategies are central also the act against domestic violence. We learned about to reducing domestic violence. A final section discusses all these acts by going to self-help group (SHG) meet- local political leadership and collective action. ings. TV also tells us about the acts,� explained a woman from Nellore (Andhra Pradesh), India. a. Equal rights for women? As expected, women viewed the new laws and their im- “ pacts more favorably than men (figure 5.6.b). “Men used I’ve heard of parental leaves for the fathers. to beat us and everything would just carry on as normal. It’s good when men take on some of the responsibility But now we can report them to the police,� announced a of caring for the children. I think they will make use of � woman from rural Ngonyameni, South Africa, when asked it in our generation. about her knowledge and views of the new gender laws in — Rural young woman, Martynice, Poland her country. Another from her group countered, however, that “we just hear about the laws on the radio, but they Facilitators asked the focus groups about their aware- do not apply in this community.� Like other progressive ness of their country’s gender laws and any impacts from forces, local awareness and passive or active support for them. A majority of groups had at least some knowledge laws may co-exist alongside normative views that accept, of one or two laws (see figure 5.6.a). Participants men- for instance, violence against women or men’s claims to tioned domestic abuse laws most often, but they also full property rights in cases of inheritance or divorce. In displayed varying degrees of knowledge about women’s some contexts, nevertheless, women perceived the laws rights to resist forced or early marriages; to obtain a di- as helping them. According to young women in Paro, Bhu- vorce, custody of children, and child support; to own and tan, “there is less harassment of women now and men inherit property; to claim equal pay and take family leave; have to think twice before they divorce their wife or have to vote and run for office; and to access family planning. extramarital affairs. Such behavior may come with a big For most, however, awareness was extremely limited. “We cost for them because of the law.� have never heard of these laws,� stated a woman from Briceni District, Moldova; another in her group added, Young women from rural Malangachilima village, Tanza- “School teachers probably know something about this.� nia, could recall several of their rights: “Yes, we all have FIGURE 5.6.A: AWARENESS OF NEW FIGURE 5.6.B: PERCEPTIONS OF NEW GENDER LAWS GENDER LAWS AND THEIR IMPACT 100 95 92 88 82 73 Frequency of mentions Frequency of mentions 57 51 46 On Norms and Agency 23 Good general Some Little or no Favorable Unfavorable knowledge knowledge knowledge views views Women Men Women Men Note: Data from 388 focus groups. Note: Data from 388 focus groups. 138 to go to school. We can inherit property like men. Men However, not all is discouraging news. There is a desire for should not beat us and, if they do, we can take them to gender equality that springs from the influence of chang- court. We can be politicians.� Yet earlier, when we asked es occurring in both the private and public spheres. For this focus group about control of assets and inheritance, some of the focus groups, the notion of gender equal- they explained that men control most possessions in a ity is simply lauded as a constraint on misbehavior and household and that “the eldest son is the chief heir.� Any harm, but the majority perceived that gender equality land a woman may inherit from her family is taken care embodies the highest ideals of their societies. For wom- of by the eldest son (or next man in line) because women en in Samte, Bhutan—Sisum’s village, whose story opens leave their village to marry, have children, and cannot Part I of the report—who reported problems of violence manage the property. In this sense, the new laws may against them, gender equality “means everything a wom- embody ideals that are quite removed from the actual an dreams of; it is a dream come true.� Others in this choices and possibilities for women. group added that gender equality would bring a “better life,� “peace and less domination,� and shared decision- Still, in Liberia and other countries in the sample, some making. Some of the focus groups, mainly rural women men openly admitted that they no longer beat women and to a lesser extent urban men, felt that the notion of because they fear going to jail. “Every day, there used to gender equality violates cherished values and traditions, be an incidence of rape, but now there is less. The use and perhaps fuels rather eases violence against women. of fast-track courts has made it more alarming for men Yet, overall the focus group narratives suggested that, to be associated with rape. Most men violated young with investment in stronger enforcement combined with girls through ‘cash violence,’� stated a young man from building greater awareness about the need for and in- the capital city of Monrovia. In Emputa village, Tanzania, tent of the laws, there is significant unrealized potential urban women proclaimed that the new laws mean they for gender legislation to alter women’s and men’s views are investing more in their assets: “They [the laws] have of gender equality. assisted us because we have worked hard and now don’t lose everything. For example, when you leave the man’s b. Do institutions work? Where to turn for help home, you divide the property and can go with some- thing to begin your new life.� But another woman from Emputa village made clear that she has been waiting for three years for the government to take any action against her violent husband. “There is no divorce, unless the husband kicks out the wife. — Urban man, neighborhood of Kragujevac, Serbia � The men’s focus groups, along with a number of the The focus groups offered discouraging accounts about women’s, expressed mixed views about the laws, often the enforcement of laws and other community resources indicating that they were ineffective or disregarded. for addressing family conflict, such as divorce or dis- “Just show me, please, a man who has been punished for agreements, that cannot be resolved behind closed his violence—no one,� affirmed a man from Moldova. In doors. When the focus groups were asked where local Levuka, Fiji, a village man contended that the laws “may couples turn to for help with marital disputes, the num- be implemented in towns and cities, but not here.� Wom- ber one response was family members (figure 5.7), fol- en in Tanzania and Sudan described weak enforcement lowed by formal state institutions or local governments, On Norms and Agency of laws prohibiting female genital cutting and, in Poland, although their presence is more marked in urban than in women reported that equal pay rules are “ignored so of- rural settings. ten. The men always receive more. They receive all the add-ons [benefits] and bonuses.� A man from Nsenene Specifically for divorce, for example, both young adult village, Tanzania, maintained that “for me, I think such and adult focus groups across the sample indicated that laws have spoiled our women. They have become big- getting a divorce is difficult or simply unacceptable for headed and unmanageable at home and in society. And couples in their communities. This local institutional real- their daughters are learning the same. If we want respect ity, where informal mechanisms and traditional gender and discipline, we need to revise such laws.� norms still prevail, is important for understanding just 139 how limited women’s possibilities are for exiting abusive Focus groups of both women and men made plain that relationships. The pressures to resort to family or other individuals or couples seeking temporary or permanent informal networks are also why central government laws, separation are strongly discouraged by peer pressure, policies, and programs—as important as they are for community opinions, and threats of unfair treatment. gender equality—often have less effect than expected. Women face complete destitution, permanent separa- Rather than using the formal justice system, which may tion from their children (“the children don’t belong to not even have a physical presence in or near many of her, they belong to her husband�), and lasting ostracism the study communities, most participants explained that from the community. Bride price and dowries that pass parents, in-laws, extended family, community elders, and assets across family networks and down to husbands local “reconciliation committees� (specific to Vietnam) or wives further complicate divorce processes. In both all try to mediate to avoid a temporary or permanent Papua New Guinea and Yemen, women referred to bride separation due to violence in the home. prices and expensive dowries as impediments to divorce: “It’s hard for a woman to get a divorce in this community In a semi-rural community of Ngonyameni, South Africa, because if the bride price is paid it makes it difficult for according to the women’s focus group, the new gender the woman to leave� (a women’s focus group in Western laws “have changed nothing here. We do not have any Highlands Province, Papua New Guinea). Religion, both job opportunities, our husbands assault us, and most of Christianity and Islam, imposes restrictions on divorce in the time the tribal court favors the man. So really nothing a number of the sample countries. has changed. These laws apply only to urban areas.� Sim- ilarly, in rural East Sepik Province of Papua New Guinea, Where official channels are an option—mainly in some women said that neither the police nor the churches will of the better-off urban communities—focus groups indi- get involved in domestic violence matters. And while cated that they are employed only as a last resort, when they mentioned that NGOs can help women get access family and other informal local institutions have been to the courts, a young woman claimed that “most women tapped and failed, and then only for the most extreme are afraid of their husbands and do not pursue claims circumstances. “A justifiable reason is if a woman was in the District Court for maintenance.� In Sigatoka, Fiji, terrorized by the husband. In this case, a woman can go women maintained that “nobody divorces in our village. to her parents’ home. She can call the police or some so- You can’t just fight with your husband and then want to cial service, but these actions are justified only in case of divorce. This is a Fijian village. There is always a way to great violence,� explained a woman from urban Kraguje- solve marital disagreements.� vac, Serbia. A young man from Hoang Mai district, Hanoi, Vietnam, hinted that the reluctance to use formal mech- anisms, until all other options had been exhausted, was less because of distrust in the formal civic institutions FIGURE 5.7: WHERE PEOPLE TURN FOR HELP WITH FAMILY CONFLICT than because of the economic cost of involving them: 160 Many couples who get divorced negotiate with each 143 134 other for child custody and property because otherwise they have to pay fees if the court is involved. In cases Frequency of mentions 106 91 where they cannot reach an agreement, they look for 79 On Norms and Agency help from relatives or friends. Women often turn to the Women’s Union for help if relatives fail. In the end, if nobody can help solve the dispute, the court makes a decision according to marriage, family, and other laws. Family State Informal institutions institutions Men and women made clear how essential local institu- Urban Rural tions are for delivering gender equality in difficult situa- tions. In the many contexts where local power inequali- Note: Data from 388 young adult and adult focus groups. 140 ties are strong, and credible threats of violence to women underpin them, women’s access to formal institutions attend community meetings and women are very much may often be tightly restricted. Yet, the continued reli- involved. ... The level of development now is enhanced ance on the traditional informal mechanisms to uphold by women’s participation.� And, in Tewor District, Liberia, laws and resolve disputes so often leaves women per- young women reported that, although their town is now sistently disadvantaged and vulnerable. Taken together, more than 20 years old, it only started building a new high these are key processes by which inequitable and exclud- school when a woman became the local commissioner. A ing power structures resist change.101 Thus, even where woman in Thimphu, Bhutan, also pointed out benefits of a nation’s constitution may guarantee equal rights, and women in political power: “Women representing their laws and regulations have been enacted that are strongly community as local leaders have also helped bring issues in the public interest and enjoy widespread support, lo- related to women into public forums.� In the ladders of cal structures and normative behaviors that perpetuate the Indian focus groups, a common attribute of women gender inequalities may nevertheless endure. Without on the top step is holding leadership positions in their strong state capacities for enforcement matched by village or neighborhood self-governance institutions. broad-based knowledge of the laws, formal legal, legisla- Beaman et al. (2009, 2012) find that a community in India tive, and regulatory reforms are unlikely to be effective. with a decade of experience with women leaders (who emerged after gender quota requirements were enacted c. Whose voice counts? in 1993) can erase the bias in men’s perceptions that they Local political leadership and civic action are automatically better leaders than women and close the gender gap in educational outcomes. “ Sometimes having better education or better access to political leaders or local officials also The contributions of civic networks to movements up � the ladders are less clear cut in our dataset. From key helps you move up the ladder [from step 1 to step 2]. informants’ reports to the field researchers, we find a — Urban woman, Nellore (Andhra Pradesh), India median of 14 different local civic groups per community in the urban sample, and a median of 10 in the rural sam- Many women’s ladders in our sample conveyed that—in ple, but the numbers vary significantly. In urban Mon- addition to participating in their local economies and gar District, Bhutan, women said that the factors, which feeling empowered by this—women are gaining more and help them move up their ladder, include “exposure [to more access to the formal political and civic institutions successful women and information],� “advocacy,� “non- in their communities. In the 97 communities of the focus formal education [such as training],� and local develop- groups, 24 urban and rural women were elected leaders, ment projects that have gender targets. As discussed in up from a total of 10 a decade ago. Opportunities for chapter 4, focus groups also mentioned informal rotat- collective action also appear to be flowering, especially ing savings groups (like the SACCOs in Malangachilima) for women, but the link between collective action and and more formal self-help groups, and leadership roles empowerment is not as clear cut as women’s political in- in them, as helpful for climbing the ladder. But overall, clusion.102 these civic supports are not that prevalent among the mobility factors. A quarter of the 24 elected women in the focus groups hailed from Liberia. The young women and girls in these With the exception of religious institutions in some On Norms and Agency focus groups frequently named their president, Ellen contexts, economic groups generally outnumber other Johnson Sirlief, as their role model. And in urban Fire- forms of grassroots organizations in our sample. Beyond stone District, men noted that 10 years ago not a single the ladder exercise, focus group responses to specific woman held a local position of authority, but now, “people questions about sources of credit and local producer or 101 See, for example, North, Wallis, and Weingast (2009). 102 In their extensive review of community-driven development programs, Mansuri and Rao (2012) find extensive problems in targeting the poor and meeting performance objectives, but they also mention more promising results from “a couple of recent studies of community-based projects, which provide microcredit, cash grants, and business- or livelihood-related training...� (page 11).Their review of community-driven development evaluations finds elite capture in communities that are “remote from centers of power, have low literacy, are poor, or have significant caste, race, and gender disparities� (page 6); and local social capital that “tends to melt away when the [project] incentives are withdrawn (page 11). 141 trade groups revealed gaps and disappointments. The The mixed reports about economic groups are also re- large majority of focus groups indicated that economic flected in performance on the ladders. At first glance, groups were absent, ineffective, or accessible to only figure 5.8 reveals that, as with civic groups in general, a few residents in their communities. And, even when there are more economic groups present in urban set- they perform well, the economic groups are often more tings. In the cities in our sample, the focus groups with helpful to women than men. In Comendador, Dominican ladders showing twin climbing described numerous Republic, for instance, the women said, “There are as- economic groups, which may also be present (albeit sociations, but they have no economic projects and they in smaller numbers) in sample locales where both men don’t help women.� In Nsenene village, Tanzania, where and women perceive disempowerment. In rural com- women are outperforming men on the ladders, a man munities where women are rising on the ladders, eco- remarked, “Women get loans from institutions, such as nomic groups are more numerous, but they practically FINCA, PRIDE, and local SACCOs. They can borrow disappear where women report disempowerment. No money and do business. We men can have nothing to do clear patterns emerge for rural men. with such initiatives, so we are left behind.� Many wom- en’s and men’s groups, in fact, commented on how the When we asked the focus groups about their hopes for men are excluded from many lending and savings oppor- their communities and their children, the call for jobs tunities. One woman in Nellore (Andhra Pradesh), India, and economic interventions emerges by far as the top stated, “Men don’t get loans these days. If they need a desire. In Nellore (Andhra Pradesh), India, a woman ar- loan, they are asked to bring their wife for a guarantee.� gued, “Life can be very different if there is marketing Another in the same group added, “Men have no cred- assistance for the products woman can make at home.� ibility these days.� And in rural Koudipally Mandal (Andhra Pradesh), India, a woman expressed hope that a factory will be built In the eight “high mobility� communities with extensive nearby, “so that the younger generation will have better twin climbing, like Jaipur and Malangachilima, there are job opportunities.� Help with farming is the wish voiced many reports about the vitality of local economic organi- by a young man of rural Sumadija District, Serbia: “A lot zations. In Cusco Province, Peru, a village with twin mo- depends on the local government, if only they would bility on the ladders of its focus groups, they described invest in agriculture and in the villages. This is where my a farmer’s group comprised of women that meets three hope lies.� In rural Kharef district, Yemen, where girls times weekly to take their produce to the nearby town are now attaining more education, women believe that market. And a governmental Vaso de Leche [Glass of “education will improve our status in the future,� and milk] program, which provides nutritional support to they hope “our children find jobs and our daughters be- poor young children and pregnant mothers, is run by come teachers.� local women. In Koudipally Mandal (Andhra Pradesh), India, another community where both women and men Focus groups also wanted better public services in are climbing in large numbers, there are 22 SHGs, which their communities, calling variously for schools, health receive support from external governmental, civic, and clinics, clean water, roads, public transportation, and private sector agencies, and are widely considered an police. In Floresti District, Moldova, like other com- important resource for local women’s awareness-raising, munities, women wished for free and accessible edu- solidarity, and economic initiatives. cation and more job opportunities for their daughters On Norms and Agency and sons. In the village of Da Wa (Parwan), Afghanistan, In Kalahandi, a tribal village in Odisha, India, also with where there is no school for girls, an adolescent girl high twin mobility on its ladders, the women said that suggested, “If there is vocational training for young they are moving up their ladder because, unlike in the women in the community, we will learn something. With past, they are now saving money and taking out loans in this, we can earn an income and change our lives.� In their SHGs. The women also reported that their SHGs River Nile State, Sudan, women hoped for legislative help create strong bonds between women of lower and policy to address problems of “women facing harass- higher castes, when before they would not sit together ment while moving around the community and region.� 142 in meetings. In East Sepik Province, Papua New Guinea, a young woman lamented that her village did not pool funds to sponsor students seeking higher education, and another FIGURE 5.8. MEDIAN ECONOMIC GROUPS one spoke of educating daughters so they “do not make (PRODUCER, TRADE, AND FINANCE) IN the same mistakes� or “get pregnant early and miss out DIFFERENT EMPOWERMENT CONTEXTS on opportunities.� In Thimphu, Bhutan, a young woman 9 pointed out that “education and awareness may be the 8 Median number of groups 7 right tools to make both men and women use more 6 family planning services.� And both younger and older 5 women expressed a need for daycare. 4 3 Some focus groups called for interventions to increase 2 awareness of and support for women’s rights, or reduce 1 the problems facing men that perhaps prevent more 0 empowerment for women. To strengthen their village, Urban (n=50) Rural (n=42) for example, women of Zabibu village, Tanzania, hoped Twin climbing Women climbing, men falling to “stand for office and participate in various forms of Twin falling Men climbing, women falling leadership in our community and country at large.� Or Note: Data from 184 adult focus groups. The Sudanese ladders are missing from this analysis because the women's focus group in Kim Dong District, Vietnam, women wanted to see did not conduct the sorting exercise. “more policies to ensure women’s rights and more job opportunities for women.� In rural Ngonyameni, South Africa, young people face heavy school dropout, job- it comes to work for pay, women’s identities and associ- lessness, drug problems, and early pregnancies. In addi- ated duties take precedence or get translated into spe- tion to more economic opportunities, the young men’s cific goals or jobs that align with such identities. focus group there thought sports or other recreational facilities may help them stop their risky behaviors and In Naw Da, Afghanistan, the women placed 80 percent take their education more seriously. Similarly, in a vil- of themselves on the bottom rungs of their ladder, be- lage of Red Sea State, Sudan, young women wanted cause women “do not have jobs, freedom, or education. their community to offer handicraft projects for them They cannot go anywhere. They do not have income and build a sports club for the men. Focus groups also sources, nor is there a school they can attend. They desired the happiness and safety of their families and cannot make decisions and their husbands are unedu- communities. In the remote village of Jahran District, cated.� None of them has found a way to climb up their Yemen, the young women’s “greatest wishes are mar- ladder in the past decade. In order to have more power riage and employment.� and freedom, they felt that they needed to be support- ed in their efforts by the “government, NGOs, village Men’s and women’s aspirations for the future and their elders, and the local mullah, who should know the rights activities have to be congruent with their gender iden- of the women and push the men to accept these rights. tity (see box 5.2). These identities develop in relation to In the present situation, we do not see any chance for each other. As discussed in chapter 1, a good husband women to change their lives.� These powerless and iso- prefers that his wife not work. If she does, his role of lated women recommended that interventions start by On Norms and Agency provider—his masculine duty and thus status—may be changing men’s mindsets. questioned by the community. “Now we need to find work and jobs. …Based on the present economic situ- ation of the community, there is no work for men. If 5. Change women need someone wants to help the community, the first step is to help the men because they are responsible for food When looking across this large dataset, it is clear that and the overall maintenance of the family� (adult men, women’s life choices remain more restricted than men’s Boyina Bagh, Afghanistan). Likewise, women’s identity is and that markets, local politics, public services, and civic built around being a housewife or a mother and, when action—in most communities—mainly reinforce rather 143 Box 5.2: Public and private power The important shifts made by women into positions of public power, where they are possible, seem to occur indepen- dently of what happens in their private lives. Women carefully hold their public power in check when they step into their homes; they find they are unable to translate potential gains from their public roles into their personal lives. A 53-year-old woman from urban Ba Dinh District, Vietnam, explained this dilemma for women: “It is not that, if you are powerful in society, you should also have power at home. Of course, if you have power in society, you gain respect from your husband and children. But a woman should use her power properly, at the right time and in the right place. You should not bring power home and apply it in your family because then you may not be happy. Some women are powerful outside, but not in their family.� In fact, gains in power outside the household do not always translate into domestic power. Sumitra is a 37-year-old woman from a village of Phulbani municipality in Odisha, India. She is the ward member of her village and a member of the local panchayat (council), and also participates at the district level. But in most ways her life is no different from any other woman in her community. Her day begins at 5:00 a.m. with all the household work; after finishing that, she goes to her daily wage work under the National Rural Employment Guarantee Scheme or to the rice paddy. Her husband also does daily wage work, but whatever he earns he spends on alcohol. The household is run on her income. Her evenings are frequently filled with verbal and physical abuse: “I don’t remember one day that has ever ended peacefully. Most of the time, I and my children go to bed with empty stomachs because my husband throws away the cooked food in anger. This is the life I am living and yet people say I have political power.� Her journey into politics started with the 2007 election, when members of the community nominated her to run for a ward seat on the council. For her own and her community’s benefit, she ran and won the position as ward mem- ber. However, she is discouraged. She says that nothing has changed in her life nor she has contributed anything to her community in the past four years. She regularly attends her ward council meetings, for which she is paid INR 30 (Indian rupees) per meeting, but she does not see herself as empowered. “We women have power in name only. We cannot contribute to improving our lives or the lives of community members.� She also noted, “If I don’t work for a day or two, my family will starve. …My elected position has not improved my life in any way.� She will not run in the next election. On the other side of the spectrum is Mrs. Latata, from Fiji, who is 71 years old and “the most senior member of the village women’s group� (which has five members). She worked as a primary school teacher for 42 years, beginning her career when she was 24, and has become an agent of change by providing information about laws and assistance for women in her community who do not know about their rights or the legal system. Latata also teaches young girls to be critical, but not aggressive, in the solution of disputes. She feels that young girls can at least be introduced to thinking about things critically. “To be a leader, they need to have the ‘know-how’ to handle problems. They need to know that when you have leadership skills, you can solve problems amicably and don’t have to make a big deal about it. You don’t have to be aggressive [to get your point across]; there is a way to deal with things and make your point, and to give your help in way that respects other views.� On Norms and Agency Latata not only promotes peaceful dialogue skills, which can be used as much in public as in private, but also teaches girls about their rights and legal issues. She recounts the story of a young woman “who, after she finished school, looked after her elderly parents. Her father eventually passed away, but told her that she would be able to survive on his pension, which was held in superannuation funds in the capital. The woman found out later that her father had not legally left his pension to her and had no clue how to access the money her father had left her. [The young woman] told me that she knew I was the only one who could help her.� Latata contacted the proper offices in the capital on the young woman’s behalf and facilitated the payments. “These are the sorts of deeds Latata feels she is able to do from the knowledge and the confidence she has learned from being part of a women’s community group,� explained another woman. 144 than ease these inequalities. Weak local institutions and political climate changes for women’s roles. As Kabeer restrictive norms blunt the effects of broader progres- (1999) notes, access to new resources may open up new sive forces on women’s power and gender equality. On possibilities for women, but the use of such possibilities balance, though, we see more evidence of rising aspira- is not necessarily uniform. Women, as they revealed in tions and relaxation and change of norms in contexts their focus groups, are still heavily influenced by the nor- where local markets are reported to be more dynamic. mative elements inherent in what affords them status Women’s local political leadership also emerges as a sig- and a good reputation with their families and communi- nificant force for women’s empowerment and gender ties, in their social relations, and in their individual (and norm change. And while community-based groups do other women’s) histories, beliefs, and identities. not receive high marks on balance from focus groups, a strong presence of economic collective action can make There seems to be no simple pattern of more flexible a difference, especially when paired with other support- gender norms routinely developing alongside the cre- ive conditions. ation of more and better economic opportunities in our study communities. Neither do we necessarily find Women’s ability to work for pay, which most women par- a stronger adherence to traditional norms among the ticipating in our study aspired to, may be one of the most sample communities that have seen fewer changes or visible and game-changing events in the life of modern experienced reversals in their economic situations. In households and all communities. Women’s work, as the the less vibrant communities, the market requirement focus groups showed, has the potential to alter tradi- for unskilled labor is opening doors for women and men tional definitions of gender roles, duties, and responsi- alike, and this can sometimes break the grip of tradition- bilities, as well as the main components of the identity of al gender norms. Gender norm change is also sometimes both men and women. Yet, empowerment and agency do possible, independent of local market and institutional not directly result from economic participation, but are forces, and women’s own actions. For example, in con- supported by what women experience when leaving the texts outside of our sample, such as in Senegal, commu- home to join the market. Women gain a greater sense of nity-based public awareness campaigns have been able self-efficacy, broaden their aspirations, and forge ways to stop the accepted practice of female genital cutting to reconcile their identity as workers with their identity (Mackie 2000); and in Afghanistan, community aware- as mothers. In Samtse, Bhutan, Sisum’s mountain village, ness projects have successfully reduced the incidence a young woman saw herself “completing law school and of early marriage (Malhotra et al 2011). looking for a job as a lawyer� and another one believed she can be “a successful government servant.� While an individual woman’s view and ability to act is relevant, sometimes a critical mass is needed to break a Throughout this study we learned from the communi- cycle. As Fogli and Veldkamp (2008) show in their study ties that the interplay of three forces influences gender of the transition of women into the United States’ la- equality: changes in the capacity to identify new oppor- bor force in the 20th century, local transmission of in- tunities and aspire to them; changes in the capacity to formation generates change. Women learn from other act and actively pursue those ideals or the use of those women’s experiences: as the women’s focus groups told opportunities; and changes in the structure of opportu- us, they learn from their mothers, their mother’s peers, nities and normative frameworks that constrain individu- and their own peers. They learn from what happens to On Norms and Agency als. These three areas do not always move together. Nor, other women’s children, and in other’s households, and necessarily, do men’s and women’s perceived trends in in communities. empowerment grow in parallel. Also, women may feel more empowered, but there is little evidence in our data The process of gender norm change appears to be un- that this perception is matched by greater influence in even and challenging. When only a few women manage their relations with men. Further complicating the linkag- to break with the established norms, uncertainty reigns es between the three forces, some norms may become and traditional norms are not challenged and may even more fluid in stressful economic times, and women’s new be reinforced. Also, the easy co-existence of new and freedoms may again narrow when their households and old norms means that households in the same commu- 145 communities become more prosperous, or when the nity can vary markedly in how much agency women can exercise, and that women often feel less empowered Beaman, L., R. Chattopadhyay, E. Duflo, R. Pande, and when the opinions and values shared by their family P. Topalova. 2009. Powerful Women: Does Exposure members and communities stay with traditional norms. Reduce Bias? Quarterly Journal of Economics 124: Moreover, lack of gain in women’s empowerment is not 1497–-540. always due to overt resistance to it; in many cases, it Beaman, L., E. Duflo, R. Pande, and P. Topalova. 2012. Fe- is due to a passive, underlying view in the community male Leadership Raises Aspirations and Educational that “this is just the way things are here.� This view com- Attainment for Girls: A Policy Experiment in India. mands adherence and we find it acknowledged again Science 335: 582–86. and again in accounts by different communities in the Dasgupta, I. 2000. Women’s Employment, Intra-House- study. Even if actual practice has changed and is differ- hold Bargaining, and Distribution. Oxford Economic ent (e.g., many women claimed some measure of eco- Papers 52(4): 723–44. nomic independence), and even if many have let go of Doss, C. 2011. Intrahousehold Bargaining and Resource the norm (men must be the sole providers), the persis- Allocation in Developing Countries. Background pa- tent belief—and reality— is that status and respect con- per written for World Development Report 2012: Gen- tinue to be governed by the traditional gender norms. der Equality and Development, World Bank, Wash- These dynamics keep change from happening. ington, DC. Duflo, E., and C. Udry. 2003. Intrahousehold Resource The women’s stories explored here in both more and Allocation in Côte D’Ivoire: Social Norms, Separate less dynamic communities unrelentingly show that, al- Accounts, and Consumption Choices. Working Pa- though many women have higher expectations and more per, no. 857. New Haven, CT, USA: Yale University, chances to work, to run for elected office, or to engage Economic Growth Center. in community organizations, all too often they must still Fehr, E., and K. Hoff. 2011. Tastes, Castes, and Culture: accommodate their public roles to conservative gender The Influence of Society on Preferences. Policy Re- norms. At the same time, other women who have limited search Working Paper, no. 5760. Washington, DC: or poor opportunities to participate in the public life World Bank. of their communities may be able to negotiate gender Fogli, A., and L. Veldkamp. 2008. Nature or Nurture? norms in ways that increase their agency, and eventually, Learning and the Geography of Female Labor Force some other outcomes, too. Participation. NBER Working Paper, no. 14097. Cam- bridge, MA, USA: National Bureau of Economic Re- search. References Guiso, L., P. Sapienza, and L. Zingales. 2006. Does Cul- ture Affect Economic Outcomes? Journal of Econom- Alderman, H., L. Haddad, J. Hoddinott, and R. Kanbur. ic Perspectives 20 (2): 23–48. 1995. Unitary versus Collective Models of the House- Morrison, C., and J. Jutting. The Impact of Social Institu- hold: Is It Time to Shift the Burden of Proof? World tions on the Economic Role of Women in Developing Bank Research Observer 10 (1): 1–19. Countries. Working Paper, no. 234. Geneva: OECD Agarwal, B. 2003. Gender and Land Rights Revisited: Development Centre. Exploring New Prospects via the State, Family, and Kabeer, N. 2000. The Power to Choose: Bangladeshi Market. Journal of Agrarian Change 3 (1): 184–22. Women and Labour Market Decisions in London and On Norms and Agency Avirgan, T., L.J. Bivens, and S. Gammage, eds. 2005. Dhaka. London: Verso. Good Jobs, Bad Jobs, No Jobs: Labor Markets and ———. 2001. Reflections on the Measurement of Women’s Informal Work in Egypt, El Salvador, India, Russia, and Empowerment. In Discussing Women’s Empower- South Africa. Washington, DC: Economic Policy In- ment: Theory and Practice. Sidastudies No. 3. Stock- stitute, Global Policy Network. http://www.facmed. holm, Sweden: Novum Grafiska AB and Sida. unam.mx/deptos/salud/censenanza/spivst/spiv/bad- Lastarria-Cornhiel, S. 2006. Feminization of Agriculture: jobs.pdf. Trends and Driving Forces. Santiago, Chile: Rimisp- Becker, G.S. 1981. A Treatise on Family. Cambridge, MA, Latin American Center for Rural Development. 146 USA: Harvard University Press. Leavy, J., and S. Smith. 2010. Future Farmers: Youth Aspi- rations, Expectations and Life Choices. FAC Discus- sion Paper, no. 13. Brighton, UK: Future Agricultures Consortium, University of Sussex. Mackie, G. 2000. Female Genital Cutting: The Beginning of the End. In Female Circumcision: Multidisciplinary Perspectives, ed. B. Shell-Duncan and Y. Hernlund, 245–82. Boulder, CO, USA: Lynne Reiner Publishers. Mansuri, G. and V. Rao. 2012. Can Participation Be In- duced? Some Evidence from Developing Countries. World Bank Policy Research Working Paper, no. 6139. Washington, DC: World Bank. Malhotra, A., A. Warner, A. Gonagle, and S. Lee-Rife. 2011. Solutions to End Child Marriage. What the Evidence Shows. Washington, DC: International Center for Re- search on Women (ICRW), Munshi, K., and M. Rosenzweig. 2006. Traditional Institu- tions Meet the Modern World: Caste, Gender, and Schooling Choice in a Globalizing Economy. Ameri- can Economic Review 96 (4): 1225–52. North, D., J.J. Wallis, and B.R. Weingast. 2009. Violence and Social Orders. Cambridge: Cambridge Univer- sity Press. Escriche, L., G. Olcina, and R. Sánchez. 2004. Gender Discrimination and Intergenerational Transmission of Preferences. Oxford Economic Papers 56 (3): 485–511. Spillman, L. 2012. Solidarity in Strategy: Making Business Meaningful in American Trade Associations. Chicago and London: University of Chicago Press. Wherry, F. 2012. The Culture of Markets. Oxford: Polity Press. Zelizer, V. 2010. Economic Lives: How Culture Shapes the Economy. Princeton, NJ, USA: Princeton University Press. Tzannatos, Z. 1998. Women and Labor Market Changes in the Global Economy: Growth Helps, Inequalities Hurt, and Public Policies Matter.� Social Protection Discussion Paper, no. 9808. Washington, DC: World Bank. World Bank. 2012. World Development Report 2012: On Norms and Agency Gender Equality and Development. Washington, DC: World Bank. 147 FINAL THOUGHTS W orld Development Report 2012 showed that en’s and men’s agency. We hoped to identify a few simple gender equality and economic development patterns suggestive of causality that could also help prac- can reinforce or hinder each other through titioners identify missing or new areas for interventions complex social and economic processes. The opportu- to encourage gender equality. In this respect, however, nities available to women and men—and their ability to our research proved too ambitious and we did not find take advantage of them—are determined by social norms, any “silver bullets� for development intervention. gender roles, beliefs about their abilities, and their as- sets, as well as by the social expectations of communities We did, however, illuminate a range of variables and and countries they live in. While we see significant prog- complex interactions that determine the space for rene- ress in many places, gender disparities are still evident, gotiating or “bending� the gender norms that influence particularly regarding empowerment and agency. behaviors and affect the pace of change and growth of men’s and women’s agency. When households and com- Data analysis from standard nationally representative munities find ways—or the need—to relax and change in- surveys has not provided much information on norms equitable gender norms, we saw that men’s and women’s and their effects, which World Developm ent Report individual and collective agency can increase and rein- 2012 identified as resistant to change and potentially force one another. hindering the advancement of gender equality. In our study, given the nature of our data, we were able to look One of the more consistent findings across the 97 re- deeper into social norms, and found that gender norms search sites is the universality and resilience of the affect decision-making at the household and individual norms that underpin gender roles. Given the diversity levels, as well as the functioning of markets and formal of the communities and people participating in the focus and informal institutions. Also, just observing changes in groups, we expected to find a wider spectrum of atti- outcomes, such as increasing female labor force partici- tudes and roles, and in particular expected to observe On Norms and Agency pation, does not always reveal the effects of these norms more flexibility in gender roles in areas with greater eco- and raises more questions. Are more women working nomic dynamism. While we did uncover some softening due to increasing labor market opportunities and their of norms in urban areas relative to rural areas, and by gains in education, despite rigid gender norms? Are gen- younger generations relative to older generations, these der norms dynamic and relatively flexible depending on differences were less striking than the similarities across the economic context? sites. In every research location, women and men of all generations identified the dominance of women’s do- We began our qualitative research into gender norms to mestic role and men’s breadwinning role as absolutely 148 explore the relationship between social norms and wom- core to female and male identities. The good husband defined by the focus groups is almost within a household, for example, if a man opts to take entirely characterized by his income-earning capacity on more domestic work.) Sometimes this tension can and household authority role, and the negative behav- become violent. Violence and the threat of violence or iors he should not engage in “too much,� such as drink- abuse play a role in reinforcing the status quo and in dis- ing, womanizing, and gambling. On the other hand, the couraging women’s efforts to challenge existing expecta- good wife is strictly defined by a long list of submissive tions and norms. qualities and household tasks, including being nurtur- ing and gentle mannered; tending to the house; caring Despite the resilience of gender norms, our study found for the children, the husband, and the elderly; cooking evidence of women’s agency increasing in some strate- well, and also contributing to the household income if gic life choices (chapter 3). The most powerful results and when needed. The idealized visualization of a good in terms of norms evolving toward gender equality wife (or a good husband) is almost impossible to find in a and resulting in greater agency are in the expansion of single woman (or man); for women such a stringent defi- education for girls (and women). However, they remain nition, and the demand for adherence to it, may be due more constrained by household preferences and strict to high levels of stress, and poverty, and for women and gender roles in how many years they can attend school men alike, the profound attachment of the man’s identity and which fields they can study. Younger generations of to a job and income are key. women are also demanding more control over marriage and child bearing than older generations, and they are Some of the focus groups gave evidence of gender norms participating more and more in these decisions. And, changing, albeit slowly and incrementally, with new eco- while in many cases this control does not translate into nomic opportunity, markets, and urbanization (chapter 2). outcomes, young women—and young men—harbor differ- In some locales and among younger age groups, partici- ent aspirations than their parents, such as having fewer pants described relaxation of gender norms where the children, getting more education, and marrying later. The structure of opportunities that increase women’s access structure of opportunities and the social pressures sur- to jobs does not curtail men’s opportunities and occurs rounding them do not always allow their aspirations to with other changes, such as improvements in public insti- materialize, but often they achieve a middle point. tutions. Economic roles for women often creep into their domestic role and, in some places, younger men even Exercises in the study to look at levels of, and changes take on some narrow domestic responsibilities. What is in, power and freedom suggest positive progress for striking is the glacial pace of this change relative to the women, but less for men (chapter 4), which is partly ex- pace of change in contextual factors. Gender norms are plained by their different starting points. Men as the tra- being contested, bent, and relaxed, but not necessarily ditional power holders are accustomed to having more broken fully and changed. Younger people may delay freedom than women. Men report fewer improvements compliance to a later point in time, but the norms and in their sense of agency, but they benchmark their gains the expectations around them do not change. and losses against a higher base line. When compared to men, women in our study in general feel more em- The process of changing gender norms is not always powered than 10 years ago. The rising availability of eco- peaceful or harmonious, and the uncertainty around nomic opportunities, increasing levels of education, and prospective change may lead to gender-based violence growing control over reproductive choices have been On Norms and Agency or new forms of gender inequality. At the household central in enabling women to gain more decision-making level, we saw that space to negotiate a shift in norms in power in their lives. order to intensify agency is most commonly expanded in the constant dialogue among men and women implicit Looking at the ladders of freedom and power created in everyday life. Tension is commonly associated when a by the women and men in the study, men tended to woman participates as an equal in these exchanges and equate power more directly and more narrowly with we chiefly focused on relaxation of norms that permit their economic success. In this respect, many have been women’s roles to expand into men’s roles in the house- disappointed in their ability to find jobs they want to do hold rather than the reverse. (There is little opposition and acquire the incomes they aspired to. Communities 149 where both men and women feel more empowered now At the individual level, education, self-efficacy, and the than 10 years ago were a small fragment of the sample, ability to aspire emerged from the focus groups as im- but generally experienced more rapid economic growth portant factors with respect to the capacity to negotiate (and consequently more job opportunities) and had more change for greater agency. In particular, the education women in decision-making positions in public institutions of boys and girls—beyond its role in building human cap- and private organizations. ital—is crucial in shaping norms. In multiple discussions, adolescent boys and girls described how education ex- Given that norm change is slow and incremental rather posed them to new ideas and knowledge, enlarging their than seismic, what role is there for development inter- capacity to analyze and encouraging critical scrutiny of ventions to influence movement toward gender equal- established gender relations and the status quo. These ity and normative change? Earlier chapters detail ex- discussions reaffirmed what is already known about the amples of the variables and combinations of variables intergenerational transfer and reproduction of norms that expand or constrain the space for renegotiating within households. Education fosters learning away from norms. The focus group discussions suggested that the household environment where gender roles are the interplay between desire for change at the indi- played out in every interaction and action. The research vidual level, the opportunities to effect change at the team realized the importance of ensuring that school household level, and the support or lack of support for curricula offer gender-neutral learning opportunities. changes in norms at the community level represent a critical set of interactions that can accelerate or de- Several sections of the research probed nearly universal lay shifts in attitudes. On the individual level, women’s patterns that emerged among young people’s aspirations. desires to operate outside prescribed norms may be They wanted higher levels of education than the current high, low, or non-existent. Within a home, depending average in their communities, better jobs than common- on the personalities and views of household members, place in their communities, marriage at an older age than tolerance for behaviors outside the norms may be con- normal in their communities, fewer children at a later age siderable or absent. A community may actively show than usual, and so on. When prompted to further de- support for or sanction those pushing against tradi- scribe what they thought were realistic outcomes, their tional norms of behavior. predictions fell somewhere between current practices and their aspirations. This capacity to visualize a differ- A motivated woman or man in a household with high ent path from the existing, accepted course to even a tolerance for bending traditional norms in a community pragmatic midpoint is a positive feature that develop- where others are doing likewise may more easily incor- ment interventions can build on. porate making decisions and choices outside the nor- mal range for their gender. The same woman or man in a Women’s and men’s success in achieving what they as- more traditional household may face considerable do- pire to often depends on factors outside the individu- mestic unrest or violence. The same man or woman in a al’s control (see the second part of chapter 4), mostly community that is resistant to behaviors that threaten in social and political structures. Evaluations of devel- traditional roles may find themselves isolated and with- opment interventions indicate that projects targeted out social capital beyond the boundaries of the house- at young people can profoundly affect their ability to hold. It is not enough to generate change in a single aspire and, by extension, to make decisions that may be On Norms and Agency area. As women learn and benefit from new gender more empowering. norms, their intra-household bargaining position must also improve at the same time. Women’s aspirations The threat of violence or abuse reinforces the status need to be accompanied by opportunities to realize quo and discourages women’s efforts to challenge ex- them and interventions that facilitate the accumulation isting norms. The team, during the research, recognized of assets and capabilities. Empowerment thus requires that action to reduce violence and abuse within the a combination of factors to become a reality. And indi- households had both short-term and long-term bene- vidual attempts to change norms may fail more often fits. Lowering the threat immediately improves physical 150 than not, especially if they try to reallocate power. and psychological welfare in the household, and also introduces a safer environment for women to partici- groups have a mixed record in supporting improvements pate in household discussions and decisions on house- in agency for women and men, despite their potential as hold actions that can shift the balance of power in the networks for change. Development interventions work household. We also found cases, however, where laws frequently with community groups, so the question aris- promoting gender equality were in place, but imple- es whether they are the best places to generate a critical mentation in practice was deficient, making women feel mass of support for women and men pushing the bound- even more disempowered. aries of entrenched norms around gender roles. The norms which uphold women’s heavy domestic We found that regulations and laws promoting gender workload are powerfully reproduced in the household, equality promoted some change when they were well- where girls mirror their mothers’ unending responsibili- publicized and well-enforced, but outreach and pub- ties and long, isolated hours at home. In contrast, men lic understanding of the laws—whether they criminal- spend work and free-time hours in activities outside the ize gender-based violence, permit divorce, or support household, which are reflected in boys’ time-use pat- women’s inheritance rights—were very uneven. In gen- terns and easy interactions in the community and wider eral, people in urban areas had more knowledge of such world. Even when girls go to school and boys help out a laws and women were more in favor of these regulations little in the house, the girls’ workload in the house does than men. In none of the sample countries did we find not change or lessen. The time that women and girls either men or women to be really well-informed of their must spend on domestic responsibilities constrains their rights, entitlements, or obligations with respect to key time available for activities outside the house: earning laws intended to promote gender equality. Clearly, the income (which gives women more voice and clout in the countries in our research need more, and more effective, household), socializing and engaging in public institu- awareness-raising campaigns to promote knowledge and tions (which puts them in contact with extended social enforcement of these laws. networks that support non-traditional behavior), and attending school (which enhances their knowledge and Overall, our research offers new evidence that increas- nurtures their aspirations). ing women’s agency involves constant dialogue between social norms, empowerment, capacity to aspire, and the At the community level, the focus groups pointed out structure of opportunities. These findings strongly sup- that the impact of moral support—whether from a com- port World Development Report 2012 and suggest that di- munity or social network—for women is critical for wom- rect intervention in all these domains can accelerate the en’s empowerment and perseverance to gain agency. improvements in agency offered by economic growth. Their efforts to bend roles in their own households are less stressful when they can talk to a neighbor and get constructive advice, for example, about “getting a hus- band or mother-in-law to agree let them work for pay outside the household.� Even when change is resisted by husbands or extended family, if other men and women in the same community are allowing women more control over assets and diversification into economic roles, then On Norms and Agency there is some sense that they can “ride a wave� in their negotiations for change. This sense that a critical mass is developing can help accelerate reforms and has growing credibility in development project design. By extension, we expected that communities with vi- brant local organizations would show a stronger sense of collective action in support of gender “norm-bend- ers.� As chapter 4 suggests, however, community-based 151 APPENDIX A Methodological Note T he field work behind On Norms and Agency that affect gender norms and individual agency across (conducted under the title Defining Gender in sets of communities (Ragin 2008). the 21st Century) reached 97 urban and rural communities of Afghanistan, Bhutan, Burkina Faso, Do- We selected countries for the global study based on minican Republic, Fiji, India, Indonesia, Liberia, Moldova, three criteria: First, at least two countries per each world Papua New Guinea, Peru, Poland, Serbia, South Africa, region.1 Second, we chose countries where World Bank Sudan, Tanzania, Togo, Vietnam, West Bank and Gaza, country units had strong interest for the work to be con- and Yemen. Nearly 4,000 individuals from three genera- ducted there and wanted to learn from the study and in- tions participated in the study between June 2010 and corporate findings into their policy analyses and guidance March 2011. activities. And the third criteria were countries where a local research team with the required expertise on quali- The rapid assessment explored trends in gender roles tative data collection could be identified. Although the and norms and what the women and men participating study samples are small and not representative of their said drives their major decisions for education, econom- general country or regional contexts, we designed them, ic participation, and family formation. We set up small, at the community level, to capture a mix of urban and same-sex discussion groups and asked their members to rural contexts, as well as more modern and traditional reflect on questions about these decisions, for example: gender norms. In every country, research teams fanned Why and how did they decide to end their education? out into both middle-class and poorer neighborhoods of Are men and women better at different jobs? Do women cities and towns, and prosperous and poor villages. The and men save differently? What makes a “good� husband final sample of communities is listed in table A.1. (Names or a “good� wife? We employed qualitative methods, of the communities in the study are pseudonyms or re- which are appropriate for examining these questions ferred to by districts.) because they permit exploration of multidimension- On Norms and Agency al factors that need to be traced over time, as well as Within the study communities, five different data collec- contextually grounded for sound interpretation of their tion tools were used: three structured focus group dis- meaning and significance in the lives of the sample wom- cussions (an interview guide was prepared for each age en and men and their communities. These methods are group: adults, young adults, and adolescents2), one key in- the preferred approach for researching “how� and “why� formant interview centering on a questionnaire about the questions given that they “allow investigators to retain community with close- and open-ended questions, and the holistic and meaningful characteristics of real-life one mini case study.3 See table I.2 in the Introduction, events� (Yin 2003, 3). Our research design also enabled which summarizes the general topics covered with each 152 us to explore commonalities and combinations of factors method.) Focus groups lasted 2.5 to 3 hours on average. Because bias can sometimes be introduced by focus were trained by the core World Bank study team to fol- group dynamics, such as when more assertive group low standard methodological use of each data collection members dominate discussions, facilitators received tool. Both facilitators and note-takers pretested all data training in additional measures to foster inclusive discus- tools, including country-level additions and adaptations. sions that would capture a range of attitudes and experi- We asked the facilitators to stay as close as possible to ences common in the community. We also set up single- the interview guides, so that they posed ask the same sex focus groups of roughly similar ages, so participants questions and created the same visuals. The facilitators would feel safe and comfortable and answer frankly. Fur- also kept to the particular sequencing of the modules for ther, on some key questions, focus group members had each countries and groups. opportunities to respond in private and then volunteer to discuss their responses in the open (same-sex) group. With our encouragement, facilitators had license to probe more deeply into specific issues that cropped up Local researchers with extensive country knowledge in the focus groups. We asked them to let focus group and qualitative field experience led the focus groups. participants reflect on and reply to more general ques- The field team members recruited to conduct the focus tions first before exploring a subject in greater detail. group discussions and interviews were generally expe- For broad questions—such as, what does it mean to be a rienced facilitators who received supplemental training powerful woman in the community?—we gave them spe- and a detailed methodology guide in preparation for cific guidance on how to allow an open flow of answers, their field work. The methodology guide reviewed the instead of guiding responses in particular directions. study’s conceptual approach and sampling procedures, If participants met questions with silence or a subject presented each of the study instruments, and discussed appeared to be too sensitive for open discussion in a documentation and analysis techniques. group, facilitators either 1) moved on to the next ques- tion and perhaps returned later to the topic, or 2) found As part of the field work in each site, facilitators inter- alternative ways to address the issue, such as allowing viewed local key informants to complete a community them to answer in private or write confidential replies. questionnaire, which provided extensive background in- We also deliberately designed the order of discussions formation about the sample community. Key informants to present more sensitive questions in later modules were community leaders, government officials, politi- when greater rapport and trust had developed among cians, important local employers, business or financial members of the group. leaders, teachers, or healthcare workers. The selection of the participants for the adolescent and young adult The facilitators introduced all focus groups and infor- focus groups was based on specific age criteria. The field mants in the study to the objectives of the assessment, teams also received instructions to compose the groups, explained the type of information being solicited and as much as possible, to reflect the range of educational how it would be recorded, emphasized that partici- and livelihood experiences common in the community pation was entirely voluntary with no adverse conse- for that age group. quences for those who did not wish to participate, and described how they would ensure the confidentiality of We reiterated to the teams from the onset that this was a their answers (the data) and the participants’ anonymity. cross-country study. The teams had to balance concerns We made certain that participants understood that the On Norms and Agency for responsiveness to the issues and pace of a specific study was not intended to directly change any policies group’s discussion with the global study’s need for a core or services affecting their communities and that they set of data that could be compared systematically across would not receive compensation, financial or otherwise, all the focus groups across all the countries. The teams for joining the study. 1 As defined by the World Bank: Sub-Saharan Africa, East Asia and the Pacific, Europe and Central Asia, Latin America and the Caribbean, Middle East and North Africa, and South Asia. 2 Adults were 25–60 years old, young adults 18–24, and adolescents 12–17. 3 This was a detailed report of a finding that emerged as important for understanding gender norms or structures shaping economic decisions in that locality. 153 Some parts of the data-collection required the study generational differences, relevant information, notable participants to reflect on earlier periods of time, and we case (or gem), rural-urban differences, and gender or identified a specific period. When facilitators asked par- generational differences. ticipants to reflect back 10 years ago, they could assist their recall by substituting a fixed reference year for the The close-ended questions (where all participants gave term “10 years ago� and linking the baseline year to a their own opinions on a set scale of possible respons- major national event to strengthen their memories (for es) were treated as a numeric dataset , where we used example, a natural disaster, a change of government, the weighed frequencies and averages. Similar treatment end of a conflict, or an epidemic). was given to the database generated by the community questionnaire. Throughout this report, we give coding We based several data collection modules on specific vi- frequencies derived from NVivo—both the number of sual displays or material to encourage richer discussion, focus groups and number of mentions of a specific study such as the the ladders of power and freedom created topic—as guidelines for findings on certain themes. by each group, list of characteristics for a good and bad wife and husband, and causes and consequences of do- In order to understandspecific pathways for explain- mestic violence. For contexts where participants’ litera- ing the change in levels of power and freedom (from cy was limited, the research teams substituted symbols the ladder of power and freedom activity presented in for text. For example, a face with a big smile was used to chapter 4), we developed a model for qualitative com- represent “very happy� to respond to a question on hap- parative analysis (QCA). QCA is based on a Boolean piness in one of the modules. method of logical comparison that represents each case (which in this study was a community) as a combi- The close-ended questions included in the focus group nation of causal and outcome conditions (Ragin 2008). guide required individual responses from the group The analysis allows identification of different combina- members, rather than a consensus response, so that the tions of conditions that produce a specific outcome; in members of a group could not bias each other’s respons- our study, this was the perceived changes in power and es. The responses to these questions were recorded in a freedom during a 10-year period for women and men standardized spreadsheet (Microsoft Excel). in a community. We also electronically taped many focus group discus- A separate tercile analysis was also undertaken with the sions, but due to limited budgets and limited time for ladder outcomes and presented in chapter 5. We ap- full transcription, note-takers attending the discussions plied a STATA tercile distribution function to the data- recorded the majority of the documentation. Regard- set of mobility indexes (calculated as the difference less of recording, all focus groups included note-takers between the ladder mean step now and mean step 10 and their notes were added to transcripts of focus years ago). The terciles were assigned on a country-by- groups findings. The note-takers, as well as facilitators, country basis. were the same sex as the participants in each focus group. The global analysis phase of the rapid qualitative as- sessment was launched with an interactive writing The final dataset from the field work is narrative and workshop that brought together 18 of the 20 country On Norms and Agency numerical data. The study’s principal findings rest on team leaders with the study’s global assessment team systematic analysis of the content of the narratives, at the Rockefeller Center in Bellagio, Italy, in Septem- comprising more than 7,000 pages of text in the global ber 2011. The opportunity to share country findings, dataset. The text was treated like a single database to collaborate intensively on identifying important and coded with NVivo9, a social science software. We themes for the global report to address, and to reflect populated thematic nodes with portions of narrative on strengths and weaknesses of the study methodolo- text following a pre-determined node tree designed by gy greatly facilitated the analysis of patterns, as well as the lead research team. In addition, free nodes were the vital context-specific dimensions of gender norms, 154 inductively coded according to specific categories: aspirations, and agency. Table A.1. Communities in the sample Country Location Name of Community Rural Boyina Bagh (Kabul province) Urban Shirabad Ulya (Kabul province) Afghanistan Urban Karta-e Bakheter (Parwan province) Rural Naw Da (Parwan province) Urban Thimphu district Rural Paro district Bhutan Rural Samtse district Urban Mongar district Urban Ouagadougou (capital city) Burkina Faso Rural Sanmantenga (province) Rural Barsalogho (province) Urban Santiago de los Caballeros (large city) Dominican Rural Comendador (municipality) Republic Urban Hato Mayor (municipality) Urban Santo Domingo (large city) Urban Suva (capital city) Rural Naitasiri (province) Rural Sigatoka (province) Fiji Urban Lautoka (large city) Urban Levuka town (old capital of Fiji, city) Rural Labasa (province) Rural Koudipally Manda/Kowdipalle (district, Andhra Pradesh state) India Urban Nellore (large city, Andhra Pradesh state) (Note: two Rural Velugodu (district, Andhra Pradesh state) states were included in Urban Hyderabad (large city) India, each one treated as a Urban Bhubaneswar (city, Odisha state) country with 4 Urban Jaipur (city, Odisha state) communities each) Rural Kalahandi district (Odisha state) Rural Phulbani (municipality, Odisha state) Urban East Jakarta (large city) On Norms and Agency Urban Tangerang (large city) Indonesia Rural Sungai Puar (district) Rural Nagari Bukik Batabuah (district) Note: Community names have been replaced with pseudonyms and/or moved to the next administrative level (municipality, district or gover- norate) when research site is smaller than 2,000 inhabitants. For cities the name of the city is used to replace the name of the specific neighbor- hood. In the case of large cities, when possible, municipality or large neighborhood other level was added as identifier. 155 Table A.1. Communities in the sample (cont.) Country Location Name of Community Rural Tchien district Rural Tewor district Liberia Urban Buchanan (large city) (Note: Liberia was a pilot Urban Greenville district country. Double Urban Firestone district the number of required com- Urban Harper district munities were surveyed) Urban Monrovia (large city) Rural Suakoko district Rural Zorzor (large city) Urban Balti (large city) Rural Briceni district Moldova Rural Floreşti¬ district Urban Ceadîr-Lunga (large city) Urban National Capital district Papua New Guinea Rural Morobe province (Note: more communities Rural Village near Buka (capital city of Bougainville district) were included Rural Milne Bay province or Alotau due to diversity of provinces in Rural Village near Wewak (East Sepik province) PNG) Rural Western Highlands province Urban Lambayeque province Rural Cusco province Peru Rural Chiclayo (large city) Urban Saylla district Rural Justynow village Rural Martynice village Poland Urban Dobrowice (city) Urban Olsztyn (large city) Rural Sumadija district Urban Belgrade (capital city) Serbia Urban Kragujevac (large city) On Norms and Agency Urban Pomoravlje district Urban Sjenica (city) Urban Umlazi township AA Urban Umlazi township BB South Africa Rural Rural community (Ngonyameni area, Kwa-Zulu Natal) Rural Semi-rural community (Ngonyameni area, Kwa-Zulu Natal) Note: Community names have been replaced with pseudonyms and/or moved to the next administrative level (municipality, district or gover- norate) when research site is smaller than 2,000 inhabitants. For cities the name of the city is used to replace the name of the specific neighbor- 156 hood. In the case of large cities, when possible, municipality or large neighborhood other level was added as identifier. Table A.1. Communities in the sample (cont.) Country Location Name of Community Urban Khartoum (capital city) Rural River Nile state Sudan (North) Rural Blue Nile state Rural Al Fashir (capital city of North Darfur state) Rural Red Sea state Urban Emputa village (Bukoba municipality) Tanzania Urban Nsenene village (Bukoba municipality) Rural Malangachilima village (Dodoma region) Rural Zabibu village (Dodoma region) Urban Assoli préfecture Rural Ave préfecture Togo Urban Lomé (capital city) Rural Tandjouaré préfecture Urban Ba Dinh district (Hanoi) Urban Hoang Mai district (Hanoi) Vietnam Rural Hung Yen district Rural Kim Dong district Urban Neighborhood, Rafah governorate Urban IDP camp (internally displaced persons), Rafah governorate West Bank Urban Neighborhood, border area, Rafah governorate and Gaza Rural Dirbas, Hebron governorate Urban University Quarter, Hebron Urban Old City, Hebron Rural Jahran district (Dhamar governorate) Rural Kharef district Yemen Urban Aden (large city) Urban Ba’adan (city) Note: Community names have been replaced with pseudonyms and/or moved to the next administrative level (municipality, district or gover- norate) when research site is smaller than 2,000 inhabitants. For cities the name of the city is used to replace the name of the specific neighbor- hood. In the case of large cities, when possible, municipality or large neighborhood other level was added as identifier. On Norms and Agency References Ragin, C.C. 2008. Redesigning Social Inquiry: Fuzzy Sets Yin, R.K. 2003. Case Study Research: Design and Meth- and Beyond. Chicago, Il, USA: University of Chicago ods. 3rd ed. Applied Social Research Methods Series, Press. vol. 5. Thousand Oaks, CA, USA: Sage. 157 The World Bank 1818 H Street, NW, Washington, DC 20433, USA. www.worldbank.org