The time women spend on paid and unpaid work is typically greater than the time men spend in the labor market (see table 1.2 for an example). Unpaid family work is rarely recorded in official statistics. It manifests itself only indirectly in the labor market in the form of gender differences in labor force participation rates. sector of employment, hours of work, and wage level.

On the whole. labor force participation rates for women are lower than those for men (figure 1.6). However. these differences are often exaggerated because the definition of the participation rate fails to capture many aspects of women's work, particularly time spent on childbearing, childbearing. and other household tasks. Men are usually in the labor force throughout the prime working years (age 20-60), and their participation rates are typically more than 90 percent in virtually every country. Female participation rates vary widely across countries. In 1990, for every ten men in the labor force there were two women in the Middle East and North Africa, three in South Asia. six in Sub-Saharan Africa, and seven in Southeast Asia (United Nations 1991). Worldwide, 41 percent of women age 15 years or over are in the labor force, but in developing countries the corresponding figure is 31 percent.

These numbers are deceptive, however, because they do not take into account the agricultural work for which women in developing countries are responsible within the family For example, the Dominican census of 1981 reported that only 21 percent of rural women participated in the labor force, but just three years later a special study suggested a figure of 81 percent. The census had omitted such activities as cultivating gardens and caring for domestic animals In India different definitions of what constitutes "work" have resulted in estimated participation rates as low as 13 percent and as high as 88 percent (Beneria 1992).

Tens of Millions of Women Suffer Female Genital Mutilation.

In Sri Lanka's Dry Zone, Women Work Longer Hours than Men.

Women are usually employed in different sectors than men. Most of women's nonagricultural employment is in the service sector. but in developing countries female employment in manufacturing has been increasing and is catching up with female employment in services (ILO/ INSTRAW 1985). Women in manufacturing tend to be concentrated in only a few sub sectors: more than two-thirds of the global labor force in garment production is female. and this subsector absorbs almost one-fifth of the female labor force in manufacturing (UNIDO 1993). Men's employment is more evenly distributed across other sectors such as mining manufacturing, construction, utilities, and transport.

Over their lifetime, women change their employment status more often than do men. They are also more likely to be self-employed or employed in occupations with flexible house such as subcontracted home work. Regardless of the sector in which they are employed, women tend to work in a narrow range of occupations. Only a few women are in high-paying jobs or in positions with significant responsibility. Nearly two-thirds of the women in manufacturing ate laborers, machine operators. and production workers; only 5 percent ate professional and technical workers and only 2 percent are administrators and managers (UNIDO 1993). However. there are regional differences: women occupy nearly 60 percent of clerical, sales. and service jobs in Latin America and the Caribbean but fewer than 20 percent of similar positions in South Asia North Africa. and the Middle East (figure 1.7).

Wages paid to women are typically about 60-70 percent of those paid to men. About one-quarter of the gender wage gap is explained by differences in educational levels, labor market experience, and other human capital'' characteristics (Psacharopoulos and Tzannatos 1992 Horton 1994). The gender wage gap can also be explained in part by women's lower participation in the labor market-a consequence of domestic and other demands on their time and, possibly, of discriminatory employment practices.

Significant changes in the global economy have affected patterns of employment and working conditions for men and women worldwide. "Globalization" is associated with the deregulation of product and labor markets, with regionalization. and with the liberalization of international trade. In turn. these processes at-e associated with increased female participation in the labor force and with the at-owing "casualization'' of employment, as seen in the growth of part-time work in industrial economies. However, the net effect of globalization on women workers is not yet clear. Growth in the international traded service sector (for example, banking and telecommunications) seems to have offered women in developing countries greater employment opportunities. In addition, the participation rate of women in manufacturing jobs has increased faster than that of men. Women's average participation in the manufacturing labor force is now around 30 percent for both developing and industrial countries.

Worldwide, the number of women employed in export manufacturing has been increasing rapidly, even though this sector employs only a small fraction of all women workers. In Mexico, for example, the number of women employed in export manufacturing rose by nearly 15 percent a year in the 1980s (Jurisman and Moreno 1990). However, employment in this sector has been increasing more rapidly for men than for women, partly because of

technological upgrading over time and partly; because women are less educated and tend to remain in low-skill occupations (Baden 1992).

Women now account for a growing share ret wage employment they tend to stay longer in the labor force than ever before

Despite persistent gender inequalities in the labor market. some recent trends are encouraging. Increasing educational opportunities and decreasing fertility rates have fed to an increase in the number of women entering the labor market. since the 1950s the female labor force has expanded twice as fast as the male labor force. Women now account for a growing share of wage employment. and they tend to stay longer in the labor force than ever before The narrowing of the gender gap in labor force participation is enabling women to accumulate the work experience necessary tot- improving their job opportunities and increasing their amines. In the formal sector more and more women are working in occupations and sectors once dominated by men. and in many countries women s wages relative to men s have increased over time

The data in this chapter illustrate some aggregate trends but they cannot tell US anything about the processes behind the persistence of gender inequality. For a more detailed look at these processes. We turn in the next chapter to a growing body of empirical evidence generated at the household and enterprise level. These studies provide a telling insight into the way in which gender inequalities are being challenged particularly by women At the same time these inequalities are reinforced by economic, legal. and cultural incentive systems that discriminate against women Discrimination continues despite compelling evidence showing that less inequality especially within the household. is associated with heftier welfare outcomes for children and better economic outcomes for the household as a v hole.

In recent years, attempts to explain persistent gender inequalities in the accumulation and use of human capital have focused on the key role of household decisionmaking and tile process of resource allocation within household Households do not make decisions in isolation. however: their decision are linked to market prices and incentives and are influenced by cultural legal and state institutions. These institutions indirectly affect not only the returns on household investment but also access to productive resources and employment outside the house hold.

Household decisions about the allocation of resources have a profound effect on the schooling, health care, and nutrition children receive. The mechanisms used to make these decisions and the effect of the decisions on the well-being of individual household members are not fully understood. Two frameworks for thinking about household decisions are the unitary household model and the collective household model (see box 3.1)

Policy interventions based on the sources and decisions about how unitary household model aim to in those resources are used within the crease household welfare, but they household. Thus, an increase in do not necessarily affect all household income may benefit some hold members in the same way. Some household members but leave others household members may be worse unaffected or worse off. The outcome off, for example, if they lose control depends on a member's ability to exover certain resources; others may excise control over resources both in be better off. in which case house-side and outside the household, and it hold welfare is not maximized. cannot be assumed that individual well

Under the collective household being increases as household income model. the welfare of individual house-rises. (Collective household models do hold members is not synonymous with not exclude the possibility that the universal household welfare. Resources unitary model may be the best approximate not necessarily pooled, and the motion of household decisionmaking household acts as a collective, with in some contexts.)

The collective household model helps explain why gender inequalities persist even though household becomes increase over time The next sections adopt a collective household framework to explain how these inequalities exact costs in forgone productivity, seduced welfare for individuals and households, encl. ultimately. slower economic growth.(fender Inequalities in Human Capital )

There are strong complementaires between education. health. and nutrition on the one hand, and increased well-being, labor productivity, and growth. On the other. Inequalities in resource allocation that limit household members' educational opportunities. access to health care, or nutrition are costly to individuals, households, and the economy as a whole.

At the household level. gender differences in access to education are closely related to inequalities in the shares of household education expenditures allocated to boys and to girls This finding stands even though private returns to girls' schooling are similar to or marginally higher than. those to boys' schooling (figure 2.1; see also Schultz 1988: Mwabu 1994). In this case. parental choice reflects the relatively greater restrictions on educational opportunities and employment choices for girls. in comparison with boys. and cultural norms on the appropriate role for girls within the household.

More importantly. the social externalities linked to female education are crucial. Evidence from a large number of countries shows that female education is linked with better health for women and their children and with lower levels (figure 2.2). This link stems from the direct effect of education on the value of a woman's time and consequently. on private returns to her labor. It also stems from the indirect effect of education on the average age at which women marry on their knowledge of basic health care and nutrition, and on reproductive choices (Rosenzweig and Schultz 1982, 1987).

Educating girls and women reduces maternal mortality and fertility rates and increases the demand for health services. A simulation study of seventy two developing countries shows that, with all other factors held constant, a doubling of female secondary school enrollments in 1975 would have reduced tile average fertility rate in 1985 from 5.3 to 3.9 children per household and would have lowered the number of births by 29 percent (Subbarao and Raney 1992. Studies for individual countries have found that one additional year of female schooling can reduce the fertility rate. on average, between 5 and 1 (i) percent (Summers 1994)

Women are more vulnerable than men to micronutrient deficiencies that aggravate poor health (World Batik 1994b). Poor health and nutrition reduce productivity and the chances of reaping gains from investment in education. Intrahousehold inequalities in consumption and nutritional allocations can therefore be a signal of inefficiency Recent estimates suggest that the com effects on morbidity and mortality of just three types of deficiency-in vitamin A, iodine. and iron-could waste as much as 5 percent of gross domestic product (GDP), yet correcting these deficiencies would cost less than 0.3 percent of low-income countries' GDP (World Bank 1994c, pp. 50-51). Studies of women tea producers in Sri Lanka and of women workers in Chinese cotton mills document the reduction in productivity associated with iron deficiencies and the positive effects of in-on supplementation on work and output (Edgerton and others 1979). A study of six villages in Andhra Pradesh, India, found that disabling conditions caused by malnutrition and the prevalence of diseases reduced female labor force participation by 22 percent (Chatterjee 1991).

Physical anti mental abuse can also have deleterious effects on the well-being and productivity of women (table 2.1). Violence against women is widespread in all cultures and cuts across all age and income groups Its consequences. include unwanted pregnancy. infection with STDs, miscarriage partial or permanent disability. and psychological problems such as depression anti low self-esteem. Recent World Bank estimates indicate that in both industrial and developing countries domestic violence and rape cause women of reproductive age to lose a significant percentage of really days. Domestic violence appears to be an example of how the relatively weaker bargaining power of women and the paucity of options for them outside the home can affect the intrahousehold distribution of welfare.

Violence against Women Starts in the Womb and Continues trough Life

Source: Heise 1993

Prospects for gainful employment. as well as the availability of basic social services such as water supply and sanitation. also influence women's well-being. The survival chances of female children in India appear to increase as the employment rate for women rises and the earnings differential between men and women decreases Battilan 1988 Dowry and marriage practices. along with household ownership of land are closely linked to women's chances of survival The effect of such practices cannot be over late. Sen's (1990) comparison of female-male mortality rates in China. India. and Sub-Saharan Africa suggest that more than 100 million women are "missing"

Public spending on social services affects the ability of individuals and households to benefit from their own private spending on human capital. In most industrial countries mortality rates decreased even before modern medical care became widely available, mainly because of improved water supplies and sanitation. The same is likely to hold for the developing world. Public interventions that address market under provision of water and sanitation facilities and shortages of health set-vices such as immunization and family planning-have a significant effect. and one that is probably greater for women than for men.

Lack of data makes it difficult to evaluate the effects of public spending on the educational attainment and health of women and men Very little empirical evidence in this area exists, but there is some indication that the proportion of public subsidies for education that benefits females is lower than the proportion that benefits males. In Kenya in 1992-93. for example. public spending on education amounted to the equivalent of an annual subsidy of 605 Kenyan shillings per capita: males received. on average. 670 shillings and females only 543 shillings. In Mexico the gender difference in public education subsidies was somewhat less: in Pakistan males received almost twice the female subsidy. Gender inequalities in the distribution of subsidies are greater at high than at primary levels. In Pakistan girls receive, on average. only 96 rupees a year for secondary school bag. compared with 58 rupees for boys. Box 2.2 explains how these figures are calculated.

Supply-side differences are partly a function of poorly targeted resources (as. for example, when more is spent on tertiary than on primary schooling). More importantly. they reflect differences in household demand for education for girls and boys. If more girls attended anti stayed in school. the proportion of public subsidy going to girls would he greater.

In health care assessing the incidence of public spending by gender is particularly difficult because of the marked differences in the health needs of women and men. These differences are related to different biological re

The main advantage of this type of analysis is that it measures how well public services are targeted to certain groups in the population, including women, the poor, and residents of regions of interest. Requirements anti to women's reproductive and childbearing roles. Some evidence suggests that because of these differences. per capita health cat-e subsidies to women are the same as and sometimes larger than those to men What is clearer is that women are less likely than men to seek health care. including in many countries. hospital care when it is netted and ate mote likely to consult a non medical health cat-e worker (lather than a qualified medical practitioner). Education tends to increase the likelihood that women will seek health care. whether public or ace the likelihood that women will seek health care. whether public or private (World Bank 1994b).

The importance of private and public investments in education and health services that will improve women's well-being is cleat. These services at-e also important for another season: these is important evidence that increases in women's well-being yield important intergenerational benefits and productivity gains in the future World Bank Living Standards Measure Study (LSMS) surveys car-tried out in Nicaragua (1993 Viet Nam 1993) Pakistan (1991), and Cote (1988) suggest that the probability of children being enrolled in school increases with their mother's educational level and that (controlling, for income and household size) girls. particularly those in non poor households. are mote likely to attend school if their mothers attended school. Other studies show that households with educated mothers lend to provide children with greater quantities of mote nutritious food often at a lower cost. than households with poorly educated mothers (Thomas 1990b)

Data from Brazil show that giving women more control over nonlabor income has a larger impact on child measures, nutritional intakes, and the proportion of the household budget devoted to human capital inputs than if men had control of this income (Thomas 1993) Other studies indicate that women spend proportionately mote of the income they control on health care for children than do men. Women also spend more on food products (as opposed to such goods as alcohol and tobacco) than men do (Duraiswamy 1992: Hoddinot and Haddad 1992) But fathers education is also important especially in interaction with mothers' education (Thomas 1990a)

Other studies have shown that children's educational attainment is lowest in households whet-e the male head has no schooling In Ghana the impact of a mother's education on her children's schooling is reduced when their father has no schooling. even after controlling for income and household size (Lavy 1992).

Material health (which is linked to education) also has important intergenerational effects. Children of mothers who are malnourished or sickly or receive inadequate prenatal and delivery cat-e face a greater risk of disease and premature death. Iodine-deficient mothers run a greater risk of giving birth to infants with severe mental retardation and other congenital abnormalities than do healthy mothers. Reduced fertility and improved health for women can increase individual productivity and improve family well-being. When good health is combined with education and access to jobs. the result is higher rates of economic growth.

Unpaid work and family responsibilities. as well as lack of investment in women's education. are strongly associated with women's relatively low rates of participation and their limited earnings in formal sector labor. Women's participation rates usually dip in the childbearing years, and earnings tend to decline following an interruption in employment. Younger on average, work more hours than older women, and married women with young children tend to work less than childless women and mothers of grown children. The correlation of marriage and childbearing with labor market outcomes can be seen even in industrial countries, where wage differences between married women and men are larger than those between single women and men Similarly, in some developing countries relative earnings decline with age (table 2.2) Children are not the only treason for interruptions in women's labor force participation; caring for ill or aged family members is often a woman’s responsibility. A study from Hungary estimates that half of all absenteeism by women workers is the direct result of the need to care for sick relatives (Einhorn 1993).

In Countries, Single Women Earn More Than Married Women and Younger Women More Than Older Women.

Sources: Blau and Kahn 1992 Sedlacek Gutierrez. and Mohindra 1993

Thus. women s labor market outcomes can be substantially poorer than those of men because women's employment opportunities are constrained by social arrangements at the family or household level. These social demands are reinforced by legal conventions. Within the labor market itself, social or employer discrimination can affect women and men differently. and these differences are reflected in the resource allocation decisions taken within the household. Although wage discrimination is illegal in many countries, employers may respond to an increase in the supply of workers by segregating jobs by gender or offering less training to women who they perceive as being temporarily attached to the labor force (even if in fact most women never drop out). For example, women in the former Soviet Union are fairly well educated and have high labor force participation. but they are concentrated in occupations requiring fewer skills and less vocational training than men, and, on average. they earn less than men (Fong 1993).

Female Wages Are Lower than Male Wages, but This Is Changing.

Note: Monthly earnings for Indonesia and Thailand rural and urban: for all others urban only

Source: Tzanntors 1995

Women's earnings relative to men's tend to increase over time. A study of six developing countries shows that female earnings relative to male earnings increased by 1 percent a year in the 1980s (table 2.3). There were two reasons for this increase: over the years women entered higher-paying sectors. and within sectors, their pay increased in relation to that of men. This gain would have been even greater had it not been for the effect on wages of increased female participation in the labor force. However, the most visible dimension of gender inequality in the formal labor sector remains the wage difference between men and women. Women's wages are. on average lower those of men by about 30 to 40 percent.

Social or employer discrimination can affect women and men difference and these difference are reflected i/? the resource allocation decisions taken within the household

A recent study of the gender wage gap in Russia shows that after controlling for education differences, the ratio of women's to men's average hourly earnings stands at just over 71 percent: it has remained at that level since the 1960s. Part of the reason for women s lower hourly earnings in Russia and many other countries lies in patterns of occupational segmentation by ,gender. Some analysts argue that women-who do most of the household work in Russian households and also have high participation in the formal labor market cope with the burdens imposed on them by taking less demanding work and devoting less time to advancing their careers (Newell and Reilly 1994).

The availability of financial services and access to them are considered important for several reasons. First. savings provide a kind of self-insurance. Second, credit helps households maintain a certain level of consumption at those times when their income fluctuates temporarily. Third, credit can be used to fund investments in capital or other inputs that will yield relatively high returns to production. if households cannot finance such investments from their own savings. A fourth and no less important reason is the role of savings and credit in increasing household members' options outside the home.

Inequalities between women and men in access to financial services- particularly credit--are widely documented. Collateral requirements, high transaction costs. limited mobility and education, and other social and cultural barriers contribute to women's inability to obtain credit (Holt and Ribe 1991). The implications tot household efficiency and individual well-being differ, however, depending on whether the household pools its financial resources. If, as the unitary household model assumes. a household pools its resources the characteristics of individual borrowers are less important than if there is little or no pooling. In the first case. credit resources will be used to meet household needs that have been jointly determined, regardless of who the borrower is. In the second case, the use credit is put to and the needs it then satisfies depend on which household member is borrowing.

A recent study of credit programs in Bangladesh sponsored by the World Bank shows that providing credit directly to women has a positive effect on variables typically associated with household and individual welfare and improved gender equality (Pitt and Khandker 1995). The study looks at three programs in Bangladesh: the Grameen Bank. the Bangladesh Rural Advancement Committee (BRAC), and the Bangladesh Rural Development Board (BRDB). in 1992, women accounted for 94 percent of Grameen Bank members. 82 percent of BRAC members. and 68 percent of BRDB members (Khandker and Khalily 1995; Khandker, Lavy, and Filmer 1994). Only the results for the Grameen Bank are presented here.

Table 2.4 presents the impact of credit obtained by women and men on a variety of social and economic indicators. The results show a clear and positive impact of both male and female borrowing on all indicators of Toward gender equality family welfare especially through an increase in per capita expenditures increases in both boys' and girls schooling, anti a reduction in fertility. Females borrowing has a greater effect on girls' schooling anti per capita expenditure than does male loot-rowing: male borrowing has a greater effect on boys' schooling and fertility than does female borrowing. Interestingly. female borrowing also results in mote female ownership of nonoland assets and an increased supply of female labor to cash - income -earning activities (Pitt and Khandker 1995).

Loans to Women and Men Have Important Welfare Implications.

The introduction of programs such as the Grameen Bank in a village has a positive effect on agricultural and nonagricultural production (Khatidketati Chowdhuty 1994: Rahman and Khandke 1994 The probability that women will be self-employed rather than work for wages increases by 52 percent. The latter finding is important because much of the wage employment open to women in rural areas is very poorly remunerated and can be quite exploitative. Self employment can bring the opportunity of higher returns for women, plus the freedom to integrate their earning activities into other work as they see fit.

Access to financial services alone cannot reduce gender inequalities in the allocation of household resources. A qualitative study reviewing several targeted credit programs in Bangladesh cautions against overgeneralizing about the benefits of giving women access to credit. The study finds that it is difficult to infer that increased borrowing alone improves women's bargaining power because in many rural Bangladesh households the question of who controls the resources is quite complex (Goetz and Sen Gupta 1994). Nevertheless. the possibility of receiving credit (or similarly of working for wages) may give women greater bargaining power within the household. This bargaining power can be used to improve child health and nutrition and may increase the likelihood that children will attend school.

The ownership of land and the distributions of land rights influence the productivity of labor and capital resources and the incentive to invest in resource management Private property rights. in particular. are associated with increased access to product and factor markets. especially credit markets. and to public services such as public utilities and agricultural extension. However. relatively little direct evidence exists to link independent owner of land by women with increased access and productivity. One obstacle to empirical work is that women s access to land and property is often mediated trough marriage (A married woman land rights are frequently limited to use rather than ownership.) Future more. complex systems of land tenure make it difficult to generalize about the effects of owner ship on productivity. None some evidence suggests that independent land rights for women could enhance both the efficiency with which resources are used and the well-being of women and their households: (Agarwal 1994).

The possibility of receiving credit may give e women greater bargaining power within the household which can be used to improve child health and nutrition.

While independent land rights may increase efficiency and household welfare, lack of secure land appears to be associated with low investments by women in land conservation. In Zimbabwe's communal areas, land that a women acquires is often allocated to her only temporarily: for example, the location of land allotment received from husbands or borrowed from neighbors is usually subject to periodic change (Jackson 1993). The same is true in parts of West Africa (David 1992 Jackson 1994). Uncertainly about the permanence of their control over the land means that women may be reluctant to invest in improvements that will benefit the landowner rather than the user.

A significant trend in recent decades in developing countries has been the move toward private ownership In some countries this trend has been encoulagecl by reforms dealing with land redistribution tenancy or land titling Such reforms are considered important for promoting long-term investments and the adoption of the latest technology They also provide the collateral people need to gain access to credit and other factor markets Ironically, evidence also suggests that perform inequalities in male and female land rights are reinforced by land reform programs For example, in Latin America most reforms are based on the premise that the man of the household its the household head. This presumption means that women (except for widows and ownership. Even where women and men benefit equally from land reform differences exist between nominal and real land rights (see box 2.3).

In the central part of European Russia and in Moldova the lands and assets of state and collective farms are being parceled out in allotments as part of wider economic reforms. Every person who lived and worked on a collective farm receives a share of the land Data show that although on average, women have received a slightly higher proportion of land shares than men, the nonland capital assets of the old farms ale being distributed as property shares on the basis of a formula heavily weighted toward an individual's wage rate and years of employment. Such criteria favor men over women and give men more valuable property chares. Dividends paid on these property shares also tend to be higher than those paid 011 land shares Holt 1995).

Land reform programs that fail to account for gender differences in rights

In the northwest and southwest provinces of Cameroon. an estimated 50 percent or more of those who claimed land within the first ten years of land registration (1974-85) were classified as public servants. Over 32 percent of all the remaining land titles went to businesses.

Women make up more than 51 percent of Cameroon s population and do more than 75 percent of the agricultural work. but they are virtually absent from land registers. Only 3.2 percent of all land titles issued in the Northwest Province were given to women; in the Southwest Province the figure was 7.2 percent. For the country as a whole. it is estimated that women obtained under 10 percent of all land certificates (World Bank 1995a). to own. use, and transfer land may actually exacerbate the insecurity of women's land claims and. as a result, harm household welfare. For example, there is evidence that land titling focused on male household heads has adversely affected women's ability to farm independently. Moreover. intrahousehold inequalities in income and decisionmaking have increased (FAO 1993) In Africa some titling programs have allowed men to take advantage of their control over land to redesignate land formerly cultivated by women as household land. This switch provided the opportunity for men to increase the amount of work they expect from women on household plots. In other cases women have received smaller and less fertile plots than they had before for their personal crops (FAO 1993).

Recognizing women's independent claims to land is therefore an important issue in property reform. In poor households. having rights to land could alleviate both women's own poverty and the household's risk of remaining poor. The season is mainly that women s access to economic resources has a positive effect on household welfare (Agarwal 1994). From the point of view of efficiency, secret land tenure increases the incentive to manage resources efficiently and expands access to formal credit markets. Because secure land tenure can mean greater productivity, it may also increase the household's incentives to invest in women's human capital.

Agricultural extension services provide information training and technology to agricultural producers. Extension services have always been regarded as necessary for agricultural modernization. Given the importance of women's labor to agriculture in most regions, providing women with access to agricultural extension services is essential for current and future productivity. Types of agricultural extension services vary, hut in most countries publicly provided services dominate. Evidence suggests that women have not benefited as much as men have from publicly provided extension services.

Given the importance of women's labor to agriculture it is in most regions providing women with access to agricultural extension services is essential for current and future of productivity.

A review of five African countries shows that extension agents are most likely to visit male farmers than female farmers (table 2.5) The impact of this inequity on female productivity depends in part on whether women and men within households pool information. There is, however. little evidence to suggest that this happens (see box 2.4). It is important to ensure that extension services reach omen directly, not only to redress gender inequalities but also to maximize productive efficiency. omen play a critical role in production of food and cash crops for the household. in postharvest activities. and in livestock care Men and omen perform different tasks they can substitute for one another only to a limited extent. and this imitation creates different demands for extension information Also, as men leave farms in search of paid employment ill urban areas. women are increasingly managing and operating farms on a regular and full-time basis. Hence. omen are becoming a constituency for extension and research services in their own right (World Bank 1994e)

The expansion of agricultural services beyond the public sector is a growing phenomenon in developing economies. The inadequacies of public funding plus the need to provide snore client-oriented services, suggest that the private sector has an important role to play However, women's limited access to land and credit put the many potential benefits offered by extension services out of reach. For example in Kenya's Meru and Maranga areas more than half the women surveyed cited a shortage of cash as their reason for not adopting, technologies that would maximize their output and increase their efficiency. The amount of education omen receive and the efficiency with which they run their farms are also closely linked This tie is particularly significant in light of the fact that one purpose of extension services is to advise farmers on use of modern technology.

Three studies of Kenya found that the gender of the farm manager was, by itself. an insignification factor in output per hectare but that the manager's educational level had a significant effect on farm productivity (Moock 1976: Saito and Spurling 1992: Bindlish and Evenson 1993). Simulations based on these studies suggest that significant gains could accrue to increased investment in women's physical and human capital. As the data in table 2.6 show, if women and men shared the same educational characteristics and input levels. farm-specific yields would increase between 7 and 22 percent. Giving women primary schooling, by itself. would increase yields by 24 percent. Thus under investment in women's education limits growth of agricultural productivity. Well-targeted extension services can help to narrow the differences in productivity that arise from educational inequalities (Schultz 1988)

Increasing Human Capital and Input Levels Would Increase tile Yield for Women Farmers.

Discriminatory labor market policies and employment laws are widespread: examples are bans on hiring married women and restrictions on the type of work pregnant women may pet-form. Labor laws may also restrict female participation in jobs and deny women access to work settings. Even legislation that seeks to promote equal opportunity can have outcomes tot women workers. For example, generous maternity and child-care benefits tot women workers may make hiring women relatively mote costly than hiring men. perpetuating the gender wage gap.

The main issue for public policy its to ensure that fair and equal employment laws exist and are enforced.

In India the Self - Employed Women's Association (SEWA) holds worker education meetings to provide members with legal information and offers one day training sessions on specialized legal information for community leaders. Members of SEWA paralegal staff research cases, prepare briefs, and argue all cases in labor courts for members. SEWA'S legal literacy efforts are part of its overall strategy of action for supporting self-employed women and women in the informal sector.

In this regard. the main issue for public policy is to ensure that fair and equal employment laws exist and are enforced. Such laws should cover employment and employment-related benefits and should not be applied in a way that restricts women's access to the labor market.

In many countries women are protected by special standards in the workplace. These safeguards are of two types. The first includes measures intended to end discrimination in the labor market by requiring equal pay for work of equal value or by prohibiting the exclusion of workers from certain jobs because of gender. The second type protects women in their role as mottles by requiring employers to pay the full cost of maternity leave and provide childcare facilities. restricting night work, and limiting work categorized as dangerous or hazardous.

The first type of labor standard is often difficult to enforce. although nearly 120 countries have enacted equal opportunity laws. In most cases the law is in place but is not enforced. Furthermore. it often applies only to the formal sector of the labor market. leaving large number of women in the informal sector with little protection. More concerted public and private action is required to devise appropriate for enforcement. In additions complementary measures are needed to increase women’s chances of entering the labor force. Such measures would include mote met-it-based hiring of women in the public sector and the creation of an appropriate regulatory framework that encourages the establishment of day-care centers private nursery schools. and kindergartens for the children of workers in both the formal and informal sectors. (Box 3.2 describes a day-care projects in Bolivia)

Many countries have laws that limit night work. over theme and the use of heavy machinery by women. While well intentioned. these laws. too. can reduce women's employment options by raising the employer's cost of hiring women and perpetuating the notion that women workers are less flexible than male workers. This adverse result occurs because women already tend to have lower levels of capital than men and are limited by traditional notions of what women’s work. Devising effective protective legislation for women is thus a difficult balancing act.

Effective laws to protect women must honor the fine line between ensuring sate working conditions for all workers, including women in the informal sector. and providing equal opportunities for all workers seeking employment. Standards relating to maternity leave and child-care benefits are common in the formal sector. To ensure that such standards do not raise the costs of female labor unnecessarily employment legislation should avoid having employers pay benefits directly. Maternity benefits should be funded through general revenue taxes

Box 3.2 day-care center in Bolivia to enhance the status of women by increasing their employment opportunities in other Latin American countries, and their knowledge of women in Bolivia are over represented , health, and nutrition. The project among the poor, and many live in will establish day-care centers to proficult circumstances. They are twice vice non formal, home-based, integrals likely as men to be illiterate, their child development for children unaccess to child care is limited, they the age of 6 in thirty-four poor subject to discrimination in hiring ban and purebred areas. Women and wages, and they are restricted benefit from the project either as by law from engaging in certain procure givers employed at the minimum ductive activities. wage or as clients of the centers. The Integrated Child Development children in day care gives women Project, financed by a credit from the more time to search for jobs or to in World Bank's International Development their current situations (Winkler opponent Association designed and Guedes 1995).

Or social security systems rather than by direct contributions from individual employers. In Eastern Europe and Latin America, however, this method of funding benefits has imposed a significant cost on the treasury, raising the question of how much should be paid and for how long.

One way of addressing this issue is for governments to encourage women and men to share the responsibility for childrearing by adopting legislation allowing either parent to qualify for the leave and benefits associated with having a child (Folbre 1994). Such legislation can be supported by changes in the tax system to ensure equal treatment of workers within the household and a margin tax rate on the earnings of additional workers in the household low enough to avoid creating a disincentive to women's participation in the labor force (Mac Donald 1994). These tax changes can be complemented by legislation that encourages absent to pay child support. Such laws are in place in many Latin American and Caribbean countries. but none systematically monitors transfers (Folbre 1994). Greater efforts to enforce the law are needed.

Legal and policy measures can have a direct bearing on the health of women. The law and its enforcement are essential for combating domestic violence against women. If the laws on violence against women are to be enforced, women need to be made aware of the legal recourse available to them and the judiciary and police need to be sensitized to the existence and

In Zimbabwe the Mussasa project provides support and counseling to abused women and sensitizes police and prosecutors to the issues of domestic violence and rape. To ensure the credibility and acceptability of the training, legal professionals are involved in the program. One branch of the criminal justice system-the police department, for example-often hosts training for another, such as prosecutors (Stewart 1992).

In Costa Rica El Instituto Legal de las Naciones Unidas y Desarrollo runs gender sensitization training for prosecutors, judges, lawyers, and other professionals who deal with crimes of violence against women (Heise, Pitanguy, and Germain 1994).

Implication of violence against women. Affirmative action policies can increase the number of women police officers lawyers, and judges. Trading police, lawyers, and judges can increase their gender sensitivity. making the legal system more responsive to women's needs. Legal literacy efforts can make women aware of their rights and show them how to use these rights to mobilize for change. Some approaches to dealing with domestic violence are illustrated in box 3.3

Such reforms generally include two kinds of necessary policies. The first emphasizes macroeconomic stability and the removal of price distortions: the second promotes labor-demanding growth in agriculture and industry and better. mote accessible basic social services. mightily in education, health care and water supply. Macroeconomic stability can be achieved by reducing large current account and government budget deficits and by curtailing excessive money and credit expansion. Correct pricing of foreign exchange and of domestic goods and services facilitates trade and investment and promotes growth Reforms of trade and price incentives encourage job creation and higher earnings.

In shots. broadly based economic reforms benefit the poor, as the experiences of diverse countries have shown. Data from the LSMS and similar studies confirm that in Costa Rica, Ethiopia. Than Indonesia, Peru, the Philippines, and Tanzania, declines in poverty rates have accompanied economic reforms. Because of underlying gender inequalities, reductions in poverty do not necessarily improve the economic status of women. Nonetheless. although empirical evidence on changes in women's welfare following reforms is scarce, data from the Philippines suggest that for women. economic reform is by improvements in social indicators and important employment gains (box 3.4). An analysis of household data from Peru indicates that reforms there have not only led to renewed growth and reduced poverty but have also improved the status of female-headed households (box 3.5).

Despite their seminal role in encouraging sustainable. long-term growth. some economic reforms involve transitional costs. for two reasons: the need to implement radical shifts in government policy, and the lag. perhaps of years, between acceleration of economic growth and the realization of the …

The percentage of the population with access to safe drinking water increased between 1980 and 1990, from 65 to 93 percent of the urban population and from 43 to 72 percent of the rural population. The figures for sanitation services showed the same increase, climbing from 81 percent with access in 1980 to 98 per cent in 1990 among urban dwellers and from 67 to 85 percent among those in rural areas. Secondary school enrollment rates for boys and girls also showed gains. Girls and boys have traditionally reached similar educational levels in the Philippines, a trend that contributed to women's gains in the labor market during adjustment in the 1980s. While labor force participation rates for men rose only slightly, the rates for women increased dramatically, from 23 to 37 percent.

At the same time, women's wages as a percentage of men's grew from 71 to 80 percent between 1978 and 1988 (Tzannatos 1995). A recent study of a low-income community in Manila found that in some private sector jobs women earn more than men, reflecting in part the high educational level among women in the sector. Earnings in the informal sector are generally low for both women and men. although some women have been so successful in their informal sector businesses that their spouses have opted to leave low-paying jobs in the formal sector to participate in their wives enterprises (Moser 1994). benefits of growth by large segments of the population. The short-term costs can include increases in food prices as subsidies are removed temporary restrictions on credit to control monetary expansion, and cuts in public spending as governments attempt to control budget deficits and reorient spending toward the most important public services. Moreover, inequalities in access to and control over resources and in entry into markets hamper some groups in taking advantage of the opportunities created by economic reform For example, people who are unable to acquire new skills may remain for some thee. Gender inequalities in access to and control over land in rural areas can keep women farmers from taking advantage of changes in the relative prices of tradable and non tradable crops.

Improvements in living standards have been widespread, and the largest gains have been made in some of the country's poorest regions, such as the urban and rural sierra (mountain areas). In most parts of the country, increases in per capita expenditure among households headed by women have been above the national average. In the urban and rural sierra, for example. consumption expenditure by these households increased by 47.7 and 44.8 percent. respectively-figures that compare favorably with increases in per capita expenditure among households headed by men (35 and 28 percent for the same regions).

Although both women and men are expected to gain from economic reform in the long run. gender inequalities at the household and market levels can mean that the benefits of reform reach women, especially poor women. only slowly. Reforms must incorporate measures aimed at counter acting shot-term transitional costs. Among the most important are safety nets specifically targeted to vulnerable groups. For example, in fiscal 1994 seventeen of the twenty-three reform programs supported by the World Bank included components aimed specifically at reducing poverty: fourteen of them targeted programs or safety net measures such as labor-intensive public works projects. targeted and social assistance programs, and unemployment and social security schemes (see box 3.6).

However, safety nets are not a substitute for a mote integrated approach to economic and social policy that includes appropriate levels of harvest in social services and infrastructure. Recognizing this fact, governments and donors are increasingly incorporating into their reform packages antipoverty initiatives and measures to reduce gender inequality. In Pakistan, for example, the FIDA-financed Public Sector Adjustment. Program provides much needed the financial support for the balance of payments account and reinforces the emphasis on government investment in the social sectors. The program promotes increased access to and improvements in basic education. health. and water supply services that benefit girls and rural areas. It also supports a shift it government education and health spending toward the primary level.

In fiscal 1994 twelve of twenty-three World Bank economic reform pro emphasized the restructuring public expenditure, primarily to and increase spending on education. health care, and other areas important to poverty reduction such as water supply and sanitation. The Burkina Faso reform program is typical; it supports an increase in the total budgeted amounts for primary education and health services, which are particularly beneficial to the poor and to women. Uganda s program seeks to protect and enhance public expenditures for basic social services. including water supply primary health care and primary education. In the long term, the reform of public spending-especially in the social sectors, physical infrastructure and agricultural research and extension-promises significant benefits) trot women.

As we have seen. policies that specifically target women or girls can address the needs or this group more efficiently and with greater cost-effectiveness than general policy measures. Female household members tend to allocate resources more directly to children. while men tend to allocate more resources to adults. In households in which resources are not pooled. targeting programs to the household as a where will not necessarily benefit all members equally.

Targeting women directly is justifiable on two grounds. First, to the extent that gender inequalities prevent an economy from realizing its full potential. targeting to women can be an effective strategy for increasing productivity and output Second, where gender differences are wide, targeting may be needed to capture social gains and to increase internal efficiency.

Targeting women is especially when doing so contributes directly to reducing poverty; or when women have particular needs-for example. when maternal mortality is very high. The exceptionally high gender gap in educational enrollments in some countries can be reduced only by policies (including subsidies) that target girls. An obvious example would be policies that affect the private costs of schooling.

Female household members tend to allocate resources more directly to children it while men

Reducing these direct costs to households Will mean setting new public spending priorities. For example. education institutions for girls especially at the primary level, might be exempted from cost recovery measures. thus increasing the implied public spending subsidy to girls. Similarly. a far; number of publicly funded scholarships can be provided for girls' as has been done in and Guatemala.

Opportunity and travel costs can discourage parents from enrolling especially daughters in school. Some counties have tried to overcome the constraints imposed by opportunity costs by introducing flexible school hours and calendar years and providing child care for younger siblings (box 3.7). In some cases girls who are responsible for looking after their younger brothels and sisters ate allowed to bring them to school.

In countries where cultural values may prevent girls from traveling alone to school. measures are needed that will increase access to safe transport.

Along with job loss women are affected by the decline in services such as day care and maternity homes.

Mothers now have to look after their children at home, which restricts their ability to participate in the labor market and, ironically, increases their dependence on welfare. When herds were privatized, priority was given to people who, according to the government, could take better care of the animals. Unfortunately, this policy adversely affected female-headed households.

The maternal mortality rate has doubled over the past three years. in part because many more babies are being delivered at home. Other factors that have contributed to the rising maternal mortality rate include a decrease in the number of ambulances and the need for patients to pay for the food they consume while they are hospitalized.

In Mongolia, targeting female-headed households, pregnant women, and children is essential for reducing poverty. Social assistance, health care. education, and help in finding employment are particularly important (Subbarao and Ezemenari forthcoming). change the geographic distribution of primary schools, of provide more boarding facilities. Projects in Pakistan are using school mapping techniques to establish criteria for placing new schools in currently underserved areas. Such programs benefit both girls and boys.

It may be necessary to target women when economic reforms of systemic transitions are occurring. For example, in Mongolia, as well as in other countries making the transition from a socialist to a market economy, women are disproportionately represented among the unemployed and otherwise disadvantaged (box 3.8). Thus, in designing safety nets to mitigate the short negative effects or economic transitions. policymakers need to recognize and evaluate the specific adverse effects on women

It is not always necessary to restrict a particular program of benefit exclusively to women: the objective can sometimes be achieved indirectly.

Techniques based on self-selection appear to be particularly efficient. In Zambia. where men have a strong preference for cash wages, offering wages in kind (food) attracted more women workers than men to public works programs.

In the financial sector. women entrepreneurs are enabled to borrow at market rates of interest when banking institutions adopt innovate collateral requirements, reduce transaction costs, and offer small loans at repeated intervals The Grameen Bank in Bangladesh and Badan Kredit Kecaman in Indonesia do not reserve loans for women specifically: instead. they adopt innovative lending policies the result of which is that women snake up the majority of participants-is 96 percent in a new branch of the Grameen Bank.