Gender shapes the lives of all people in all societies. It influences all aspects of our lives, the schooling we receive, the social roles we play, and the power and authority we command. Population processes-where women and men live, how they bear and rear children, and how they die-are shaped by gender as well.
Gender refers to the different roles men and women play in society, and to the relative power they wield. While gender is expressed differently in different societies, in no society do men and women perform equal roles or hold equal positions of power. The impact of this inequality on women's lives varies tremendously. In the United States, for example, this inequality is reflected by a glass ceiling, which keeps most women from advancing to top levels of management. In some Asian societies, gender inequality can compromise the basic health of women in poor families because they are the last in the household to receive food and medical care.
Gender equality has gained wide acceptance as an important goal for many countries around the world. The growing support for and attention to gender equality is bolstered by the assumption that it would improve the lives of most people, especially those of women and children. Many now consider enhancing gender equality to be a vital component of population policies, as expressed by participants from 180 countries at the 1994 International Conference on Population and Development (ICPD) in Cairo. The participants agreed on the principle "that advancing gender equality and equity and the empowerment of women, and the elimination of all kinds of violence against women, and ensuring women's ability to control their own fertility, are cornerstones of population and development-related programmes."' Whether and how these declarations are put into action through policies and programs remains to be seen in the next decades.
Most countries now acknowledge that gender and the extent of gender equality will influence the timing and shape of demographic change. Recent research has explored these relationships and has highlighted potential policy implications. In general, as the differences between men's and women's roles diminish, women gain status and power within a society and begin to control their reproductive lives. When women have more autonomy, maternal and child health tends to improve, fertility and childhood mortality tend to decline, and population growth slows.
When women have frequent and numerous births, their life choices are often restricted. Their options increase if they have access to safe, effective means of controlling the timing and number of their births. When women have fewer children, they face fewer years of child care and they are freer to participate in activities in the public sphere, such as paid employment or political office.
Decreasing the number and planning the timing of pregnancies reduce women's risk of dying of pregnancy-related illnesses, especially in low-income countries. Child health also tends to improve with more widely spaced births. Babies are less likely to have a low birth weight if the mother has waited two or more years between pregnancies; and mothers are better able to care for the physical needs of their children if births are fewer and further apart.
Many governments expect population programs that promote gender equality also to influence the timing and shape of demographic change. The growing adoption of these policies makes it imperative to understand the relationship between gender and population processes. However, research tells us that this relationship is not exact and is not the same everywhere or at all times. Gender equality is not likely to translate automatically into mortality or, especially, fertility decline, and both mortality and fertility declines have occurred in places with little gender equality. Social scientists have learned that they must look at the links between fertility, mortality, and gender differences within a social and cultural context.
This Population Bulletin will explore the connections between gender and the levels and trends in fertility and mortality. Its goal is to highlight how gender interacts with population processes in the less industrialized world. The focus is on fertility and mortality, although the third source of population change-migration-is intricately connected with gender as well. Who moves, how far they move, how long they stay away, and how closely they maintain ties and fulfill responsibilities to their families and home communities are all influenced by gender.
Gender involves men as much as it does women, and the international commitment to greater gender equity affects men and men's relative power. However, this Bulletin will concentrate on women and women's roles in the society and economy.
What Is Gender?
Gender is more than the differences between women and men, and the term means different things to different people (see Box 1). In this Population Bulletin, we use a definition that focuses on the socioeconomic aspects of gender. Three aspects are particularly relevant to population change: 1. Gender is a social institution. Gender is central to the way a society is organized. Like the family, religion, race, and other social institutions, gender affects the roles men and women play in a society.
Gender also establishes patterns of behavior through interaction with other institutions, just as social class and the economy-also social institutions-influence each other. In most western societies, economic growth depends on people of different classes working in different jobs and accepting different rewards and benefits. Gender functions in a similar manner to organize society. For many years, the U.S. labor force operated under the assumption that most workers were men with wives and children at home to support. Salaries, hours, employer expectations, and the structure of businesses all reflected this assumption, and shaped the way that women were hired, fired, or compensated.
2. Gender involves differences in power.
Like race and social class, gender orders social relationships in such a way that some individuals have greater power than do others. In every society, the roles women and men assume accord women fewer opportunities and privileges.
Gender affects both "power to" and "power over." "Power to" refers to the ability to act and often requires access to social resources such as education, money, land, or time. Women usually have less "power to"-go to school, inherit land, or enter or refuse a marriage. Women are less likely to develop individual characteristics (such as higher levels of education) that would give them access to betterpayingjobs or political office and enhance their power.
Individuals with "power over" are able to assert their wishes and goals even in the face of opposition from others. Women generally have less "power over" than men in all facets of society. They usually have less say than their husbands in family decisions and less authority than men in the workplace. Because women hold far fewer positions in governing bodies, they have little impact on decisionmaking or public policies.
Gender inequality may also be structured and perpetuated by the economy, the political system, and other social institutions. Civil law and religious customs in various countries, for example, may restrict a woman's ability to own property, work in certain occupations, or serve as a religious leader.
3. Gender is a cultural construct. Gender is organized differently in different societies. Accordingly, the expectations for women and men vary throughout the world. These differences are perhaps most clearly illustrated in intercultural comparisons of what is considered "male" and "female" work. In many areas of Latin America, for example, women are not involved in most aspects of agricultural work, especially plowing (in some areas, for women to touch a plow is considered bad luck). Farming is considered men's work. But in sub-Saharan Africa, "female" farming systems predominate, and women are involved in most aspects of agricultural production? In some African countries, women are entrepreneurs and actively participate in the public markets. In South Asia, however, market roles are considered men's domain, and women who participate in the public sphere out of economic necessity are not respected.
Differences Between Women and Men There is no single measure of gender inequality, nor of the relationship between gender inequality and population processes. Many demographers borrow methods and incorporate research from other disciplines to enrich their understanding of these complex relationships. Our knowledge of social processes in Egypt, for example, has been enhanced by the combination of several large-scale fertility surveys and anthropological studies of women's lives (see Box 2, page 8).
Most current research about gender, power, and population change, however, is based on surveys and administrative data that contain a number of indicators that appear to reflect women's relative status-and power-within a community. The research has focused on education, marriage patterns, employment outside the home, and participation in political systems.3 These variables cannot completely describe gender roles in a society, but they can help us gauge the breadth of gender inequality in countries and regions. Because these same indicators show a strong association with fertility and mortality change, these comparisons can also teach us about the vital links between gender, power, and demographic change.
Education Education is one of the most important sources of opportunity in any society. The ability to read and write gives individuals access to a wide body of knowledge. Formal education and its related activities expose people to social life outside the family. School provides entry into other opportunities. Political office, social preeminence, and jobs with higher status and better pay are nearly always associated with higher levels of education.
In nearly all developing countries, women are more likely to be illiterate and to have completed fewer years of education than men. Although literacy rates have climbed in recent years, women represent a large proportion of the illiterate population in developing countries, especially in Africa.
Older women in many developing areas had limited opportunities for schooling. They are not likely ever to learn to read because most national efforts to increase literacy focus on younger women and school-age girls. But illiteracy was higher among women than men even among young adults in many countries in 1990 (see Table 1). Whereas 55 percent of men in Bangladesh ages 15 to 24 were illiterate, 73 percent of women in the same age group were. In China, illiteracy has decreased greatly across all population groups, but a large gap remains between women and men. Among those over age 25, 23 percent of men and 54 percent of women were illiterate; among Chinese 15 to 25 years old, 4 percent of men but 13 percent of women were illiterate.
Girls are still much less likely to be enrolled in school than boys in many less industrialized countries, which suggests that the gender gap in education will be repeated in the next generation. Girls' representation decreases as they move up the educational ladder, and often drops precipitously at the university level. In Mexico, for example, about as many girls as boys were enrolled at the secondary level, but only 75 women per 100 men were enrolled at the university level in 1990.
Girls are also more likely than boys to be chronically absent from school, or to drop out altogether, especially in less industrialized societies where parents rely on their children to care for younger siblings, perform farm work, or earn wages. In many societies, girls help with household tasks and child care from an early age.
In China, the number of girls kept out of school to work at home is reportedly increasing because of economic reforms. Chinese families may now supplement their meager incomes from state-run enterprises by selling surplus produce or working in a family business. Parents expect children, especially daughters, to help out, which often means leaving school. Girls living in low-income rural areas are especially likely to drop out of school. Nationally, 70 percent of school-age drop-outs are girls.4
Official statistics do not adequately reflect the lower school attendance of Chinese girls. Girls make up nearly half (47 percent) of children enrolled in primary school in China, but they also comprise about three-fourths of the children who are enrolled in school but seldom attend.
Parents in China, and in many other countries, are more willing to let a daughter than a son drop out of school because they see less value in a daughter's education. A son's education, in contrast, may be viewed as an investment that will bring his parents a higher income in the future, and will ensure them a more secure old age. Parents are willing to sacrifice the current income a son could provide so he can continue his education and enhance his employability. They are more likely to keep a young daughter out of school and expect her to contribute to the family income.
Marriage practices can also devalue the investment in a daughter's education, and affect parents' decisions about their children's schooling. In some Asian societies, daughters move away from their parents' community at marriage and transfer their allegiance to their in-laws. Parents essentially lose their investment in their daughter's education.
Girls and boys often have very different education experiences, beginning with when (or whether) they enter school and extending to the kinds of opportunities that schooling provides. Girls are given less attention inside and outside the classroom, and often are directed toward fields of study that limit their career choices. Girls are less likely to receive family support to stay or excel in school-for example, through release from household chores or provision of tutors. When they enter the labor force, women gain a smaller return on their educational achievements than do men.5
The educational differences between females and males are among the most significant indicators of gender inequality. Recent international programs to raise women's literacy and educational levels are important steps toward reducing the gender equality gap. However, programs need to go beyond ensuring that girls stay in school as long as boys and also must address other aspects of educational differentials. They will need to focus on giving girls and women equal opportunities within school, and equal access to the rewards of educational achievement.
Work In most societies, women's working lives are quite different from those of men. Women often put in more hours of work than men, but many of these hours are unpaid or are not counted in official labor statistics. In most lowincome countries, women make up the majority of unpaid family workers, while men make up the majority of the self-employed and other paid employees (see Table 2). These differences reflect both the lower status of women and the difficulty that women have in using work to increase their access to societal resources.
Even when women do work for pay, their compensation and status are likely to be lower than men's. Women's wages for nonagricultural jobs are less than three-fourths those of men in many industrialized countries, including the United Kingdom and Canada, as well as in many less industrialized countries, such as Thailand, Syria, Argentina, and Bolivia (see Figure 1). In Bangladesh, women earn less than half as much as men in nonagricultural jobs.
In most societies, sex segregation by occupation and gender inequality within the workplace are the core reasons why women's jobs often have lower pay and status. Women are less likely than men to hold higher-paying professional, technical, and managerial jobs in most countries. Women usually have a much higher representation in clerical, sales, and service positions. They outnumbered men in clerical jobs in Mexico and the Philippines, for example, even though a relatively small percentage of women are in the labor force in these countries. In Brazil, more than twice as many service jobs were filled by women as by men, while only 18 women per 100 men worked in production or craft jobs.
Women hold more professional jobs than do men in Brazil and the Philippines-but these are exceptions. In Cameroon, for example, only 32 women per 100 men were professionals. In contrast, women's representation is fairly high in agricultural jobs in Asia and sub-Saharan African countries and it exceeded that of men in Bangladesh (see Table 3).
Women are less likely than men to be promoted to jobs with higher status and pay, even if their qualifications are similar.fi Part of this inequity is accounted for by the responsibilities that women have at home. Family and home commitments often limit the time and energy women can devote to work, which makes them appear less committed or competent than men, who can spend more time at their jobs. But in many cases, women's lower positions reflect simple gender discrimination. Employers are reluctant to promote women to managerial jobs because they worry that women's household responsibilities might interfere with their work performance, or simply because they believe women are not as good as men in leadership or administrative positions.
Employment is an important source of social contacts outside the family. These contacts enrich women's lives and can expose them to new ideas and knowledge. However, women receive fewer rewards from employment than men, and work does not always enhance women's status.
Family Roles Women's strong attachment to family and household responsibilities means that they have fewer opportunities to assume public roles in the society. Nearly all women in all societies marry and have children. They must fit any schooling, jobs, or political activity around those family roles. However, women usually have less power than men even within the family.
Women usually begin married life at a younger age than men. The average age at first marriage in the United States in 1994 was 24.5 for women and 26.7 for men, for example. The gap tends to be much wider in many less industrialized countries, especially in Asia and Africa. In Niger, Africa, the average marriage age is 16 for women and 24 for men. In Bangladesh, it is 18 for women and 26 for men.7
Women's younger age at marriage and the age differences between spouses help perpetuate their weaker authority within marriage and family. The younger her age at marriage, the less likely a woman will have had an opportunity to develop a career, to create support networks beyond her family, or to complete higher levels of schooling. In many societies, childbearing starts soon after marriage, which further impedes women's ability to develop resource or support networks outside the household.
Dowry and brideprice-the payments that brides carry with them or receive at the time of marriagecan hinder their acceptance in their new families. In parts of India, brides are treated poorly by their husbands' families if the dowry payment was considered too low.
In many societies, women have fewer inheritance rights, which can also compromise their position in the family. Women are less likely to bring land or wealth to their families through inheritance. If they do bring new property into the marriage, they sometimes lose ownership and control to their husbands.
Political Representation Across the world, women occupy a small fraction of positions with decisionmaking power. In South Korea, where women have made remarkable gains in educational achievement, women held only 1 percent of parliamentary seats, and only 4 percent of high-level appointed positions at the ministerial level in the early 1990s (see Table 4). Women occupied less than 10 percent of parliamentary seats in Brazil, Mexico, and many African countries.
Because gender roles accord women less power, women must overcome considerable discrimination to attain leadership positions. People commonly assume that women lack leadership potential or that they cannot hold a high-level job because of demanding family responsibilities.
The consequences of this underrepresentation are many. Women's exclusion from high-level decisionmaking keeps them from gaining leadership experience and perpetuates their lower status. And, although women share political interests with men in many areas, women's unique social roles and responsibilities are likely to give them alternative priorities for public policies and actions. When women's voices never reach the state, ministerial, or corporate decisionmakers, these agendas are less likely to deal with issues that specifically concern women. Women are more likely than men to be committed to ending gender inequality in the workplace, for example, or to policies that help them juggle home and work responsibilities.
Gender and Fertility and Mortality Change Gender and gender inequality are reflected in demographic processes to varying degrees, and in different ways. The strength of the gender/demographic change relationship usually depends upon the gap in power signified by male and female differences in education, employment, and other spheres. In many societies, the children women produce are an important source of power. But when women have alternative sources of power-from jobs, money, or political representation, for example-their dependence on children as a source of power decreases. Fertility is likely to decline.
How these processes are linked differs from society to society and can change over time. Researchers are still exploring the nature of these links, but existing evidence suggests that the way that gender is organized in a society may influence the timing and pace of fertility and mortality change. Certain elements of gender equality seem to be especially important in facilitating such change. In the case of mortality, a woman's position in the family or in the larger society will influence her access to health knowledge and technology that might influence her family's exposure to illness, or once ill, her use of health interventions.
Mortality The extraordinary increases in living standards and health care that occurred throughout much of the world in the past four decades brought greater improvements in life expectancy to women than to men. In the 1950s, women's life expectancy was three to four years longer than men's in places like Brazil, Mexico, South Korea, and the United States. In the 1990s, the female advantage was six to eight years in these countries. In most modern societies, in fact, women live longer than men.8
Some of the gender gap in life expectancy reflects a difference in gender roles: young adult men, for example, are more likely to engage in risky behavior than are young women. High death rates from accidents and violence in the young adult ages explain some of the lower male life expectancy. But the risk of mortality is greater for males than females at every age in most modern societies, which suggests an apparent female biological advantage.9 Medical research suggests that this advantage begins even before birth-women pregnant with male fetuses are more likely to miscarry.'o In some societies, howeverparticularly in South and East Asiavast discrepancies in gender roles cause higher mortality for girls and women. In Afghanistan, Bangladesh, India, Iran, Nepal, and Pakistan, among other countries, female life expectancy is about the same as that for males in the 1990s. In the 1950s, average life expectancy was lower for women than men in most of these countries. The female mortality rates in some of today's developing countries more likely reflect discrimination against women based on gender, not just unhealthy living conditions. In communities in which women occupy a much lower status than men, women may consistently receive less food than men, have less access to medical attention, and perform long hours of arduous labor. Part of the higher-than-expected mortality for women can be explained by the health risks associated with frequent pregnancies and births among women living in countries with high fertility rates. Women face a higher risk of death during pregnancy and childbirth if they are severely malnourished, live in unsanitary conditions, and lack access to basic maternity care. Unfortunately, pregnant women often face such circumstances in low-income countries.
The direct link between women's inequality and mortality is seen in the lower survival rates of young girls in some Asian countries, such as India, Korea, and China, where there is a strong preference for sons. In these societies, bearing a son gains parents greater status and prestige than bearing a daughter, and sons receive better care.
The strong son preference is dramatically illustrated in birth statistics from several Asian countries, which may reflect differential mortality. In every society, about 105 boys are born for every 100 girls. In some countries, and in some regions within countries, this sex ratio at birth is highly skewed toward males, which signals a problem either with the statistics or with the survival chances of female infants. The sex ratio was 112.5 in Korea, and 113.8 in China in 1990. In China alone, these skewed sex ratios suggest that 550,000 infant girls per year were unaccounted for during the 1980s.11
There are several possible explanations for the unnaturally high sex ratio in these Asian countries. First, many girls' births are never reported in national birth statistics. Parents may be less conscientious about reporting a girl's birth than a boy's because they consider the birth less significant. In more extreme cases, parents fail to report a girl's birth because they abandoned or even killed the infant. Second, some parents choose to abort female fetuses because they only want a son. Ultrasound and other prenatal tests that can reveal the sex of the fetus are becoming more widely available in Asia. Parents are using the tests to actualize their preference for male children (see Box 3, page 14).12
Gender differences can also mean that girls fare worse in the first few years of life. In the United States and most countries, girls' survival at ages one to five is similar to boys'. But in Singapore, Egypt, Pakistan, Guatemala, and the other countries shown in Table 5, death rates are higher for girls than boys. Here again, the higher mortality for girls reflects their gender roles, specifically the lower value and status of women relative to men.
In rural Bangladesh, females have higher death rates than males at all ages. This discrepancy can be traced to two differences in the treatment of Bangladeshi boys and girls. First, girls eat less protein and less food than boys, which contributes to their undernourishment. Tradition dictates that men and boys are served food first, and there often is insufficient food left for the female family members in poor rural households. Second, although girls and boys are equally likely to contract serious diseases, boys are taken to health clinics much more often than are girls. Parents make a greater effort to obtain health care for their sons than for their daughters, even when treatment is free.13
Gender inequality can also be manifested in how long mothers breastfeed their male and female babies. Breastfeeding in the first year of life ensures better nutrition and conveys the mother's natural defenses against some infectious diseases to her infant. In low-income households, longer periods of breastfeeding are associated with better infant health. In some developing countries, sons are breastfed longer than daughters, another reflection of the better parental care that many sons receive. This differential can translate into higher mortality for girls.14 Fertility
Reproduction is a central factor in all women's lives. It helps shape the organization of most social institutions, and gender in particular. Whether birth rates are high or low, and whether individual women have many children or none, women's roles in family and society are inextricably linked with childbearing. Although we do not fully understand how gender and fertility interact, we know their relationship extends far beyond the simple biological facts of pregnancy and childbirth into all aspects of society. A change in fertility levels may alter gender roles within a society. Alternatively, changes in gender roles might lead women to change the timing and number of children they have. The direction of influence is probably not linear or straightforward, and we may never understand the genderfertility relationship completely. However, it has significant implications for future population growth.
Fertility rates throughout the world have declined substantially in the last several decades. In the less industrialized regions, declines began in the 1960s and 1970s, and continue to the present. In the industrialized countries, fertility started to decline in the early 20th century, or even earlier, and slowly fell to the current low levels. In England and Wales, the total fertility rate (TFR), or average number of births per woman at the prevailing birth rates, fell by 1.4 percent annually over 25 years. The pace of fertility decline has been much more rapid in today's developing countries. In Costa Rica, the TFR fell by 2.9 percent annually between the 1960s and early 1980s.'5 In the 1990s, TFRs in developing countries range from around two children per woman in China and Korea to about six children per woman in Iran, Kenya, and Pakistan.16
Women's motherhood roles are likely to dominate their lives even where birth rates are low. But childbearing consumes much more time from women in countries with high fertility rates. If women have five or more children during their reproductive lives, a significant portion of their time will be spent caring for children. Women who have only two children, in contrast, are likely to finish their childrearing responsibilities earlier in their lives.
Explaining Differentials and Change
After decades of demographic research in developing countries, we have identified a variety of factors that appear to influence demographic change. Urbanization, the shift from extended to nuclear family patterns, the expansion of industrial and service economic sectors, better access to health services and information, and rising levels of education have all been associated with decreasing mortality and fertility.
Gender also plays a role. It both affects and is affected by these broader social and economic transformations. Women tend to have more power relative to men in societies with low rather than high mortality and fertility. Education and employment, for example, often accord women wider power and influence, which enhance their status. But attending school and working often compete with childbearing and childrearing. Women may choose to have fewer children in order to hold ajob or increase their educations.
One of the central issues involved in evaluating the effects of women's education or work on demographic variables is whether and how they give women power. The ways that education and work affect a woman's power to make decisions about health care, about contraception, or about the timing and number of children they have, will influence child mortality or fertility. In societies in which women's major source of power comes from their role as mothers, increased female education and employment may actually lessen women's power. In these situations, the income and opportunities a woman gains from employment might not compensate for the loss of power she suffers because she has fewer children.
Women's educational levels and labor force participation do not necessarily explain gender's effects on demographic outcomes (see Box 4). However, knowing what proportion of the female population is illiterate, gets a high school education, or is engaged in waged labor provides important insights into the kinds of lives that women lead. Because female education and work have been repeatedly associated with lower fertility and mortality in demographic surveys, these variables can help explain or predict demographic change. Education's Influence In nearly all societies, the amount of education a woman achieves affects the number of children she has and the way she cares for her children.
Fertility levels are usually lowest among the most highly educated women within a country. In Brazil, for example, women with no formal schooling have an average of 6.7 children while those with secondary or higher schooling have only 3.2 children on average. In Guatemala, where few women stay in school beyond the primary level, the TFR drops from 6.9 for women with no education to 5.4 for women who have attended a few years of school (see Table 6, page 19).
Even stronger than the relationship between women's education and fertility is the relationship between child health and mortality and women's education. Mothers with a primary or higher education witness fewer deaths among their children than women with less education. In Burkina Faso, the infant mortality rate was 111 deaths per 1,000 births for women with no education, compared with 90 infants deaths per 1,000 for women who had completed primary school, and 53 deaths per 1,000 for children of mothers with secondary or higher education (see Table 7). In Indonesia, the death rate for children under age five drops from 131 per 1,000 births for mothers with no formal education to 51 for the most educated mothers. In Peru, the infant mortality rate sinks from 100 to 34, and the under-five mortality rate from 150 to 45, as the mother's education increases from none to secondary or higher.
The mother's education also has an effect on a child's nutritional status. Children whose mothers completed secondary or higher education are much less likely to be short or underweight for their age than are children with less educated mothers, which indicates that they were more likely to receive adequate food (see Table 8, page 22).
Women's education influences fertility and child mortality at the societal as well as the individual level. Figure 2 illustrates education's effect on fertility in Botswana and Nigeria. Women in Botswana are relatively well educated by African standards. Just over half of the women have a primary or higher education. In sharp contrast, Nigerian women have extremely low levels of educational attainment. Over half of Nigerian women completed no formal education at all, and only one-third attained a primary or higher education.
In both countries, the fertility rate declines as women's educational levels increase, reflecting the usual relationship between the education and childbearing of individual women. In Botswana, the TFR was 4.9 according to the 1988 Demographic and Health Survey (DHS) in that country. The TFR drops from 5.9 children for women with no education to 3.4 children for women with a secondary or higher education. In Nigeria, the overall TFR was 6.0. It dropped from 6.5 for the large majority of women with no education to 4.2 for the small percentage with a secondary or higher education.
Contraceptive use increases with the educational level of women in both countries, but the percentage of women using contraception is below 28 percent for Nigerian women at the highest education levels, while it is 41 percent for women in Botswana who went beyond primary schooling.
A country's fertility level appears to be affected by the general level of education. A higher general level of education often means lower fertility and higher contraceptive use throughout the society.l7 This societal effect of general education levels may explain why women with no formal education have lower fertility and higher contraceptive use in Botswana than in Nigeria.
Pathways of influence Education's influence on demographic processes extends through many aspects of life, and is not limited to what women learn in the classroom.
Women's education can influence fertility by raising the age at marriage; by providing women with new job opportunities; by introducing women to new values or ideas; and by serving as a proxy, a marker of a group of characteristics, such as higher socioeconomic status, urban living, or others that are known to be associated with lower fertility. Education probably influences infant/child mortality through similar mechanisms.18
Raising the marriage age Education affects women's age at marriage, which has an important bearing on overall fertility. Women who are more highly educated marry (and have their first birth) later than those with less education, as Figure 3 indicates. In Brazil, for example, more than half of the women ages 15 to 49 with no formal education married before age 20; over 40 percent became mothers while still in their teens. In contrast, only 14 percent of women who stayed in school beyond the primary level married before age 20, and less than 10 percent became teenage mothers. Women who delay marriage and motherhood shorten the number of years that they are likely to bear children, and they usually have fewer children than women who begin marriage and childbearing sooner.
Part of this relationship is explained by the fact that women may postpone marriage until after they finish schooling. Where levels of schooling are high, education is likely to be an important factor in the timing of marriage. In Thailand, for example, where about a third of women were still in school at age 15, age at marriage is strongly associated with women's educational attainment. The average age at marriage for women with no education is just under 21, while it is over age 25 for those with a secondary or higher education." But the relationship between women's education and later marriage age is found even in societies in which women tend to leave school at much younger ages, which suggests that their schooling acts as a catalyst that delays marriage.
Two aspects of a woman's educational experience that might change her fertility patterns are the greater likelihood that she will work before marriage (which she might be reluctant to end in order to marry) and changes in ideas about marriage or marriage timing.
Changing attitudes In many societies, as little as one year of schooling affects fertility and, even more consistently, child health outcomes. In some countries, completing a few years of school is associated with higher fertility-Kenya and Indonesia are examples of this phenomenon, as shown on Table 6, page 19-however, these are exceptions to the general pattern.20
In most countries, including Burkina Faso, Colombia, and Indonesia, children of women who completed any primary school have lower mortality. Some studies suggest that women who attend school learn skills that help them notice and read health messages and make use of health services.2' When women attend school, they may learn about new ways to promote sanitation and health.
Other research results suggest that education is less about a structural change, such as increasing job opportunities or even literacy skills, than about a change in attitudes, values, or goals. Women might change the number of children they see as desirable because of school experiences. For many women, attending school might be a catalyst that changes their attitudes about how to handle health problems or how many children they should have. Women who attend school are exposed to and learn about environments and institutions outside their families, and develop the confidence to interact effectively with them.
Demographer John Caldwell's research in Nigeria found that the influence of mother's education exceeded that of health care availability or fathers' education, and several other important variables.22 Caldwell and other researchers have found that educated women, even if "educated" is defined as one or two years of school, are more capable of maneuvering in the modern world than are less educated women. Although health care might be available to all, mothers with knowledge about the modern world are more likely to use such services, to be assertive in that use, and to see health care as a right rather than a gift.
Caldwell also argued that the education of women might change intrafamily relations in profound and subtle ways. Other family members may equate a woman's education and her exposure to these modern institutions with superior knowledge on her part and defer to her decisions on whether to seek medical care. Caldwell's research suggests that education not only changes an individual woman's interactions with institutions, but also changes the way others look at her and what they expect her to do. Children of educated mothers are healthier because their mothers are more likely to obtain better health care for their children, even if it means challenging the authority of mothers-in-law, husbands, or medical personnel. When women obtain an education, they are likely to use their new roles to protect their children.
Networks One pathway through which education might influence fertility or health-seeking behavior is informal or formal networks.23 People everywhere talk to one another about the latest events or the goings-on in their families. These interpersonal links, however, may also be seen as a pathway on which information about and acceptability of a new method of controlling pregnancy might travel. Women are particularly likely to discuss issues related to reproduction with their friends and relatives. Such networks-whether in a local church, around a village well, or at a lunch break in a factory-can reinforce existing norms and behavior, or they can raise questions about them. For example, in Taiwan, village women form informal but powerful groups that arbitrate many aspects of women's lives, from marriage decisions to attitudes about pregnancy or children. Women use these groups to mediate a difficult situation at home or to get advice about a difficult child or mother-in-law.24
Women often use networks to test out community attitudes about the appropriateness or potential sanctioning of behavior. Thus networks can discourage as well as encourage new attitudes and behavior. A study of a small town in Mexico inhabited mostly by an Indian population revealed some of the inhibiting effects of social networks.25 Although the women there express a desire to have fewer children, they do not use contraception for fear of social sanction. That disapproval would come from men in the town who believe that a large population is needed in order to give their indigenous group strength against the government. In the face of such intense social disapproval, which is delivered through networks in both men's and women's communities, women do not dare try to limit their family size.
Education may play a vital role in these networks because through education, a woman learns about new ideas. Her education has the potential to influence not only her own behavior but others' as well-she may act as a source of new ideas to both formal and informal women's groups. The more that women are integrated into a variety of venues and roles in the society, the more diverse their networks are likely to be, and the more likely it is that new ideas about fertility or health behavior will be widely spread.26
Education as a proxy Lastly, schooling may be a proxy for other important aspects of women's lives. This issue is especially important in child health outcomes, but is relevant to fertility as well. Women who go to school may be different from other women in important ways, especially in societies in which women's school attendance is rare. These women may come from families or themselves have characteristics that would push them toward both education on one hand and lower fertility and better child health on the other. Economics plays a role. Women from wealthier families are more able to afford schooling, contraceptives, doctors' visits, or medicines than women from poor families. Researchers have found it especially hard to separate the effects of income from those of education in explaining differences in child health.
Individual attitudes and personality are also important. Women who attend school may be more likely than other women to hold a set of "modern" attitudes that favor smaller families, assertive use of public institutions, and faster interventions in child illnesses.27
The acknowledged link between women's education and fertility and mortality outcomes has led many countries to incorporate education into population policies. The Programme for Action signed at the 1994 ICPD called for increasing women's literacy and education around the world. Not only will education give women the tools they need to improve many aspects of their lives, policy analysts argued, it will also lead to lower fertility and healthier children.
Women's Work The relationship between women's work outside the home and demographic processes has been a common theme in demographic research. But the research has yielded less conclusive results than those in education, and it has revealed several surprises. The effect of women's employment on fertility, health, and mortality is not the same from country to country or from one period of time to another. While employment might signal increased resources and status for women in one context, it may be synonymous with poverty and arduous physical labor in another. The key to whether female employment affects demographic change lies in whether work translates into increased power for women, power that could influence demographic outcomes by enabling women to make informed decisions about health care, contraception, or the timing and number of children they bear.
In most countries, women who work for cash have fewer children than those who do not work for cash. This relationship conforms to the theory that paid employment increases the "opportunity costs" of having children, and thus encourages women to limit the number of children they have. "Opportunity costs" in this instance include the potential wages foregone when women leave (or refrain from entering) the labor force to have and care for children.
Education plays an important role in the interaction between fertility and labor force participation because schooling theoretically increases a woman's work opportunities. Thus it augments the opportunity costs of having children. In most, but not all, societies, women who are educated are more likely to be economically active than women with lower levels of education. The general fertility decline that usually occurs as countries develop economically, especially when this includes expanded education and job opportunities for women, appears to support this theory. But the relationship between employment and fertility is not straightforward.
Within the same country, the fertility differences between women who do and do not work can be quite small. In Ghana, for example, women working for cash average only 0.7 fewer children than those not working for cash (see Figure 4). But in other (and most) societies, the difference is substantial. In Brazil, for example, women who work for pay average nearly two fewer children than those who do not. Women's employment has the opposite effect on fertility levels in some countries. The DHS in Mali showed no fertility difference by whether or not a mother worked. In Nigeria, women who work for cash had higher fertility than those who do not.
Women's labor force participation does not have a consistent effect on child health either (see Table 9). In Uganda and Guatemala, children of employed mothers are less likely to have stunted physical growth from severe malnourishment, but in Sri Lanka the opposite appears to be true. There is also no uniform relationship between mother's employment and the percentage of children who are underweight for their age, another measure of malnourishment.
Even within a country, mother's paid employment does not have a consistent effect on child health. In Bolivia, for example, children whose mothers work are more likely than children whose mothers do not work to have received all three DPT immunizations (against diphtheria, pertussis, and tetanus). But both sets of children are about as likely to have been immunized against measles or to have had an episode of diarrhea in the two weeks preceding the survey (see Table 10).
The social context of women's employment, the difficulty in obtaining comparable measures of employment, and mediating factors, such as gender and the type of economy, make it difficult to generalize about the relationship between mother's employment, fertility, and mortality.
Measuring Women's Work The measurement of women's labor force participation has always been problematic. The methods commonly used to collect information on women's working lives are inadequate and sometimes misleading. Whereas estimates of male economic activity are fairly consistent and rates are fairly similar across societies, women's economic activity rates vary widely from country to country, as Table 2 shows (page 10). According to official statistics in China, for example, the vast majority of women work, while in Egypt, only 22 percent of women are reported to be employed.
Estimates of women's economic activity depend largely on how the data are collected and what definition of employment is used.29 Many labor market surveys collect work information by asking about a person's "main" activity. Women often juggle two or more roles, which often include economic activities, but they may list childrearing and housekeeping as their main activity because these are the most important socially. Defining women's participation in economic activity is especially difficult in rural agricultural economies. The following observation about women's work and family roles in Africa rings true for many societies:
"Given the pervasiveness of family-based economic enterprises in the region, the assumption cannot be made that occupational roles can be studied meaningfully in isolation from familial roles. It is not simply that all workers have family responsibilities but that a large proportion of productive resources, including land and labor, are still in the control of kin. This is particularly so for women, because they remain more enmeshed in traditional forms of economic activity, food production, processing and distribution, which typically occur on family farms and in the domestic domain. 30
Accurate measurement of women's labor force participation is also plagued by the heavy representation of women in informal economic sectors and in seasonal and temporary jobs, all of which tend to be undercounted in surveys. Because of time-consuming responsibilities at home, lower educational levels, and cultural barriers that limit job opportunities, women's employment is often intermittent.
The Social Context The difficulty of measuring whether women work outside the home is linked to the meaning and value of women's employment in a particular society. In a society where women are expected to work, work may reflect positively on women's selfworth and signal that women have opportunities in an important sphere of society. In another setting, women's labor force participation may be a sign of lower status or her husband's failure to earn enough to allow her to stay home and attend to family. These factors affect the reliability of the measurements of work and the effects that work has on demographic variables. Among some societies, for example in India and in upper-income social classes in Mexico, families demonstrate their affluence and membership in a privileged socioeconomic group by not having wives and daughters in the labor force." In these societies, the poorest women, those in economic distress, work for pay. Employment will not necessarily improve their status.
In general, however, employment outside the home operates much like education in that it exposes women to new ideas, new norms of behavior, and the influences of people outside the family. Women might learn about contraceptive methods from coworkers, for example. Earned income might award them a larger say in family decisions. Women's attitudes about their own roles and relative power may change when they work, so that they see their position in the family differently. Other family members may also accept a changed role for such women. Because of interactions with people outside the family and a greater sense of power within it, women may feel they should have a larger role in decisions about how many children they should have, how to treat a sick child, whether and what kind contraception to use, or whether to move to another area.
In many areas of China, women are pressured by their in-laws to bear several children. But Chinese women who work in successful small businesses are able to negotiate down the number of children they are expected to have. They fulfill their obligations to their extended families through substantial economic contributions rather than through the traditional route of childbearing. Their economic power gives them unusual leverage to go against kinship rules. Chinese women who have decisionmaking power in their businesses usually have similar power in childbearing and childrearing as well.32 Whether employment brings women more power depends on a variety of factors, including the kind of work involved, relative wage levels, social norms about the acceptance of women working, and the way that work transforms women's home and social lives. The effect of women's employment on fertility and child health depends partly on how much power work brings, but the effect is also mediated by the social, economic, and political environment in which a woman lives. These conditioning factors include such diverse issues as the availability of adequate, affordable child care; the economic health of the society; gender segregation within the labor force; and the distribution of power, resources, and responsibilities within the household.
Family Responsibilities Women are usually responsible for most of the housework and child care, even when they are employed full-time outside the family. This "double day," as it has been termed in many societies, means that women often work more hours than men. In industrialized and nonindustrialized countries, women contribute a larger share of total work hours than men, and women are less likely to be paid for their efforts (see Figure 5).
The setting in which a woman works and the organization of her home life will condition the decisions she makes about fertility and determine whether she will be able to implement these decisions. If a woman can find adequate and affordable child care, for example,33 she may be able to balance her (and her family's) desired number of children with work outside the family.
Women's employment is likely to encourage lower fertility where it is difficult or impossible to combine work and family responsibilities. Employment in professional and clerical jobs has the strongest negative association with fertility in the most industrialized countries, where childrearing and work roles are in greatest conflict.
When women can combine childbearing and work, such as in rural agricultural economies, or where parent substitutes are more available, work is not highly correlated with lower fertility.4 In many less industrialized countries, mothers rely on their older children and other family members to care for young children while they work at a paid job-or they care for their own children while they work. In Niger, for example, 74 percent of employed women with children under age five cared for their own children while they worked and another 11 percent relied on older children (see Figure 6). Relatively few women in these countries use formal child-care centers or schools to care for children.
Social Class Even within a society, the context of work is likely to differ from group to group. Class status is one of the most important mediators of the effects of women's work on fertility and mortality outcomes (see Box 5, page 28). In their study of poor Indian households, Sonalde Desai and Devaki Jain found that part of the negative relationship between women's paid work and child health is actually a class difference. Women in poorer families are more likely to work outside the family than are women from wealthier households. Thus poorer women are more likely to work, but less able to provide their children adequate food or medical intervention. Desai and Jain found that children of poor mothers, whether or not the mothers were employed, were less likely to be immunized and more likely to have suffered stunted growth. Extreme poverty, not employment, was central to child health in these communities.35 Gender and Power
To explain why women's education and work do not always influence fertility or mortality, we need to return to our definition of gender. Two aspects of gender are especially important: gender as representative of power differences, and gender as a social construct that varies from society to society. Investigating women's education and work through these lenses underscores just how important power and social construction are to understanding gender and the role of gender in demographic processes (see Box 6, page 32).
One important reflection of power is a woman's control over the money that she or other family members earn. Control over household money appears to give women more say over family decisions related to that money, and grants them a more central role in other decisions about reproduction and contraceptive use, among other issues. Control over money is especially empowering for women if they have control over more than just subsistence spending."
In a study of families in Mexico in which women worked and where the household members pooled their income, there was a significant relationship between the proportion of pooled income contributed by the wife and her leverage in fertility decisions: Where women's wages made up over 40 percent of the family's total income, women reported more say in decisions about contraceptive use or family size than did women who contributed a smaller proportion to their family's income.37 The power that work can bring to women affects child health as well. In many situations, when women work, they are not able to spend as much time with their children. In poor households, which may not be able to afford adequate child care, such a situation might result in the neglect of children." However, in some situations, negative effects of mothers' work seem to be at least partly balanced by the positive contributions that mothers' income can make. In fact, many studies have shown that relative to men, women contribute a higher proportion of their wages to household needs. Men are more likely to use a larger proportion of their income on such things as alcohol, tea, or cigarettes. A study of several villages in South India found that women always contributed a higher proportion (and often all) of their income to the household than did men."
Such findings suggest why women's paid employment might be associated with better nutrition and medical care, and consequently lower mortality, for children. In Nicaragua, toddler-age children of working mothers tended to be healthier (according to measures such as weight-for-height) than those whose mothers did not work. However, more refined data from Guatemala revealed that what seemed to be important in children's health was not whether a mother worked, but what proportion of the household income she earned; the higher the proportion of family income earned by the mother, the healthier the children.40
Economic and political change affects gender roles and determines whether employment empowers women and generates fertility decline. For example, women's productive roles can be devalued in the shift from an agricultural- to industrial-based economy. A study of a rural fishing village in Portugal found that women's lives were noticeably worsened with the introduction of factory work into the villages in the 1970s. Even though poor, women were a vital, even powerful, part of the fishing industry before industrialization. They had control over their hours, their resources, and their children; and they had close ties with other women in the village. After factory work became dominant and production moved out of the home, women lost many of those aspects of their lives they valued most. They had little control over their work hours, they had trouble balancing work and home responsibilities, they became more dependent on a cash economy, and their work was valued less because they received low wages. These women became dependent on outside sources of income and, consequently, lost much of their autonomy. They began to see their role as mothers and housewives as their main source of status. 41
In the Margoli area of Kenya, fertility declined at the same time as women gained in political representation, education, and employment. But researchers found that what appeared to be a clear relationship between improvements in vital aspects of women's lives and fertility decline was spurious. They pointed out that a sharp increase in violence against women occurred at about the same time. In Margoli, widespread poverty appeared to be at the root of both the motivation for fewer children and the increase in violence against women. Expanded educational and work opportunities did not bring women additional power in this situation. Rather, they placed women in direct competition with men for scarce jobs, and apparently generated a resentment that occasionally erupted into violence.42
For some women, work itself carries a great risk of violence. Thousands of Burmese women working in the sex trade in Thailand were recruited and transported there against their will, and constantly face the threat of physical and sexual abuse. They represent the many women across the world who are forced by others or by circumstances to move to or work in places where they lack the power to protect themselves against maltreatment (see Box 7, page 34).
Thus women's employment does not always reflect an increase in power. But in those societies where employment offers women increased opportunities that give them greater power over and access to key social resources, we can expect opportunities for women to increase and gender inequalities to decrease. As women turn their attention to tasks other than childrearing, and as they contribute to the better health of their family members, fertility and mortality are likely to decline.
Gender in Population Policy The role of gender in the development, implementation, and evaluation of population-related programs in the developing world has undergone significant change in recent years. After demographers and international development specialists began to document the high fertility and rapid growth of developing-country populations in the 1950s, programmatic efforts focused on bringing down fertility rates as quickly and efficiently as possible. Many in the international family planning field argued that good family planning programs could overcome the upward pressure that poverty or lack of industrial development might exert on fertility and population growth rates. Family planning programs in developing countries began in the 1950s, under support from the International Planned Parenthood Federation (IPPF) and the Population Council. With national and international support, programs were introduced in many developing areas during the 1960s, and they gained additional strength and numbers in the ensuing decades. In the 1990s, 95 percent of the world's population lives in countries with government or private support for family planning.43 Although many governments believe that family planning programs are a necessary source of health and other benefits for women and children, the principal motivation for most of these programs is to lower fertility and slow population growth.
There is no doubt that family planning programs have been instrumental in lowering fertility in many countries throughout the world. According to one estimate, these programs accounted for 43 percent of the fertility reduction in less industrialized countries between 1960 and 1990.44 But even with widespread recognition of the success of family planning programs in lowering fertility, there has been disagreement over whether these programs should be the central way that governments or agencies attempt to effect demographic change. For example, at the 1974 UN World Population Conference in Bucharest, many less industrialized countries argued that "development is the best contraceptive," and supported reducing population growth indirectly, through efforts in social and economic development. By the time of the next International Population Conference in Mexico City in 1984, however, many of these same countries believed that family planning programs were necessary for reducing fertility. However, at the 1994 ICPD and the 1995 UN Conference on Women in Beijing, policymakers from all over the world agreed that both family planning and social and economic development were needed for countries to achieve sustainable rates of population growth, avoid damage to the environment, and enhance human rights.45
Addressing Cultural and Gender Influences Family planning program evaluators have found that a variety of factors influence a program's success, such as media outreach for family planning and well-developed administrative and political support, especially from the government. In addition, many argue that goals to lower fertility rates at the societal level need to be complemented by a commitment to meet various needs of individuals who are seeking family planning services.46 Family planning workers have long noted that programs need to be tailored to the social and cultural environment of the population they serve.
Family planning program administrators have also begun to look more critically at the quality of care they provide. In a key article on this topic, Judith Bruce argues that standards of quality of care can be instituted and evaluated even as family planning programs try to balance demographic objectives with individuals' needs.47 She favors assessment methods that use standards of program evaluation other than the number of contraceptive users recruited or the decline in fertility. If evaluators assess the knowledge, health, and satisfaction of clients, for example, they would learn about elements of program quality such as interpersonal relations between program workers and clients, the kinds of services and choice of methods offered, and the mechanisms of follow-up. Attention to and evaluation of these program outcomes will encourage programs to be closely tied to their social, political, and economic surroundings.
Gender, too, is seen as an important component of family planning programs (see Box 8, page 38). Programs have been redesigned to address the cultural norms and values of a specific community. The injectable contraceptive Depo-Provera, for example, is avoided by women in many cultures because of the bleeding it can cause.48 In a village in the Gambia, however, women reported liking this method because it allowed them to keep their contraceptive use secret from their husbands, who would have disapproved of such family planning.49
In Tunisia, the national family planning program implemented a program to improve postpartum health services.50 New mothers and their babies are invited to visit the hospital on the 40th day after birth. In many Muslim cultures like Tunisia's, the 40th day marks the end of a postpartum recovery period in which women's needs are carefully provided for by family and friends. In this way, the program successfully incorporated an important day in Islam-one that is easily remembered by women and their families. This approach has resulted in better provision of postpartum care for mothers and their new babies.
The development of culture- and gender-specific programs does not preclude a focus on fertility reduction. Indeed, one of the arguments for addressing these issues of women's needs or cultural differences is the belief that such attention would make programs more efficient and that more women (and perhaps men) would use contraceptives and fertility would fall.5' Widening the Focus
Revamping family planning programs has led to a broader definition of reproductive health and, more recently, discussions about whether family planning and contraception use should be at the center of women's health at all.52 Many policy analysts have begun to call for new health program goals. They argue that concern for the global effects of high population growth is not useful for guiding the development of programs and that improving human welfare should also be part of a solution to the "population problem."
Proponents of this new, broader focus point out that "population resources have been invested primarily in family planning programs, rather than in creating the conditions that facilitate people's use of those programs."53 They argue that because many social, political, and economic factors influence women's (and men's) desire for children and their ability to achieve their reproductive goals, population programs that seek solutions beyond simple family planning programs are more likely to be effective. This new approach would require substantial adjustments in population policies. Family planning might be subsumed under "reproductive health" programs, and those in turn might be placed under broader health programs. The goals of these programs would involve reducing mortality and morbidity as much as lowering fertility.54 In many (although not all) of the discussions of these kinds of programs, the underlying argument is that by expanding the focus and being more inclusive, these programs will not only be better for individual women, but will also encourage and strengthen contraceptive use and fertility decline.55 Examples from Bangladesh Some other new directions are suggested by Bangladesh's family planning program. In recent years, Bangladesh has been held up as an example of successful fertility intervention. Still a poor country by any standard-with low levels of education for women and high mortality-Bangladesh has nonetheless seen its fertility rate decline from an average of 7 births per woman in the 1970s to 3.4 births per woman in the mid-1990s. Much of this decline is attributed to the rapid increase in contraceptive use, which rose from 3 to 45 percent of married women of reproductive age over this period.
Bangladesh's extensive and wellplanned family planning program is credited with much of the success in lowering fertility and increasing contraceptive use. That program, which began on a large scale in the late 1960s and early 1970s, has become increasingly culture and gender sensitive. Since the late 1970s, family planning workers have made home deliveries to address the difficulty women had in getting to clinics because of norms of female seclusion.56 By accommodating many aspects of Bangladeshi society, the program was both more successful and less intrusive than more conventional programs.
Some analysts argue, however, that family planning programs can do more for women. While they recognize the contributions of the Bangladeshi program to women's awareness of contraception and ability to realize their reproductive goals, these analysts question the very aspects of the program that probably contributed to its success. Because the program works within the cultural and social boundaries of the society, it may strengthen, and certainly does not challenge, women's inequality. Some analysts claim, for example: "In its intensive focus on family planning services for women, however, the program fails to disturb and may even reinforce the patriarchal structures that keep women isolated and vulnerable... Contraception cannot solve the larger problem of women's subordination, which we believe should be addressed more directly. 57
These analysts suggest that programs that reduce women's economic dependence on men and draw women into the public arena could significantly challenge gender inequality in Bangladesh. Such programs include credit programs to aid women in business and employment programs to help women enter market-based jobs. Bangladesh again provides a useful example of programs that have successfully increased women's empowerment.58 Programs underwritten by two organizations, the Grammeen Bank and the Bangladesh Rural Advancement Committee have given poor women new economic opportunities by providing financial credit and support for consciousnessraising groups that encourage women to adopt economic roles. Although neither organization provides family planning services, both have had an effect on women's contraceptive use. Usage is notably higher among women involved in the programs and in the areas in which these programs operate. Program administrators argue that these programs are effective because they give women opportunities to earn cash incomes, allow them more interactions in the public sphere, and increase their political and legal awareness. These changes can increase wives' bargaining positions within their households and make it more likely that they will play a role in reproductive decisions.
International Attention Many participants at the 1994 ICPD argued forcefully for widening the scope of population programs. They called for a complete change in development and population agendas, contending that programs with an exclusive focus on fertility decline can harm the very women they are supposed to serve.59 These activists were not necessarily calling for the abolition of family planning programs. Many women's groups in developing countries have supported increased access to contraceptives and expanding family planning services.
In 1987, for example, Philippine reproductive rights activists were instrumental in blocking a proposed executive order that would have banned most contraceptives in their country60Many women's rights activists around the world see reproductive freedom-the right to make and carry through on one's reproductive choices-as fundamental to women achieving equality." But many women's groups now call for broadening the focus of development programs. Women's situations, these activists argue, are more likely to improve if women's lives and well-being, rather than fertility reduction, are the main focus of development and health programs. The ICPD Programme of Action states, for example, that governments should address issues such as women's rights to property ownership, gender discrimination in employment and pay, and violence against women.62 Ensuring women the means to achieve their desired reproductive goals, whatever those might be, would be only one plank among many in the program, as suggested below. "...the population movement needs to delink family planning programs from the societal fertility reduction objective. This linkage has adversely affected the quality of contraceptive services and, in some circumstances, has led to dependence upon undesirable means to induce fertility decline, including community-wide incentives, quotas, and coercion. Instead, the design of family planning programs should be guided by the individual's right to regulate her or his fertility safely and effectively.63
Many organizations and governments are already developing policies that could have widespread effects on women's lives. Foremost among these are laws that give women equal legal rights, which can further women's positions in other aspects of the society. A number of countries currently have laws that discriminate against women in the right to own or manage property and in the right to use contraception without their husband's knowledge.64
Other policies are aimed at more specific aspects of women's lives. There has been increasing interest in making it easier for, or even requiring, girls to enter and stay in school. Policies to increase women's education have particularly broad benefits to individual women and their children, to the local community, and ultimately to the country.
In China, the push for women's education has made delaying marriage a cornerstone of both family and marriage reforms and family planning policies. Later marriage for girls means that they have more opportunity for schooling or job training, that they will bring skills into their married life, and that they will begin child bearing at a later age. The Chinese government has argued that all these outcomes benefit both the individual young women and the larger social community.
Some societies are also working to develop support programs for working parents, in the form of programs like maternity leave or child care provision (see Table 11, page 41). China provides very good maternity leave coverage by giving women time off with full pay. Other countries, such as Kenya, offer new mothers much less help. Programs that provide substantial support make it easier for mothers to combine home and work responsibilities, and ultimately benefit their children.
In the policy arena, women's health programs have been under close scrutiny. Population and health programs have moved away from a simple focus on family planning and fertility reduction. Programs have become more sensitive to their cultural environment. In the process, they have become more efficient, and contributed in a positive way to women's lives. Most recently, policymakers and analysts have begun to consider how widening their approach to health and development programs, and lessening the emphasis on family planning, can promote gender equality in a society more effectively. More reproductive health programs are being designed like the one in Bangladesh, whose aims are women's empowerment and gender equality as much as reducing birth rates.
Looking Toward the Future The international population conference in Cairo and the women's conference in Beijing highlighted the importance of the relationship between gender and population issues. Programs and policies to address the issues women face must deal with the central role that reproduction plays in women's lives. Programs to increase literacy and school enrollment among girls and women, for example, must account for the ways that childbearing and childrearing could interfere with women's opportunities-from seclusion of women, which makes school attendance logistically difficult, to the need for child care while attending classes.
The central role that gender plays in fertility and mortality is equally compelling. On a daily basis, in all aspects of their lives, women and men confront the ways that gender shapes their opportunities and outlooks. The number of children women and men desire, and decisions about whether to use contraception or to seek medical help for ill children all reflect gender's influence at the individual, household, and societal level.
Social scientists and policy analysts are paying increasing attention to the connections between gender and demographic processes. Work in this area is likely to go in at least two directions, guided by the different goals of researchers and policymakers. One explores how social institutions have been influenced by gender and how that affects demographic processes. Another avenue of investigation considers how demographic processes affect social structures and individuals.
Both approaches recognize the centrality of gender to individual lives, social institutions, social organization, and policies. And, they acknowledge the need to employ a variety of approaches to capture the complexity and strength of gender in any society. Although women are obviously central to any study of fertility, developing an understanding of the role of gender requires looking beyond women's lives and their activities.
Research that does not focus on a particular demographic outcome but seeks to understand how gender is involved in all social processes is likely to raise new research questions and to challenge us to find innovative ways to examine gender and population change. For example, the way that women gain strength and power through childbirth and motherhood in some societies but not others suggests that assessing the impact of fertility decline on women's status will require a deep understanding of the particulars of a given society. In this case, a first research question must be directed toward understanding the extent and sources of women's (and men's) power in a particular society. In some places, economic power might be instrumental in expanding women's status, while in another, that kind of economic power might reduce the power and status derived from motherhood.
These questions are leading us to develop new research methods and strategies that enable as full an understanding of women and gender as possible. In addition to the extensive survey data that demographers collect in countries around the world, we are increasingly relying on smallscale, intensive projects that examine a society or village in depth to develop an understanding of the complex processes that underlie any social phenomenon. Using a variety of approaches and borrowing methods, understanding, and theory from other disciplines will inevitably lead us to new research questions and agendas that can enhance our knowledge of how gender interacts with fertility, mortality, and migration.
For those developing population policy, widening the lens also promises more effective strategies for improving women's lives or developing programs to address women's reproductive or health needs. As researchers and policymakers discuss ways to implement the programs of action from the Cairo and Beijing conferences, discussions will inevitably lead to the questions still unanswered, and sometimes not yet asked. For example, now that it is clearer that increasing women's labor force participation does not necessarily enhance women's position or power in a society, what should programmatic goals be in this area? Should we develop new ways to address women's subordinate status? Are there new ways to increase women's economic activity that might ensure changes in their status?
Programs aimed more directly at demographic outcomes are also undergoing scrutiny. The Cairo and Beijing conferences brought attention and legitimacy to claims that reproductive health programs can perpetuate women's lower status at the same time that they serve women's health needs. We now see that the attempt to make Bangladesh's family planning program attentive to women's roles in that society reinforced female subservience and oppression because it did not challenge existing gender inequalities. Can reproductive health programs promote women's equality with men and still maintain a society's values? These and other questions remain for those striving to develop programs that are both effective and culturally appropriate-and that enhance women's lives and health.
These are huge challenges. That gender has been getting increased attention from a variety of constituencies is encouraging. This attention is likely to result in new strategies that will augment our understanding of the role gender plays in demographic change and will enhance the development of policies and programs as well.
1. Identify some of the reasons why the child mortality rate for girls is lower than that of boys in some countries and higher in other countries. Compare the biological and gender factors responsible for mortality differentials among children.
2. Discuss some of the long-term social and demographic consequences of the strong son preference in India, China, and other countries. 3. Assess the role of educational attainment in determining the status of women. Why are women less likely to attain a secondary-level education in some countries?
4. How does educational attainment affect demographic change (fertility, mortality, migration)?
5. Critique the use of women's work status as the dependent variable in the analysis of fertility and child health. What is included in the definition of work? What are the drawbacks to this measure in various societies? Consider other factors that may contribute to the effect that work seems to have on demographic variables (for example, the type of work, number of hours worked, and social factors).
6. Does employment have a clear relationship to women's status? Explain with examples.
7. What are some of the ways that formal and informal networks can affect fertility, mortality, and migration?
8. Compare the ways that gender influences population processes in industrialized and nonindustrialized countries. Identify two ways in which the influences are similar, and two ways in which they differ. 9. Discuss the concept of "power" as it relates to gender and demographic change. What kinds of power are at issue at the family level? At the community level? At the national level?
10. How is an increase in women's power in a society likely to affect the men in the same society?
11. Some women's groups have called for broadening the focus of family planning programs to encompass a broad array of factors that might improve women's lives and well-being. How do you think this will affect fertility rates among the women served by more broad-based programs? Prepared by Kimberly A. Crews
1. United Nations Development Programme, Human Development Report 1995 (New York: Oxford University Press, 1995): 9.
2. Lynne Brydon and Sylvia Chant, Women in the Third World: Gender Issues in Rural and Urban
Areas (New Brunswick, NJ: Rutgers University Press, 1989). 3. Karen O. Mason, "The Status of Women: Conceptual and Methodological Issues in Demographic Studies," Sociological Forum 1, no. 2 (1986): 284-300. 4. Elizabeth Croll, Changing Attitudes of Chinese Women (London: Zed Books, 1995): 134. 5. Jill Ker Conway and Susan C. Bourque, "Introduction," in The Politics of Women's Education: Perspectives from Asia, Africa, and Latin America, eds. J.K. Conway and S.C. Bourque (Ann Arbor: University of Michigan Press, 1993): 1-11.
6. Harriet Bradley, Men's Work, Women's Work (Minneapolis: University of Minnesota Press,
7. UN, The World's Women 1995: Statistics and Trends (New York: United Nations, 1995). 8. UN, World Population Prospects: The 1996 Revision (New York: United Nations, 1996). 9. Xinhua Steve Ren, "Sex Differences in Infant and Child Mortality in Three Provinces of China," Social Sciences and Medicine 40, no. 9 (May 1995): 1259-69; and I. Waldron, "The Role of Genetic and Biological Factors in Sex Differences in Mortality, in Sex Differences in Mortality, eds. A.D. Lopez and L.T. Ruzicka (Canberra, Australia: Department of Demography, Australian National University, 1983).
10. The female advantage in mortality is high even before birth. In the United States, the sex ratio of fetal deaths was 113 males to 100 females in 1991. NCHS, Vital Statistics of the United States 1991: Vol lI, Mortality, Part A (Washington, DC: Public Health Service, 1996): table 3-10.
11. Chai Bin Park and Nam-Hoon Cho, "Consequences of Son Preference in a Low-Fertility Society: Imbalance of the Sex Ratio at Birth in Korea," Population and Development Review
21, no. 1 (March 1995): 59-84; and Zeng, Tu Ping, Gu Baochang, Xu Yi, Li Bohua, and Li Yongping, "Causes and Implications of the Recent Increase in the Reported Sex Ratio in China," Population and Development Review 19, no. 2 (1993): 283-302. 12. Zeng Yi, et al., "Causes and Implications of the Recent Increase in the Reported Sex Ratio in China"; and Sten Johansson and Ola Nygren, "The Missing Girls of China: A New Demographic Account," Population and Development Review 17, no. 1 (1991): 35-52. 13. Lincoln Chen, Emdadul Huq, and Stan D'Souza, "Sex Bias in the Family Allocation of Food and Health Care in Rural Bangladesh," Population and Development Review 7, no. 1 (March 1981): 55-70.
14. N.I. Sabir and GJ. Ebrahim, "Are Daughters More at Risk than Sons in Some Societies?" Journal of Tropical Pediatrics 30, no. 4 (Aug. 1984): 237-9. 15. Alberto Palloni, "Fertility and Mortality in Latin America," in World Population: Approaching the Year 2000, ed. S. Preston, The Annals of The American Academy of Political and Social Science, vol. 510 (Newbury Park, CA: Sage Publications, 1990): 126-44. 16. Carl Haub and Diana Cornelius, World Population Data Sheet 1997 (Washington, DC: Population Reference Bureau, 1997).
17. Mead T. Cain, "Patriarchal Structure and Demographic Change," in Women's Position and Demographic Change, eds. N. Federici, K.O. Mason, and S. Sogner (Oxford: Clarendon Press, 1993): 43-60.
18. John Cleland and C. Wilson, "Demand Theories of the Fertility Transition: An Iconoclastic View," Population Studies 41 (1987): 5-30; Helen Ware, Effects of Maternal Education, Women's Roles and Child Care on Child Mortality," in Child Survival: Strategies for Research, eds. W.H. Mosley and L. Chen, supplement to Population and Development Review 10 ( 1984): 191-214; and John Knodel and Gavin Jones, Population and Development Review 22, no. 4 (Dec. 1996): 683-702. Knodel and Jones argue that social class differences in access to education might have a more important influence on fertility than gender differences.
19. John Knodel, Apichat Chamratrithirong, and Nibhon Debavalya, Thailand's Reproductive Revolution: Rapid Fertility Decline in a Third World Setting (Madison, WI: University of Wisconsin Press, 1987): 75.
20. Some research suggests that women who enter school give up traditional practices, such as prolonged breastfeeding of their infants and a lengthy period of postpartum sexual abstinence that delay their next pregnancy without using a contraceptive method. Those women with just a few years of schooling are not as likely to use family planning methods, so fertility actually increases. Women with higher educations will have given up the traditional birthspacing practices, but will have replaced them with use of modern contraceptives.
21. Robert Levine, Emily Dexter, Patricia Velasco, Sarah Levine, Arun Joshi, Kathleen Stuebing, and Medardo Tapia-Uribe, "Maternal Literacy and Health Care in Three Countries: A Preliminary Report," Health Transition Reviev 4, no. 2 (1994): 18691.
22. John Caldwell, "Education as a Factor in Mortality Decline: An Examination of Nigerian Data," Population Studies 33 (Nov. 1979): 395-413.
23. Susan Watkins, "More Lessons from the Past: Women's Informal Networks and Fertility Decline, a Research Agenda" (Paper delivered at IUSSP Seminar on Fertility in SubSaharan Africa, Harare, Zimbabwe, Nov. 19-22, 1992).
24. Margery Wolf, Women and the Family in Rural Taiwan (Stanford, CA: Stanford University Press, 1972).
25. Carole Browner, "The Politics of Reproduction in a Mexican Village," Signs 11, no. 4 (1986): 710-24.
26. Watkins, "More Lessons from the Past." 27. Ware, "Effects of Maternal Education." 28. UN, The World's Women 1995: 34.
29. Richard Anker, "Measuring Women's Participation in the African Labour Force," in Gender, Work and Population in Sub-Saharan Sub-Soharen Africa, eds. A. Adepoju and C. Oppong (London: James Currey for International Labour Office, 1994): 68. 30. Christine Oppong, "Introduction," in Gender, Work, and Population in Sub-Saharan Africa, eds. A. Adepoju and C. Oppong (Geneva: James Currey for International Labour Office, 1994): 2.
31. Larissa A. Lomnitz and Marisol Perez-Lizaur, A Mexican Elite Family, 1820-1980 (Princeton, NJ: Princeton University Press, 1987); and Patricia Jeffrey, Frogs in a Well: Indian Women in Purdah (London: Zed Books, 1979).
32. Hill Gates, Cultural Support for Birth Limitation Among Urban Capital-Owning Women," in Chinese Families in the Post-Mao Era, eds. D. Davis and S. Harrell (Berkeley, CA: University of California Press, 1993): 251-74.
33. Guy Standing, "Women's Work Activity and Fertility," in Determinants of Fertility in Developing Countries 2, eds. R. Bulatao and R. Lee (New York: Academic Press, 1983): 54789; Christine Oppong, "Women's Roles, Opportunity Costs, and Fertility," in Determinants of Fertility in Developing Countries 2, eds. R. Bulatao and R. Lee (New York: Academic Press, 1983): 547-89; and A.I. Garey and N. Townsend, "Sharing the Costs of High Fertility: Support and Child Care in a Village in Botswana" (Paper delivered at the Annual Meeting
of the Population Association of America, Miami, FL, May 5-7, 1994). 34. Cynthia Lloyd, "Understanding the Relationship Between Women's Work and Fertility: The Contribution of the World Fertility Surveys," Population Council Working Papers, no. 9 (New York: 1990); Oppong, "Introduction."
35. Sonalde Desai and Devaki Jain, "Maternal Employment and Family Dynamics: The Social Context of Women's Work in Rural South India," Population and Development Reviev 20, no. 1 (March 1994):115-36.
36. Rae Lesser Blumberg, "Income Under Female Versus Male Control: Hypotheses from a Theory of Gender Stratification and Data from the Third World," in Gender, Family, and Economy: The Triple Overlap, ed. R.L. Blumberg (Newbury Park, CA: Sage Publications, 1991): 97-127; and Rae Lesser Blumberg and Marion Coleman, "A Theoretical Look at the Gender Balance of Power in the American Couple,"Journal of Family Issues 10, no. 2 (1989).
37. Martha Roldan, "Renegotiating the Marital Contract: Intrahousehold Patterns of Money Allocation and Women's Subordination Among Domestic Outworkers in Mexico City," In A Home Divided: Women and Income in the Third World, eds. D. Dwyer and J. Bruce (Palo Alto, CA: Stanford University Press, 1988): 229-47.
38. Alaka M. Basu and Kaushik Basu, "Women's Economic Roles and Child Survival: The Case of India," Health Transition Review 1, no. I (April 1991): 83-104. 39. Joan Mencher, "Women's Work and Poverty: Women's Contribution to Household
Maintenance in South India," in A Home Divided, eds. Dwyer and Bruce: 99-119. 40. Patricia Engle, cited in A Home Divided, eds. Dwyer and Bruce: I-19. 41. Sally Cole, Women of the Pr-aia (Princeton, NJ: Princeton University Press, 1991). 42. Candace Bradley, "Women's Empowerment and Fertility Decline in Western Kenya," in Situating Fertility: Anthropological and Demographic Inquiry, ed. S. Greenhalgh (Cambridge: Cambridge University Press, 1995): 157-78.
43. John Bongaarts, W. Parker Mauldin, and James Phillips, "The Demographic Impact of Family Planning Programs," Studies in Family Planning2l, no. 6 (1990): 299-310. 44. John Bongaarts, "The Role of Family Planning Programs in Contemporary Fertility Transitions, Population Council Working Papers 71 (1995). 45. Lori S. Ashford, "New Perspectives on Population: Lessons From Cairo, Population Bulletin 50, no. 1 (Washington, DC: Population Reference Bureau, 1995); Wolfgang Lutz, "The Future of World Population," Population Bulletin 49, no. 1 (Washington, DC: Population Reference Bureau, 1994); and "Beijing and Beyond: Toward the Twenty-First Century of Women,Women's Studies Quarterly 24, nos. 1 and 2 (Spring/Summer 1996).
46. PRB Staff, Family Planning Programs: Diverse Solutions for a Global Challenge (Washington, DC: Population Reference Bureau, Nov. 1994); Farzaneh Roudi and Lori Ashford, Men and Family Planning in Africa (Washington, DC: Population Reference Bureau, July 1996); and Lori S. Ashford, Implementing Reproductive Health Programmes (Plymouth, UK: USAID and UKODA, Dec. 1995).
47. Judith Bruce, "Fundamental Elements of the Quality of Care: A Simple Framework," Studies in Family Planning21, no. 2(1990): 61-91.
48. Rachel Snow, "Each to Her Own: Investigating Women's Response to Contraception," in Power and Decision: The Social Control of Reproduction, eds. G. Sen and R. Snow (Cambridge, MA: Harvard University Press, 1994): 233-54.
49. Caroline Bledsoe, Allan Hill, Umberto D'Alessandro, and Patricia Langerock, "Constructing Natural Fertility: The Use of Western Contraceptive Technologies in Rural Gambia," Population and Development Review 20, no. 1 (1994): 81-114. 50. Francine Coeytaux, "Celebrating Mother and Child on the Fortieth Day: The Sfax, Tunisia Postpartum Program," Quality/Calidad/Qualite, no. 1 (New York: The Population Council, 1989).
51. Anrudh K. Jain, "Fertility Reduction and the Quality of Family Planning Services," Studies in Family Planning 20, no. 1 (1989): 1-15. 52. Ashford, "New Perspectives on Population."
53. Adrienne Germain, Sia Nowrojee, and Hnin Hnin Pyne, "Setting a New Agenda: Sexual and Reproductive Health and Rights," in Population Policies Reconsidered: Health, Empowerment, and Rights, eds. G. Sen, A. Germain, and L. Chen (Cambridge, MA: Harvard University Press, 1994): 27-46.
54. Anrudh Jain and Judith Bruce, "A Reproductive Health Approach to the Objectives and
Assessment of Family Planning Programs," in Population Policies Reconsidered, eds. Sen et al.: 193-210.
55. Steven Sinding, John Ross, and Allan Rosenfield, "Seeking Common Ground: Unmet Need and Demographic Goals," International Family Planning Perspectives 20, no. 1 (March 1994): 23-27, 32.
56. Bonnie Kay, Adrienne Germain, and Maggie Bangster, "The Bangladesh Women's Health Coalition," Quality/Calidad/Qualite 3 (New York: The Population Council, 1991);James Phillips, Mian Bazle Hossain, Ruth Simmons, and Michael Koenig, "Worker-Client Exchanges and Contraceptive Use in Rural Bangladesh," Studies in Family Planning 6, no. 1 (1993): 329-43; and Ruth Simmons, Laila Baqee, Michael Koenig, and James Phillips, "Beyond Supply: The Importance of Female Family Planning Workers in Rural Bangladesh," Studies in Family Planning 19, no. 1 (1988): 29-38. 57. Sidney R. Schuler, Syed Hashemi, and Ann H. Jenkins, "Bangladesh 's Family Planning Success Story: A Gender Perspective," International Family Planning Perspectives 21, no. 4 (1995): 137.
58. Sidney R. Schuler and Syed Hashemi, "Credit Programs, Women's Empowerment, and Contraceptive Use in Rural Bangladesh," Studies in Family Planning 25 (1994): 65-76. 59. Ruth Dixon-Mueller and Adrienne Germain, "Population Policy and Feminist Political Action in Three Developing Countries," in The New Politics of Population, supplement to Population and Development Review 20 (1994): 197-220.
61. Ruth Dixon-Mueller, Population PoLicy and Women's Rights: Transforming Reproductive Choice
(Westport, CT: Praeger, 1993). 62. Ashford, "New Perspectives on Population." 63. Jain and Bruce, "A Reproductive Health Approach."
64. Julie Mertus, "State Discriminatory Family Law and Customary Abuses," in Women's Rights/ Human Rights: International Feminist Perspectives, eds. Julie Peters and Andrea Wolper (New York: Routledge, 1995): 135-48; and Lori Heise, "Freedom Close to Home: The Impact of Violence Against Women on Reproductive Rights," in Women's Rights/Human Rights, eds. Peters and Wolper: 238-55.
Adepoju, Aderanti and Christine Oppong, eds. Gender, Work, and Population in Sub-Saharan Africa. London and Portsmouth, NH: James Currey and Heinemann for the International Labour Office, 1994. Caldwell,John. "Routes to Low Mortality in Poor Countries." Population and Development Review 12 (1986): 171-220.
Desai, Sonalde and Devaki Jain. "Maternal Employment and Family Dynamics: The Social Context of Women's Work in Rural South India." Population and Development Review 20, no. 1 (March 1994): 115-36. Dixon-Mueller, Ruth. Population Policy and Women's Rights: Transforming Reproductive Choice. Westport, CT: Praeger,1993.
Emecheta, Buchi. The Joys of Motherhood. New York: George Braziller,1979. Federici, Nora, Karen Oppenheim Mason, and Solvi Sogner, eds. Women's Position and Demographic Change. Oxford: Clarendon Press, 1993.
Ginsburg, Faye, and Rayna Rapp, eds. Conceiving the New World Order: The Global Politics of Reproduction. Berkeley: University of California Press,1995. Greenhalgh, Susan. "Controlling Births and Bodies in Village China." American Ethnologist 21 (1994): 3-30.
Greenhalgh, Susan. "The Social Construction of Population Science: An Intellectual, Institutional, and Political History of 20th Century Demography." Comparative Studies in Society and History 38, no. 1 (1996): 26-66. Inhorn, Marcia. Quest for Conception: Gender, Infertility, and Egyptian Medical Traditions. Philadelphia: University of Pennsylvania Press, 1994. Jejeebhoy, Shireen. Women's Education, Autonomy, and Reproductive Behaviour: Experience from Developing Countries. Oxford: Clarendon Press, 1995. Johansson, Sten, and Ola Nygren. "The Missing Girls of China: A New Demographic Account." Population and Development Review 17, no. 1 (1991): 35-52.
Lorber, Judith. Paradoxes of Gender. New Haven: Yale University Press, 1994. Mason, Karen Oppenheim. "Gender and Demographic Change: What Do We
Know?" IUSSP Paper,1995.
Riley, Nancy E. "Similarities and Differences: Anthropological and Demographic Perspectives on Gender." In Anthropological Demography: Toward a New Synthesis, eds. D.I. Kertzer and T. Fricke. Chicago: University of Chicago Press, 1997. Scheper-Hughes, Nancy. Death Without Weeping: The Violence of Everyday Life in Brazil. Berkeley: University of California Press,1992.
Wolf, Margery. Women and the Family in Rural Taiwan. Palo Alto: Stanford University Press,1972.
Zeng Yi, Tu Ping, Gu Baochang, Xu Yi, Li Bohua, and Li Yongping. "Causes and Implications of the Recent Increase in the Reported Sex Ratio in China." Population and Development Review 19, no. 2 (1993): 283-302.
Nancy E. Riley is assistant professor in the Sociology/Anthropology Department at Bowdoin College, Maine. Her research interests include feminist demography and issues relating to gender, family, and population in China. Dr. Riley is currently involved in a project examining the effects of factory and service work on women's power in China. She began work on this Population Bulletin while she was the Mellon Visiting Scholar at the Population Reference Bureau.
The author acknowledges the valuable contribution of James McCarthy to this manuscript and thanks S. Philip Morgan and Sunita Kishor for their careful reading and helpful comments on an earlier draft of this Population Bulletin.…