THE WORLD BANK：world development report 2012 Gender Equality and Development

Overview 15 TABLE 1 Almost 4 million missing women each year Excess female deaths in the world,by age and region,1990 and 2008 (thousands) Total women girls at birth girls under 5 girls 5-14 women 15-49 women 50-59 under 60 19902008 19902008 19902008 19902008 19902008 19902008 China 890 1.092 259 71 21 208 56 9 30 1.470 1254 India 265 257 428 251 45 388 228 吗 子 1,255 856 Sub-Saharan Africa 42 53 183 203 61 77 302 751 50 99 639 1,182 High HIV-prevalence countries 0 0 6 39 18 38 328 4 53 416 Low HIV-prevalence countries 53 177 163 57 264 423 96 68 586 766 South Asia (excluding India) 0 1 99 72 20 176 161 37 51 346 305 East Asia and Pacific(excluding China) 3 4 14 1 14 9 137 113 48 6 216 179 Middle East and North Africa 5 6 13 43 24 15 小 80 52 Europe and Central Asia 7 14 3 0 0 2 4 4 3 27 23 Latin America and the Caribbean 0 0 11 20 10 17 17 51 33 Total 1,212 1427 1.010 617 230 158 1,286 1,347 343 334 4,082 3,882 Source:WDR 2012 team estimates based on data from the World Health Organization 2010 and United Nations Department of Economic and Social Affairs 2009. Note:Totals do not necessarily add up due to rounding. overt discrimination in the household,result- South Asia.High maternal mortality is the main ing from the combination of strong preferences contributor to excess female mortality in the re- for sons combined with declining fertility and productive years.In Afghanistan,Chad,Guinea- the spread of technologies that allow parents to Bissau,Liberia,Mali,Niger,Sierra Leone,and know the sex before birth.32 This is a particular Somalia,at least 1 of every 25 women will die issue in China and North India(although now from complications of childbirth or pregnancy. spreading to other parts of India),but it is also And a much larger fraction will suffer long-term visible in parts of the Caucasus and the Western health consequences from giving birth.33 Balkans. Progress in reducing maternal mortality has Missing girls during infancy and early child- not been commensurate with income growth. hood cannot be explained by a preference for In India,despite stellar economic growth in re- sons alone,although discrimination against girls cent years,maternal mortality is almost six times may contribute to it.It is a result not so much of the rate in Sri Lanka.In the past two decades, discrimination as of poor institutions that force only 90 countries experienced a decline of 40 households to choose among many bad options, percent or more in the maternal mortality ra- particularly regarding water and sanitation. tio,while 23 countries showed an increase.The Markets and households cannot compensate for main problem is,again,that households are be- these poor services. ing asked to make many decisions in the face of Missing women in the reproductive ages re- bad options-a result of multiple service deliv- flect two main factors.First,stubbornly high ery failures.In many parts of the world,this sit- rates of maternal mortality persist,especially in uation is reinforced by social norms that influ- much of Sub-Saharan Africa and some parts of ence household behavior and make it difficult

Overview 15 South Asia. High maternal mortality is the main contributor to excess female mortality in the re￾productive years. In Afghanistan, Chad, Guinea￾Bissau, Liberia, Mali, Niger, Sierra Leone, and Somalia, at least 1 of every 25 women will die from complications of childbirth or pregnancy. And a much larger fraction will suffer long-term health consequences from giving birth.33 Progress in reducing maternal mortality has not been commensurate with income growth. In India, despite stellar economic growth in re￾cent years, maternal mortality is almost six times the rate in Sri Lanka. In the past two decades, only 90 countries experienced a decline of 40 percent or more in the maternal mortality ra￾tio, while 23 countries showed an increase. The main problem is, again, that households are be￾ing asked to make many decisions in the face of bad options—a result of multiple service deliv￾ery failures. In many parts of the world, this sit￾uation is reinforced by social norms that infl u￾ence household behavior and make it diffi cult overt discrimination in the household, result￾ing from the combination of strong preferences for sons combined with declining fertility and the spread of technologies that allow parents to know the sex before birth.32 This is a particular issue in China and North India (although now spreading to other parts of India), but it is also visible in parts of the Caucasus and the Western Balkans. Missing girls during infancy and early child￾hood cannot be explained by a preference for sons alone, although discrimination against girls may contribute to it. It is a result not so much of discrimination as of poor institutions that force households to choose among many bad options, particularly regarding water and sanitation. Markets and households cannot compensate for these poor services. Missing women in the reproductive ages re- fl ect two main factors. First, stubbornly high rates of maternal mortality persist, especially in much of Sub-Saharan Africa and some parts of Source: WDR 2012 team estimates based on data from the World Health Organization 2010 and United Nations Department of Economic and Social Aff airs 2009. Note: Totals do not necessarily add up due to rounding. TABLE 1 Almost 4 million missing women each year Excess female deaths in the world, by age and region, 1990 and 2008 (thousands) East Asia and Pacific (excluding China) China India Sub-Saharan Africa High HIV-prevalence countries Low HIV-prevalence countries South Asia (excluding India) Middle East and North Africa Europe and Central Asia Latin America and the Caribbean Total girls at birth girls under 5 girls 5–14 women 15–49 women 50–59 1990 2008 1990 2008 1990 2008 1990 2008 1990 2008 Total women under 60 1990 2008 890 1,092 259 71 21 5 208 56 92 30 1,470 1,254 265 257 428 251 94 45 388 228 81 75 1,255 856 42 53 183 203 61 77 302 751 50 99 639 1,182 0 0 6 39 5 18 38 328 4 31 53 416 42 53 177 163 57 59 264 423 46 68 586 766 0 1 99 72 32 20 176 161 37 51 346 305 3 4 14 7 14 9 137 113 48 46 216 179 5 6 13 7 4 1 43 24 15 15 80 52 7 14 3 1 0 0 12 4 4 3 27 23 0 0 11 5 3 1 20 10 17 17 51 33 1,212 1,427 1,010 617 230 158 1,286 1,347 343 334 4,082 3,882

16 WORLD DEVELOPMENT REPORT 2012 for women to get maternal health care quickly in the quality of institutions-in the provi- enough even where it is available.And high fer- sion of clean water,sanitation,and maternal tility,partly reflecting low incomes,compounds health care.Because there is only a single point the problem in parts of Sub-Saharan Africa. of entry-through better institutions-for ad- Second,the impacts of the HIV/AIDS pan- dressing female mortality,solving the problem is demic on the mortality of women in many East- hard-much harder than getting girls to school. ern and Southern African countries have been But for any basic notions of human justice,the dramatic.The reason for the greater prevalence global development community must make ad- of HIV/AIDS among women relative to men is dressing this problem a priority. their greater susceptibility and the greater like- lihood that their sexual partners are older and Gender segregation in economic activity thus more likely than younger men to have HIV. and earnings gaps In addition,countries that have had a low-lying Although women have entered the labor force civil conflict (such as Democratic Republic of in large numbers across much of the developing Congo)have also seen an increase in the num- world in the past quarter century,this increased ber of“missing”women.This is in contrast to participation has not translated into equal em- other countries that have had outright wars- ployment opportunities or equal earnings for like Eritrea,where men who went“missing”in men and women.Women and men tend to the years of war increased. work in very different parts of the "economic An examination of the historical experience space,"with little change over time,even in of northern and western European countries high-income countries.In almost all countries, and the United States reveals that similar pat- women are more likely than men to engage in terns of excess female mortality in infancy and low-productivity activities.They are also more the reproductive years existed there but disap- likely to be in wage or unpaid family employ- peared between 1900 and 1950.These reductions ment or work in the informal wage sector.In occurred primarily because of improvements agriculture,especially in Africa,women operate smaller plots of land and farm less remunerative crops.As entrepreneurs,they tend to manage smaller firms and concentrate in less-profitable FIGURE 8 Women and men work in different sectors sectors.And in formal employment,they con- distribution of female/male employment across sectors centrate in“female'”occupations and sectors (figure 8).These patterns of gender segregation 31% Communication Services 169% in economic activity change with economic de- velopment but do not disappear. 21% Retail,Hotels,and Restaurants 17% As a result of these differences in where 13% Manufacturing 12% women and men work,gender gaps in earn- ings and productivity persist across all forms of 4% Finance and Business 4% economic activity-in agriculture,in wage em- ployment,and in entrepreneurship(map 1).In 5 Electricity,Gas and Steam,and Water 1% almost all countries,women in manufacturing 0.5% Mining 2% earn less than men.In agriculture,farms oper- ated by women on average have lower yields 2% Transport and Telecommunications 7% than those operated by men,even for men and women in the same households and for men 27% Agriculture,Hunting,etc 29% and women cultivating the same crops.34 Fe- 1% Construction 11% male entrepreneurs are also less productive than male entrepreneurs.35 In urban areas in Eastern 100% All Sectors/All Occupations 100% Europe and Central Asia,Latin America,and Sub-Saharan Africa,the value added per worker is lower in firms managed by women than in Source:WDR 2012 team estimates based on International Labour Organization 2010(77 countries). those managed by men.36 For firms operating in Note:Totals do not necessarily add due to rounding. rural Bangladesh,Ethiopia,Indonesia,and Sri

16 WORLD DEVELOPMENT REPORT 2012 in the quality of institutions—in the provi￾sion of clean water, sanitation, and maternal health care. Because there is only a single point of entry—through better institutions—for ad￾dressing female mortality, solving the problem is hard—much harder than getting girls to school. But for any basic notions of human justice, the global development community must make ad￾dressing this problem a priority. Gender segregation in economic activity and earnings gaps Although women have entered the labor force in large numbers across much of the developing world in the past quarter century, this increased participation has not translated into equal em￾ployment opportunities or equal earnings for men and women. Women and men tend to work in very different parts of the “economic space,” with little change over time, even in high-income countries. In almost all countries, women are more likely than men to engage in low-productivity activities. They are also more likely to be in wage or unpaid family employ￾ment or work in the informal wage sector. In agriculture, especially in Africa, women operate smaller plots of land and farm less remunerative crops. As entrepreneurs, they tend to manage smaller fi rms and concentrate in less-profi table sectors. And in formal employment, they con￾centrate in “female” occupations and sectors (fi gure 8). These patterns of gender segregation in economic activity change with economic de￾velopment but do not disappear. As a result of these differences in where women and men work, gender gaps in earn￾ings and productivity persist across all forms of economic activity—in agriculture, in wage em￾ployment, and in entrepreneurship (map 1). In almost all countries, women in manufacturing earn less than men. In agriculture, farms oper￾ated by women on average have lower yields than those operated by men, even for men and women in the same households and for men and women cultivating the same crops.34 Fe￾male entrepreneurs are also less productive than male entrepreneurs.35 In urban areas in Eastern Europe and Central Asia, Latin America, and Sub-Saharan Africa, the value added per worker is lower in fi rms managed by women than in those managed by men.36 For fi rms operating in rural Bangladesh, Ethiopia, Indonesia, and Sri for women to get maternal health care quickly enough even where it is available. And high fer￾tility, partly refl ecting low incomes, compounds the problem in parts of Sub-Saharan Africa. Second, the impacts of the HIV/AIDS pan￾demic on the mortality of women in many East￾ern and Southern African countries have been dramatic. The reason for the greater prevalence of HIV/AIDS among women relative to men is their greater susceptibility and the greater like￾lihood that their sexual partners are older and thus more likely than younger men to have HIV. In addition, countries that have had a low-lying civil confl ict (such as Democratic Republic of Congo) have also seen an increase in the num￾ber of “missing” women. This is in contrast to other countries that have had outright wars— like Eritrea, where men who went “missing” in the years of war increased. An examination of the historical experience of northern and western European countries and the United States reveals that similar pat￾terns of excess female mortality in infancy and the reproductive years existed there but disap￾peared between 1900 and 1950. These reductions occurred primarily because of improvements Source: WDR 2012 team estimates based on International Labour Organization 2010 (77 countries). Note: Totals do not necessarily add due to rounding. FIGURE 8 Women and men work in different sectors Communication Services Retail, Hotels, and Restaurants Manufacturing Finance and Business Mining Agriculture, Hunting, etc. Construction All Sectors / All Occupations Transport and Telecommunications Electricity, Gas and Steam, and Water 31% 21% 13% 4% 0.5 0.5 % % 2% 27% 1% 100% 16% 17% 12% 4% 1% 2% 7% 29% 11% 100% distribution of female / male employment across sectors

Overview 1) MAP 1 Earnings gaps between women and men (female earnings relative to SI of male earnings) Germany 62c Iceland 69c /A.R.of Egypt 82c Mexico 80c Benin 80c Nigeria 60 Ethiopia34c Sri Lanka 50c Malawi 90 Salaried Workers ⊙6 Farmers Entrepreneurs Sources:Data for Benin come from Kinkingninhoun-Medagbe and others 2010:for Malawi from Gilbert,Sakala,and Benson 2002:for Nigeria from Oladeebo and Fajuyigbe 2007:for Bangladesh,Ethiopia,and Sri Lanka from Costa and Rijkers 2011;and for Egypt,Georgia,Germany,Iceland,India,and Mexico from LABORSTA,International Labour Organization. Lanka,the differences in profitability are signifi- disproportionate responsibility for housework cant between female-owned and male-owned and care,while men are responsible mostly for businesses.37 market work(figure 10).When all activities are So,what explains this persistent gender seg- added up,women typically work more hours regation in economic activity and the resulting than men,with consequences for their leisure gaps in earnings?The Report argues that gender and well-being.And everywhere they devote differences in time use,in access to assets and more time each day to care and housework credit,and in treatment by markets and formal than their male partners:differences range institutions (including the legal and regula-from one to three hours more for housework, tory framework)all play a role in constraining two to ten times the time for care (of children, women's opportunities.These constraints are elderly,and the sick),and one to four hours shown in figure 9 as wedges blocking progress less for market activities.Even as women take toward greater gender equality.Income growth up a bigger share of market work,they remain has some influence in shifting these patterns but largely responsible for care and housework. does not eliminate them.The mutually reinforc-And these patterns are only accentuated after ing interactions between these different factors marriage and childbearing. make the problem particularly difficult to break. A second factor driving segregation in em- Consider each in turn. ployment and earnings gaps is differences in The differing amounts of time that men and human and physical endowments (including women allocate to care and related household access to assets and credit).Despite increases in work are one factor driving segregation and women's education,there are still differences in the consequent earnings gaps.In most coun- human capital between women and men.These tries,irrespective of income,women bear a include a gap in years of schooling among older

Overview 17 disproportionate responsibility for housework and care, while men are responsible mostly for market work (fi gure 10). When all activities are added up, women typically work more hours than men, with consequences for their leisure and well-being. And everywhere they devote more time each day to care and housework than their male partners: differences range from one to three hours more for housework, two to ten times the time for care (of children, elderly, and the sick), and one to four hours less for market activities. Even as women take up a bigger share of market work, they remain largely responsible for care and housework. And these patterns are only accentuated after marriage and childbearing. A second factor driving segregation in em￾ployment and earnings gaps is differences in human and physical endowments (including access to assets and credit). Despite increases in women’s education, there are still differences in human capital between women and men. These include a gap in years of schooling among older Lanka, the differences in profi tability are signifi - cant between female-owned and male-owned businesses.37 So, what explains this persistent gender seg￾regation in economic activity and the resulting gaps in earnings? The Report argues that gender differences in time use, in access to assets and credit, and in treatment by markets and formal institutions (including the legal and regula￾tory framework) all play a role in constraining women’s opportunities. These constraints are shown in fi gure 9 as wedges blocking progress toward greater gender equality. Income growth has some infl uence in shifting these patterns but does not eliminate them. The mutually reinforc￾ing interactions between these different factors make the problem particularly diffi cult to break. Consider each in turn. The differing amounts of time that men and women allocate to care and related household work are one factor driving segregation and the consequent earnings gaps. In most coun￾tries, irrespective of income, women bear a MAP 1 Earnings gaps between women and men (female earnings relative to $1 of male earnings) Sources: Data for Benin come from Kinkingninhoun-Mêdagbé and others 2010; for Malawi from Gilbert, Sakala, and Benson 2002; for Nigeria from Oladeebo and Fajuyigbe 2007; for Bangladesh, Ethiopia, and Sri Lanka from Costa and Rijkers 2011; and for Egypt, Georgia, Germany, Iceland, India, and Mexico from LABORSTA, International Labour Organization. Mexico 80¢ A.R. of Egypt 82¢ Iceland 69¢ Benin 80¢ Nigeria 60¢ Malawi 90¢ Ethiopia 34¢ India 64¢ Germany 62¢ Georgia 60¢ Sri Lanka 50¢ Bangladesh 12¢ Salaried Workers Farmers Entrepreneurs

18 WORLD DEVELOPMENT REPORT 2012 FIGURE 9 Explaining persistent segregation and earnings gaps INFORMAL INSTITUTIONS Social norms on care/market work ECONOMIC MARKETS OPPORTUNITIES Differential access to HOUSEHOLDS labor/credit/land Differential markets,and allocation of networks time/resources AGENCY ENDOWMENTS FORMAL INSTITUTIONS GROWTH Biased laws/regulations, and limited infrastructure Source:WDR 2012 team. cohorts as well as differences in what women large.Data for 16 countries in five developing and men choose to study in younger cohorts- regions indicate female-headed households are differences that affect employment segregation, less likely to own and less likely to farm land.38 especially in countries where most young people More generally,where evidence is available go to college.In agriculture and entrepreneur- for all farmers,women seldom own the land ship,large and significant gender disparities in they farm.For example,in Brazil,women own access to inputs(including land and credit)and as little as 11 percent of land.And their land- in asset ownership are at the root of the gender holdings are systematically smaller than those productivity gap.Indeed,yield differences for owned by men.In Kenya,women account for 5 female and male farmers disappear altogether percent of registered landholders nationally.39 when access to productive inputs is taken into And in Ghana,the mean value of men's land- account(figure 11).Differences in access to in- holdings is three times that of women's land- puts may be further compounded by differences holdings.40 Similarly large gaps are observed in in the availability of“market time,”as noted use of fertilizers and improved seed varieties in above,which can make the same investment agriculture,and in access to and use of credit less productive for women than for men.Jointly, among entrepreneurs. these constraints mean that women entrepre- Third,market failures and institutional con- neurs and farmers are often restricted to busi- straints also play a role.Labor markets often do nesses and activities that are less profitable and not work well for women,especially if their pres- less likely to expand. ence is limited in some sectors or occupations. How big are gender differences in access to When few women are employed,employers may assets (especially land),credit,and other in- hold discriminatory beliefs about women's pro- puts?A variety of data sources suggests they are ductivity or suitability as workers-these beliefs

18 WORLD DEVELOPMENT REPORT 2012 large. Data for 16 countries in fi ve developing regions indicate female-headed households are less likely to own and less likely to farm land.38 More generally, where evidence is available for all farmers, women seldom own the land they farm. For example, in Brazil, women own as little as 11 percent of land. And their land￾holdings are systematically smaller than those owned by men. In Kenya, women account for 5 percent of registered landholders nationally.39 And in Ghana, the mean value of men’s land￾holdings is three times that of women’s land￾holdings.40 Similarly large gaps are observed in use of fertilizers and improved seed varieties in agriculture, and in access to and use of credit among entrepreneurs. Third, market failures and institutional con￾straints also play a role. Labor markets often do not work well for women, especially if their pres￾ence is limited in some sectors or occupations. When few women are employed, employers may hold discriminatory beliefs about women’s pro￾ductivity or suitability as workers—these beliefs cohorts as well as differences in what women and men choose to study in younger cohorts— differences that affect employment segregation, especially in countries where most young people go to college. In agriculture and entrepreneur￾ship, large and signifi cant gender disparities in access to inputs (including land and credit) and in asset ownership are at the root of the gender productivity gap. Indeed, yield differences for female and male farmers disappear altogether when access to productive inputs is taken into account (fi gure 11). Differences in access to in￾puts may be further compounded by differences in the availability of “market time,” as noted above, which can make the same investment less productive for women than for men. Jointly, these constraints mean that women entrepre￾neurs and farmers are often restricted to busi￾nesses and activities that are less profi table and less likely to expand. How big are gender differences in access to assets (especially land), credit, and other in￾puts? A variety of data sources suggests they are FIGURE 9 Explaining persistent segregation and earnings gaps HOUSEHOLDS FORMAL INSTITUTIONS MARKETS INFORMAL INSTITUTIONS ENDOWMENTS ECONOMIC OPPORTUNITIES AGENCY Differential allocation of time/resources Biased laws/regulations, and limited infrastructure Differential access to labor/credit/land markets, and networks Social norms on care/market work GROWTH Source: WDR 2012 team

Overview 19 FIGURE 10 Across the world,women spend more hours per day on care and housework than men FOR SALE 数 分 Market activities Housework Child care Pakistan 0.6 4.7 5.5 2.5 1.2 0.2 Cambodia 2.7 3.8 4.4 33 0.9 0.1 South Africa 2.1 3.8 4.2 18 0.5 0.0 Bulgaria 29 3.9 4.7 2.6 0.4 0.1 Sweden 3.2 4.6 3.2 2.3 0.6 0.3 Italy 2.1 4.8 4.9 1.4 0.6 0.2 women ①=12hous men Source:Berniell and Sanchez-Paramo 2011. can persist if there are no mechanisms in place measures,prevent women from entering some to correct them.Access to information about sectors or occupations. jobs,and support for promotions and advance- In sum,whether women are farmers,en- ment,often occur in gendered networks,hurt- trepreneurs,or workers,many are caught in a ing women trying to enter a male-dominated productivity trap:working hard on an uneven field (or equally hurting men trying to enter a playing field with unequal access to productive female-dominated one,such as nursing).And inputs.This trap imposes significant costs on sometimes,legal barriers,framed as protective women's welfare and economic opportunities

Overview 19 measures, prevent women from entering some sectors or occupations. In sum, whether women are farmers, en￾trepreneurs, or workers, many are caught in a productivity trap: working hard on an uneven playing fi eld with unequal access to productive inputs. This trap imposes signifi cant costs on women’s welfare and economic opportunities can persist if there are no mechanisms in place to correct them. Access to information about jobs, and support for promotions and advance￾ment, often occur in gendered networks, hurt￾ing women trying to enter a male-dominated fi eld (or equally hurting men trying to enter a female-dominated one, such as nursing). And sometimes, legal barriers, framed as protective FIGURE 10 Across the world, women spend more hours per day on care and housework than men Source: Berniell and Sanchez-Páramo 2011. Pakistan Cambodia South Africa Bulgaria Sweden Italy Market activities Housework Child care 0.6 4.7 5.5 2.5 1.2 0.2 2.7 3.8 4.4 3.3 0.9 0.1 2.1 3.8 4.2 1.8 0.5 0.0 2.9 3.9 4.7 2.6 0.4 0.1 3.2 4.6 3.2 2.3 0.6 0.3 2.1 4.8 4.9 1.4 0.6 0.2 women = 12 hours men FOR SALE