57627 World Development Report 2012: Gender Equality and Development Outline World Development Report 2012: Gender Equality and Development Overview 1. The evolution of gender equality across the world over the past quarter century is a story of contrast. There has been progress on many fronts. More countries than ever before guarantee women and men equal rights under the law in areas such as ownership of property, inheritance and marriage. One hundred and thirty six countries now have explicit guarantees for equality of all citizens and non- discrimination between men and women in their constitutions. On average, there are now more women than men in universities across the globe, with women's tertiary enrollment having risen more than sevenfold since 1970 (Figure 1). Women are drawn in ever larger numbers to the wage labor market in every region of the world. Figure 1 Tertiary education gender ratio, Figure 2 Adult literacy male-female gap, 2008 1970-2007 female literacy rate malefemale literacy gap 90 80 70 60 adult literacy rate, % 50 40 30 20 10 0 Source: UNESCO (2009). Source: World Bank (2010a). 2. In other parts of the world and for some aspects of gender equality, though, the situation has changed slowly if at all. In many countries of South Asia and Sub-Saharan Africa (SSA), girls are still much less likely than boys to attend, much less complete, basic education and become literate adults (Figure 2). The likelihood of women dying during childbirth in many of these countries is still at levels comparable to Northern and Western Europe in the 19th century. The social status of women and girls continues to be inferior to those of men and boys. Son preference in some societies in Africa and Central, South and East Asia still results in a sex ratio at birth that favors males ("missing girls"). The returns to women's labor and entrepreneurship continue to be lower in almost all countries--developing and developed--while most women spend longer hours than men in paid and unpaid work. Women continue i to cluster in sectors and occupations that are characterized as "female"--many of them lower-paying-- while in rich and poor countries alike men are overrepresented in dangerous professions like mining, fishing and fire-fighting. The representation of women in politics and in senior managerial positions in businesses remains far lower than men in almost all countries. And, almost everywhere, women and girls are more limited in their ability to move about and are more likely to be the victims of gender-based violence at home. According to available Demographic and Health Survey (DHS) data, the percentage of currently married women ages 15 to 49 who have experienced violence by their husband or partner in the 12 months prior to the survey is as high as 30 percent in Bangladesh and Kenya and around 15 percent in Haiti and Moldova. 3. New aspects of disparity have also emerged in recent years that work against men and boys. In many countries, boys are more likely to drop out of secondary school than girls and have higher repetition rates. Girls' test scores are higher than those of boys in certain subjects across the world. And in some advanced economies, women outperform men in university entrance examinations and enrollment in elite universities. In many countries in Eastern Figure 3 Life expectancy in ECA, 1980-2008 Europe and Central Asia, men bear a greater share of the burdens of ill health and women outlive men on average by 9 years--a larger ECA male ECA femalemale gap World female World male gap than elsewhere in the world (Figure 3). 80 And where conflict and fragility are life expectancy at birth, years 75 pervasive, as in much of Central Africa or 9.4 9.5 9.3 9.2 9.1 70 10.1 parts of the Caribbean, adolescent boys are 8.9 8.3 8.6 particularly vulnerable to violence and 65 crime. 60 Focus of the Report 55 50 1980 1985 1990 1995 2000 2005 2006 2007 2008 4. This Report will look at these facts and trends regarding the various dimensions of gender equality in the context of the Source: World Bank (2010d). development process. It will argue first that gender equality is a core development issue--a primary objective of development in its own right. Following Sen (1999) and others, the point of departure for the Report is a view of development as a process of expanding freedoms equally for men and women. It will then go on to analyze and delineate the two-way relationship that exists between gender equality and development. On the one hand, it will assess the extent to which sustained income growth--a key precondition for broad-based development-- contributes to greater gender equality, and identify where these positive impacts are greatest. On the other, it will explore whether and how greater gender equality can itself contribute to economic growth and development (see Box 1 for a summary of how gender equality is reflected in the Millennium Development Goals). In looking at these links, the Report will examine the evidence and draw on a growing body of analytical and empirical work in both developing and developed countries. ii Box 1 Gender and the Millennium Development Goals (MDGs) The 2010 MDG Summit concluded with the adoption of a global action plan to achieve the eight MDGs by 2015. It also adopted a Resolution calling for action to ensure gender parity in education and health, economic opportunities and decision-making at all levels through gender mainstreaming in the formulation and implementation of development policies. The Resolution and the action plan reflect the belief of the international development community that gender equality and women's empowerment are development objectives in their own right (MDG3 and 5), as well as serving as critical channels to the achievement of the other MDGs. Gender equality and women's empowerment help to promote universal primary education (MDG2), reduce under-five mortality (MDG4), improve maternal mortality (MDG5), and reduce the likelihood of contracting HIV/AIDS (MDG6). And it is on several of those goals--notably those for reductions in maternal and infant mortality and the incidence of HIV/AIDS--where the world and most low-income countries will still fall short of MDG targets. The inequalities that are reflected in the poorer performance of women and girls across several MDGs are compounded among women belonging to groups that are disadvantaged or excluded on the basis of other characteristics, such as income, location, race, ethnicity or disability. For instance, the female disadvantage in education and literacy outcomes is larger in rural areas and among lower-income households, particularly in parts of Sub-Saharan Africa, South Asia, and Middle East and North Africa. And, in several Latin American countries, indigenous and Afro-descendant women are less likely to be employed in nonagricultural paid employment. The 2010 Resolution also stresses that the achievement of the MDGs would require coordinated interventions that target women and other vulnerable groups across sectors that include: · Taking action to improve the numbers and active participation of women in all political and economic decision-making processes, including by investing in women's leadership in local decision-making structures and creating an even playing field for men and women in political and Government institutions; · Expanding access to financial services for the poor, especially poor women; · Investing in infrastructure and labor-saving technologies, especially in rural areas, that benefit women and girls by reducing their burden of domestic activities; and · Promoting and protecting women's equal access to housing, property and land, including rights to inheritance. 5. Gender equality will be defined in the Report in relation to three key dimensions: the accumulation of endowments, the use of those endowments to access economic opportunities and generate returns, and their application to take actions, or agency, affecting individual and household well- being. Endowments encompass education, health, land and other assets such as financial resources that women and men accumulate during their lifetimes. Many of these, such as basic education, are amassed at early stages in life reflecting mostly decisions of parents. Others, such as material assets, are acquired later in life through a variety of mechanisms such as productive labor and the exercise of inheritance rights. Access to economic opportunities determines how endowments and time are combined to generate income and consumption, which are a key dimension of well-being. The decision concerning time allocation between home and workplace, productive activity and leisure takes place in the household and will depend on both the value placed on women's potential contribution to the overall well-being of the household vis-à-vis other household members as well as views of gender roles and women's preferences and needs.1 Finally, agency will be understood as the ability to make choices and take actions related to oneself, one's situation in the household, and one's situation in the public sphere. Autonomy to influence 1 In line with Sen's "perceived contribution response" (Sen (1990); Agarwal (1997)). iii one's life is not only a key dimension of well-being; it also helps determine outcomes such as the accumulation of endowments or the use of those endowments. 6. This definition illustrates how gender equality is both similar to and differs from inequality based on other attributes such as race or ethnicity. Three differences are of particular relevance to the measurement of gender equality. First, the welfare of women and men living in the same household is conceptually difficult to measure separately, which is compounded by the paucity of data concerning outcomes within the household. Second, preferences can differ systematically between men and women reflecting both innate factors and "learned" behaviors while needs and constraints also vary across gender because of both biological and social factors. And, third, gender cuts across distinctions of income and class. While all women may share similar disadvantages relative to men, disparities between rich and poor women in the same country may be larger than those between men and women of similar incomes in that country. These characteristics raise the question as to whether gender equality should be measured in terms of equality of outcomes, or in terms of equality of opportunity. As discussed in Box 2, the economic and philosophical literature on this issue is divided. In light of this, the Report will take a pragmatic approach, focusing mainly on equality of outcomes when talking about gender equality in endowments and agency, but on equality of opportunity when talking about productive activities and economic choices. We will emphasize equality of outcomes for endowments and agency because, following Phillips (2004), we believe that outcomes in these domains convey essential information about opportunities. However, when it comes to gender equality in productive activity--for example, as reflected in choices of where, how and how much to work outside the home--we will choose to focus instead on equality of opportunity, acknowledging that differential preferences may lead men and women to make different choices. Box 2 Equality of what? Opportunities, outcomes and women's welfare This Report takes at its premise that the welfare of an individual should not depend on gender. Although the principle is clear, measuring welfare at an individual level is difficult. Unlike the literature on inequities across different groups, where there is a clear concept of consumption poverty based on household aggregates, the welfare of individual men and women is intrinsically an intra-household issue. And attempts to break into the household have not met with much success. One line of enquiry has tried to attribute consumption to a particular member of the household, but the results have never squared with reality. Even in many contexts where women's welfare is obviously low relative to that of men, these measures have shown little or no female disadvantage (Case and Deaton (2003)). A second line of enquiry asks whether other measures of welfare, notably happiness, can serve as a `proxy' for welfare. However, looking across countries at the percentage of women who report being very happy reveals little, or no, relationship with numerous indices that aim to measure female disadvantage. This suggests that the issue of how to measure women's welfare in a reliable and consistent manner remains a question that is very much open--and one that this Report is not going to take any steps to resolve. Having made that decision, we are still left with two options: should we define equality in terms of opportunities or in terms of outcomes? This issue is still the subject of much debate among philosophers, economists and development practitioners more broadly. An advantage of framing gender equality in terms of equality of opportunity is that it allows the analysis to distinguish between differences or inequalities that arise as a result of circumstances beyond the control of individuals, from those that exist due to differences in preferences and choices. A substantial body of research, much of it drawn from experimental economics (Croson and Gneezy (2009)), has documented systematic male-female differences in risk aversion, social preferences and attitudes towards competition. It follows that if men and women differ, on average, in attitudes, iv preferences and choices, then not all observed differences in outcomes can be attributed to differences in opportunities. However, some of the experimental literature also suggests that observed differences in preferences and attitudes are largely "learned" and not inherent--that is, they are the result of culture and environment due to which men and women internalize social norms and expectations leading to the behavioral differences observed in practice (Booth and Nolen (2009); Gneezy, Leonard, and List (2009)). In this context, an argument in favor of focusing on equality of outcomes is that persistent differences in power and status between men and women can over time become internalized in behaviors, aspirations, and preferences that perpetuate the inequalities. So, it is difficult to define equality of opportunity without also taking into consideration how actual outcomes are distributed. It is only by attempting to equalize outcomes that one can break the vicious circle of low aspirations and low opportunity. This Report will recognize the ambiguity on this issue by looking both at opportunities and outcomes. The approach we take is pragmatic. We are going to think of gender equality in two complementary ways. First, when referring to gender equality in endowments and agency, we will focus primarily but not exclusively on equality of outcomes. For us, increasing female primary and secondary enrollments to 100 percent is important as is making sure that girls do not face a disadvantage in what they are learning. This is because we believe that outcomes in these domains are the best possible measure of realized opportunities (Phillips (2004)). This is especially true if the way that opportunities are translated to outcomes differs for men and women, as we know it does for health. Or, as in the case of political voice, where equality of outcomes is an intrinsically desirable goal. Second, when it comes to gender equality in productive activity--as reflected, for instance, in economic choices of where, how and how much to work outside the home--the Report will recognize that differential preferences may lead men and women to make different choices with respect to their productive lives. And that these preferences may be very different across countries and regions. In this case, we will not argue that equal outcomes between women and men are better; we cannot show this since the only way to define "better" is to measure women's welfare, which is hard to do. Instead, we ask whether there is a way for policy to be implemented in an environment where consumer preferences are unknown. This question has a well established answer. As long as we can demonstrate areas or sectors that are subject to market failures, even without knowing individual preferences, removing these market failures will always improve aggregate welfare. Thus, in these cases the Report will focus on the identification of market failures, policies that can be used to alleviate such failures and evaluations of such approaches when they have been tried. A concrete example: In recent years, India enacted a quota for elected female leaders. In certain constituencies, reassigned every 3 years (so-called "reserve constituencies"), only women could stand for elections. Recent work (Beaman and others (2009); Bhavnani (2009)) shows that in constituencies that were first reserved and then unreserved, (a) male perceptions of the ability of female leaders improved dramatically and (b) women are now equally likely to be elected to leadership posts as men. The act of reservation led to a completely new outcome once the reservation was removed. This result suggests that there was a failure in the political market such that information about women as leaders was inaccurate. Our view is that solving this market failure (although this may not necessarily be through quotas) will increase welfare, even though the final outcome may well not be equality and will depend on the preferences of the voting population, which are unknown. v Key messages 7. The Report will emphasize four main messages: During the last quarter century, sustained growth in many countries has been particularly effective in reducing disparities on some dimensions of gender equality. And, the pace of change in these outcomes has been much faster in today's developing countries than it had historically been in high-income countries. 8. At the household level, higher incomes help relax the constraints that force parents to choose which children to educate, feed or provide with medical care, especially in response to adverse shocks. Richer societies find it easier to allocate public resources to expanding and improving the delivery of services such as education and health, which will disproportionately benefit girls because they are likely to have been excluded from these. And a growing economy supports the expansion of economic opportunities in the labor market, which is likely to benefit women who are initially shut out of the formal labor market. 9. The faster pace of change in today's developing countries can be seen in indicators such as female education and literacy, life expectancy and fertility. In the parts of the world that have seen significant income growth over a protracted period, the improvements in these indicators have been dramatic. For instance, the magnitude of the fertility decline in India over the past 50 years took close to 150 years to happen in the United States. However, economic growth--even at a pace more rapid and sustained--will not by itself be enough to bring about improvements across all dimensions of gender equality. 10. The benefits of even broad-based growth on gender outcomes do not go far enough in three respects. First, even in areas such as girls' education where disparities have generally been responsive to income growth, relying primarily on the economic development process will mean that it will take many decades to equalize the remaining gaps, such as those in parts of Sub-Saharan Africa and South Asia. Second, even in countries that grow fast, poor women tend to benefit far less than their richer counterparts. So, large gender disparities tend to persist for those in the lower income brackets. Third, for improvements in some dimensions of gender equality, such as those related to women's access to economic opportunities, control over resources and participation in politics, higher incomes per se are unlikely to make a significant difference because other constraints likely will continue to bind. The persistence of many of these gender disparities even in high income economies today, despite large gains in the civil and economic rights of women during the last century, suggests that these outcomes are the result of complex, slow-moving processes and deep structural factors that growth alone will not address. The costs to countries of maintaining gender inequality are likely to have become even larger in the globalized world of the 21st century. At the same time, globalization offers opportunities for more rapid changes in norms and beliefs concerning gender equality. 11. The increased momentum of globalization and growing integration of the world economy means that goods, labor, capital, ideas, and information move across the globe much more easily and quickly than before. This world of increasingly footloose capital and talent is also more competitive with small differences in costs and opportunities likely to loom large. Societies where a large part of the population does not have access to the same opportunities merely because they are of the wrong gender are likely to vi pay an even greater price in foregone efficiency and competitiveness than was the case in the world of the 1960s and 1970s. 12. Globalization also means that countries and societies, and their values and norms, are increasingly exposed to new influences. References and role models are more global than ever before and these changes can have a profound impact on the expectations and aspirations of young men and women. As evidenced by United Nations resolutions and the focus on gender equality in the Millennium Development Goals, this flow of ideas has paved the way for a growing global consensus on the importance of gender equality both in its own right and as a contributor to development. It has also meant that social norms concerning what is suited to men and women, and which get passed on from one generation to the next, shift more rapidly than in the past. Responses from the World Values Surveys over several years point to the evolution of views with respect to gender equality. For example, while on average people across the world express the view that men make "better" political and economic leaders than women, this view is much less pervasive among men today than it was a decade ago. This can be seen in Figure 4 where the observations for most countries lie below the 45 degree line for all age cohorts. Figure 4 Attitudes towards men and women leaders (percentage who believe that men are better political leaders than women) Source: Staff calculations based on World Values Surveys. There is an important role for policies targeted towards reducing the most costly gender disparities that are not responsive to growth, and closing the most egregious gender gaps has become more urgent now than it was two or three decades ago. 13. Countries across the world already recognize the need for such policies and intervene in a range of areas. These interventions include enacting laws that treat wives and husbands equally in cases of divorce, forbid gender-based discrimination in labor markets, and reserve a share of political representation for women. There are efforts to design the delivery of education and health services so as to equalize access across genders by taking into account cultural concerns and matching girls and women with female teachers or health workers. Other measures, such as gender-specific education or wage vii subsidies, target girls and women directly if they are thought to be disadvantaged by persistent inequalities. However, such policies have often been implemented without sufficient consideration to their cost or efficacy. 14. The design of such policies needs to take into account not only the economic costs of gender inequality but also the potentially transformative role of women in bringing about social, economic and political change. When women are free to seek an education or work outside the home, for example, their ability to contribute to their own development and that of their families, communities and societies is sharply enhanced. When men and women are able to take up jobs and occupations according to their ability and preference rather than gender norms, aggregate productivity may increase. And when countries are able to mobilize the equal participation of men and women in social and political life, governments and policy making are likely to gain in accountability and diversity of views and voices. 15. But it is also clear that not all interventions will be as effective and that some policies will be much more costly than others. This Report will look at how countries might evaluate alternative policies to correct gender disparities. In particular, we will show that there are measures that can advance both gender equality and economic development jointly. Where countries can mobilize the necessary political consensus behind such interventions, the dichotomy between economic development and gender equality becomes a false one--it should be possible to achieve both. Nevertheless, for some dimensions of gender equality, especially those rooted deeply in social norms and customs, there will be costs associated with corrective policies, especially in the short-term, and these costs need to be recognized in policy design and choice. What the Report will do 16. In looking at how gender outcomes are shaped over time and across countries in the context of the development process, the Report will address the following questions: What has been the nature and pace of progress in different dimensions of gender equality over the last quarter century? To what extent do these patterns vary across regions and countries as well as within countries? What explains the progress or the lack of progress in specific aspects and what has been the impact of economic growth? What factors and constraints have slowed down or prevented progress in certain areas even in the presence of economic growth? What are the most significant opportunities and new challenges regarding gender equality associated with the process of global economic integration? What is the rationale for public action to enhance gender equality? What types of policies can best promote greater gender equality in different dimensions? 17. In addressing these questions, the Report will propose a simple conceptual framework (Figure 5). Gender-related outcomes are determined by how gender roles, behaviors and aspirations play out in a number of spheres throughout the lifecycle. These spheres include the household and the decision making within it; access to labor, land and other markets and the ways in which these markets operate; and social norms, legal rights and how other formal and informal social institutions function. Outcomes at the individual and household level (endowments, economic participation and agency) are a reflection of viii the preferences of those who make decisions and choices within the household and of the incentives and constraints each of them faces in making these decisions. In this way, household decision-making, social institutions and markets interact to determine gender-related outcomes whether these relate to endowments, access to economic opportunities or agency. 18. These interactions take place Figure 5 Conceptual framework in the context of a broader environment. Government policies can shape the functioning of markets and regulate social conduct. Collective action can change traditional formal institutions or social norms. External shocks (systemic or idiosyncratic such as a financial crisis or a natural disaster) can affect the functioning of markets. Global economic integration and technological change introduce new opportunities and bring about new challenges that require adaptive responses from households, markets and institutions alike. 19. The interplay of these three domains--households, markets and institutions--provides important analytical and policy insights. Analytically, this approach provides a way to understand why there has been significant progress in certain dimensions of gender equality but only limited success in others. In particular, the Report will argue that institutional constraints and market failures can slow and even stall progress towards greater equality, even in the presence of strong and sustained economic growth. Focusing on the interactions between households, markets and institutions, rather than on each one of these elements independently, also helps clarify that multiple constraints may exist across different markets and institutions. Consequently, improvements on a particular dimension of gender equality may require coordinated interventions on a number of fronts. In terms of policy, the proposed framework makes it clear that public action can affect household behavior through its impact on existing markets and institutions, both formal and informal. And, it provides a rationale for public action in those cases where institutional constraints or market failures exist. 20. The Report will be structured into three parts. Part I will provide the factual foundation to the rest of the Report by assembling existing and new data to map out patterns and trends in gender equality over time, and across regions and countries. The goal here is threefold. First, we want to take stock of changes across and within countries in the last 25 years. Second, we aim to assess how economic growth has impacted gender equality and identify where (in what countries, among which groups, and in what domains of equality) progress is lagging. Third, we want to consider what these changes imply for the future. 21. Part II will be analytical. This part of the Report will delve into what drives the patterns of gender inequality described in Part I, and explore why progress has been faster in some domains than ix others. It will look at how decisions within households; their access to and the functioning of markets; and the formal and informal institutions interact to influence outcomes. In doing this, it will analyze the market and institutional failures that contribute to perpetuating gender inequalities and illustrate how addressing those failures can generate both equity and efficiency gains. It will also look at whether and when income growth itself may have adverse impacts on gender equality. Finally, it will try to draw new insights regarding the evolution of gender equality in a changing world. The value added of this part of the Report will come from drawing the links between markets, institutions and household decision making; approaching market and institutional failures as interconnected; and adopting a dynamic lens. 22. Part III will focus on policies. It will examine various rationales for public action to enhance gender equality. It will draw on empirical evidence and the impact evaluation literature, including the World Bank's work (see Box 3), to assess the effectiveness of different types of interventions and the role of collective action. The value added of this Part will come from first setting out a typology of policies that distinguishes between two types of policies. The first set aims to level the playing field between men and women whether by enhancing the equalizing effect of economic growth or by redistributing assets and opportunities. The second set of policies is targeted to gender-differentiated market or institutional failures that underlie specific inequalities. Because gender inequalities are often the result of multiple, often mutually reinforcing, market and institutional failures, this section will also examine the need to combine and/or sequence policy interventions to address political economy bottlenecks. It will also emphasize the potential for learning across countries. Finally, it will pay particular attention to issues of policy formulation, sequencing and coordination. 23. The rest of this outline describes the three-part structure and provides a brief chapter-by-chapter description of the Report. Box 3 The World Bank's recent work on gender equality The Bank recognizes the importance of gender equality for poverty reduction and development effectiveness. During the last decade, in particular, the Bank has worked actively to integrate gender into development policy and practice. The Bank's Policy Research Report (PRR), Engendering Development ­ Through Gender Equality in Rights, Resources, and Voice (2001), made a significant contribution to integrating gender issues into the Bank's analytical and operational work. Through a synthesis of research and empirical evidence, the PRR assessed the extent of existing gender disparities in rights, resources and political voice and their implications for poverty reduction and development. It concluded that gender disparities are costly and that economic development and targeted policy interventions can be mutually reinforcing in addressing these inequalities effectively. In the decade since the PRR, there has been a range of research and analysis focused on gender issues within the Bank. This work has included country-level analytical work, including self-standing Gender Assessments, components of Poverty Assessments and Country Economic Reports, and regional reports. And, on lending, almost half of all Bank projects approved in FY08 addressed a gender issue in their design, up from about a third two years earlier. Moreover, coverage of gender issues in economic sector loans expanded faster than in social sector loans. The Bank's gender policy framework (OP/BP 4.20) in 2003 and the four-year Gender Action Plan (GAP) in 2007 have helped to move gender issues into the mainstream of discussions of country strategy and policy. The GAP made the case for gender equality as "smart economics" because women's economic empowerment can help promote shared growth and accelerate the implementation of Millennium Development Goal 3 on gender equality. The GAP has focused, therefore, on key economic sectors such as agriculture, infrastructure, finance, x private sector development and labor markets. By providing financial support (about $70 million to over 200 activities), its goals are to intensify gender mainstreaming in Bank and IFC operations and country-level analytical and advisory work; to improve knowledge and data concerning women's economic participation and the relationship between gender equality, growth, and poverty reduction; and to strengthen the Bank's partnerships for women's economic empowerment with governments, other multilateral organizations, the private sector and civil society. An important focus of the GAP has been on measuring and documenting results from pilot interventions aimed at increasing women's economic empowerment, including through the use of impact evaluations. Building on the momentum generated by the GAP, efforts are underway to further strengthen the Bank's gender work. These actions aim at: improving the integration of gender issues in Country Assistance Strategies and in the Bank's project portfolio; deepening partnerships around issues of women's empowerment, including with the private sector (such as the Adolescent Girls Initiative); clarifying accountabilities within the institution regarding the analysis and incorporation of gender issues within Bank programs; and strengthening the results framework for the Bank's gender activities. xi Part I Dimensions of gender equality I.1 Part I provides the factual foundation to the Report by assembling existing and new data at the country and regional level to map out how different dimensions of gender equality have evolved over time and across the world. First, it highlights areas where progress has occurred and relates these changes to economic growth. Second, this section reviews domains in which progress on gender equality has been limited because income growth has been insufficient to bridge these gaps or there has been a compounding effect between gender and other forms of inequality. Third, it presents evidence of emerging challenges and opportunities in achieving gender equality in a dynamic, globalized economic environment. I.2 Equality between men and women starts with their access to and accumulation of human and physical endowments. Chapter 1 focuses on how the developing world has fared in this regard, with special reference to education, health, land and financial assets. Chapter 2 then looks at whether or not there are systematic differences by gender in economic opportunities and returns to these endowments. Specifically, it focuses on how individuals allocate time and resources to reap returns on their endowments, mostly in labor and asset markets. Chapter 3 examines the differential agency of men and women--their abilities to make choices and take actions. The emphasis is particularly on voice, political participation and mobility. Chapter 1 Endowments Over the last three decades, there has been a marked improvement in gender equality in human endowments across the world--both in the absolute status of women and in their status relative to men--and in some instances, the gender gap has reversed, with women now surpassing men. 1.1 Across nations, educational attainment had been on the rise for males and females. In 1950, women over the age of 15 in the developing world had gone to school for a mere one and a half year on average. By 2010, the average years of schooling in developing countries had risen to 6.5 for women, compared to 7.6 for men. In advanced economies, the average years of schooling for the adult population are 10.9 for women and 11.2 for men (Barro and Lee (2010)). And, at the primary level, gender parity has been reached in two out of three countries with available data. 1.2 Enrollments at secondary and tertiary education levels have also converged across genders in many countries, although significant gaps remain in some regions (Figure 1.1). The exception is Sub- Saharan Africa where more boys gained secondary school access relative to girls between 1999 and 2008, worsening gender disparities in this respect during this period. During the last decade, tertiary education enrollments surged in all regions of the world. On average, we now observe a gender gap favoring females. Female tertiary enrollment rates in 2008 lagged behind in 36 countries (out of the 121 with available data), but exceeded males in 79 countries. Across regions, MENA, and to a lesser extent South Asia, experienced the greatest relative growth in female tertiary student enrollments. 1.3 As countries move up the income ladder, gender disparities against females tend to first shift across educational levels, from primary, to secondary, to tertiary education, and ultimately, they reverse. The largest female gender gaps tend to occur in nations with low income and low overall participation in 1 schooling; the gender gap tends to turn in favor of women in nations with higher income and high overall participation in schooling. By 2008, boys in 51 out of 156 countries lagged significantly behind girls in primary/secondary school enrollment. Figure 1.1 Gender gaps in secondary and tertiary education, 2008 100 School enrollment, tertiary, female (% gross) 150 School enrollment, secondary, female (% net) 80 100 60 40 50 20 0 0 0 20 40 60 80 100 0 50 100 150 School enrollment, secondary, male (% net) School enrollment, tertiary, male (% gross) EAP ECA LAC MENA SAR SSA PRT1 45° line EAP ECA LAC MENA SAR SSA PRT1 45° line Source: Staff calculations based on World Bank (2010d). There has also been notable progress in life expectancy and child survival, and declines in fertility levels. 1.4 Biological, social and cultural factors interact with the environment and affect the health of men and women in different ways. Thus both sex and gender influence health outcomes. Sex differences in biology--genetics, physiology and hormones--play an important role in the onset and progression of disease. Men and women face unique risks and disease burdens due in part to different roles in reproduction and attitudes and responses to risk. At the same time, gender differences can shape the likelihood of seeking and receiving quality health care or compliance with treatment. 1.5 There has been a sharp and consistent improvement over time in life expectancy for both men and women worldwide. Female life expectancy at birth in developing countries increased from around 48.0 years in 1960 to 68.8 years in 2008, while male life expectancy at birth grew from 46.0 to 65.0 during the same period. In 1960, women had shorter life expectancies than men (42.0 vs. 43.6 years) in South Asia. Women in South Asia today are expected to outlive men on average by 2.7 years (65.5 vs. 62.8 years). 1.6 High maternal mortality rates (MMR)--deaths of women during pregnancy, childbirth, or in the 42 days after delivery--contribute significantly to excess female mortality in developing countries among women in childbearing age. On average, maternal mortality rates have experienced a substantial decline of over a third from 1990 levels (WHO and others (2010)), but many countries in Sub-Saharan Africa and South Asia still have extremely high rates. 1.7 Improvements in nutrition, health care and standards of living also brought about significant reductions in under-five mortality rates equally for boys and girls across the world. Between 1990 and 2006, there has been a reduction from 13 to 10 million in deaths among children under the age of 5 due to diseases such as diarrhea, malnutrition, pneumonia, AIDS, malaria and tuberculosis. In SSA, the probability of dying by age 5 in 2006 was 156 and 144 per 1,000 live births for boys and girls 2 respectively. On average, there are no regional differences in relation to underweight prevalence between boys and girls. Figure 1.2 Total fertility rate, 1962-2007 1.8 There is great variability in fertility rates East Asia and the Pacific around the world, but rates have steadily decreased on Europe and Central Asia Latin America and the Caribbean average (Figure 1.2). And the fertility decline in many Middle East and North Africa 10 developing countries has been much faster than it had South Asia SubSaharan Africa Births per woman been in developed countries. In India, for instance, fertility levels were high and stable through 1960 and 5 then sharply declined from 6 to 2.3 births per woman. On the other hand, in the United States, fertility declined gradually in the 1800s and early 1900s and it 0 took around 200 years to go from 7 to 2.1 births per 1962 1965 1967 1970 1972 1975 1977 1980 1982 1985 1987 1990 1992 1995 1997 2000 2002 2005 2006 2007 woman. However, fertility rates in some African countries, such as Nigeria, are still rising. Source: Staff calculations based on World Bank 1.9 While contraceptive knowledge is widespread, (2010d). usage remains low. SSA has the lowest contraceptive prevalence in the world, where only 21 percent of women ages 15 to 49 reported using modern contraception in 2007. The greatest gains in contraceptive prevalence during the last two decades have been in MENA, from 42 percent to 62 percent, and in South Asia, from 40 percent to 53 percent (WHO and others (2010)). Modern contraceptive use ranges from 2 percent of currently married women in Chad to 70 percent in Brazil. Progress has not been even across or within countries. 1.10 The speed of convergence in gender gaps in education appears to have accelerated over time. But some countries are lagging behind, as are population sub-groups within countries. Sub-Saharan Africa-- and particularly countries in West Africa--and South Asia have the largest gender gaps at present. In 2008, there were approximately 91 girls for every 100 boys in primary school in Sub Saharan Africa, and in South Asia there were 95 girls in primary school for every 100 boys. In many countries, there is a cumulative effect of gender bias against females over the educational lifecycle. In Pakistan, a girl in the 5-9 age group is 14 percent less likely to attend school than a boy, and her odds worsen to 24 percent as she moves into the 10-14 age group (Aslam and Kingdon (2008)). In 2008 in SSA, there were only 66 female tertiary students for every 100 male students. The lack of progress in these areas is all the more striking given progress achieved by other regions and countries at similar income levels. 1.11 Constraints to educational attainment also operate differentially across genders and new patterns have begun to emerge. On average, school repetition and, to a lesser extent, dropout rates are higher among boys than among girls. In many countries, girls outperform boys in school. In the United States, girls get better school grades in all major subjects, including math and science (Perkins and others (2004)). At the tertiary education level, women are equally or more likely to graduate from university than men in all regions, except SSA and South Asia. But there is also considerable sorting across fields of study. Men tend to be concentrated in the engineering, manufacturing and construction fields, while women tend to predominate in areas such as education, health and welfare, humanities and arts. 3 1.12 In health, while key outcome indicators among men and women have converged considerably, progress has not been uniform either. The age profile of female mortality varies significantly across countries and regions, indicating that different mechanisms are at play across societies that manifest themselves at different points during the female lifecycle. This is illustrated in Figure 1.3, which shows female excess mortality ("missing women") over the lifecycle for Sub-Saharan Africa, China and India. In SSA, female mortality spikes around their prime reproductive years (24-35 years) and as a result, it accounted for almost two thirds of all global maternal deaths in 2008 (WHO and others (2010)). Maternal mortality rates in Chad and Liberia are comparable to those in Sweden in the 18th century (Figure 1.4). Figure 1.3 "Missing women" distributed by Figure 1.4 Comparative perspectives on incidence of age maternal mortality 1400 Afghanistan 1200 Chad Sweden England & Wales Trendline (Sweden, 17501850 1000 Liberia Maternal Mortality Ratio 800 Tanzania SSA Sulfa drugs 600 Cameroon Burkina introduced in 1930's Malawi 400 Bangladesh Cambodia India 200 Indonesia DR LAC Poland China 0 Europe ECA Sri Lanka 1690 1700 1710 1720 1730 1740 1750 1760 1770 1780 1790 1800 1810 1820 1830 1840 1850 1860 1870 1880 1890 1900 1910 1920 1930 1940 1950 1960 1970 1980 1990 Source: Staff calculations based on WHO and others Source: Anderson and Ray (2009). (2010). 1.13 Globally, HIV/AIDS is the leading cause of death among women of reproductive age. Women and girls account for about 60 percent of the estimated infections in SSA. Most women with HIV contracted it through unprotected sex with an infected man. Women's vulnerability to HIV is linked to physiological differences, but is exacerbated by a variety of sociocultural factors, such as access to information and ability to negotiate condom use. In Côte d'Ivoire, HIV prevalence among women more than doubles prevalence among men (UNAIDS and WHO (2009)). Life expectancy in Botswana grew at pace with other regions around the world (with a small and constant female advantage) until 1990 but has fallen since then by almost 20 years due to deaths from AIDS. The greater longevity of women has also been eliminated. 1.14 Similarly, under-five mortality rates are usually higher for boys than for girls due to biological differences. Yet, also captured in Figure 1.3, in many parts of Africa, East, South and Central Asia, under-five mortality rates are significantly larger for girls, in part reflecting continued son preference in many of these societies. On the other hand, during the course of the last two decades men suffered a sharp and unexpected drop in life expectancy in many ECA countries, most critically in Russia, partially related to increased risky behavior, including smoking and alcoholism (World Bank (2007)). Much of the disease burden and premature mortality related to tobacco use affects poor men in particular (Guindon and Boisclair (2003)). 4 1.15 On average, within countries, gender gaps in education and health are small compared to gaps in outcomes across ethnicity, socioeconomic class or geographic location. In general, there is a compounding effect between gender and other forms of inequality. It is estimated, for instance, that approximately two thirds of out-of-school girls globally belong to ethnic minority groups in their countries (Lewis and Lockheed (2006)). And, in both India and Pakistan, while wealthier boys Figure 1.5 Median grade attained by income quintile and girls participate in school at similar rates, and gender, latest available year there is a gender gap of almost 5 years between them at the bottom quintile of the income distribution. Similar patterns can be observed in Benin, Democratic Republic of Congo, The Gambia and Togo (Figure 1.5). Likewise, differences in fertility levels between the wealthiest and poorest women tend to be small in countries with generally low levels of fertility--to the order of 0.5 to 1 child. But the gap grows dramatically at higher levels of fertility. In Zambia, the average fertility of a woman in the poorest quintile is 8.5 children Source: Staff calculations based on ICF Macro (2010). (the highest in the world); for a woman in the richest quintile, it is just over 3 children--closer to countries like Indonesia and Colombia. Gender disparities in access to physical capital and assets remain large and significant. 1.16 In many countries, women have limited control over assets. Land ownership and inheritance may be restricted to men only, both by tradition and by law. In some parts of Africa, according to customary law, women are unable to acquire land titles without their husband's authorization. Single women and single mothers cannot own land and widows lose access to their husband's land (Katz and Chamorro (2003); Doss (2006)). 1.17 Availability of cross-country data on land ownership at the individual level is at best limited. Deere and Doss (2006) document sharp disparities in Latin America, where women own as little as 11 percent of land in Brazil and 27 percent in Paraguay (Table 1.1). 1.18 In most countries, female-headed households (FHHH) are less likely to own land than male- headed households (MHHH).2 The gap between male- and female-headed households in land ownership reaches about 20 percentage points in Albania, Bangladesh, Tajikistan and Vietnam. FHHH also own less livestock in almost all countries with available data, except Nepal. Among landowning households, the 2 There is significant regional variation in whether land ownership is joint or individual. As part of our analysis, we will look at this question and hence the discussion of land ownership will use data both by the head of the household (for which there is a significant amount of data) and at the individual level (for which there is much less data) and, where possible, compare the two. 5 average size of landholdings tends to be Table 1.1 Land ownership by sex comparable across male- and female-headed Country/year Women Men Couple households, but since most households are male-headed, the vast majority of land is Brazil (2000) 11 89 n/a owned or controlled by men. Honduras (2001) 26 74 n/a 1.19 Studies on overall asset differences at the time of marriage in South Africa, Mexico (2002) 22.4 77.6 n/a Indonesia, Ethiopia, Guatemala and Nicaragua (2000) 22 78 n/a Bangladesh find male assets universally higher--13 times larger in Bangladesh and 7 Paraguay (2001) 27 69 3.2 times larger in Guatemala (Quisumbing and Maluccio (2003)). Hallman and others (2005) Peru (2000) 12.7 74.4 12.8 show that gender gaps on assets at the time of Source: Deere and Doss (2006). marriage have at best held constant or increased over time. 1.20 Although notable gender disparities in access to physical capital and assets remain de facto, legal inheritance and property rights for women have improved recently across a broad range of countries. For instance, in Tanzania, the Land Act 1999 and Village Land Act 1999 secured women's rights to acquire, hold, use, and deal with land, and can supersede customary law. In Eritrea, the 1994 Land Proclamation secured land ownership and inheritance rights to women. Nonetheless, legislative changes have in many cases proved to be insufficient to elicit changes in observed practices. The current distribution in land and other capital assets ownership speaks to a persistent female disadvantage. 1.21 The Report will assemble additional data on land ownership and other assets at the individual level for a broader set of countries and document longitudinal trends wherever possible. It will also document gender differentials and trends in financial assets and access to credit, an area where considerable expansion has occurred since the introduction of microfinance programs targeting poor women.3 Income growth and new technologies may also have unanticipated adverse effects on gender equality. 1.22 Economic growth may initially reinforce and deepen traditional gender differentials in endowments in the short term. In China, for instance, new rural industrial wage work opportunities led families to initially prioritize junior secondary education of males, leading to a considerable widening of the gender gap (Hannum (2005)). Access to education was expanded later to girls as the economy continued to grow. 1.23 In countries where parents continue to favor sons over daughters, gender bias in the form of sex- selective abortions, female infanticide or neglect of girl children is believed to account for millions of "missing girls" in early childhood. Anderson and Ray (2009) estimate that 184,000 girls in India and 3 Women account for 83 percent of the poor clients reported by over 3,500 microfinance institutions in 143 countries (Daley-Harris (2009)). 6 644,000 girls in China were "missing" at birth in 2000. The abuse of new technologies--such as cheap mobile ultrasound clinics--for sex-selective abortions accounted for much of this shortfall, despite such practices being outlawed in India in 1994 (Jha and others (2006)). Chapter 2 Access to economic opportunities and returns 2.1 This chapter will compile existing data from multiple sources to examine whether there are systematic differences across gender in access to economic opportunities and productivity to labor, entrepreneurship and capital. We will focus on how households--or rather the men and women that form households--allocate their time across different uses of labor (farming, self employment, wage labor, home production) and will highlight where choices and returns differ for men and women. The female-male participation gap in the labor market has declined and more women work outside of agriculture than ever before, but progress has been slower than with access to education and health. 2.2 Growing economic opportunities have Figure 2.1 Changes in female-male labor force drawn large numbers of new female workers participation gap, 1980-2008 into the market. Between 1980 and 2008, increases in female labor force participation (FLFP) were observed in all regions (with the exception of ECA) and across all age groups, with growth generally being faster in LAC, MENA and the developed world and among young adults. In Argentina, FLFP jumped from 39 to 51 percent during this period, while in Benin the share of female workers grew from 58 to 68 percent. On average, growth in participation rates has been higher among women than men and, consequently, the female-male participation gap has diminished Source: Staff calculations based on International Labor across the world (Figure 2.1). Organization (2010). But important differences still exist in terms of the distribution of men and women across industries and occupations. 2.3 Women tend to be concentrated in particular areas of the economic space and, within these areas, women are more likely to engage in lower-productivity activities than men. Women are more likely to be self-employed and work in the informal salaried sector than men (Figure 2.2). Female workers are also more likely to transition between the informal sector and being out of the labor force, and are less mobile between the formal and informal sectors (Bosch and Maloney (2009), Gong and van Soest (2004)). 7 Figure 2.2 Employment status and share in the informal sector Status in employment Informal employment in nonagricultural employment 100 80 100 80 Percent 60 percent 60 40 40 20 20 0 Guinea Guatemala Indonesia Benin Chile Kenya Brazil Honduras Philippines Egypt Dominican Rep. Tunisia Syria Algeria Venezuela Mexico India Morocco Thailand Costa Rica Colombia South Africa Bolivia Chad El Salvador 0 F M F M F M F M F M F M F M EAP ECA LAC MENA SAR SSA PRT1 Salaried employees Contributing family workers Own account workers Member of producer's cooperatives Employers Women Men F: Female ; M:Male. Years 1998-2008 Data for years 1994-2000. Source: Staff calculations based Source: International Labor Organization (2002). on International Labor Organization (2010). 2.4 Among the self-employed, women in the non-agricultural sector tend to operate small, informal businesses, often out of their homes. Eighty percent of industrial homeworkers in some developing countries, such as Thailand and Chile, are women (Beneria (2001)). Women working in the agricultural sector have lower yields than men on average. And, in both agricultural and non-agricultural sectors, self-employment is more likely to be a part-time activity for women than for men. 2.5 Similarly, there is an inverse correlation between the size of firms and the share of women who are managers (Bruhn (2009))--i.e. in most countries the majority of household enterprises are run by women, although not necessarily owned by women (Mead and Liedholm (1998)). On the other hand, the vast majority of large, formal firms are run by men (Klapper and Parker (2010), Sabarwal and Terrell (2008)). 2.6 In the agricultural sector, differences in Figure 2.3 Gender distribution across productivity between men and women are associated industries with differences in access to productive assets such as land and fertilizer, as well as with lower levels of mechanization. Outside the agricultural sector, differences in productivity (earnings) are associated with firm size (and capital intensity), lack of access to credit and lower levels of human capital among female entrepreneurs. There remains extensive industry and occupational segregation between men and women in the formal sector. Specifically women are more likely to be employed in agriculture, public administration, education, health and social work. Source: Staff calculations based on International These patterns are similar across countries and Labor Organization (2010). regions and do not appear to correlate with development or level of income (Figure 2.3). In turn, men are overrepresented in dangerous professions--like mining, construction, transport or heavy manufacturing--with high occupational injury 8 rates in all countries. Similarly, the burden of defense and maintenance of public order also falls heavily on males in all societies. Gender differentiated patterns in sector of employment and occupations contribute to a persistent gender gap in earnings. 2.7 Available data suggests that there Figure 2.4 Wage gap differences across nations, 1980-1990 remains a significant gender wage gap in favor of males, although this gap has Changes in Wage Gap in Manufacturing diminished slowly over time (Figure 2.4). Ratio Female:Male Wages in 2002-2007 1.4 Differences in average wages by gender range from 20 percent in Pakistan and 1.2 Mozambique to over 80 percent in countries 1 LKA BEL NZL FIN NLDFRA SLV AUS NOR DNK SWE as diverse as the Slovak Republic, Cote .8 GBR CHE BHR TWN EGY LUX CRI PRT d'Ivoire, Latvia and Jordan. SGP .6 JPN JOR HKG KOR CYP PRY 2.8 The size of the gender earnings gap .4 .4 is not related to income or level of .6 .8 1 Ratio Female:Male Wages in 1980-1990 Ratio Female:Male Wages 45° line development. Recent reductions in this gap can be partially attributed to overall Source: Staff calculations based on International Labor improvements in endowment levels among women relative to men, changes in the concentration of women in particular sectors and occupations, and shifts in work experience patterns and career interruptions linked to greater control over women's fertility (Chioda, Garcia-Verdu, and Muñoz Boudet (2010)). The control and allocation of women's time remain highly determined by the demands related to provision of care and reproductive roles within the household. 2.9 There are significant differences Figure 2.5 Average number of hours a day allocated to in male and female patterns of paid and domestic tasks for adult population in Latin America unpaid work (Antonopoulos (2009)). Women usually carry the time burden of Men Women domestic responsibilities--such as 6.4 7.3 8 5.6 5.5 collection of fuel or water--as well as 6 4.5 4.3 4.2 4.1 4.8 3.9 3 3.3 2.5 2.6 4 1.8 1.7 1.7 2 1.5 1.8 caring for children. Home activities 2 compete and crowd out other possible 0 uses of women's time (Gammage (2010), Blackden and Wodon (2006)). In most countries, women dedicate more time to unpaid work than men do (Figure 2.5). Source: Staff calculations based on Time Use Surveys. 2.10 A consequence of this uneven sharing of responsibilities among men and women is that women usually enter the labor market with the additional responsibility of domestic work, which affects and shapes the extent of their participation in labor markets and the returns on their endowments. Reducing 9 the burden of domestic chores on women's time can open up opportunities in other areas. Dinkelman (2010), in examining the employment effects of an electricity roll-out program in South Africa, finds that women's employment rates increased by about 9.5 percentage points, with even larger impacts for women with fewer child care responsibilities. In preparing the Report, household data on time use across different activities for men and woman will be documented and analyzed in selected countries. Chapter 3 Agency 3.1 This Chapter will document a set of stylized facts on selected dimensions of agency that are important both in their own right as well as determinants of gender equality in endowments and returns. This Report will focus on three specific areas of agency. First, it will cover the ability of women and men to make choices and take actions related to themselves. Second, it will focus on the ability to make choices and take actions within households. Finally, a third area of agency will focus on the ability to make choices and take actions in the political and economic spheres. Agency can change over the lifetime--with gains and with potential reversals in levels of agency during a person's life. Particular moments (shocks) when changes in agency levels can take place relate with lifecycle events such as marriage, divorce, death of spouse, births, and old age, and with labor market transitions. Household formation patterns are changing in part due to women's increase in agency. 3.2 Much of the world is seeing a delay in the Figure 3.1 Desired and observed fertility rates, age of marriage, an increase in cohabitation as well latest available year as a drop in the age of the first child for most 8 women. In South and Southeast Asia and West and Recent UGA89 NER06 7 Before Unwanted Fertility (number of children per woman) UGA06 MLI87 Central Africa, substantial declines have been MLI06 NER92 KEN89 MWI92 LBR86 ZMB07 ZMB92 BFA93 MWI04 GHA88 TGO88 SEN86 BFA03 6 RWA92 TZA92 observed in the proportions of women who are BEN06 NGA08 TZA05 MDG92 BEN96 NGA90 RWA08 CMR91 SEN05 JOR90 TGO98 LBR07 GTM87 CMR04 ZWE88 NAM92 5 MDG09 married between the ages of 15 and 24. These BOL89 HTI95 NPL96 PAK07 EGY88 KEN09 MAR87 GTM02 PAK91 GHA08 ZWE99 HTI06 declines are most marked in China and the Middle 4 BOL03 NPL06 PER86 JOR09 ECU87 NAM07 PHL08 PHL98 ECU04 SLV85 EGY08 BRA86 IND93 BGD94 East (Mensh, Singh, and Casterline (2005)). DOM91 3 PER00 SLV02 BGD07 COL86 IND06 IDN87 IDN07 BRA96 MAR04 DOM07 COL05 Moreover, women appear to have gained more 2 control over their fertility, partly closing the gaps 1 between actual and desired fertility levels (Figure 0 0 1 2 3 4 5 6 7 8 3.1). Evidence from household-level data also Total Fertility (number of children per woman) suggests that, on average, women desired fewer Source: Staff calculations based on ICF Macro children and, concomitantly, have had fewer (2010). children to match their expectations during the past decade. Freedom of movement and circulation remains a major constraint to women's agency. 3.3 Physical mobility relates directly to the ability of individuals to access services and more generally to take advantage of opportunities offered in a given society. Hence, mobility is important not only in its own right, but also in determining time allocations and the balance between tasks within and outside the household. Not surprisingly, therefore, freedom of movement and circulation for women and men as a right is widely protected under most national constitutions and international statutes. 10 3.4 But women still face significant restrictions on their mobility. These limitations have different origins--sometimes they are aimed at protecting women and girls from harassment, violence, and rape. For other women, mobility restrictions reflect limited resources--from vehicles to money to the safety of public transportation systems. In some cases, restrictions respond to societal proscriptions. For instance, in Pakistan, poor women's higher rates of unaccompanied mobility are associated with a loss of prestige and susceptibility to sexual violence (Mumtaz and Salway (2005)). In rural Yemen, norms dictate that women may ride cars only accompanied by a male family member, and may not use bicycles, motorcycles or donkeys (Aljounaidi (2010)). Norms bind behaviors for girls and women, but also for boys and men. 3.5 While reproduction has been central in defining female gender roles, social notions of "masculinity" place men in the role of providers and emphasize a code of conduct characterized by self- reliance and valor. The expression of these social norms in men's agency can lead to differential gender outcomes. In the United States and the United Kingdom, men are less likely than women to seek medical care for conditions such as depression, substance abuse or physical disabilities (Galdas, Cheater, and Marshall (2005)). And the inability to fulfill social expectations as breadwinners, for instance due to large rates of unemployment (compounded by social norms that frown upon male expression of vulnerability) can increase men's risk to psychological stress leading to destructive behavior such as violence or alcoholism. Men's higher rates of mortality and morbidity have also been linked to higher propensity of risk-taking behavior, such as reckless driving (Courtenay (2000)). In most countries, extensive gains in the granting of equality of rights to women have not been matched by practice. 3.6 In 1985, about 40 percent of countries in SSA, East Asia, South Asia, and MENA had ratified the Convention against all forms of Discrimination against Women (CEDAW), while almost 70 percent of Latin American and Caribbean and 65 percent of OECD countries had done so. At present, almost all countries recognize equality de jure and non-discrimination is guaranteed in their constitutions and other legal statutes.4 3.7 On the other hand, gains on paper do not necessarily translate into de facto rights because of norms, lack of information on granted rights or poor enforcement. For example, laws might mandate joint titling for couples, but land registrations take place under the sole name of the husband. Banks may require a male signature to grant a woman a loan. Constitutional rights may also sometimes be at odds with customary law where discriminatory practices under customary law can supersede constitutional rights (Figure 3.2). 4 In spite of being one of the treaties with the highest number of ratifications in the UN system, CEDAW has the largest number of reservations (by 22 states) to any human rights treaty. Most of these reservations focus on Articles 2 and 16. Article 2 requires States Parties to (a) condemn discrimination against women in all its forms and embody this principle in their national constitutions and legislation; (b) establish legal protection to the rights of women; and (c) modify discriminatory laws. Article 16 covers marriage and family matters. 11 With regard to political activity and Figure 3.2 Constitutional recognition of customary law (CL) in SSA countries voice, after more than half a century of voting 100% 1 rights, women's participation in politics still Recognizes CL allows lags. discrimination 80% 7 12 Silent about CL 3.8 Participation rates are lower for women 60% 1 than for men, in a general framework of 1 2 Recognizes CL silent on diminishing overall participation, particularly discrimination 40% 4 among youth. While the number of women Recognizes CL prohibits 10 discrimination with no parliamentarians has increased over time, mostly 20% 8 clear test due to legislation mandating quotas, change has Recognizes CL prohibits discrimination with clear been slow even with such policies. In some 0% test Civil law countries Common law countries instances, such as Argentina or Belgium, there has been steady improvement with a lag Source: Hallward-Driemeier (forthcoming). following the policy being instituted. In others, such as Brazil, there has been little change. Similarly, few countries have seen increases in women's participation in governments, regardless of the structure and size of cabinets. Even in 2010, only 6 developing countries and 11 higher income countries had cabinets in which more than 25 percent of the ministers were women. Part II: Analysis II.1 The discussion in Part I highlighted that while there has been progress on some dimensions of gender equality in many developing countries, significant gender disparities remain, particularly with regard to women's economic participation and productivity, and agency. The main objective of Part II is to identify the factors that have underpinned progress or the lack thereof in specific aspects of gender equality, including by examining how these changes relate to economic growth. II.2 To achieve this goal, Part II will apply the framework presented in the Overview, which posited that gender differences in outcomes result from interactions between households, markets and institutions. These interactions mediate the effects of economic growth, technological change, globalization and policies on gender equality. We will use this framework to answer four overarching questions: (i) How are gender-differentiated outcomes in endowments, access to economic opportunities and agency shaped by household decisions and their relationship with formal and informal institutions and the functioning of markets?; (ii) How do these relationships between households, markets and institutions mediate the impact of economic growth on gender outcomes?; (iii) How does the process of global economic integration affect both outcomes and their pace of change?; and (iv) What does this analysis tell us about the role for policies and their impacts? II.3 Addressing these questions is relevant to the policy discussion in Part III of the Report. First, it will help us understand how economic growth plays out in terms of gender outcomes. In those domains where economic growth has particularly beneficial effects on gender equality, there will be less need for public policies targeted at improving outcomes. But where gender inequalities are not likely to improve significantly as a result of growth, the design of specific interventions will be aided by a deeper 12 understanding of exactly how these outcomes are determined. Second, it will help us anticipate how global changes, such as those driven by technology or growing inter-dependence of countries, are likely to affect gender equality. Third, it will reveal to us where and how public action will be most effective, laying the foundation for the more in-depth discussion of policies in Part III. II.4 The empirical approach to be used here will rely on a combination of new analytical work and a critical review of the existing literature with a particular emphasis on causal evidence arising from policy evaluations and other studies. This will be complemented by a new multi-country qualitative assessment of differences in economic decision-making (Box II.1). Proposed conceptual framework II.5 Our approach starts from the premise that households make decisions that determine both individual and household outcomes--whether in terms of education, health, asset ownership, economic participation, labor productivity, or agency--on the basis of preferences, incentives and constraints. Preferences, incentives and constraints are themselves shaped by markets (which determine the economic returns to household decisions), by formal institutions (including laws and service delivery systems), by Figure II.1 Interactions between households, markets and institutions informal social institutions (such as social norms, values and social networks), and by the multiple interactions between all three. The process, moreover, is not static. Over time, the decisions and outcomes across all households (for example, more women entering the labor market) in turn feed back into shaping markets and institutions (including social norms) themselves, underpinning a process of dynamic change. This framework and process is captured in Figure II.1 and further elaborated below. II.6 A key element of this framework is the household decision making process, which is shaped by the preferences and bargaining ability of individuals within the household. Gender roles, behaviors and aspirations of household members play an important part in this process. Preferences related to gender are either innate or constructed (that is, shaped by social norms and values, as well as rooted in different notions of social identity). They may also be shared by or differ across individuals within the household. Evidence from evaluations of transfer programs, such as pensions and conditional cash transfers, which shows that spending decisions differ depending on whether money is given to women or men within the household, 13 has been interpreted as suggestive of differences in preferences by gender (Duflo (2003), Lundberg, Pollak, and Wales (1997)). II.7 The bargaining ability of individuals within the household is determined by two distinct sources of bargaining power: economic and social.5 Economic bargaining power is primarily a function of the amount of wealth and assets controlled by each individual. Social bargaining power, which embodies one aspect of agency, is shaped by existing institutions including social norms and values. Both are influenced by the possibility of the individual leaving the household ("exit"). For instance, a woman's bargaining power over time allocation within and outside the household may be weakened by her lack of ownership of household assets or by the presence of strong negative social proscriptions against female work outside the home. Similarly, where custody of children in the case of divorce is based primarily on the financial capacity to care for the children, women's capacity (and willingness) to use the threat of "exit" as a bargaining tool may be limited. Box II.1 Gender differences and decision making? A multi-country qualitative assessment To complement the quantitative analysis in the Report, a multi country qualitative assessment to explore gender differences in economic decision making has been launched. The aim is to understand how contextual factors, power relations and norms, come into play in women's and men's decisions. Looking at the differences between women and men, between youths and adults, and between urban and rural communities, this analysis explores decisions surrounding school to work transitions, job choices, family formation, the accumulation and control of major assets, gender roles and power relations. The questions also look into changes over the past ten years in women's and men's roles and views, and whether or not the gender norms shaping these roles have been changing. Countries considered in the assessment represent all regions and include the following: Liberia, South Africa, Tanzania, Burkina Faso and North Sudan; Indonesia, Vietnam, Papua New Guinea and East Timor; Moldova, Serbia and Poland; Peru and Dominican Republic; Palestine and Yemen; Afghanistan and Bhutan. While this work is still ongoing, some preliminary results suggest that: Views on gender roles are changing although less for older men than women or young men: When placed in the hypothetical situation of a woman willing to start a business, for instance, Serbian young men are more optimistic than older men about the woman's possibility to start up her own business without her partner/husband's support. Among older men, women are perceived as lacking both connections and capability to start a successful business, especially because this business would negatively affect the women's attention to their `household and family duties.' Women's success is evaluated on the basis of her capability to balance the roles of mother, wife and worker; while a `good husband' is equated with earning an income. While women often say that the husband will support the woman's business if it proves to be successful (e.g. "my husband would be enthusiastic if things developed well, he would also be better off, he would join me in the business"), men tend to disagree and often articulate the view that a successful enterprise launched by their wives will lead to divorce. There are gender differences regarding time allocations: When asked about what a woman does during the day, young men in Liberia did not include time for women to allocate for paid work; in their view women's days are divided into "housework and relaxation." When asked if they would make any changes to their daily 5 The bargaining process should not be interpreted as being always confrontational. In fact individuals within a household are expected to share goals and aspirations and hence to engage in cooperative behavior regarding a large number of decisions. And, in certain instances, individuals can be assumed to make decisions about specific dimensions of their lives independently of other household members. 14 routine, some young men suggested they would not do housework (considered to be a woman's job), but others said that they would like to reduce their paid work time to help their wives. When women in the same town were asked what they would like to change about their time allocation during the day they had a difficult time answering: not working, not cooking or not doing chores was rarely seen as an option. Men and women have similar views on good and bad jobs: When asked to rank the best and worst way to make a living, women and men in Peru see the best jobs as those related to having an education, being independent (having your own business), and formal jobs (like waiting tables or working in a supermarket). Women added secretarial jobs. When asked about the worst jobs, both men and women identified these as "male jobs" such as construction or truck loading: their perceived lower status was associated with high physical effort, long hours, difficult working conditions or relatively low pay. II.8 Besides relative bargaining power, decisions that determine gender outcomes at the household level are also influenced by returns to resources today and those anticipated in the future. And these in turn also depend on markets and on institutions, both formal (such as legal rights) and informal (such as social norms). So, for instance, the decision about whether to send a boy or a girl to school will to some extent depend on how much parents benefit from a son's or daughter's education, which depends not just on market returns to education but also on customs regarding marriage and elder care, among other factors. Similarly, time allocations by women and men to different activities (including work within the home) will respond to existing returns to work in and outside the household, but also to perceptions regarding gender roles. II.9 As noted earlier, the impact of economic growth on gender outcomes is mediated by these relationships. For instance, economic growth could promote greater integration of women into the economy through increased access to employment opportunities and higher returns to market work. This process could be further aided by shifts in the productive structure--such as the expansion of employment in the service sector; or dampened by them--if for example, growth is driven by sectors traditionally more closed to women such as the extractive industries. Similarly, higher income levels may alleviate budgetary constraints among relatively poorer households. And if household budget constraints initially limited schooling for daughters more than for sons, the former would benefit relatively more from this process. In other areas, the net impact of economic growth may be to accentuate gender biases--for example, in parts of East Asia the initial response to fast income growth was to prioritize secondary education for boys over girls because wage employment opportunities favored males. The point is not that growth is necessarily good or bad for gender equality--but rather that its impact can only be fully understood in the context of how it works through the relationships between household decision-making, markets, and social institutions, as shown in Figure II.1. II.10 The gender-differentiated outcomes (e.g., education, assets and productivity) that result from this decision making process can be efficient or inefficient,6 as well as equal or unequal. These distinctions are important for thinking about policy, as interventions can target inefficiencies or they can target inequalities. Policies targeted at inefficient outcomes have the potential to make everyone better off while those that address inequality directly do not. Nevertheless, because some efficient outcomes could also 6 When outcomes are efficient, no single individual can be made better off without making somebody else worse off (subject to a constant level of resources/income). In contrast, when outcomes are inefficient at least one individual can be made better off without making somebody else worse off (again subject to a constant level of resources/income). 15 be highly gender unequal, societies that value gender equality may still choose to put in place policy interventions aimed at making outcomes more equal between men and women, regardless of their efficiency costs. The policy discussion in Part III will elaborate further on this notion of policy interventions that target inefficiencies and policy interventions that target inequalities. The Report will further argue that both types of policies will be needed in all societies that value development. II.11 In the context of the conceptual framework laid out in Figure II.1, inefficiencies can arise both in the decision-making process within the household as well as in the actual and anticipated market returns to which households respond in making investments and allocating time across genders. And in both respects, the inefficiencies reflect institutional constraints and market failures. Institutional constraints result from failures either in the determination of rights or in enforcement of rights due to customary practices, social norms and other informal institutional arrangements. Market failures, on the other hand, are associated with information asymmetries, externalities, enforcement problems or limitations in the type and nature of contracts available.7 II.12 Although we have described the proposed framework in terms of a snapshot at a point in time, we will also use it to look at changes over time. Allocation decisions and agency formation within the household take place in a dynamic context, so that outcomes today can affect both the decision-making process and returns in the future. For example, as individuals work and earn income, they can accumulate assets and this, in turn, will affect their bargaining power and thus household allocation decisions in the future. In addition, increased access to assets may mean that the household or certain individuals within the household can also command higher returns and/or increase their productivity and, ultimately, earn higher income. II.13 Moreover, while the capacity of single individuals and households to change the way in which markets and institutions function is limited, the nature of institutions and markets is shaped and conditioned by society (understood as the sum of individuals and households). And it is precisely this process that allows institutions, including social norms and values, and markets to evolve over time, as a result of policy interventions, collective action, or exogenous changes brought about, for instance, by exposure to norms, values and practices in other societies due to global integration. II.14 Moving forward, Chapters 4 and 5 will apply the framework described above to analyze the process underlying the determination of endowments (education, health and physical and financial assets) and access to economic opportunities. Chapter 6 will examine the determinants of and constraints to agency formation, as well as analyze how agency itself shapes the resource and time allocation processes. It will also explore issues related to the evolution of social norms and values, and examine social and political participation as reflecting potential institutional constraints that create and/or perpetuate gender inequality. Finally Chapter 7 will re-examine the proposed framework in the context of increased globalization (both economic and social) to ask how greater global integration is likely to influence and change the relationships described above. 7 With information asymmetries, market participants are unable to gather or have differential access to necessary information. Externalities imply that the behavior of certain market participants affects the outcomes of others in the same--or another--market. 16 Chapter 4 Choosing endowments 4.1 This chapter will apply the above framework to explore how markets, formal institutions, and social institutions interact with household decision-making to determine the endowments of health, education, and assets of different household members. In particular, it will focus on two interlinked decisions made by households, both of which have implications for current and future gender outcomes: endowment decisions relating to investments in children in the household and endowment decisions relating to adults in the household, including parents themselves. In both instances, the chapter seeks to shed light on how gender differences in social and economic bargaining power, returns and income shape decisions and hence outcomes. 4.2 These decisions define future paths and directions for individual household members in ways that, once determined, are difficult to change. So, gender-specific choices by households regarding the education, health care and nutrition of its male and female children will determine in large measure their endowments of human capital and their ability to earn income and exert control over their lives, as well as their initial survival. Decisions relating to adults concern not only their survival, but also their ability to rebound from shocks (health), their effectiveness in the workplace (education/training), and their ability to run a business or farm (assets). Finally, it is important to keep in mind that there are clear trade-offs across the endowments of adults and children, and this will be especially evident in environments with particularly limited resources due to poverty and/or shocks. 4.3 This chapter will focus on endowments in health, education and assets in sequence. However, one should not think of these endowments as totally separable: outcomes in each of these dimensions are linked and reflect a common underlying decision-making process. For instance, the decisions to invest in a child's education and to bequeath assets to that child could both reflect expectations of benefits from the child's future earnings. Income levels and growth shape all dimensions of endowments 4.4 A critical factor for the decisions regarding investments in human capital is income, which matters both at the household level and at the national level. Indeed, as the higher income levels of countries increase the supply of services by government or other service providers tends to expand and households' financial constraints are relaxed through higher returns in labor markets. The overall effect at the country level is documented by the closing of education gaps as income rises (as presented in Part I). At the household level, an example of the impact of increased income is found in the decline in child labor and increase in schooling from the expansion of social pensions in South Africa (Edmonds (2004)). 4.5 In some regions, however, the income constraint can be more binding for girls than for boys. For instance, girls in China and Turkey were found to be more likely than boys to drop out when income is a constraint (Brown and Park (2002) and Tansel (2002)), and boys in Pakistan and India were more likely to be enrolled in fee-levying schools than girls (Aslam (2009) and Kingdon (1996)). Part I also presented evidence on the occupational segregation of boys and girls in higher levels of education, driven both by market failures in the labor market and by social norms. As economies grow, these can change, but the direction of that change can vary. For instance, while rural industrial wage work opportunities for boys led families to initially prioritize junior secondary education of males in China, Sri Lanka experienced a pro-female bias linked to the higher returns to secondary education for girls (Himaz (2010)). Similarly, 17 the higher opportunity cost of education for boys who can command earnings from a younger age can result in lower investments in education (see, for Brazil, Duryea and Arends-Kuenning (2003)). 4.6 Evidence from households experiencing income shocks also points to differentiated impacts on boys' and girls' enrolment--but not always to the detriment of girls and not always in a negative direction. Indeed, while evidence from Indonesia, Cote d'Ivoire or Ethiopia points to an overall decrease and a disadvantage for girls, children in Mexico or Nicaragua during various economic crises since the early 1990s saw increased schooling, as their earning potential decreased (Chaudhury, Christiaensen, and Asadullah (2006); Cameron and Worswick (2001); Jensen (2000); Maluccio (2005); Mackenzie (2003)). This suggests that where consumption levels are much closer to subsistence levels and where credit constraints are more likely to be binding, shocks result in reductions in schooling. 4.7 As was the case in the context of education, in countries at higher levels of income, women appear to benefit from significantly increased access to sexual and reproductive health services. This is likely a reflection of greater availability and accessibility of services (formal institution of government and private providers) as well as higher household incomes. Within countries, access is also typically lower for poorer women (World Bank (2003)). 4.8 As incomes rise, however, men seem to show a disproportionate incidence of health problems such as obesity, smoking and alcoholism, typically associated with risky behaviors (although there is evidence of similar patterns emerging for women in many middle and high-middle income countries, especially in Europe and Central Asia). Thus, the nature and extent of differences in health problems experienced by males and females may change with economic growth. Labor market structure, social norms and service delivery also determine education outcomes 4.9 Besides income, social norms can play a large role in education decisions. For example, girls may be withdrawn from school when they reach puberty (due to purdah in some countries) or because of early marriage. Also, the social norms that dictate the allocation of housework can act as barriers for girls, who are responsible for the care of younger siblings or fuel and water collection. Improved access to water or the presence of adult women in the household result in significant increases in enrolment for girls (see Koolwal and van de Walle (2010), Canagarajah and Coulombe (1997) and Assaad, Levison, and Zibani (forthcoming) for examples from Egypt, Ghana, Morocco, Nepal, Pakistan and Yemen). On the other hand, boys can be at a disadvantage in some contexts, when gender norms push boys into risky behaviors or force them to assume the role of breadwinners at an early age (Bannon and Correia (2006)). 4.10 These social norms can also carry through to the quality of education. Research into teachers' attitudes towards students in schools in Ethiopia and Guinea indicate that they are more positive in general about the participation, interest and intelligence of boys rather than that of girls (Colclough, Rose, and Tembon (2000)). In another context, simply reminding women of these stereotypes or highlighting their gender--the "social identity threat"--lowered their outcomes on math and engineering tests compared to those of a control group (Bell and others (2003)). Thus, social factors that might affect the parental decision over whether or not to educate girls and boys can also bind on children directly during their education. 18 4.11 Finally, formal institutions also play a role in the determination of investments in children's education. For instance, patrilocal marriage traditions could provide incentives for greater investments in boys, while countries where parents can obtain benefits from their daughters could exhibit more gender equality in educational outcomes. The introduction of legislation that granted daughters equal ownership rights in joint property in India in 1994 not only increased their likelihood of inheriting land, but also in greater educational attainment for girls (Deininger, Goyal, and Nagarajan (2010)). 4.12 The formal institutions that provide education services (governments or private sector) are also critical in the investment decisions. In addition to the income effect mentioned earlier (whereby greater GDP per capita results in greater education spending), the quality and form of service delivery can have a gender-differentiated impact. For instance, evidence on the negative impact of distance to school on girls' attendance (relative to boys) exists in various countries, and points to either safety issues or to the opportunity cost of transport in terms of reduced time for household chores (Aslam (2009) and Vuri (2008)). The expectations of teachers with respect to girls' performance influence learning and education quality differentially. Market structure, social norms and service delivery also matter for health outcomes 4.13 Health is an area where biological differences create different needs for males and females and thus, in some of its dimensions, simple equality does not provide an adequate benchmark. One obvious example of this is in reproductive and sexual health care. These patterns of biological differences creating different needs are also evident in HIV/AIDS, where biology places women at greater risk of infection. 4.14 While access to health care for women and girls has improved with economic growth in many settings, health outcomes themselves have at times failed to improve. Part of this lack of responsiveness can be due to issues around the quality, relevance and accessibility of health care. The adaptation of services to women's needs and to social norms regarding gender (with female physicians or with greater privacy for instance) could be a key factor in ensuring increased access results in better outcomes. 4.15 Investment in women's health and their access to health care (and in particular to reproductive health care) may also be constrained by social norms related to women's control over their own body and to the value of women's health. A recent study in Zambia (Ashraf, Field, and Lee (2010)), for instance, found that women were much more likely to use (concealable) contraception and have fewer births if they received family planning services alone than if they were with their husbands. Social norms and women's control over their own sexuality (in her ability to influence her partner's behavior), as well as issues of security play a strong role in women's ability to avoid HIV/AIDS and other sexually transmitted diseases. US Agency for International Development (2008) highlighted the lack of autonomy of women in decisions related to their own health care. Estimates from the most recent Demographic and Health Surveys in 40 countries reveal that overall, only 20 to 25 percent of women decide on their own health care in South Asia and Sub-Saharan Africa, another 30 to 40 percent decide jointly with their husbands or another household member, while over 40 percent have no say at all in the matter. Greater understanding and knowledge of the importance of women's health for her children's health can be critical to shift social norms around the needs of pregnant and nursing women and promote healthy nutrition and prenatal and delivery care. 19 4.16 One final issue that the Report will examine with respect to health is that of son preference. While this is clearly driven by social norms, the modes of achieving this can change as new technology becomes available--shifting from differential child mortality to selective abortions. The allocation of assets is strongly shaped by formal legal institutions and social norms 4.17 The third area of endowments the Report will examine is assets, including physical and financial assets. For children, the decision on assets is primarily a bequest decision made by parents. And, these decisions will be shaped by social norms and laws on inheritance, which establish constraints and expectations over what is appropriate for boys and girls to inherit and, to the extent preferences differ, by the relative voice of each parent. 4.18 Figure 4.1 shows the diversity across countries in equality of inheritance rights to property. A third of the countries presented here (total of 95 countries) have a Figure 4.1 Inheritance rights across the world score of 5, which denotes equal 35 rights in the formal legal system (and customary practices that do 30 not take precedence in ways that 25 disfavor women). At the other extreme, a third of the countries 20 have formal legal frameworks 15 that do not grant equality in 10 inheritance to women and men. The rest of the countries have 5 systems which lie in between 0 these two extremes. Inheritance 5 4 3 2 1 practices can also affect elderly women upon the death of their spouse, especially in cases when the assets (land and housing) are passed to children or to the deceased's family and the spouse finds herself without a home (Strickland (2004) and Human Rights Watch (2003)). Similarly, some of the rules around the inheritance of pensions upon the death of the pension holder can leave elderly women in vulnerable situations, when children can claim large shares of the pension. The decision to bequeath particular assets to children will be also be influenced by social norms, in particular around family structure, with investments in children likely to be driven by future benefit of parents from those investments. The Report will review existing rules and practices in these areas. 4.19 For adults, asset allocation decisions are shaped by returns in the labor market (with all of the attendant market and institutional constraints shaping these) as well as by social norms and laws that impinge on their ability to acquire and control property. Looking at laws across countries, the disadvantage of females relative to males is not as stark as it is for inheritance, but there are significant differences in these rights as well across different regions of the world. Eleven countries exhibit unequal property ownership rights (thirty four presented differences in terms of inheritance). These differences range from countries like Cameroon, where husbands have administrative rights over property and wives cannot conduct legal transactions without their husbands' approval, to Indonesia where husbands manage 20 their wives' personal assets, to Chile where the default marital regime gives husbands administrative rights over conjugal and their wives' property, and to Nepal where land tenancy rights are more limited for women (World Bank (2010c)). 4.20 More limited ownership of assets (either through limited inheritance or lack of control over their own assets) can also partly explain women's lower access to financial services (e.g. due to lack of collateral). Out of nine countries for which data is available, rural female-headed households have less access to credit than their male counterparts in seven of them (exceptions were Ghana and Panama). As a counterpoint, female-headed households in China appear to have the same access to credit as male-headed households, in a context of equal access to land and labor (de Brauw and others (2008)). More generally, when access is compounded by the amounts and terms of the loans, evidence suggests that rural women have significantly lower access. 4.21 Overall, this chapter will show how social norms and formal institutions can affect the allocation of endowments both explicitly (for example, laws that constrain inheritance patterns), or more subtly when they shape individual preferences and expectations (for example views on sending girls to school). Moreover, they can bind at different points, not only affecting the amount invested, but also the quality of that investment. Households either end up reinforcing or overcoming the constraints set by institutions. The returns to economic activity defined in the markets also provide powerful signals on which allocations are made. Distortions in the labor market can lower the incentive to invest in a daughter's health while inheritance patterns can intensify the preference for sons over daughters. These returns, as we will see in the next chapter, are, in turn, also shaped by institutional constraints and market failures. Chapter 5 Accessing economic opportunities 5.1 The evidence in Part I highlighted two empirical regularities across countries and over time. First, despite recent increases in female labor force participation, women across the world are still more likely to be responsible for household chores and child care and less likely to engage in market-oriented activities than are men. And second, women who participate in the labor market tend to be disproportionately concentrated in self-employment and the informal sector, and are less productive on average than their male counterparts. So, on average, women still tend to have lower earnings than men. 5.2 This chapter will examine the factors and forces that underlie these observed gender differences in access to economic opportunities and productivity. In doing this, it will take account both of the determinants within households of the decision to allocate time across various activities as well as the factors that lead to differential productivity and earnings between men and women. Following the framework in Figure II.1, the decision about allocating the time of different individuals within a household, both to activities in and outside the home, is made in light of their preferences, incentives and constraints. Time, allocated in this manner, is then combined with other inputs (such as capital and land) to generate income, which if generated from market activities, is determined by prices. 5.3 Four broad activities are considered across which time is allocated at the household level: housework (which is non-remunerated), self-employment in the agricultural sector, self-employment in the non-agricultural sector (including activities operated from home), and wage employment. The difference between housework and work outside the home is that only the latter is remunerated. In this setting, the amount of time allocated to each of these activities by different individuals in a household will 21 be determined by two factors. First, this allocation will depend on the economic returns that each individual can command in these various activities given his or her endowments (determined as discussed in Chapter 4). And, second, it will depend on individuals' relative economic and social bargaining power, and their preferences. 5.4 The time allocation decision is hence thought to follow and be determined by the resource allocation decision discussed in the previous chapter. So, a more educated woman is likely to command higher returns in the labor market and hence may, other things equal, devote more time to remunerated activities outside the household than a less educated one. For the same reason, her social bargaining power within the household may also be stronger so that she is able to work more outside the home if she prefers to do so. 5.5 Since this decision on allocating time has implications in terms of income, it will also influence both the future returns and the economic and social bargaining power of different members. For example, if a young woman who engages in market-oriented activities has the capacity to devote part of her earned income to accumulate productive assets, she could as a result not only increase her productivity but also strengthen her economic bargaining power within the household. Time allocation, productivity and income growth 5.6 Given the focus both on gender differences in the allocation of time and gender gaps in earnings and productivity, this chapter will consider individuals' choices of whether or not to participate in market activities jointly with their decisions regarding sectors in which to work. This view of time allocation has important implications because it is typically necessary for households to engage in multiple activities at the same time. For instance, Goodin and others (2008) note that households need a minimum amount of time to survival-related activities,8 such as cooking, sleeping, fetching water or ensuring a minimum amount of consumption. Only after these tasks are taken care of can time be devoted to other activities (so-called "discretionary time"). Gender differentials in the allocation of survival-oriented activities within households (which could reflect social norms) and, consequently, in the availability of discretionary time, will then translate into differences in the capacity of women and men to engage in all discretionary activities, including market-oriented activities such as wage employment. A similar situation occurs when there are no effective insurance mechanisms available to households. As a result, they may decide to engage in multiple activities as a way of coping with risk, even though productivity of labor and other inputs may be relatively higher when applied to one activity only. So, most households in rural areas devote time to both farm and non-farm activities even if the average productivity in the non- farm sector is higher. In such cases, those with stronger bargaining power, access to resources and/or time availability within the household (often men) will be more likely to participate in higher productivity activities. 5.7 Economic growth has significant implications both for gender differences in the allocation of time and for gender differentials in productivity and earnings. Higher incomes will alter the returns to market activities depending on how the growth process evolves. And, these changes in turn will affect the decisions within the household as to the allocation of the time of various individuals across different activities. So, for instance, an increase in the availability of outside employment opportunities for women 8 This concept is similar to what the feminist economics and sociology literature refer to as "reproductive activities". 22 in the service sector due to greater demand for these services associated with higher per capita incomes could raise the returns to work outside the household and affect women's time allocation and their earnings, including through changes in their sectoral participation. (Goldin (2006); Mammen and Paxson (2000)). 5.8 These impacts of growth are also conditioned by context, and in particular, by norms, beliefs (including personal experience) and culture. This can be seen, for example, in the recent work of Fernandez (2007) and Fernandez and Fogli (2009), whose analysis of the labor market participation of American women concludes that cultural proxies have quantitatively significant effects on women's work outcomes. Similarly, Fernandez, Fogli, and Olivetti (2004) present cross-sectional evidence that the wives of men whose mothers worked are themselves significantly more likely to work. Other work finds that gender-differentiated patterns in labor market transitions are linked closely to processes of household formation. Using data from Argentina, Brazil and Mexico, Bosch and Maloney (2009) find that the patterns of women's transitions between formal and informal employment and being out of the labor force can be explained largely by household formation decisions with single women far more similar to single men in transitions and over-represented in the formal sector. The implications of institutional and market failures 5.9 Following this analysis, the chapter will turn to how institutional and market failures can lead to persistent gender differences in time allocations and productivity. Specifically, failures in the formulation of rights and rules (the de jure environment) and in their enforcement (the de facto environment) on the one hand, or some combination of information asymmetries, externalities and "lumpy" contracts on the other hand, affects individuals' time allocations through their impact on market returns to different activities as well as on social and economic bargaining within the household. Rights and enforcement 5.10 Evidence suggests that lower access to productive inputs, such as land or capital, is one of the root causes of the productivity gender gap both in the farm and non-farm sector. Lower access appears to be associated with differential rights and weak enforcement. Goldstein and Udry (2008) argue that insecure land property rights (i.e. weak enforcement of property rights) among women lead to underinvestment and underutilization of productive inputs, which in turn translate into lower yields (returns) among female farmers. Similarly, Goetz and Gupta (1996) show that despite rapid increases in access to micro-credit among women, less progress have been made in terms of their ability to actually control these funds despite the fact that the loan is under their name. 5.11 Many studies also point to factors that constrain women's ability to work. For instance, many countries (e.g. Swaziland, Lebanon, Afghanistan) require by law that a woman acquire her father's or husband's permission in order to work outside the home, while in other places, women cannot open or operate an individual bank account (Freedom House (2010)). Such practices were also common in the not too distant past in some OECD countries, including Spain and Chile. Gender-differentiated rights of this sort will likely translate into gender-differentiated social and economic bargaining power within the household, and ultimately, time allocation decisions and productivity differentials. 23 Information asymmetries, externalities and lumpy contracts 5.12 Barriers to knowledge and learning about women's performance arising from limited presence of women in particular markets may reinforce initially low levels of female labor participation. For instance, in many countries, low female participation rates in formal private employment make it difficult for employers to adequately form expectations about female workers' productivity. As a result, employers may continue to be reluctant to hire women, hence perpetuating the bias against female employment. Similarly, in the credit market, limited access among women may preclude learning about their potential performance as borrowers, including their ability to repay. Perceived cost or other differentials between men and women and social norms that reinforce the existing situation will only exacerbate this problem. 5.13 Increasing women's participation in such a situation with information asymmetries will require interventions that are able to reduce the impact of existing expectations. Affirmative action policies in the United States and other developed countries have played this role by promoting learning among employers about the performance of under-represented groups. In the absence of affirmative action policies, female employment in the public sector in relatively large numbers can also have such a demonstration effect. And, in credit markets, the expansion of micro-finance may have played a similar role. By actively targeting women and designing delivery mechanisms that promote good performance and high repayment rates among borrowers, micro-finance institutions have generated enormous amounts of information about women's performance as borrowers. 5.14 Low female participation in specific occupations or professions can also adversely affect the performance of women who are already employed in them or aim to enter them, especially if it is costly to gather information or if networks that are centered on gender are important. In both cases, additional female participation will benefit the involvement of other women. So, for instance, Fernandez (2007) shows that when small proportions of women work, learning is very slow and the changes in female labor force participation (LFP) are also small, while when the proportion of women working is close to half the total, rapid learning and rapid changes in female LFP take place. She also provides evidence that learning of this sort can explain the evolution of female labor force participation in the US over the last century. Moreover, Jensen (2010) makes the case that investment and participation decisions are driven by perceived, rather than actual returns. This suggests that, in the absence of critical group mass in a specific market, imperfect information can slow learning even farther. 5.15 Interestingly, high female participation rates in specific occupations and the existence of significant (or "thick") networks can also have negative effects (Schelling (1971)). For instance, in the presence of differential costs in information gathering and learning, new entrants will be more likely to cluster where others from the same group are already present, hence perpetuating occupational segregation. This argument has been used in the education literature to explain gender segregation by field of study, as well as the feminization of the teaching profession (England and others (2007)). Furthermore, the presence of information asymmetries, such as the ones discussed above, would only exacerbate this phenomenon. For example, evidence from employment audits that employers discriminate against men in "traditionally female" jobs (nursing) and against women in "traditionally male" jobs (Booth and Leigh (2010)) is more likely to reflect discrimination arising from imperfect information rather than differential hiring preferences across sectors. 24 5.16 The third and final type of market failure to be considered in this chapter arises from the characteristics of tasks and how workers will perform over the length of the workday, which can result in contracts that specify a certain number of minimum hours on the job, rather than offer a continuum of hours that can be chosen by the employee. This "lumpiness", both in contractual options for wage employment and in time requirements for production, decreases the range of jobs open to women. Furthermore, in many countries part-time employment is not legally recognized and even in most countries where part-time work is indeed available, the vast majority of job openings in the formal sector are in fact full-time positions. Given that women tend to be disproportionally responsible for housework and, as a result, can devote relatively less time to other activities, the lack of availability of reduced-hours employment options de facto diminishes their ability to participate in the formal sector and increases the likelihood that, conditional on labor market participation, they become active in the informal sector--as noted in Part I. 5.17 Although to some extent higher female participation in the informal sector can be attributed to a stronger preference for flexibility among women, evidence on the impact of increased availability of daycare options and on the introduction of part-time contracts suggests that "lumpy" contracts are indeed a constraint to female participation in formal employment. Access to daycare services is associated with increases in both the number of hours worked and participation in formal employment among female workers, while it is found to have no impact on overall female labor force participation rates (Baker, Gruber, and Milligan (2008); Havnes and Mogstad (2009); and Fitzpatrick (2010)). Similarly, Bosch and Maloney (2009) show that both female labor force participation and employment in the formal sector increased in Argentina after the introduction of part-time contracts, and that female formal employment grew more in sectors with a higher incidence of part-time workers. 5.18 The chapter will draw on existing evidence and new analysis to be undertaken using information from household and labor force surveys. Most of the new analysis will focus on those areas where knowledge is scarcer by combining survey information with time use data and working with panel data where available. A strong emphasis will be put on causality so as to guide the policy discussion that follows in Part III. Chapter 6 Forming agency 6.1 This chapter will analyze how formal institutions, markets and informal institutions jointly influence decision making processes within households and determine agency. In doing so, it will also analyze the role of agency in shaping the impact of formal institutions, markets and informal institutions on the allocation of endowments and time in the household. Finally, it will point to the role of agency in changing formal and informal institutions as well as markets, a point that will be further explored in Part III of the Report. 6.2 This Report focuses on agency defined as: (i) the ability to make choices and take actions related to oneself (autonomy in allocating time and in movement, aspirations and sense of worth); (ii) autonomy and voice within the household (reflection of men's and women's preferences in household formation and allocation processes within the household); and (iii) voice within society and the public sphere (including labor markets and political institutions). Each of these aspects is important both in its own right as well as a determinant of outcomes in other domains such as assets or returns to wage labor. 25 Agency is a key dimension of welfare as well as a key determinant of other outcomes 6.3 The ability to influence one's life by making choices and taking actions is a key dimension of well-being in and of itself, as noted by Sen (1999). Gaining agency fosters power in people for use in their own lives, their communities and their society, by being able to act on issues that they define as important (Luttrell and others (2009)). 6.4 Increases in agency through greater autonomy and voice can also contribute to improving gender equality in other dimensions. For example, evidence on the importance of "soft skills" in the labor markets, including self-confidence, suggests that such skills play a greater role in determining an individual's employability than previously thought. Since women's labor market participation and earned income increases their bargaining power in the household, greater self worth can also have ripple effects on other aspects of autonomy. Many elements of autonomy and voice are key factors influencing decisions concerning the allocation of time and endowments themselves. For instance, limited control over fertility can result in women having more children than they desire, which in turn will affect decisions concerning endowments and time use. And, to the extent that political representation can be an important platform for broader changes in social norms and institutions, slow progress in this regard can represent a missed opportunity to improved gender equality. Voice and sense of worth are largely shaped by autonomous assets 6.5 Autonomous income, assets, Figure 6.1 Household decisions made by men only abilities and education are important according to wife's education level, latest available year determinants of the voice of individuals and of their sense of worth (see Agarwal (1994); 90 Rahman and Rao (2004)). Indeed, assets 80 Visits to family, relatives, friends Making large purchases Own health care grant autonomy and shape such aspects of 70 Fraction of women with at least primary education agency as age of marriage, childbearing, 60 mobility, access to justice and safety. For 50 instance, educational attainment and 40 autonomous income are important factors in 30 determining women's decision making 20 ability regarding their households, their 10 children's education and health, and their own well-being. According to available 0 0 10 20 30 40 50 60 70 80 90 Fraction of women with no education survey data, married women are most likely to make decisions (alone or together with Source: ICF Macro (2010). their partners) if they earn an income or have higher educational levels (Figure 6.1). They are also less likely to conform to traditional gender roles with regards to working within the home as opposed to participating in the labor market. And, the more education a woman has, she is less likely to agree that wife beating is justified for any reason and more likely to think that she has the right to refuse sex with her husband (Vadnais, Kols, and Abderrahim (2006)). 6.6 Control over resources--physical, human and financial--also impacts on women's ability to take action despite unfavorable social norms or formal institutions. For instance, ownership of assets can help 26 individuals leave a relationship even when legal regimes or practices are not conducive (e.g. where domestic violence is not prosecuted). The need to control assets in order to go against established social norms will likely diminish as these grant more equal rights, and as formal institutions provide the kind of services required to enforce those rights. As rights are established, they can in turn increase women's voice within households. For instance, increased alimony rights in Brazil and a change in the divorce law in Chile resulted in increased women's agency (see Rangel (2006); Heggeness (2010)). Lack of equal rights and their enforcement constrains agency 6.7 Formal institutions can constrain agency by preventing men and women from making certain choices. For instance, in some countries as described earlier, laws restrict women's rights to own land, which limits their ability to have autonomous income and assets, and hence decreases their voice and weight in household decisions. To date, while equal rights have been granted in many domains in many countries, inequality in rights persists between women and men, and is most commonly observed in the spheres of family formation (marriage, divorce, child custody), assets (ownership, inheritance), and rights to representation. This chapter will review in detail the situation in these areas, as well as the evolution of rights over the past decades to explore factors which help promote equal rights. 6.8 In addition, even in areas where countries have made progress in granting equal rights to all (de jure), the constraint has often shifted to the actual enforcement of these rights (de facto). Factors that limit enforcement can emerge at different levels. Some of them relate to service delivery by formal institutions including the lack of appeals mechanisms when rights are not respected, the lack of accountability or knowledge in law enforcement agencies, or the lack of legal assistance services for those with limited resources. For example, although different countries have passed legislation to protect women against domestic violence, they have often not provided the necessary training for judicial, health and police officers on how to deal with victims, including avoiding `secondary victimization' practices. 6.9 Other constraints related to the lack of knowledge of rights and enforcement mechanisms and, more generally social norms that dictate what is acceptable, irrespective of the prevailing legal framework. Finally, women may face barriers such as lack of money, transport, or lack of information about legal assistance services.9 6.10 Social norms are sometimes enshrined in customary laws, which supersede the formal legal system in some countries. For example, some African constitutions oblige courts to apply customary law where applicable. Elsewhere, even in situations where constitutional law, statutory law and common law have largely superseded it, customary law may nevertheless govern in certain areas, such as family relations. For example, in many places, the requirements for marriage, the rights and duties of husbands and wives, the obligations toward and custody of children, the ownership of property acquired during marriage, and many other aspects of family life are governed by customary law. In 1994, Eritrea amended the Civil Code and made a new Land Proclamation that gave women the legal right to own and inherit land, and disallowing discrimination based on sex, ethnicity, or religion. In practice, however, women's access to land is still restricted by customary practices. In Botswana, the civil court ruled in 9 Fugate and others (2005) documents this for the US, while Camerer and others (1998) and Rasool and others (2002), do so for South Africa. 27 2006 that traditional marriages are best dealt with by the customary courts and denied divorce to a woman under the Matrimonial Causes Act with the argument that the Act does not apply to customary marriages. Physical mobility is limited by a complex web of constraints 6.11 Physical mobility can be limited as a result of constraints ranging from social norms on what is viewed as acceptable for a woman (because of their expected roles or perceived vulnerabilities); to time restrictions and work burdens inside and outside the home; to lack of control over resources (control over income to cover the cost of transportation and control over household transportation assets themselves); and to lack of adequate infrastructure and transportation services. In rural settings, several studies have concluded that the state of rural infrastructure (including markets, roads, and electricity) is one of the most significant factors influencing women's mobility outside the home (Ahmed, Quisumbing, and Hoddinott (2007) and Batliwala (1994)). A recent study in Yemen and the West Bank shows that mobility of girls was largely related to security issues, highlighting the importance of formal institutions (service delivery) relative to social norms (World Bank (2010b)). 6.12 Physical mobility can also be limited by safety considerations, which themselves depend on a series of factors. These include formal institutions, such as laws defining violence against women as a crime and criminal system that promotes the enforcement of these laws. They also include service delivery in the form of the provision of streets, roads, and transportation systems that are safe for women. A recent study in Lima, Peru, highlighted that women's priority on public transportation was for services that were safe (both personal and road safety), while men prioritized speed and reliability before personal safety (Gomez (2000)). Social norms and information failures shape political participation 6.13 Factors affecting low political participation and representation vary across countries but most likely reflect a combination of unfavorable social norms (which convey the message that women have no role to play at that level leading to self-censorship), weak formal institutions (which result in lack of information on political processes and their importance and could also affect men's participation), and the various institutional and market constraints on women's mobility discussed earlier. For instance, as highlighted in the overview, there is still an overall perception that men are "better" political leaders than women, although trends suggest this is becoming less prevalent. As in the case of labor market participation, this likely reflects both prevalent social norms on gender roles and an information failure, resulting from the low number of women representatives and hence the lack of knowledge of their ability to perform. Affirmative action policies have been put in place in some countries, such as India, to remedy this information failure. Bhavnani (2009) shows that reservations for women resulted in increased female representation, even after they were removed, partly because they introduced women into politics who were able to win elections even after the reservations were withdrawn and partly because they allowed parties to learn that women can win elections. 6.14 Similarly, scarcity of political candidates can result from a low sense of worth and self-efficacy, a lack of access to networks, as well as a lack of financial resources for campaigning. For instance, Fox and Lawless (2004) analyzed the process by which women and men emerge as candidates for public office. They find that, even with similar characteristics, women have lower levels of political ambition 28 because, first, they are less likely than men to be encouraged to run for office and, second, they are significantly less likely than men to view themselves as qualified to run. Changes in individuals' agency can influence broader societal change 6.15 Within each household, social norms play a role in the processes of allocation of endowments and time or work, for instance, by defining options open to men and women. These influences can be either further reinforced or dampened by the voice of individuals within each household. Fernandez and Fogli (2009), for instance, show that personal experience plays a role in shifting social norms and decision making through the importance of mothers as role models for their sons and daughters in terms of labor market participation. In addition, within each household, changes in the allocation of endowments and time/work will in turn feed back into the decision-making processes, and influence subsequent decision making because a woman's increased income will give her more weight within her household. 6.16 Individual men's and women's agency can evolve as a result of changing institutions, markets and household allocation processes. In turn, these changes, when aggregated across households at the level of a society, are likely to be key determinants of changes in social norms and institutions--through formal political representation, mobilization of civil society, or as role models shift options open to both men and women. For instance, when choices are made or behaviors adopted by a sufficiently large number of individuals, social norms can shift, reflecting the effects of fewer information asymmetries. Learning is typically slow at first but intensifies as a critical mass is reached, as has been documented for labor market and political participation. Men and women develop "adaptive preferences," adjusting needs and preferences to possibilities and stereotypes (Sen (1990); Sen (1995); Kabeer (1996); Nussbaum (2000)). 6.17 Just as individual households can mitigate or exacerbate social norms, interventions or exogenous changes such as globalization or technology, can also change social norms, sometimes in a rapid way. For example, in the context of labor markets, the rapid development of the manufacturing sector in South Korea contributed to a reversal in the preference for sons in one generation. Similarly, societies can accept and endorse new gender roles quickly, as has been observed with respect to the role of men in child rearing in some Nordic countries. Chapter 7 Endowments, economic opportunities and agency in a changing world 7.1 The world is changing and the pace at which changes are occurring has accelerated in recent decades. What started as increased trade openness has translated into growing global economic integration and interdependence as transnational movements of people and capital have grown and information has become a global, more accessible commodity. In addition, the world population is concentrating rapidly in medium and large cities, due to both migration from rural to urban areas and population growth in the latter. Technological developments have also led to rapid changes in the way people work. As these changes have taken place, the level and speed of economic growth have risen to unprecedented levels in most areas of the world and particularly among some developing countries (Figure 7.1). In addition to changing the way people work, technological changes have changed the way people communicate and learn. Growing integration has also brought about higher volatility and increased uncertainty, from which countries are less able to isolate themselves--as illustrated by the 29 recent series of food, economic, and financial crises. Finally, in the absence of substantial changes in modes of production, higher levels of economic activity are also likely to exacerbate the impact of climate change, which could place additional stresses on households. How households adjust to these shocks is rarely gender neutral, although patterns are perhaps more difficult to predict than is commonly assumed. 7.2 Using the framework presented earlier, the discussion in this chapter will focus on the interplay between different aspects of global integration and the key nodes of our analytical framework--formal and informal institutions, markets, and households. This chapter will simultaneously look at the three broad sets of gender outcomes that relate to endowments, economic participation and agency. The analysis will be centered on the question: how are markets, formal institutions and informal social institutions affected by the process of global integration and how do these changes impact household allocation decisions? The chapter will also explore whether these processes may have corrected or exacerbated some of the institutional and market failures presented in the previous three chapters or may have led to the emergence of new ones. Figure 7.1 Per capita GDP growth by region 8 Annual growth in GDP per capita 10year moving average (%) East Asia and Pacific 6 4 South Asia Middle East and North Africa Latin America and Caribbean 2 Highincome OECD 0 1970 1975 1980 1985 1990 1995 2000 2005 2 SubSaharan Africa Europe & Central Asia 4 Source: World Development Indicators (2009) 6 Gender in a globalized economy: Opportunities and challenges 7.3 Three salient features of the globalization process will constitute the focus of the discussion in this section: trade openness, technological change (in production), and factor (particularly labor) mobility (both within and across countries). 7.4 The last few decades have been witness to significant relocation of industrial and, increasingly, services production from the developed to the developing world, as well as within the developing world. Often this process has been driven by the cost advantages of abundant and relatively cheaper labor in the developing world, subject to certain minimum skills and transportation and infrastructure constraints. In this context, women have been regarded and sometimes targeted by business as a large potential source of labor, particularly in places with relatively high average education levels. They may even have been 30 given preference over male workers at times due to perceived differential work-related traits such as higher docility and ability to operate in strictly rule-bound environments (themselves potentially shaped by social norms and expectations), or better fine motor skills (Chant (1997); Sklair (1993); Tiano (1994); Pearson (1997)). 7.5 Important changes in both institutions and markets have taken place in the developing world to accommodate and benefit from these trends, including the emergence of maquilas (factories for the assembly of imported component parts for re-export) and similar businesses, the creation of Export Processing Zones (EPZs), and the appearance of outsourcing. These developments have often translated into increased economic opportunities and higher wages for women (Aguayo-Tellez, Muendler, and Poole (2010), Black and Brainerd (2002), Artecona and Cunningham (2002); Fleck (2000)). Moreover, because women's attachment to the labor market and existing networks is weaker in places, they have been better able to benefit from new opportunities in non-traditional areas of employment than men (Luke, Munshi, and Rosenzweig (2004)). Expanded economic opportunities are likely to have provided incentives for higher female participation in paid work outside the home, as well as for higher investments in girls and women, through a combination of both higher potential economic returns to female labor force participation and increased female voice within the household. 7.6 The last few decades have also seen a revolution in information technology (IT). IT has transformed production in two important ways relevant to gender outcomes: first it has contributed to the decline of the relative importance of manual and physical labor in production; second it has allowed production to be much more mobile over time and space than in the past. Both developments hold promise for working women around the world and the chapter will analyze the impact that IT and other technological advances have had on economic production processes. Recent instances of a similar phenomenon can be found in OECD countries with the advances in household production and the rise in clerical work that took place between the 1950s and the 1970s, which, in combination with higher education levels and changing social attitudes towards women's work outside the home, catapulted women into the labor market (Goldin (2006)). 7.7 Finally, higher levels of economic integration have been accompanied by increased migratory flows both within and across countries. Within countries, rural-to-urban migration has primarily been driven by and contributed to the urbanization process and the economic opportunities associated with agglomeration (World Bank (2008)). Although women stand to benefit from this process, particularly in the form of increased access to labor markets, they may also face new constraints in the areas of mobility and safety that could seriously undermine potential benefits. In addition, existing institutional and market failures may in some cases be more binding in the urban context. For instance, if the probability of informal appropriation of property and/or squatting is higher in urban areas, insecure land and housing property can actually lead to lower labor market participation rates among women as they are forced to remain at home to protect it from invasion. In this vein, Field (2003) and Field (2007) present evidence of positive impacts of land and housing titling on female labor force participation in urban Peru. Across countries, although both men and women migrate internationally, important factors regarding the migration decision--including the rationale for migration, country of origin and destination, and sector of work--differ between them. 31 7.8 In addition to looking at the direct impact of globalization on markets and institutions, both formal and informal, this chapter will focus on the role that connectivity across multiple markets may play in propagating and transforming the impact of some developments. For instance, as new economic opportunities open up in areas traditionally dominated by men, sectors predominantly served by women (such as education and health) may experience both a shortage of labor and increasing wage costs as women exit the sector and/or are able to command higher wages. Such a trend could have important implications for basic service delivery (where services are provided by female nurses and teachers) in the absence of substantial changes in the way in which these services are provided. Similarly, there is an emerging literature on both developed and developing countries that explores the impact that increased female labor force participation and, particularly, increased number of hours devoted to market production by women has on children's health and, more broadly, on development outcomes (Miller and Urdinola (2010)). Increased information flows can change social norms 7.9 The impact of personal computers, the internet and all other changes related to this revolution extends well beyond the economic sphere, touching and changing people's lives in a million different ways. Information is shared more frequently and faster than ever before between and within regions and countries around the world. With information comes knowledge about other people and places, making it harder to maintain cultural isolationism. Comparators and reference groups are more global and these changes have the potential to radically alter expectation formation mechanisms among men and women, as well as (eventually) individual preferences and social norms. However, increased information flows can also lead to backlash. Indeed, when external social norms and preferences are perceived as a threat to a particular culture or set of practices, they could be used as a rationale to maintain deeply unequal norms. 7.10 This chapter will focus on the effect on access to information on gender equality, mostly through its impact on social norms and preferences on what is acceptable for men and women, both in the private and in the public sphere. For example, recent work has documented the effect of the spread of television programs, with their attendant role models in reducing fertility in Brazil and in decreasing the acceptability of domestic violence and son preference as well as reduced fertility in India (La Ferrara, Chong, and Duryea (2008); Jensen and Oster (2009)). Risks and shocks in a changing world 7.11 Most of the developments discussed above have the potential to significantly improve the life of women and men around the world through higher economic growth and access to productive economic opportunities. However these same developments are also associated with higher levels of risk and uncertainty as the global economy becomes more integrated and shocks in one part of the system rapidly translate to the rest, as illustrated by the recent series of inter-related food price and financial crises. Finally, as production continues to increase, the impacts of climate change are likely to affect different groups differently. The question then arises as to whether increased levels of risk and uncertainty could have gender-differentiated impacts, and if so, through which mechanisms. 7.12 In a world of highly mobile capital, and less mobile labor, often the brunt of adjustment to shocks--whether external or internal in origin--falls on labor earnings and hence directly on households (Paci, Revenga, and Rijkers (2009)). The evidence presented elsewhere in the Report indicates that 32 women and men tend to work in different industries and occupations and that women are often less productive than men on average. In addition, it is commonly believed that women tend to hold more precarious jobs, operate in smaller and less capitalized firms or are engaged in more vulnerable economic activities than men. These factors would suggest that women are more likely to be affected by economic shocks. The existing empirical evidence to this effect is mixed, however. Using data from the Asian crisis, Hallward-Driemeier, Rijkers, and Waxman (2010) show that female employment was not hit the hardest, primarily due to the fact that women were disproportionately employed in firms more resilient to the crisis. On the other hand the gender earnings gap increased, particularly in larger firms--implying that the smaller loss in women's employment came at the expense of a larger reduction in earnings. 7.13 In contrast, work on the distributional impacts of the more recent financial crisis shows that impacts were quite heterogeneous across countries and sectors with no common patterns by gender and no evidence that women were more affected than men (Habib and others (2010)). Other evidence also suggests that gender impacts vary significantly by country and are hard to generalize (Sabarwal, Sinha, and Buvinic (2010); Newhouse and Cho (2010); Turk and Mason (2009)). On the one hand, relatively high female participation rates in export-oriented industries which were the initial casualties of the crisis, such as light manufacturing, may have led to strong first-round negative employment effects. On the other hand, lower female participation rates in pro-cyclical sectors, such as construction, or industries dependent on external demand, such as tourism, implied that the aggregate employment effects for women relative to men were muted once these second-round effects are taken into account. The direction of impacts and their gender differences, as well as the channels through which they occur are likely to depend on the specific nature of the crisis at hand. 7.14 On a more micro level, gender differences in assets ownership and control can increase women's vulnerability to shocks. For example, women have lower access to agricultural assets, such as land, and when they do have access, the quality of these assets is often lower than that of assets controlled by men. This frequently leads to underinvestment (including protective investments) in female-held assets which could exacerbate the negative impact of climate shocks (World Bank (2009)). In addition, to the extent that both of these serve as insurance mechanisms (particularly in environments where access to credit is constrained and/or insurance markets are underdeveloped), these gaps would indicate that women are potentially more vulnerable to shocks and, more importantly, have a more limited set of coping strategies at their disposal. 7.15 Finally income shocks can exacerbate gender differentials in other spheres. As presented earlier in Chapter 4, for instance there is a large literature that documents that these shocks can magnify the impact of gender differentials in schooling preferences. For instance, in Indonesia, Cameron and Worswick (2001) show a drop in female school participation in response to crop losses. In Ethiopia, while an adverse weather-induced crop shock has no discernible impact on the schooling of boys, the same adverse shock has a deleterious impact on both the probability of enrollment and completion of schooling for girls (Chaudhury, Christiaensen, and Asadullah (2006)). On the other hand, Rose and Al- Samarrai (2001) find that, in Ethiopia, while boys have greater access to schooling, in times of economic crisis, they may also be the first to be withdrawn when waged employment is available. Also, in countries where consumption levels are higher, children can actually be kept in school during crises, as the opportunity cost of their schooling decreases with lower earning potentials, as in Mexico and other Latin American countries (Maluccio (2005), Mackenzie (2003)). 33 7.16 In order to examine these questions, the analysis in this chapter will build on a critical review of the existing literature on globalization and its impacts on labor market outcomes, risks and vulnerability, and more broadly on social norms and values, as well as new conceptual and analytical work to be commissioned on the gender impacts of globalization and the gender differences in terms of vulnerability and risk mitigation options. Finally, the chapter will build on cross-country qualitative work currently underway (Box II.1). New work will primarily focus on the impact of economic integration, technological change and access to information, while the discussion on the gender impacts of urbanization, migration and climate change will rely mainly on existing evidence. Part III Policies III.1 This section of the Report will examine the rationale for policy intervention in light of the facts and analysis presented in Part I and Part II. The discussion will build on the premise that gender equality is intrinsic to development. So, while public policies aimed at improving gender equality may often generate gains in terms of growth, that is not the only motivation that policymakers should consider. III.2 One rationale for policies aimed at improving gender equality has been that such policies, if successful, will yield a large dividend in terms of economic growth. The Report will review the evidence for this view, paying particular attention to where the evidence is strong and where it is not. Our focus will not, however, be limited to interventions that are expected to have positive impacts on economic growth. In keeping with the broader view of development in this Report, we will argue that a change of focus is needed so as to think of gender equality as both a contributor to the development process and a key component of this process. Consequently, the Report will also discuss policies that directly address gender inequality even if these are expected to have no or only small positive impacts on economic growth or efficiency. Rationale for public action III.3 The rationale for public action is threefold. First, as mentioned above, the Report will argue that gender equality is a development objective in its own right. This view follows Sen (1999) and others who see development as a process of expanding freedoms equally for men and women. III.4 Second, economic growth can and has been good for gender equality over the last quarter century, but is not enough. Three observations emanating from the evidence presented in Part I and the analysis in Part II underpin this idea: The pace of progress will not be fast enough if tied only to economic growth. While many gender gaps are being closed as a consequence of economic growth, at the current pace of growth it would take a long time for the remaining gaps to close fully. Effective and well-chosen policies can help enhance the speed at which outcomes for women catch up to those of men even where these are highly responsive to income growth. Growth does not explain all progress observed in the last quarter century. In some domains, such as maternal mortality, progress cannot be explained by economic growth alone. Other factors or policies/interventions have mattered and it is important to understand what these were. 34 Moreover, the dynamics of changes in gender equality within countries have not always been smooth. In some countries, there have been relatively few changes for many decades followed by significant progress in narrowing many gender disparities in a short period. Spain's experience in the last quarter of the 20th century is a case in point. Understanding what public actions drive such changes and contrasting these with situations in other countries where over many years little has changed offers lessons. For some gender disparities, there has been little progress despite growth. Convergence is being achieved much faster in education than in political participation. And while women may be going to college more than men, the courses of study that they choose remain very different. Finally, where female labor force participation has increased, the representation of women in senior management and the board rooms of large corporations has risen more slowly. In such domains, other measures are needed to complement income growth for disparities to shrink. As noted in Part II of the Report, many of the reasons for the relative lack of progress in these areas have to do with market failures and institutional constraints that bind differently on men and women. These failures and constraints are typically multiple and mutually reinforcing so that progress on one dimension of gender equality (such as tertiary education) does not assure that outcomes on another dimension (in the wage labor market, for instance) will necessarily improve in tandem. III.5 Third, gender inequality is likely to be costlier in a globally-integrated world. With internationally-mobile capital and labor, the relocation of manufacturing and services across countries is relatively easy and commonplace. And with greater competition, small cost differences are likely to weigh larger in these decisions. So, countries that succeed in reducing gender-based inequalities, especially in secondary and tertiary education and in labor market participation, will have an advantage over those that delay action. Chapter 8 Addressing gender inequalities: A policy typology 8.1 Two different types of policy interventions will be considered for the purpose of the discussion in this section. The first type includes policies that attempt to enhance the equalizing power of economic growth, as well as redistributive measures that aim at shifting relative power from men to women or vice versa. In other words, these are policies that seek to level the playing field between men and women. The second broad type of policies are those that directly address market failures, institutional constraints and inefficiencies in household bargaining that contribute to gender inequalities. These policies target sources of inefficiency that also result in inequality. Leveling the playing field 8.2 Economic growth has the potential to be a great equalizer. However, existing differences in endowments, access to assets and economic opportunities, and other important dimensions of gender inequality will to a large extent determine how the gains from growth are actually distributed. In addition, so-called "gender-blind" reforms that seek to accelerate economic growth can sometimes have a negative (albeit unintended) impact on gender equality. 35 8.3 Policies to be discussed in this section will aim at addressing these inequalities by leveling the playing field for men and women. Within this broad category, a distinction will be made between policies that are primarily concerned with the distributional impact of economic growth and those that directly address gender inequalities through the explicit redistribution of opportunities or assets. This distinction is important because in the first case economic growth generates new surplus (a bigger pie) that can then be shared by different groups (i.e. the focus of the policy can be on shaping how the surplus is actually shared including in terms of reducing opposition), while in the second case there is no additional gain (the size of the pie is given), at least in the short-term, so that some individuals or groups will lose in order for others to gain. So, the focus of the policy is on the amount of redistribution and the mechanism to be used). Some illustrative examples of both types of policies follow. 8.4 Higher levels of education are generally associated with better social and economic outcomes at the individual level and, ultimately, with increased economic growth at the aggregate level. However, the benefits of public investments in education do not always accrue equally to different groups of the population, including men and women. For instance, the construction of new schools in underserved areas could have a differential impact on enrolment rates among boys and girls if the two groups face different constraints to schooling (e.g. girls cannot travel unaccompanied, while boys can) or different opportunities in the labor market (e.g. boys have access to profitable employment opportunities more frequently and/or at an earlier age than girls). In this context, measures intended to make it easier and safer for girls to go to school or that provide incentives for boys to remain in school as they approach working age would fall under the first type of policies discussed above--that is, policies that aim at leveling the playing field so as to enhance the equalizing potential of economic growth. 8.5 There may also be "gender-blind" policies that seek to accelerate economic growth, where there are negative (albeit unintended) impacts on a specific dimension of gender equality. For instance, interventions seeking to increase the use of fertilizers would tend to disproportionately favor men since they are more likely to use fertilizers to start with, hence widening the productivity gap between men and women in the agricultural sector. In addition, these interventions could result in the shifting of other complementary inputs from women to men. This shift would in turn have an absolute negative impact on the productivity of female farmers and widen the gender productivity gap even further.10 In these situations, the government may chose to pursue compensatory or corrective measures that mitigate the gender-discriminatory impact of growth-enhancing policies, as with targeted fertilizer subsidies for female farmers. 8.6 Finally, some policy interventions may aim at explicitly redistributing relative power between men and women. Take, for example, legislative changes that force equal distribution of inheritance among all children, normally at the expense of the oldest son (or all sons). As in the previous examples, the ultimate goal of such policies is to promote a more equitable distribution of (future) economic growth along gender lines through more equitable access to opportunities or assets, but the means of achieving 10 Of course, it is possible that increased agricultural productivity among male farmers could ultimately benefit the women (and men) in these workers' households by increasing household income and consumption levels. The focus here is on the first-round impacts on employment, productivity and earnings, (which may also shift household bargaining power away from women), but second-round impacts also need to be recognized. 36 this goal is different. Specifically, a decision is made here to reassign a particular opportunity or asset across groups and therefore to directly impact gender equality in the short term. 8.7 The Report will argue that in implementing both types of policies, policymakers should pay attention to three important issues. First, while in some cases men and women can benefit equally from the same intervention once existing constraints are removed, in others men and women will actually require different interventions to achieve the same results. For instance, biological differences between men and women imply that the composition of basic health care packages will need to be different across the sexes. In these cases, leveling the playing field will not necessarily involve equal treatment across genders. 8.8 Second, identity matters in decision making. This implies that at any given time women and men in a society may behave differently when underlying conditions change, and policies can take advantage of such differences in behavior (Lundberg, Pollak, and Wales (1997); Duflo (2003)). For instance, there is considerable work that shows that a mother's education leads to positive benefits for children in terms of education and health. There is also evidence that women and men spend income they control differently, with women focusing more on expenditures related to children. In such cases, gender itself can be used as the means of delivering better outcomes on gender equality and other development goals. 8.9 Finally, when the policy intervention in question explicitly redistributes relative power between men and women, it will be important to evaluate short-term losses against long-term gains. For instance, the introduction of political quotas for women may initially translate into lower quality of the legislative process as newly-elected female representatives may be on average less experienced. However, over a longer horizon, policies that improve gender equality in political representation will bring in a new set of individuals with heretofore underutilized skills, and this is likely to result in long-term gains for society as a whole. In these cases, implementation should pay special attention to minimizing short-term losses and addressing political economy issues associated with the emergence of short-term winners and losers. Addressing inequality in the presence of inefficiencies 8.10 The analysis in Part II highlighted that the lack of progress in certain dimensions of gender equality over the last quarter century can often be linked to the existence of market failures (information asymmetries, externalities and incomplete contracts), institutional constraints (incomplete formulation of rights and weak enforcement) and inefficiencies in household bargaining. This section focuses on policies that address these failures and constraints with some illustrative examples. Such measures clearly provide the potential to achieve both greater gender equality and faster growth. 8.11 As argued in Part II, information asymmetries are often the root cause of low participation equilibriums (be it in markets or politics). Moving from a low- to a high-participation equilibrium in the presence of such asymmetries will then require interventions that promote experimentation and learning at a large scale and/or generate sufficient information on average group performance as to ultimately decrease the uncertainty surrounding performance. Affirmative action policies in the United States and other developed countries have played this role. By promoting the employment of initially under- represented groups including women, these policies have promoted learning among employers as to the performance of these individuals. While these policies were widely discussed at the time of their implementation and concerns were expressed about both their potential negative impact on efficiency and 37 their long-term effect, and they remain controversial even today, empirical evidence for the United States seems to suggest that they did indeed redistribute employment from white males to women and minority groups at no significant efficiency costs (Holzer and Neumark (2000)). In addition, although little evidence is available on the long-term impact of affirmative action in labor markets once the policy is removed, empirical evidence on the effects of political quotas in some contexts provides interesting insights. Specifically, it appears that female quotas in political representation have lasting effects, with more women elected to public office even after the quotas are removed (Beaman and others (2009); Bhavnani (2009)). This suggests that once voters learn about the performance of female politicians they are more likely to vote for women than before. 8.12 Similarly, the co-existence of multiple legal or regulatory systems is often associated with failures in the enforcement of rights. In many African countries, customary and formal laws co-exist and are applied, sometimes interchangeably, by lower courts in cases related to family and inheritance matters. Because customary law is generally more discriminatory against women than formal law, uncertainty regarding the legal framework in place, combined with lack of information about the content of the law and its application in practice frequently translate into limited access to justice for women. In such cases, measures to clarify the legal framework or strengthen the formal system where it clearly overrides the customary system, would improve outcomes. 8.13 The Report will also evaluate the extent to which multiple failures and constraints act simultaneously to determine a particular gender outcome. In these cases, policies that target some, but not all existing constraints are likely to yield insufficient results. For instance, improvements in public transport infrastructure intended to facilitate access to school or jobs may not suffice to deliver the desired results due to the presence of additional constraints on mobility, particularly for women. Such constraints could relate to issues of safety or to prevailing norms regarding women's ability to travel alone. Similarly, a policy aimed at reducing differences in productivity between female and male entrepreneurs through micro-credit schemes may succeed in attracting large numbers of women borrowers (as many of these schemes have), but this increase in numbers need not translate into higher productivity if these women systematically undervalue the time they spend on these enterprises due to the lack of employment opportunities outside the household. This suggests that improvements in productivity would likely require interventions in both the credit and labor markets. 8.14 Recognizing the co-existence of these multiple failures and constraints, as well as the need for coordinated action is a crucial but difficult step in the reform process. Most governments around the world are organized largely in sectoral ministries or departments, making it difficult to develop a shared view of the underlying problems and the required solutions. Linking these entities and strengthening coordination between them is, therefore, particularly important for policy making on gender equality. 8.15 Finally, all the interventions to be discussed in Part III could have spillovers beyond their intended sphere of influence. For example, conditional cash transfer programs have not only had a positive impact of school enrolment and attendance, especially among girls, but have also increased women's empowerment by placing cash in the hands of mothers. In some cases, these interventions have led to significant declines in early marriage, teenage pregnancy, and self-reported sexual activity among program beneficiaries (Baird and others (2009)). Spillovers can also sometimes be negative. Policies that promote higher female labor force participation could have unintended consequences in terms of child 38 health and development in the absence of compensatory measures. Policy design and implementation should aim to capitalize on positive spillovers, while minimizing negative ones. Chapter 9 Learning by doing and from others 9.1 In this chapter, the focus of the discussion will move from specific policy interventions to the process of policy formulation and, more broadly, the dynamics of policy change and reform as these pertain to gender equality. The Report will argue that understanding and taking account of these processes is especially critical in this area since these reforms relate to attempts at changing the status quo and its underlying power structure. 9.2 Three main themes will receive particular attention. The first relates to the timing and sequencing of different policy interventions (i.e. the path to reform), and their relationship to the social, institutional and economic context of each specific country. The second examines what developing countries can learn from the experience of developed countries regarding gender equality and vice versa. In particular, we ask whether countries can learn from each others' experiences so as to make even more rapid progress on gender equality (so-called "leapfrogging"). The final theme is concerned with how informal and formal institutions interact in affecting reforms. Path to reform 9.3 By generating incentives for and eroding resistance to change, economic growth is among the most powerful drivers behind reform efforts in the area of gender equality around the world. Economic growth results in greater overall welfare and allows for improvements in equality to a cure without reducing the welfare of the traditionally better off group. Fernandez (2010) uses the U.S. historical experience to show that, as growth proceeded and fertility fell, men granted more property rights to women because they valued the future welfare of their daughters more.11 9.4 Progress on many gender outcomes, however, is not always smooth. As with women's access to different levels of education and their participation in wage labor markets, changes may initially be slow before accelerating once a critical mass is reached. And, in other instances, change may be rapid in one domain, such as completion of tertiary education but limited in other respects, such as managerial or occupational representation. 9.5 Developed and developing countries have also followed very different routes to gender equality. In some countries, policies aimed at increasing gender equality focused first on the sphere of rights and legislation, particularly regarding the family. In others, equality in rights of ownership and control of property was given priority. Similarly, some countries have placed the emphasis on influencing market outcomes, such as wages and working conditions, while in other cases, government aim to change social norms. 11 This finding also illustrates the implication of another difference between inequality based on gender compared to other characteristics, in that men and women live in the same household and care about each other's welfare and that of their children. 39 9.6 The Report will use these experiences in illustrating how there is no single path to or blueprint for reform. Instead, the timing and sequencing of different policy interventions emerges from and responds to the social, institutional and economic context of each specific country, including their growth and development experience. The discussion will draw on a series of in-depth country studies to be commissioned as part of the background work for the Report. These country case studies will describe the reform process in different countries, paying special attention to how the social, political and institutional context and economic developments in each interacted with the timing and pattern of legislative reform in shaping gender outcomes and social norms concerning gender. These studies will allow us to address such questions as: what areas where tackled first?; were reforms sequenced or bundled?; and what were the main constraints to reform and how were these dealt with over time? (see Box 9.1). Box 9.1 Country case studies: Understanding the dynamics of change A number of in-depth country case studies will be undertaken to explore the key factors that have underpinned changes in various dimensions of gender equality. The purpose of these country case studies is to consider: (i) how economic, social and political factors have interacted with legislative and policy reforms in changing gender relations and outcomes in these countries; and (ii) how these impacts have been transmitted through changes in gender norms around economic choices, such as investing in education, accumulating assets and seeking wage employment, as well as perceptions of women and men within these societies. Since the focus of these case studies is on the underlying dynamics of change, each of them will take a medium-term perspective on gender equality (dating back to the 1960s or early 1970s). Over this time period, the case studies will document the main legal and policy changes undertaken by the country with regard to gender equality. While particular attention will be paid to measures that concern the economic sphere, such as legislation aimed at eliminating gender-based discrimination in labor markets and gender parity in property rights, reforms in other areas, such as family law and political participation, will also be covered. Attention will also be focused on the extent to which these laws and policies have been enforced. Learning from others' experience 9.7 The evidence presented in Part I highlights how change in such dimensions of gender equality as education and fertility has been quite rapid in developing countries compared to the historical experience of today's developed countries. The analysis in Part II aims to show that as global integration proceeds, even faster progress will be needed because some gender inequalities could be more costly than before. 9.8 The question then arises as to whether developing countries can learn from developed countries' experiences and from each other so that they can implement measures that lead to greater gender equality even faster. In other words, is there a possibility that these countries can "leapfrog" with respect to gender equality, in the same way that many have done with technological change? 9.9 On the one hand, the world today is different from the way the world was 150 years ago. Education levels are higher in every country and women have the right to vote and participate in the political process in almost all countries. In addition, increased globalization implies that the process of economic growth and structural transformation in the developing world will differ from that observed in developed countries when they were earlier in their growth process. These factors suggest that there should be a possibility of such "leapfrogging." 40 9.10 In addition, because developing countries are not constrained by many of the barriers to gender equality encountered by developed countries years ago, they may face lower costs to innovate in their efforts to promote gender equality. Advances in information and communication technology have opened up new spaces for women both socially and economically, in ways that are able to accommodate existing social norms and values. For instance, women can gather information and participate in markets using cell phones, thereby avoiding the need to travel long distances or significantly saving on the time they need to devote to household chores. Similarly, women in developing countries are joining the labor force at a much faster pace than counterparts did in the developed world at comparable stages of development and are finding employment in sectors that have traditionally been thought of as "male sectors," such as light manufacturing. As a consequence, levels of labor market segregation are now lower in some developing countries than in developed countries. Important lessons can be drawn from these experiences and both developing and developed countries can learn from and capitalize on them. Interrelation between social and formal institutions 9.11 The discussion in Part II paid attention to the role of both formal and informal institutions in the determination of gender outcomes, through their impact on individual preferences and bargaining power, as well as existing constraints and incentive structures. For this purpose, formal institutions were characterized as the combination of the State, the law, and legally-recognized institutions such as labor unions while informal institutions referred to social norms and social networks. 9.12 The discussion in this last section will adopt a broader perspective to examine how these institutions and their interaction affect policy formulation and reform. In some cases, the impetus for change comes from the formal sphere (e.g. legislative change allowing women to hold property in marriage), while in others change will first occur in the informal sphere (e.g. suffragette movements in the U.S and Europe in the 19th century). In turn, changes in formal institutions can be undertaken with the aim of transformational impact in the informal sphere (e.g. the introduction of mixed education) or merely echo and acknowledge changes that have already happened (e.g. formal recognition of divorce or equalizing child custody rights). Ultimately, the tension surrounding reform and the likelihood that it will have a real impact will reflect and depend on the extent to which formal and informal institutions are aligned with each other. 9.13 This section will explore these issues paying special attention to the role of individuals and communities as well as to social movements as actors in the reform process, and relate these to the role of the State and other formal institutions in pushing forward change. As in the previous two sections, the discussion will draw extensively from case material. 41 References The word processed describes informally reproduced works that may not be commonly available through libraries. Agarwal, Bina. 1994. A Field of One's Own: Gender and Land Rights in South Asia. Cambridge: Cambridge University Press. --------. 1997. "'Bargaining' and Gender Relations: Within and Beyond the Household." Feminist Economics 3(1):1­51. Aguayo-Tellez, Ernesto, Marc-Andres Muendler, and Jennifer P. Poole. 2010. "Globalization and Formal Sector Migration in Brazil." World Development 38(6):840­56. Ahmed, Akhter U., Agnes R. Quisumbing, and John F. Hoddinott. 2007. Relative Efficacy of Food and Cash Transfers in Improving Food Security and Livelihood of the Ultra-Poor in Bangladesh. Washington, DC: International Food Policy Research Institute (IFPRI). Aljounaidi, Lamis. 2010. "Gender and Transport in MENA: Case Studies from West Bank Gaza and Yemen." Washington, DC: World Bank, MENA, Knowledge and Learnig Quick Notes Series 21. Anderson, Siwan, and Debraj Ray. 2009. "Missing Women: Age and Disease." Review of Economic Studies 77(4):1262­300. Antonopoulos, Rania. 2009. "The Unpaid Care Work - Paid Work Connection." Geneva: International Labour Office, Policy Integration and Statistics Department Working Paper 86. Artecona, Raquel, and Wendy Cunningham. 2002. The Effects of Trade Liberalization on the Gender Wage Gap. Washington, DC: World Bank, Background Paper for the 'Engendering Development' Policy Research Report. Ashraf, Nava, Erica Field, and Jean Lee. 2010. "Household Bargaining and Excess Fertility: An Experimental Study in Zambia." Harvard University. Cambridge, MA. Processed. Aslam, Monazza. 2009. "Education Gender Gaps in Pakistan: Is the Labor Market to Blame?" Economic Development and Cultural Change 57(4):747­84. Aslam, Nonazza, and Geeta Gandhi Kingdon. 2008. "Gender and Household Education Expenditure in Pakistan." Applied Economics 40(20):2573­91. Assaad, Ragui, Deborah Levison, and Nadia Zibani. Forthcoming. "The Effect of Domestic Work on Girls' Schooling: Evidence from Egypt." Feminist Economics. Baird, Sarah, Ephraim Chirwa, Craig McIntosh, and Berk Ozler. 2009. "The Short-term Impacts of a Schooling Conditional Cash Transfer Program on the Sexual Behavior of Young Women." Washington, DC: World Bank Policy Research Working Paper Series 5089. Baker, Michael, Jonathan Gruber, and Kevin Milligan. 2008. "Universal Child Care, Maternal Labor Supply, and Family Well-Being." Journal of Political Economy 116(4):709­45. Bannon, Ian, and Maria C. Correia, eds. 2006. The Other Half of Gender: Men's Issues in Development. Washington, DC: World Bank. Barro, Robert J., and Jong-Wha Lee. 2010. "A New Data Set of Educational Attainment in the World, 1950-2010." Cambridge, MA: National Bureau of Economic Research Working Paper Series 15902. Batliwala, S. 1994. "The Meaning of Women's Empowerment: New Concepts from Action." In G. Sen, A. Germaine, and L. Chen, (eds.), Population Policies Reconsidered: Health, Empowerment and Rights. Boston, MA: Harvard University Press. Beaman, Lori, Raghabendra Chattopadhyay, Esther Duflo, Rohini Pande, and Petia Topalova. 2009. "Powerful Women: Does Exposure Reduce Bias?" Quarterly Journal of Economics 124(4):1497­ 540. Bell, Amy E., Steven J. Spence, Emma Iserman, and Cristine E. R. Logel. 2003. "Stereotype Threat and Women's Performance in Engineering." Journal of Engineering Education 92:307­12. Beneria, Lourdes. 2001. "Changing Employment Patterns and the Informalization of Jobs: General Trends and Gender Dimensions." Geneva: International Labor Office Working Paper ??? Bhavnani, Rikhil R. 2009. "Do Electoral Quotas Work after They Are Withdrawn? Evidence from a Natural Experiment in India." American Political Science Review 103(1):23­35. Black, Sandra, and Elizabeth Brainerd. 2002. "Importing Equality? The Impact of Globalization on Gender Discrimination." Industrial and Labor Relations Review 57(4):540­59. Blackden, C. Mark, and Quentin Wodon. 2006. "Gender, Time Usse, and Poverty in Sub-Saharan Africa." Washington, DC: World Bank Working Paper 73. Booth, Alison Lee, and Andrew Leigh. 2010. "Do Employers Discriminate by Gender? A Field Experiment in Female-Dominated Occupations." Economic Letters 107(2):236­8. Booth, Alison Lee, and Patrick J. Nolen. 2009. "Gender Differences in Risk Behavior: Does Nurture Matter?" London, UK: Centre for Economy Policy Research 7198. Bosch, Mariano, and William F. Maloney. 2009. "Comparative Analysis of Labor Market Dynamics Using Markov Processes: An Application to Informality." Labour Economics 17(2010):621­31. Brown, Philip H., and Albert Park. 2002. "Education and Poverty in Rural China." Economics of Education Review 21(2002):523­41. Bruhn, Miriam. 2009. "Female-Owned Firms in Latin America. Characteristics, Performance, and Obstacles to Growth." Washington, DC: World Bank Policy Research Working Paper Series 5122. Camerer, Lala, Antoinette Louw, Mark Shaw, Lillian Artz, and Wilfried Scharf. 1998. Crime in Cape Town: Results of a City Victim Survey. Cape Town: Institute of Criminology, University of Cape Town. Cameron, Lisa, and Christopher Worswick. 2001. "Education Expenditure Responses to Crop Loss in Indonesia: A Gender Bias." Economic Development and Cultural Change 49(2):351­63. Canagarajah, R. Sudharshan, and Harold Coulombe. 1997. "Child Labor and Schooling in Ghana." Washington, DC: World Bank Policy Research Report Working Paper Series 1844. Case, Anne, and Angus Deaton. 2003. "Consumption, Health, Gender, and Poverty." Washington, DC: World Bank Policy Research Working Paper Series 3020. Chant, Sylvia. 1997. "Gender and Tourism Employment in Mexico and the Philippines." In M. T. Sinclair, (eds.), Gender, Work and Tourism. London: Routledge. Chaudhury, Nazmul, Luc Christiaensen, and Mohammad Niaz Asadullah. 2006. "Schools, Household, Risk, and Gender: Determinants of Child Schooling in Ethiopia." Oxford, UK: Centre for the Study of the African Economies, Working Paper 250. Chioda, Laura, Rodrigo Garcia-Verdu, and Ana Maria Muñoz Boudet. 2010. "Regional Study on Gender in LAC: Linking Labor Market Outcomes and Intra-Household Dynamics." World Bank, Latin America and the Caribbean Region, Office of the Chief Economist, Poverty and Gender Unit. Washington, DC. Processed. Colclough, Christopher, Pauline Rose, and Mercy Tembon. 2000. "Gender Inequalities in Primary Schooling: The Roles of Poverty an Adverse Cultural Practice." International Journal of Educational Development 20(1):5­27. Courtenay, Will H. 2000. "Constructions of Masculinity and Their Influence on Men's Well-being: A Theory of Gender and Health." Social Science & Medicine 50(2000):1385­401. Croson, Rachel, and Uri Gneezy. 2009. "Gender Differences in Preferences." Journal of Economic Literature 47(2):448­74. Daley-Harris, Sam. 2009. State of the Microcredit Summit Campaign Report 2009. Washington, DC: Microcredit Summit Campaign. de Brauw, Alan, Qiang Li, Chengfang Liu, Scott Rozelle, and Linxiu Zhang. 2008. "Feminization of Agriculture in China? Myths Surrounding Women's Participation in Farming." China Quarterly 194(June):327­48. Deere, Carmen Diana, and Cheryl R. Doss. 2006. "Gender and the Distribution of Wealth in Developing Countries." New York, NY: United Nations University, World Institute for Development Economics Research, Research Paper 115. Deininger, Klaus, Aparajita Goyal, and Hari Nagarajan. 2010. "Can Changes in Inheritance Legislation Improve Female Asset Ownership? Evidence from India's Hindu Succession Act." Washington, DC: World Bank Policy Research Working Paper Series 5338. Dinkelman, Taryn. 2010. "The Effects of Rural Electrification on Employment: New Evidence from South Africa." Princeton, NJ: Princeton University, Woodrow Wilson School of Public and International Affairs, RPSD Working Paper 272. Doss, Cheryl R. 2006. "The Effects of Intrahousehold Property Ownership on Expenditure Patterns in Ghana." Journal of African Economies 15(1):149­80. Duflo, Esther. 2003. "Grandmothers and Granddaughters: Old-Age Pensions and Intrahousehold Allocation in South Africa." World Bank Economic Review 17(1):1­25. Duryea, Suzanne, and Mary Arends-Kuenning. 2003. "School Attendance, Child Labor and Local: Labor Market Fluctuations in Urban Brazil." World Development 31(7):1165­78. Edmonds, Eric V. 2004. "Does Illiquidity Alter Child Labor and Schooling Decisions? Evidence from Household Responses to Anticipated Cash Transfers in South Africa." Cambridge, MA: National Bureau of Economic Research Working Paper Series 10265. England, Paula, Paul Allison, Su Li, Noah Mark, Jennifer Thompson, Michelle J. Budig, and Han Sun. 2007. "Why Are Some Academic Fields Tipping Toward Female? The Sex Composition of US Fields of Doctoral Degree Receipt, 1971-2002." Sociology of Education 80(2007):1­23. Fernandez, Raquel. 2007. "Women, Work and Culture." Journal of the European Economic Association 5(2-3):305­32. --------. 2010. "Women's Rights and Development." Milan, Italy: Fondazione Eni Enrico Mattei Research Paper Series 57. Fernandez, Raquel, and Alessandra Fogli. 2009. "Culture: An Empirical Investigation of Beliefs, Work, and Fertility." American Economic Journal: Macroeconomics 1(1):146­77. Fernandez, Raquel, Alessandra Fogli, and Claudia Olivetti. 2004. "Mothers and Sons: Preference Formation and Female Labor Force Dynamics." Quarterly Journal of Economics 119(4):1249­ 99. Field, Erica. 2003. "Property Rights and Investment in Urban Slums." Journal of the European Economic Association 3(2-3):279­90. --------. 2007. "Entitled to Work: Urban Property Rights and Labor Supply in Peru." Quarterly Journal of Economics 122(4):1561­602. Fitzpatrick, Maria Donovan. 2010. "Preschoolers Enrolled and Mothers at Work? The Effects of Universal Prekindergarten." Journal of Labor Economics 28(1):51­85. Fleck, Susan. 2000. A Gender Perspective on Maquila Employment and Wages in Mexico. Washington, DC: World Bank. Fox, Richard L., and Jennifer L. Lawless. 2004. "Entering the Arena? Gender and the Decision to Run for Office." American Journal of Political Science 48(2):264­80. Freedom House. 2010. Women's Rights in the Middle East and North Africa. Washington, DC: Freedom House. Fugate, Michelle, Leslie Landis, Kim Riordan, Sara Naureckas, and Barbara Engel. 2005. "Barriers to Domestic Violence Help Seeking: Implications for Intervention." Violence Against Women 11(3):290­310. Galdas, Paul M., Francine Cheater, and Paul Marshall. 2005. "Men and Health Help-Seeking Behaviour: Literature Review." Journal of Advanced Nursing 49(6):616­23. Gammage, Sarah. 2010. "Time Pressed and Time Poor: Unpaid Household Work in Guatemala." Feminist Economics 16(3):79­112. Gneezy, Uri, Kenneth L. Leonard, and John A. List. 2009. "Gender Differences in Competition: Evidence from a Matrilineal and a Patriarchal Society." Econometrica 77(5):1637­64. Goetz, Anne Marie, and Rina Sen Gupta. 1996. "Who Takes the Credit? Gender, Power, and Control over Loan Use in Rural Credit Programs in Bangladesh." World Development 24(1):45­64. Goldin, Claudia. 2006. "The Quiet Revolution that Transformed Women's Employment, Education, and Family." American Economic Review 96(2):1­21. Goldstein, Markus, and Christopher Udry. 2008. "The Profits of Power: Land Rights and Agricultural Investment in India." Journal of Political Economy 116(6):981­1022. Gomez, Liza. 2000. Gender Analysis of Two Components of the World Bank Transport Projects in Lima, Peru: Bikepaths and Busways. Washington, DC: World Bank. Gong, Xiaodong, and Arthur van Soest. 2004. "Wage Differentials and Mobility in the Urban Labor Market: A Panel Data Analysis for Mexico." Labour Economics 9(2002):513­29. Goodin, Robert E., James Mahmud Rice, Antti Parpo, and Lina Eriksson. 2008. Discretionary Time: A New Measure of Freedom. Cambridge, UK: Cambridge University Press. Guindon, G. Emmanuel, and David Boisclair. 2003. "Past, Current and Future Trends in Tobacco Use." Washington, DC: World Health Organization, HNP Discussion Paper, Economics of Tobacco Control Paper 6. Habib, Bilal, Ambar Narayan, Sergio Olivieri, and Carolina Sanchez-Paramo. 2010. "Assessing Ex-ante the Poverty and Distributional Impact of the Global Crisis in a Developing Country." Washington, DC: World Bank Policy Research Working Paper Series 5238. Hallman, Kelly, Agnes R. Quisumbing, Marie Ruel, and Benedicte de la Briere. 2005. "Mothers' Work and Child Care: Findings from the Urban Slums of Guatemala City." Economic Development and Cultural Change 53(4):855­85. Hallward-Driemeier, Mary. (eds.) Forthcoming. "Expanding Women's Opportunities in Africa." Washington, DC: World Bank. Hallward-Driemeier, Mary, Bob Rijkers, and Andrew Waxman. 2010. "Women in Crisis: Firm-Level Evidence on the Labor Market Impacts of the Indonesian Crisis." World Bank. Washington, DC. Processed. Hannum, Emily. 2005. "Market Transition, Educational Disparities, and Family Strategies in Rural China: New Evidence on Gender Stratification and Development." Demography 42(2):275­99. Havnes, Tarjei, and Magne Mogstad. 2009. "The Irrelevance of Subsidized Child Care for Maternal Employment: The Norwegian Experience." University of Oslo Economics Department. Oslo. Processed. Heggeness, Misty L. 2010. "Does the Right to Divorce Affect Bargaining Power within Marriage? The Case of Chile." Minnesota Population Center, University of Minnesota. Minneapolis, MI. Processed. Himaz, Rozana. 2010. "Intrahousehold Allocation of Education Expenditure:The Case of Sri Lanka." Economic Development and Cultural Change 58(2):231­58. Holzer, Harry J., and David Neumark. 2000. "Assessing Affirmative Action." Journal of Economic Literature 38(3):483­568. Human Rights Watch. 2003. Double Standards: Women's Property Rights Violations in Kenya. New York, NY: Human Rights Watch. ICF Macro. 2010. "Demographic and Health Surveys". Calverton, MD, Measure DHS. International Labor Organization. 2002. Women and Men in the Informal Economy: A Statistical Picture. Geneva: International Labour Organization (ILO). --------. 2010. "Key Indicators of the Labour Market". Geneva, International Labour Organization (ILO). Jensen, Robert. 2000. "Agricultural Volatility and Investments in Children." American Economic Review 90(2):399­405. --------. 2010. "The (Perceived) Returns to Education and the Demand for Schooling." Quarterly Journal of Economics 125(2):515­48. Jensen, Robert, and Emily Oster. 2009. "The Power of TV: Cable Television and Women's Status in India." Quarterly Journal of Economics 124(3):1057­94. Jha, Prabhat, Rajesh Kumar, Priya Vasa, Neeraj Dhingra, Deva Thiruchelvam, and Rahim Moineddin. 2006. "Low Male-to-Female Sex Ratio of Children Born in India: National Survey of 1.1 Million Households." The Lancet 367:211­18. Kabeer, Nalia. 1996. "Agency, Well-Being and Inequality: Reflections on the Gender Dimensions of Poverty." IDS Bulletin 27(1):11­21. Katz, Elisabeth, and Juan Chamorro. 2003. "Gender, Land Rights, and the Household Economy in Rural Nicaragua and Honduras." Paper presented at the Annual Conference of the Latin American and Caribbean Economics Association. Puebla, Mexico, Kingdon, Geeta. 1996. "The Quality and Efficiency of Private and Public Education: A Case Study of Urban India." Oxford Bulletin of Economics and Statistics 58(1):57­82. Klapper, Leora F., and Simon C. Parker. 2010. "Gender and the Business Environment for New Firm Creation." World Bank Research Observer 25(2):1­21. Koolwal, Gayatri, and Dominique van de Walle. 2010. "Access to Water, Women's Work and Child Outcomes." Washington, DC: World Bank Policy Research Working Paper Series 5302. La Ferrara, Eliana, Alberto Chong, and Suzanne Duryea. 2008. "Soap Operas and Fertility: Evidence from Peru." London, UK: Centre for Economic Policy Research Discussion Paper 6785. Lewis, Maureen A., and Marlaine E. Lockheed. 2006. Inexcusable Absence. Washington, DC: Center for Global Development. Luke, Nancy, Kaivan Munshi, and Mark Rosenzweig. 2004. "Marriage, Networks, and Jobs in Third World Cities." Journal of the European Economic Association 2(2-3):437­46. Lundberg, Shelly J., Robert A. Pollak, and Terence J. Wales. 1997. "Do Husbands and Wives Pool Their Resources? Evidence from the United Kingdom Child Benefit." Journal of Human Resources 32(3):463­80. Luttrell, Cecilia, Sitna Quiroz, Claire Scrutton, and Kate Bird. 2009. "Understanding and Operationalizing Empowerment." London, UK: Overseas Development Institute Working Paper Series 308. Mackenzie, David. 2003. "How Do Households Cope with Aggregate Shocks? Evidence from the Mexican Peso Crisis." World Development 31(7):1179­99. Maluccio, John A. 2005. "Coping with the 'Coffee Crisis' in Central America: The Role of the Nicaraguan Red de Protección Social." Washington, DC: International Food Policy Research Institute (IFPRI), FCND Discussion Paper 188. Mammen, Kristin, and Christina Paxson. 2000. "Women's Work and Economic Development." Journal of Economic Perspectives 14(4):141­64. Mead, Donald C., and Carl Liedholm. 1998. "The Dynamics of Micro and Small Enterprises in Developing Countries." World Development 26(1):61­74. Mensh, Barbara S., Susheela Singh, and John B. Casterline. 2005. "Trends in the Timing of First Marriage Among Men and Women in the Developing World." New York: Population Council, Policy Research Division Working Paper 202. Miller, Grant, and B. Piedad Urdinola. 2010. "Cyclicality, Mortality, and the Value of Time: The Case of Coffee Price Fluctuations and Child Survival in Colombia." Journal of Political Economy 118(2):113­55. Mumtaz, Zubia, and Sarah Salway. 2005. "'I Never Go Anywhere': Extricating the Links between Women's Mobility and Uptake of Reproductive Health Services in Pakistan." Social Science & Medicine 60(8):1751­65. Newhouse, David, and Yoonyoung Cho. 2010. "How Did the Great Recession Affect Different Types of Workers? Evidence from 17 Middle-Income Countries." World Bank. Washington, DC. Processed. Nussbaum, Martha C. 2000. Women and Human Development: The Capabilities Approach. Cambridge, UK: Cambridge University Press. Paci, Pierella, Ana Revenga, and Bob Rijkers. 2009. "Coping with Crises: Why and How to Protect Employment and Earnings." Washington, DC: World Bank Policy Research Working Paper Series 5094. Pearson, Ruth. 1997. "Global Change and Insecurity: Are Women the Problem or the Solution?" In Isa Baud and Ines Smyth, (eds.), Searching for Security: Women's Responses to Economic Transformations. London: Routledge. Perkins, Robert, Brian Kleiner, Stephen Roey, and Janis Brown. 2004. The High School Transcript Study: A Decade of Change in Curricula and Achievement, 1990-2000. Washington, DC: U.S. Department of Education, National Center for Education Statistics. Phillips, Anne. 2004. "Defending Equality of Outcome." Journal of Political Philosophy 12(1):1­19. Quisumbing, Agnes R., and John A. Maluccio. 2003. "Resources at Marriage and Intrahousehold Allocation: Evidence from Bangladesh, Ethiopia, Indonesia, and South Africa." Oxford Bulletin of Economics and Statistics 65(3):283­327. Rahman, Lupin, and Vijayendra Rao. 2004. "The Determinants of Gender Equity in India: Examining Dyson and Moore's Thesis with New Data." Population and Development Review 30(2):239­68. Rangel, Marcos A. 2006. "Alimony Rights and Intrahousehold Allocation of Resources: Evidence from Brazil." Economic Journal 116(July):627­58. Rasool, Shahana, Robyn Pharoah, Antoinette Louw, Kerry Vermaak, and Aki Stavrou. 2002. Violence Against Women: A National Survey. Pretoria: Institute for Security Studies. Rose, Pauline, and Samer Al-Samarrai. 2001. "Household Constrains on Schooling by Gender: Empirical Evidence from Ethiopia." Comparative Education Review 45(1):36­63. Sabarwal, Shwetlena, Nishta Sinha, and Mayra Buvinic. 2010. "The Global Financial Crisis: Assessing Vulnerability for Women and Children." World Bank. Washington, DC. Processed. Sabarwal, Shwetlena, and Katherine Terrell. 2008. "Does Gender Matter for Firm Performance? Evidence from Eastern Europe and Central Asia." Bonn: Institute for the Study of Labor (IZA) Discussion Paper 3758. Schelling, Thomas C. 1971. "Dynamics Models of Segregation." Journal of Mathematical Sociology 1:143­86. Sen, Amartya. 1990. "Gender and Cooperative Conflict." In I. Tinker, (eds.), Persistent Inequalities. Oxford, UK: Oxford University Press. --------. 1995. Inequality Reexamined. Cambridge, UK: Oxford University Press. --------. 1999. Development as Freedom. New York: Random House. Sklair, L. 1993. Assembling for Development. San Diego: University of California Center of US-Mexico Relations. Strickland, Richard S. 2004. "To Have and to Hold: Women's Property and Inheritance Rights in the Context of HIV/AIDS in Sub-Saharan Africa." Washington, DC: International Center for Research on Women (ICRW), Working Paper June 2004. Tansel, Aysit. 2002. "Determinants of School Attainment of Boys and Girls in Turkey: Individual, Household and Community factors." Economics of Education Review 21(2002):455­70. Tiano, Susan. 1994. Patriarchy on the Line: Labor, Gender, and Ideology in the Mexican Maquila Industry. Philadelphia, PA: Temple University Press. Turk, Carolyn, and Andrew Mason. 2009. "Impacts of the Economic Crisis in East Asia: Findings from Qualitative Monitoring in Five Countries." Paper presented at the The Impact of the Global Economic Slowdown on Poverty and Sustainable Development in Asia and the Pacific Meeting. September 28. Hanoi, September 28. UNAIDS, and WHO. 2009. AIDS Epidemic Update: November 2009. Geneva: Joint United Nations Programme on HIV/AIDS (UNAIDS) and World Health Organization (WHO). UNESCO. 2009. Global Education Digest 2009: Comparing Education Statistics Across the World. Paris: UNESCO. US Agency for International Development. 2008. "Understanding Women's Empowerment: A Comparative Analysis of Demographic and Health Surveys (DHS) Data." Washington, DC: US Agency for International Development (USAID), DHS Comparative Reports 20. Vadnais, Daniel, Adrienne Kols, and Noureddine Abderrahim. 2006. Women's Lives and Experiences: Changes in the Past Ten Years. Calverton, MD: ORC Macro Calverton. Vuri, D. 2008. "The Effect of Availability and Distance to School on Children's Time Allocation in Ghana and Guatemala." Rome, Italy: Understanding Children's Work (UCW Project), Working Paper 40. WHO, UNICEF, UNFPA, and World Bank. 2010. Trends in Maternal Mortality: 1990 to 2008. Washington, DC: World Health Organization (WHO). World Bank. 2003. World Development Report 2004: Making Services Work for Poor People. New York, NY: Oxford University Press. --------. 2007. Global Monitoring Report 2007: Confronting the Challenges of Gender Equality and Fragile States. Washington, DC: World Bank. --------. 2008. World Development Report 2009: Reshaping Economic Geography. Washington, DC: World Bank. --------. 2009. World Development Report 2010: Development and Climate Change. Washington, DC: World Bank. --------. 2010a. "Education Statistics (EDSTATS)". Washington, DC, World Bank. --------. 2010b. Gender and Transport in the Middle East and North Africa Region: Case Studies from the West Bank and Yemen. Washington, DC: World Bank. --------. 2010c. "Women, Business and the Law". Washington, DC, World Bank. --------. 2010d. World Development Indicators. Washington, DC: World Bank.