Female Education and Fertility: Examining the Links
Ian Diamond, Margaret Newby, and Sarah Varle
In the past 20 years a large amount of both theoretical and empirical research has investigated the relationship between women's education and fertility. The results of this work suggest that, unlike the relationship between a woman's years of schooling and infant mortality, which tends to be linear (Cleland and van Ginneken, 1988), the relationship between education and fertility is much more complex. The underlying pattern most commonly shows a negative relationship, although positive relationships at very low and very high levels of schooling have been demonstrated.
In a comprehensive review of empirical research on the relationship between women's schooling and fertility, Jejeebhoy (1995) usefully categorizes education-fertility relationship into four types, based on the results of 59 studies (see Figure 2-1). The first pattern (a) is one in which fertility falls monotonically with increased years of schooling. This pattern occurs in 26 of the 59 studies. There is a similar pattern in (c), but rather than a constant decrease in fertility, the relationship is ''seven-shaped," and the first few years of schooling have either no effect on or produce a slight rise in fertility (13/59). Pattern (b) shows an inverted U or J-shaped curve, indicating that a few years of schooling increases fertility initially, but eventually fertility declines (13/59). The final relationship (d) is one in which there is no relationship or fertility rises monotonically with education (7/59), although it should be noted that there are no examples of this latter relationship in recent studies.
Comparisons based on a large number of international studies are risky given
compositional differences in the demographic and socioeconomic structure of the populations, as well as in the period to which the studies refer. Nevertheless, Jejeebhoy's review demonstrates clearly that for women with just a few years of education, there is little evidence of a systematic relationship between education and fertility, although a number of recent studies indicate a moderate negative effect. At the secondary level or above, however, the relationship is always negative. The countries in which there is either no or a positive relationship are almost exclusively in sub-Saharan Africa (6/7), as are the majority of inverted U
or J-shaped relationships (8/13) and seven-shaped relationships (6/11). Therefore, with the exception of sub-Saharan Africa, the education-fertility relationship has a strong tendency to be negative. It should also be noted that a number of the sub-Saharan African studies included in Jejeebhoy's review are now relatively dated, and the incidence of a positive effect is now much rarer than was observed using World Fertility Survey data. For example, Muhuri et al. (1994) find a positive association between fertility and a small amount of education in only 4 of 14 sub-Saharan African countries in which Demographic and Health Surveys were conducted during 1985-1992.
An important question is the extent to which there is a simple macro-level relationship between Jejeebhoy's categorization of the education-fertility relationship and the country's level of economic and educational development (see Singh and Casterline, 1985; Entwisle and Mason, 1985). Jejeebhoy's data show that negative relationships characterize countries that have both higher per capita income and a higher level of female literacy. This finding supports the view that both economic development and mass education influence childbearing behavior. However, it would be extremely naive not to recognize the large amount of heterogeneity at the individual level.
Therefore, it is important to consider individual-level data when attempting to identify the ways in which women's education influences fertility. Throughout the 1980s and 1990s, a large number of studies, often based on data from the World Fertility Surveys and Demographic and Health Surveys, have addressed the issue. Many of these studies have tried to explain the impact of education on fertility through the proximate determinants of fertility, most notably by quantifying education's relationship with age at marriage and use of contraception. These studies typically show, not surprisingly, that educated women tend to marry later and are more likely to use modern methods of contraception.
Indeed, educated women generally also have a lower desired family size. Cleland and Kaufmann (1993) point out, however, that after controlling for age and existing family size, educational differentials in desired family size are relatively modest and much smaller than the actual fertility differentials. This finding leads them to conclude that the key question to be addressed is why more-educated women are more likely to act on their reproductive preferences, rather than why they want smaller numbers of children (see also Sathar, 1996). Cleland and Kaufmann (1993) go on to argue that attempts to explain the education-fertility relationship should focus more on identifying the determinants of reproductive decision making. The studies that look at proximate determinants are all extremely useful, then, but ultimately one needs to know what it is about education that leads to these behavioral changes. Answering the question of the pathways through which education influences fertility is the primary aim of this paper.
Many of the pathways identified relate to how educated women are, for example, better able to appreciate and interpret media messages, deal with bu-
reaucracy, enter the labor market, and have better dialogues with their spouses than their less-educated counterparts. More-educated women also (tend to) belong to groups with different social norms than those of the less educated. Each of these factors has implications for effective reproductive decision making, and each is associated what can be described as 'women's autonomy' (although see Jeffery and Basu  for a critique of the education-autonomy viewpoint).
In her review, Jejeebhoy (1995) develops a framework in which education influences fertility first through five types of autonomy—knowledge, decision making, physical (in interacting with the outside world), emotional, and economic and social—and then through the proximate determinants. All five types of autonomy are influenced by education, and all have an important effect on fertility behavior, particularly through ensuring that women have an increased ability to make decisions, to move freely, and to have control over their actions. It should be noted that there are a number of complexities within this framework. For example, it is often assumed that there is a close positive relationship between autonomy and status. However, while the ability to move freely is often cited as a key indicator of autonomy (see, for example, Steele et al., 1996), freedom of movement may not always be associated with increased status. Caldwell (1986) argues that women who move around of their own volition may have lower status in their society, and Amin (1996) describes how in Bangladesh, family planning workers often wear an elaborate burqa or veil to compensate for any loss in status due to moving around.
The framework proposed by Jejeebhoy (1995) is essentially of the path analytic type. A comprehensive illustration of such a framework is provided by Jeffery and Basu (1996) and is adapted in Figure 2-2. This framework could be used as the base for a multivariate analysis aimed at identifying the factors that influence fertility directly as opposed to doing so only indirectly. For example, Mason (1984) suggests that women's education may affect fertility only through other factors. However, those who have undertaken individual-level analyses of large data sets have typically found that there remains an effect of years of schooling even after controlling for many other socioeconomic and behavioral factors (Rodriguez and Cleland, 1980; United Nations, 1987; Cleland and Rodriguez, 1988; Rodriguez and Aravena, 1991; Castro Martín and Juárez, 1995; United Nations, 1995), although as Jejeebhoy (1995) points out, few of these analyses have addressed the pathways through women's autonomy.
Another important concern is the impact of national and regional norms and culture. To take an example, a particularly strong state could alter childbearing norms across all social strata with little individual choice; one could argue that China provides such an example. Yet it is rare that a state can have such a strong and widespread impact on the norms and values of its people, nor is it likely that individuals will be completely unaffected by their normative and cultural context. Moreover, the impact of such contextual factors may vary with the individual's level of educational attainment. For example, an educated woman may be more
inclined to act counter to local norms or more likely to marry at a social level at which lower childbearing norms prevail (see Basu, this volume).
In summary, there is no global relationship between education and fertility; rather, the linkages are both variable and complex. Yet given this essential caveat, it can be said that the relationship between the two is largely negative, and some common associations between different levels of education and reduced fertility can be identified. Before the discussion of these associations proceeds, however, four additional caveats must be noted. First, the education to which we refer is formal academic schooling received by children and young people. The suggested associations with fertility should not be generalized to special adult education programs or training within the workplace, although they clearly have implications for these other forms of education. Second, the discussion focuses on female education. Although we recognize that the influence of men is extremely important in reproductive decision making, this is a large topic that merits separate treatment and is not addressed in this chapter. Third, while years of schooling is used as the measure of education, it is essential to recognize that there is great variability in the quality and content of education that is acquired during a year of schooling in different parts of the world; hence the categorization should be seen in broad terms. Finally, this chapter focuses primarily on the linkages between education and fertility at the individual level, while bearing in mind that the education system also helps shape societal norms that may affect the fertility of women who do not themselves receive formal education.
The next section summarizes the ways which the education-fertility relationship varies with level of education (years of schooling). The chapter then examines specific factors that appear to influence the association between education and fertility. The final section presents a summary and conclusions.
The Education-Fertility Relationship And Years Of Schooling
In almost all countries, primary education is associated with lower numbers of children ever born than are found among those with no education, although, as noted earlier, a positive relationship has been observed in some countries, particularly for only 2 to 3 years of primary education (see Singh and Casterline, 1985). The relationship between primary education and desired family size, on the other hand, is almost universally negative, whether or not the primary education is completed (Jejeebhoy, 1995). Both incomplete and complete primary education also tend to be associated with later age at marriage and increased contraceptive use (see Westoff et al., 1994).
In contrast with primary education, almost every study reported by Jejeebhoy (1995) shows a marked decline in children ever born for women with secondary education. In many of the studies, the negative effect is substantial. The difference in total fertility rates (TFRs) between women with secondary education and
those with no education is more than 50 percent in Latin America. In Asia the differential is smaller because the TFR among women with no education is relatively low (3.36 with no education versus 2.52 with secondary education in Indonesia; 4.90 and 3.64, respectively, in Pakistan). In many sub-Saharan African countries the differential is smaller still. Women in that region with secondary plus education have higher fertility than do women with no education in Asia.
It should be noted here that the association between secondary education and fertility may be attributable to the fact that in many developing countries, those girls who attend secondary school form a highly select group that might be expected to have lower fertility for other reasons, such as higher socioeconomic status. In the majority of developing countries, a woman who reaches and completes secondary education is rather a special case. Very few countries have achieved mass secondary education for women, and thus the influence on fertility will be largely at the level of the individual. For a girl to attend secondary school, her family must believe strongly in the value of education, which implies an investment in their daughter's future, and expectations and aspirations for her in terms of further education, employment, and marriage will be high. All this may act to delay marriage and reduce fertility. In addition, the woman may already be exposed to low family-size norms since, in general, the high cost of secondary schooling means that one would expect it to be biased toward small families in which there are fewer siblings competing for resources.
The relationship between higher education (beyond the secondary level) and fertility is undoubtedly negative. Jejeebhoy's (1995) data show that higher education is associated with large reductions in completed family size (percentage reductions over no schooling range from almost 20 to over 70 percent) and with substantially lower desired family size and higher age at marriage. In none of the studies is 10 or more years of education associated with higher fertility relative to women with no education.
At the same time, it is difficult to draw clear conclusions about the influence of higher education on fertility since the proportion of the female population involved is very small, and those women who have benefited from tertiary education in most developing countries are predominantly in the younger age groups. Thus it remains to be seen what influence higher education has on fertility over the full span of the childbearing career. As the work by Kiernan and Diamond (1983) in Great Britain suggests, it may be that in relatively low-fertility societies (as many societies will already be when they have widespread higher education), higher education influences the timing of childbearing as much as it affects completed family size. Further understanding will be possible only when the more highly educated cohorts complete their childbearing careers. It is possible to speculate that their completed fertility may be relatively high since they will have the economic resources to care for and educate a large number of children. It is equally possible to speculate that their completed fertility will be low be-
cause of the costs of rearing children, along with a shift toward social norms related to low fertility.
However, since educating these few women to the tertiary level will have small macro-demographic effects on fertility, the importance of higher education lies more in how these women will use their education in a position of influence than in the effects of that education on their personal fertility. This influence may be informal, with these women serving as role models for other family members, friends, and neighbors, or it may be institutionalized through their role as family planning workers, government ministers, and civil servants—careers to which their education gives them access. The informal role is more likely among upwardly mobile women from relatively disadvantaged backgrounds and the latter among the socioeconomic elite.
An important pathway through which higher education can influence declines in fertility is by delaying marriage. Choe (1996) has shown that highly educated women in Korea are much less likely to be married at any age. The reasons for this are that attendance at university usually delays marriage until the degree is completed, and higher education increases career aspirations and entry into employment. In addition, highly educated women may not be well positioned in the marriage market because some men may not wish to marry women who are more educated than themselves.
It is also important to note that those women from relatively lower socioeconomic groups who do receive tertiary education are likely to be extremely motivated and independently minded and to have high career aspirations, qualities that are inconsistent with the traditional norms and ways of life found in many developing countries. In addition, higher education usually can be obtained only in cities, so that those with higher education will be urban born or rural-urban migrants; in almost all countries, both these groups have lower fertility than rural dwellers. Thus women with higher education come from a very select socioeconomic group or tend to have exceptional personal characteristics and experience of urban living. As a result, it is difficult to be sure whether the lower fertility of highly educated women can be attributed to their education per se or to their other characteristics.
Finally, it must be noted that education does not work in isolation to affect fertility. There will probably be a certain degree of direct influence, but in many ways, education will also serve as a proxy for other factors. Increased education rarely occurs without concomitant changes to a society, such as increased health services, enhanced communication, and better infrastructure. The effects of these factors cannot easily be disentangled. The direct effects of education at the individual level, such as the increased employment prospects it affords, the transfer of knowledge about the costs of children, or an increase in social skills enabling better use of health services, are meaningless if other societal changes are not occurring in parallel. Thus, for example, health services cannot be used if they are not available, and one cannot aspire to different jobs or lifestyles if there
is not some way, however remote, of being able to achieve them. Moreover, as discussed further below, "the notion that there is a linear relationship between length of schooling and degree of modernity or westernization is a gross oversimplification" (Cleland and Kaufmann, 1993:24).
Factors Influencing The Education-Fertility Relationship
The available evidence suggests that the following factors help shape the relationship between education and fertility: key contextual factors, the skills and knowledge imparted by schooling, social and ideational influence, and enhanced employment opportunities for the educated.
Two contextual factors influence the relationship between education—particularly incomplete primary education—and fertility: the existence of mass education1 and the presence of an active family planning program. It is likely, of course, that countries with a strong program of mass education will also have a strong family planning program, and thus it may be difficult to separate out the influence of each (Mauldin and Ross, 1991).
Presence of Mass Education
Countries in which a small amount of education (less than completed primary) is associated with a decline in fertility at the individual level tend to be those countries with mass education. This may be the case in part because in the latter countries, the influence of education extends beyond the individuals who attend school to the society at large. In general, a shift toward smaller families tends to occur in parallel with the introduction of mass education. Caldwell (1982) argues that this shift takes place as a result of not only the direct influence of education on individuals, but also the interaction among knowledge, ideas, and increased opportunities afforded by mass education through the restructuring of family and community relationships. Thus women with just a few years of schooling will tend to adopt norms of lower fertility and higher age at marriage that are characteristic of societies with mass education even if the influence of education on them as individuals is very small.
With a globally increasing emphasis on education, countries in which large proportions of the population have completed primary education are becoming more common. This is the case in many parts of Latin America, the Caribbean, and the Far East. By the 1980s, for example, gross primary enrollment rates for females exceeded 90 percent in all the countries of East Asia listed by King and Hill (1993). Among those who were of school age in the 1960s and 1970s, however, when completed primary education was rare in developing countries, individuals with completed primary education are likely to have been educational "pioneers" within their communities. In India in 1992-1993, for example, only 11.7 percent of women interviewed in the National Family Health Survey had completed primary schooling, and this percentage was even lower in some rural areas. At about the same time in Kenya, only 15.3 percent of women interviewed for the Demographic and Health Survey (National Council for Population and Development, 1993) had completed primary education. In such contexts, the influence of education on fertility is likely to be limited to those individuals who have completed primary schooling. As noted above, such contexts can be expected to become increasingly rare.
As the level of education increases within a community, norms concerning childbearing within that community will change. Thus, in situations where there is mass education, women who complete primary school may have lower than expected fertility because their reproductive decisions are influenced by lower community childbearing norms. Lower fertility will then result from changing societal norms, which will have occurred only partly as a result of mass education. Other changes that often occur in parallel with the development of education are that a nation moves toward a monetarized economy and becomes more industrial, less agricultural, and more urbanized. Thus the advent of mass education is just one aspect of a changing society that will influence the individual effects of education.
Presence of an Active Family Planning Program
Entwisle and Mason (1985) found that in poor countries, the relationship between education and fertility is negative when there is a family planning program, but positive when there is not. In most cases, countries in which women with incomplete primary education have a substantially lower total marital fertility rate than women with no education are those classified by Mauldin and Ross (1991) as having a strong or moderate family planning program effort in 1989 (see Table 2-1).2 In Botswana, for example (Lesetedi et al., 1989), which was
TABLE 2-1 Mean difference between total marital fertility rate of women with no education and those with incomplete primary education by family planning program effort as a percentage of the total marital fertility rate with no education
% Difference in TMFRb
Very weak or none
a Program effort data from Mauldin and Ross ( 1991).
b Total marital fertility rate (TMFR) of women married 0-19 years in the 0-4 years prior to the survey. Data are for 31 countries in which Demographic and Health Surveys were conducted during 1985-1992 (Muhuri et al., 1994).
c Numbers in parentheses are the number of countries in that category.
classified as having a strong family planning program effort, among women married for 0-19 years, those with incomplete primary education had a total marital fertility rate 8.8 percent below that of women with no education. In countries such as Nigeria, Burundi, and Liberia, where the marital fertility of women with incomplete primary education was found to be substantially higher than that of women with no education, family planning effort was classified as weak, very weak, or none. A strong family planning program will provide both the means to reduce fertility and the messages to encourage this reduction.
It is important to note, however, that family planning program effort may be related to a country's level of economic development. Thirty years ago, Carleton (1967, cited in Cochrane, 1979) argued that since none of the ways in which education affects fertility is completely independent of economic development, care must be taken in assigning causality to the family planning program. In addition, as family planning can influence all educational groups (Cleland and Kaufmann, 1993), it is possible that the influence of education on fertility may be weaker in the presence of a strong family planning program. An example here would be China, where a strong state family planning program influenced fertility across all socioeconomic and educational groups. At the same time, for reasons explained below, it is likely that the impact of a family planning program will be stronger in societies in which at least a small amount of education is usual,
and that one effect of small amounts of education is to facilitate the effective use of family planning programs.
Skills and Knowledge Imparted
The quality of education provided at all levels (primary, secondary, and beyond) varies greatly among countries; moreover, the level of competence achieved varies greatly among countries, schools within countries, and pupils in the same school. Nevertheless, certain basic skills can be expected to be imparted by primary schooling. The most obvious skills transferred in the first few years of schooling are basic literacy and numeracy. The skills imparted are unlikely to meet high standards, and many of those who have received only 2 to 3 years of education may be functionally illiterate. Yet children who have had some education are better equipped than their noneducated counterparts to recognize written text and the fact that it is likely to contain some sort of message, and those who have completed primary education will probably have achieved a substantial level of literacy. These children are thereby brought into contact with health education, family planning, and other media information.
LeVine et al. (1994) examined the effects of literacy on the ability to read and listen to health announcements. A study in Nepal among a sample of women with a mean level of schooling of 1.4 years found comprehension of a health interview to be positively correlated with maternal schooling. This increased comprehension also appeared to have an effect on practice since there was a positive correlation between maternal schooling and a number of health outcomes, including contraceptive use. LeVine and colleagues also found that in Mexico and Zambia, where their samples had a higher mean number of years of schooling, literacy increased the ability to understand radio broadcasts through better comprehension of decontextualized language. They offer the following conclusion (p. 188):
Literacy is not a dichotomous trait acquired in the first five years of primary schooling that permits the adult population of a country to be divided into literates and illiterates, but a package of cognitive and language skills which make it possible to participate in literate discourse and communication and which can improve during primary school.
Not only does primary education increase the ability to understand messages transmitted by various forms of mass media, but there is also evidence that the exposure to mass media increases with even a few years of education. The 19931994 Demographic and Health Survey in Bangladesh showed that 9.5 percent of women with no schooling watched television weekly, as opposed to 15.8 percent of those with incomplete primary education. Among women with no education, 28.7 percent listened to the radio weekly, as opposed to over 43 percent of those with incomplete primary education (Mitra et al., 1994), and the differential is still
greater for those with completed primary education. At the same time, it should be noted that part of this effect may not be attributable to primary education increasing access to the media, but to a correlation between both education and radio and television ownership and socioeconomic status.
At the secondary level of schooling, literacy will be achieved to a relatively high standard. The individual is exposed to many forms of the written word, including information on health and family planning. For example, in the 19931994 Bangladesh Demographic and Health Survey, it was found that among ever-married women, 38.7 percent of those with secondary or higher education read a newspaper at least once a week, compared with 8.2 percent of those with primary education. The probability of watching television or listening to the radio once a week was also significantly higher among the former women (Mitra et al., 1994), although, as discussed above, this may in part be an effect of socioeconomic status rather than education per se. Moreover, secondary education can help develop cognitive skills needed to evaluate information and form personal opinions accordingly.
Primary education increases the information available to children not only through improved access to the media and written messages, but also through the knowledge imparted at school. For example, basic primary education generally includes some health education, although the amount and quality of the information presented varies greatly. Cleland and van Ginneken (1988:90) argue that while specific evidence is sparse, it is likely that primary schooling ''imparts sufficient understanding of health matters to guide maternal behaviour in later life." Even though health education may constitute a small portion of the curriculum, the fact that it is provided means the overall concept of being biologically able to control fertility and health is likely to become apparent to the child.
In communities that have health systems and family planning programs, primary education establishes a foundation for further information received within the community, and thus can have an indirect impact through an increased ability to "hear the message" of family planning programs or through improved health for individuals or their children. As Caldwell (1994) argues for mortality, the school is the best place to inculcate a society with norms that are congruent with modern health systems.
Further, there is much evidence to show that improvements in infant and child health occur even with very low levels of maternal education (Preston, 1978). For example, in the 10-year period up to 1993-1994 in Bangladesh, infant mortality was 113/1000 among women with no education and 92/1000 among those with incomplete primary education. A number of commentators (e.g., Basu, 1994; Cleland and Streatfield, 1992; Preston, 1978) have pointed out that reduced infant and child mortality can lower fertility since it decreases the perceived need for replacement and insurance births to ensure that a certain number of children will survive to adulthood. Reduced mortality can also increase birth spacing since the death of a child truncates the period of breastfeeding, shorten-
ing the length of amenorrhea. This tendency for even small amounts of education to reduce infant and child mortality may thus represent an indirect influence of education on fertility. A caution in this regard is expressed by Basu (1994), who points out that the decline in mortality may itself be a consequence of lower fertility since there will be fewer high-risk, high-parity births to older women. This observation illustrates the complexities involved in trying to untangle the education-fertility relationship.
Formal health education in secondary schools increasingly includes family planning. For example, in some schools in East Africa, secondary school children are shown a video called "Consequences" that explores the difficulties faced by a teenage girl who becomes pregnant before she finishes her education. The effects on future fertility of childbearing at very young ages are highlighted. The aim is to discourage early marriage and childbearing, as well as premarital pregnancy, which is the focus of the film. Another important issue in the film is the negative effect of pregnancy and motherhood on employment prospects and careers, perhaps introducing to secondary school girls the idea that education is a valuable asset for improving life opportunities and offering access to a rewarding career.
In terms of cognitive skills, primary education will improve the ability to approach decision making and problem solving in a more abstract way. In terms of childbearing decisions, not only will the ability to weigh costs and benefits be more likely, but there will also be more information available with which to make such decisions. However, it should be noted that the ability to obtain and process information will reduce childbearing only where the marginal costs of childbearing at the existing level of fertility outweigh the benefits. Another condition for lower fertility is that women be able to act on their rational choices, perhaps through the increased autonomy education appears to provide (see Easterlin, 1975; Cochrane, 1979; Jejeebhoy, 1995).
It is also worth noting that childbearing is not, of course, always a planned rational decision; a large number of births are the result of unplanned or mistimed pregnancies (for example, to teenage girls). Cleland and Kaufmann (1993) observe that one way education contributes to reduced fertility is by reducing the percentage of such pregnancies. Schooling has traditionally been thought to increase children's ability to think rationally as it assists them in moving toward abstract and reasoned thought. Indeed, there is a literature from many areas of the world indicating that the incidence of unplanned pregnancy is higher among less-educated women. On the other hand, the assumption that schooling facilitates rationality is not universally accepted. Cross-cultural studies have shown that schooling does not increase rationality, but rather teaches new skills that change the way problems are approached and hence the way they are solved and decisions are made. Schooling may also emphasize the negative economic implications of large families or the health benefits of smaller/better-spaced families, or it may promote the advantages of large families. It thus alters which pieces of
information and which values go into the decision, not the fact that a rational decision is made.
Social and Ideational Influences
Perhaps more important than the cognitive effects of a few years of schooling are its social and ideational influences. Through the latter influences, schooling acts as a "catalyst of modernisation" (Martin and Juarez, 1995).
The social values imparted in school are an important influence on the direction of the education-fertility relationship. Streatfield (1989), for example, describes how low-fertility norms were incorporated into the education system of Bali. He uses this example to argue that within a culture in which one is taught to obey and trust one's teachers and other authority figures, such as political and religious leaders and parents, even a small amount of schooling is likely to have an important influence on childbearing. Attitudes toward childbearing are likely to change in a downward direction because the perspectives promoted by schooling are generally those of the middle classes and the West, as found by Caldwell (1982) in a study of school textbooks in Kenya, Ghana, and Nigeria. The norm suggested is generally that of the nuclear family with few children.
Conversely, education is likely to influence fertility positively if the educational message is strongly traditional and pronatalist. In some societies, religious schooling provides such messages, teaching about traditional values that are inconsistent with widespread family planning. In Bangladesh, for example, particularly in rural areas, a significant number of children attend madrassas (religious schools) and schools attached to mosques, where they are unlikely to learn western or modern values (Bangladesh Task Forces, 1991). At the same time, however, religious leaders in some countries have become involved in spreading messages about family planning and the benefits of small families.
Schooling may also exert a social influence by exposing students to the outside world and the media, either while they are attending school or later, especially if schooling results in the opportunity to live away from the home and the immediate family environment, as often happens in parts of West Africa. As Cleland and van Ginneken (1988:91) suggest, "education provides a wider social network, new reference groups and authority models and a greater identification with the modern world." Thus Cleland and Kaufmann (1993) posit a process of transferred social values that differ from those of the traditional home environment. Although this transfer may begin to occur with a few years of education, it can probably be expected to increase with continued exposure.
Students in secondary school are particularly prone to identify with social groups whose norms differ from those held in the home community (see de Vries,
1992). Those norms are likely to include low fertility. Indeed, it is important to recognize that individuals play an active role in choosing a social group (see Coggans and McKellar, 1994), and secondary education may motivate women to seek peer groups in which low fertility is a norm.
Attendance at school often brings children into an environment where they are exposed to new authority figures and role models and to ideas other than those held in their home. Schooling also exposes children to different ideas, whether through the study of particular subjects, such as history, geography, or religion, or simply through the reading of storybooks. As a consequence of the increased information and exposure to new ideas and authority figures that full primary schooling brings, ideational change occurs.
An example of ideational change attributed to education has to do with the relationship between mother and child. LeVine et al. (1991) argue that the better educated the mother is, the more she will interact with her children. Verbal interaction, in particular, increases, and the child's upbringing becomes more child-centered, both factors serving to encourage greater individuality on the part of the child (though see Carter, this volume, for a critique of this point). The increased reciprocal verbal and nonverbal interaction makes childrearing more labor-intensive, with the result that the mother is likely to want fewer children.
On the other hand, not all the ideas to which children are exposed in school are conducive to lower fertility. The "modern" ideas encountered by children in school might include, for example, shorter breastfeeding durations and erosion of traditional practices such as postpartum abstinence that are particularly prevalent in sub-Saharan Africa. Shorter breastfeeding durations lead to shorter periods of postpartum amenorrhea and thus to an increase in natural fertility, and without a compensating increase in contraceptive prevalence, total marital fertility is also likely to rise. The evidence on this issue, however, is mixed. On the one hand, in Botswana (Lesetedi et al., 1989), the mean duration of breastfeeding and postpartum abstinence was found to be higher among women with incomplete primary education than among those with no education, although this finding could reflect misreporting among the group with no education. On the other hand, in the 19931994 Bangladesh Demographic and Health Survey (Mitra et al., 1994), the duration of full and exclusive breastfeeding was shorter among women with incomplete primary education than among those with no education, but the median duration of any breastfeeding was over 36 months for both groups. Thus at low levels of schooling, it cannot be assumed that all aspects of education will have a negative influence on fertility.
With regard to secondary education, an important influence of education on fertility comes from its close relationship to the processes of urbanization and integration into the international economy. These processes tend to be associated
with reduced son preference, a rise in the opportunity costs of children, higher expectations for living standards and increased life choices as the result of better prospects and opportunities for work, and increasing acceptability of female employment.
As women achieve secondary education, the gender gap in education will narrow, and there will be an increased probability of a woman's husband being highly educated. Better-educated husbands are in turn more likely to hold ideals that are consistent with smaller families, improved health, and family planning (see also Basu, this volume). Secondary education may also delay marriage, both because it becomes more difficult for women to find a suitable husband and because women may wish to pursue cash employment. Cochrane (1979) argues that high levels of education reduce the pool of acceptable spouses, given the traditional pattern that men marry women less educated than themselves. She found that increased education reduced the probability of women entering a formal union, while it increased the probability of men doing so. Recently, similar results were found for Korea by Lee (1994) and Choe (1996). Smock and Yousseff (1977, cited in Stichter and Parpart, 1990) found that in Egypt, high levels of education or employment could jeopardize chances of a good marriage match because such women were seen as loose or immoral—another way education can make finding a spouse more difficult.
As noted earlier, one way in which education is posited to influence fertility negatively is by increasing women's sense of autonomy. Jejeebhoy (1995:184) believes that "the impact of women's education on their fertility is greatest when education offers women an expanded role in family decisions and control over, or access to, resources." In exposing women to new ideas and allowing them to gain cash employment, thus taking them out of the home, education may affect the autonomy of women in a number of ways. For example, it can enhance a woman's position in the community; give her increased confidence and skill, thereby enhancing her negotiating power in the household; or give her a degree of economic independence that reduces her dependence on her spouse and his family. Moreover, if education and subsequent employment have also secured the women a more-educated husband, he is likely to be more liberal in his attitudes toward women and childbearing (see Basu, this volume). Where women are more educated and autonomous, they are likely to have a greater say in their choice of husband, which has been shown to contribute to improved communication between spouses, and this in turn aids in women's control over fertility (Kabir et al., 1988).
Although this effect of increased autonomy clearly does occur in many societies, one must note that it is not universal. Cleland (1995) reports that in Bangladesh, women with schooling do not appear to have any more autonomy
than their uneducated counterparts. Fertility decline in Bangladesh, he argues, has not been coupled with a change in female employment, a decline in child labor, or the evolution of alternative forms of risk insurance. Other forces, such as the decline in child mortality, are playing a more important role in the fertility transition of some South Asian nations (although education levels are, of course, a principal predictor of child mortality rates).
Moreover, greater autonomy does not guarantee that desired family size will be smaller. Where desired fertility is lower, however, women with greater autonomy are more likely to be able to implement their family-size desires and will be less influenced by other family members. The probability of ideal family size decreasing will depend to some extent on the strength of a country's family planning program. The stronger the program, the more likely educated women will be to adopt contraceptive methods—and the more quickly—in order to realize their desires for a small family size. Education leads to greater self-confidence and interpersonal skills (Caldwell, 1986; Cleland and van Ginneken, 1988), empowering women to seek medical and family planning advice and act upon it in a way that is beneficial for them. For this to occur, however, these services must be available.
Enhanced Employment Opportunities
Both the new cognitive skills and the ideational change brought about by primary education will help increase women's chances of employment outside the home. The nature and extent of this effect will depend on the overall level of education in the society. If there is very little education, a small amount of schooling will give women a relative advantage over their peers, thus increasing their employment opportunities substantially. If, on the other hand, education is not uncommon, it will not have as great an effect in this regard. Salaff (1981) showed that in Hong Kong in the mid- 1970s, young women with about 3 years of formal schooling entered employment in factories producing plastic bags, wigs, and garments. In contrast, in Sri Lanka, with high levels of education relative to the rest of South Asia, Rosa (1989) found there was an unofficial requirement of 8-10 years of education in order to enter work in garment factories. In Tanzania, primary education is adequate for entry into teacher training college, whereas in other societies, such as Taiwan or Singapore, the best job available is likely to be factory employment. Thus if one of the important ways education influences fertility is through employment, the strength of that effect will depend on the nature of the labor market and whether the individual views employment as a lifetime career or a temporary phase in the life cycle before marriage.
It is important to note that the relationship between primary education and access to cash employment is not universal. For example, in South Asia the traditional norm is for women to practice seclusion and be excluded from cash employment unless absolute economic necessity dictates otherwise (Cleland and
Kaufmann, 1993; Shaheed, in Afshar and Agarwal, 1989). For the majority of women, primary education will have little impact on their job opportunities and may well serve to reduce their market employment, since families that can afford to educate their daughters to completed primary level may not feel the economic necessity of sending them to work. Indeed, primary education may increase the possibility that a woman will gain a more traditional husband, one able to provide sufficient economic resources so that the woman will not need to work. This observation may help explain why some studies, such as that of Cleland and Rodriguez (1988), conclude that education does not work through employment to affect fertility. However, with increasing industrialization and gradual weakening or adjustment of the norms of female behavior, the degree to which South Asia appears to be an exception may be reduced.
Secondary education provides the qualifications for a fairly good job, perhaps even a career for women. Eight years of education is sufficient for a career as a teacher in many developing countries (Bellew and King, cited in King and Hill, 1993). A study of 1,923 households in metropolitan Dhaka, Bangladesh, in 1992 (Mahmud, 1995) revealed that among secondary-educated women who were working, 22 percent had managerial or professional jobs, 52 percent were in skilled labor, and only 2 percent were engaged in manual labor. Among women with 5 years or less of education, 53 percent performed manual labor, and fewer than 1 percent were in managerial or professional occupations. Thus it can be seen that secondary education can give women access to an entirely different range of employment and to jobs that are far more likely to be economically and psychologically rewarding. The fact that families invest in secondary education for girls suggests that higher-level employment is a goal to which they want girls to aspire. As de Vries (1992) argues, higher levels of education mean that employment may come to replace the home-based reproductive role as the predominant activity in women's lives.
Moreover, if teachers are women, they will serve as role models for the type of career and lifestyle pupils may hope to emulate. It has been suggested that as important or perhaps more important than the overt curriculum in schools is the hidden curriculum by which pupils interpret in differing ways the unconscious messages of their teachers. These messages may concern notions of what behavior and roles are appropriate for women and how work can interact with childbearing careers and family roles.
The effect of employment on fertility depends largely on the nature of the employment and whether it is viewed as a lifetime career or a temporary phase in the life cycle before marriage. The main ways in which employment influences fertility are by delaying marriage, increasing the opportunity costs of the woman's time within marriage, and increasing the costs of childbearing as aspirations increase. The higher costs of childrearing may also be due to the urban residence of parents if they are dependent on an income they will lose if their daughter marries and leaves the labor force. The worker herself may be reluctant to give
up a lifestyle that affords her relative economic and social freedom as compared with married life (Salaff, 1981). However, evidence from factory workers in Sri Lanka shows that employment in low-paid work can accelerate marriage if women are eager to leave a life of tedious factory work at low pay (Rosa, 1989).
Access to cash employment changes consumption patterns or aspirations through increased choice. Easterlin (1978, 1980) argues that as education levels in a society rise so, too, do the opportunity costs of large families. Children essentially become an "old good" as education, cash employment, and the mass media offer new consumption patterns and lifestyles. As aspirations for higher living standards come to predominate, large families become increasingly less attractive. Not only are there alternative demands on family resources, such as improved housing and leisure activities, but also the cost of each child will rise as preferences shift from "quantity" of children to ''quality." For example, the more educated a couple are, the more likely they may be to want their children to be educated, generally to a higher level than they themselves were (see, e.g., Shah, 1986). Education, particularly secondary education, is expensive, and parents with such ambitions for their children may thus have fewer children.
However, employment opportunities may not exert a negative influence on marital fertility in some settings. Where the only opportunities are in low-paid, low-status jobs for young women, the only effect may be to delay marriage, after which fertility may be no different from that of women who had never worked.
Moreover, as with primary education, secondary education does not necessarily work to affect fertility through employment since a significant proportion of educated women may not work. In Bangladesh, where norms opposed to women's cash employment mean that women often work only out of economic necessity, Mahmud (1995) found that 47 percent of secondary-educated women were nonmarket workers (i.e., confined to a reproductive and productive role in their home), and their labor force participation was only 21 percent. Thus the importance of education for employment clearly is not the same in all countries, and care should be taken before assuming that secondary-educated women will enter a career. In economic terms, some women will have invested in cultural rather than economic capital through their secondary education.
Summary And Conclusions
In general, the influence of education on fertility varies greatly between countries with different levels of schooling. Yet it is fair to say that the relationship between education at all levels and fertility has in recent years been, on the whole, negative. Societies in which there is a positive relationship, particularly with small amounts of education, are now relatively rare and confined largely to specific social contexts.
The context in which education is received and in which the woman subsequently lives is fundamental in mediating the effects of education on fertility.
Three aspects of the national context have been hypothesized to be important: the presence or absence of mass education, the strength of the family planning program, and employment opportunities for women. In addition to these, the extent to which education is valued within a society and the extent to which the social structure prevents women from achieving the full economic and social benefits of their schooling should not be overlooked. It is also essential to recognize that all effects of education are influenced by such factors as insufficient opportunities for employment and the quality of the schooling (see Glewwe, this volume).
The importance of the national context applies particularly to the influence on fertility of just a few years of education. The influence is most likely to be negative when there is mass education, which typically means that broad social norms are shifting toward smaller family size. Similarly, if the country has a strong family planning program, it will be relatively easy for women to hear messages about contraception and obtain the means to control their fertility. A few years of education can also influence fertility downward where it gives access to a job that may offer an alternative to early marriage and childbearing. The influence of a few years of education on fertility may be attributable to two primary factors. First, the basic literacy skills that are conferred bring the concept of the written word into the realm of consciousness and enable the individual at the very least, to hear the message about family planning or health practices. Second, basic primary education exposes the individual to new authority figures and ideas. In situations in which there is mass education and societal norms toward low fertility, these new ideas are likely to be consistent with smaller families.
Differences in the influence on fertility of 1 or 2 years of education versus completed primary education depend heavily on context. Where there is mass education, a strong family planning program, and prospects for cash employment as a result of very basic education—for example, in Brazil—2 to 3 years of education can be associated with lower fertility as much as, or more than, completed primary education in countries where these contextual factors do not exist, such as Pakistan. Between these two extremes are countries with varying levels of these contextual factors. For example, in the Middle East, mass education is relatively prevalent, but women's opportunities for cash employment are very poor; in Bangladesh, there is a strong family planning program, but a relative absence of mass female education. Generally, however, completed primary education appears to have a stronger negative relationship with fertility than does incomplete primary education. Furthermore, the relationship between completed primary education and fertility is less dependent on mass education than is that between partial primary education and fertility because completed primary education can act to reduce fertility by making women educational pioneers within their communities. Other effects of primary education on fertility include increased likelihood of employment that may delay marriage and increase the opportunity costs of childbearing. Another effect of women's education, directly
and through employment and urban living, is to increase aspirations for their own children, which in turn increases the costs of childbearing. The result is an incentive for lower fertility and a shift in childbearing aspirations from quantity to quality of children.
The effects of secondary and higher education are probably more universally generalizable since it appears that context is less important at these levels. As a result of skills and knowledge gained from secondary and higher education and the greater prospects that result, women may be better able to make independent decisions and to implement fertility control even in the absence of a family planning program. In addition, women with secondary and tertiary education usually belong to social groups with less traditional norms that generally favor lower fertility. Such women may also belong to a different social group with regard to marriage prospects. In countries with low levels of secondary and higher education, those prospects may be relatively limited because of the tradition that men marry women less educated than themselves and the perception that educated women are too westernized and independent minded. Marriage is therefore delayed, and in some cases may even be foregone.
Employment opportunities for women with secondary education—and the associated influences on fertility, such as later age at marriage—do vary among countries and appear to be closely related to cultural traditions. In South Asia, only a minority of secondary-educated women work for cash, while in Latin America and Southeast Asia, many women work. In the latter cases, secondary education is likely to provide access to more rewarding white-collar employment, such as teaching or a wide range of clerical or semiprofessional occupations that are not particularly compatible with childbearing and perhaps offer an attractive alternative to early marriage and childbearing.
In most developing countries, the number of women with higher education is very small. Therefore, the influence of higher education on national fertility is significant mainly in terms of the extent to which highly educated women act as role models for their less-educated peers and thus contribute to changing social norms with regard to childbearing.
Cleland and Rodriguez (1988:442) argue that "a recognition that fertility behaviour is strongly conditioned by culture, albeit crudely labelled by region or language is an essential first step towards future elucidation." This chapter has shown that context is extremely important in explaining the effects of education on fertility. For women with little education, the social policy and employment context (which is conditioned by culture) is of most importance; for women with higher levels of education, culture and its effects on employment opportunities continue to be important, though to a lesser extent. With low levels of education, a woman's behavior is still very much dependent on community norms, and thus what is important is whether those norms are shifting toward lower fertility. At higher levels of education, the woman's attitudes toward marriage and childbearing become more important, yet her behavior will also be affected by the norms
of the social groups to which she belongs. No individual's behavior can ever be divorced from the social, economic, and cultural context in which she is situated, but as educational attainment rises, the individual moves away from community-based childbearing norms toward a more individual set of beliefs and behaviors and, perhaps, into social networks with lower fertility norms.
The contributions of Sarah Varle and Margaret Newby were funded by United Kingdom Economic and Social Research Council studentships. The authors are grateful to Caroline Bledsoe, Jenna Johnson-Kuhn, and two anonymous referees for extremely helpful comments on the first draft of this chapter.
Afshar, H., and B. Agarwal 1989 Women, Poverty and Ideology in Asia: Contradictory Pressures, Uneasy Resolutions. London: Macmillan.
Amin, S. 1996 Female education and fertility in Bangladesh: The influence of marriage and family. Pp. 184-204 in R. Jeffery and A. Basu, eds., Girls' Schooling, Women's Autonomy and Fertility Change in South Asia. New Delhi: Sage Publications.
Bangladesh Task Forces 1991 Report of the Bangladesh Task Forces, Volume 3, Developing the Infrastructure. Dhaka, Bangladesh: Dhaka University Press Ltd.
Basu, A. 1994 Maternal education, fertility and child mortality: Disentangling verbal relationships. Health Transition Review 4:207-215.
Caldwell, J. 1982 Theory of Fertility Decline. London: Academic Press.
1986 Routes to low mortality in poor countries. Population and Development Review 12:171-220.
1994 How is greater maternal education translated into lower child mortality? Health Transition Review 4:224-229.
Castro Martin, T. 1995 Women's education and fertility: Results from 26 Demographic and Health Surveys. Studies in Family Planning 26(4)187-202.
Castro Martin, T., and F. Juárez 1995 The impact of women's education on fertility in Latin America: Searching for explanations. International Family Planning Perspectives 21(2):52-57.
Choe. M.K. 1996 Changing marriage patterns in Korea with reference to Japan and the United States. Paper presented at the Nihon University International Symposium on Life and the Earth in the 21st Century. Tokyo, Japan. March 4-7.
Cleland, J. 1995 Fertility transition in South Asia. Paper presented at British Society for Population Studies Fertility Transition Meeting. London School of Hygiene and Tropical Medicine. July 1995.
Cleland, J., and G. Kaufmann 1993 Education, fertility and child survival: Unravelling the links. Paper prepared for the International Union for the Scientific Study of Population Committee on Anthropology and Demography Seminar. Barcelona, Spain. November 10-14.
Cleland, J., and G. Rodriguez 1988 The effect of parental education on marital fertility in developing countries. Population Studies 42:419-442.
Cleland, J., and K. Streatfield 1992 The Demographic Transition: Bangladesh. Staff Reference Series: 1/92. UNICEF Dhaka Programme Planning Unit.
Cleland, J., and J. van Ginneken 1988 Maternal education and child survival in developing countries: The search for pathways of influence. Social Science and Medicine 27:1357-1368.
Coggans, N., and S. McKellar 1994 Drug use amongst peers: Peer pressure or peer preference? Drugs: Education, Prevention and Policy 1:15-26.
Cochrane, S.H. 1979 Fertility and Education: What Do We Really Know? Baltimore, Md.: The John Hopkins University Press.
de Vries, R.F. 1992 The Importance of Education for Differences in Female Labour Force Participation and Fertility Behaviour. Postdoctorale Onderzoekersopleiding Demografie [PDOD] Paper No. 13. Amsterdam, Netherlands. University of Amsterdam.
Easterlin, R. 1975 An economic framework for fertility analysis. Studies in Family Planning 6:54-63.
1978 The economics and sociology of fertility: A synthesis. In C. Tilly, ed., Historical Studies of Changing Fertility. Princeton, N.J.: Princeton University Press.
1980 Population and Economic Change in Developing Countries. Chicago, Ill.: National Bureau of Economic Research.
Entwisle, B., and W. Mason 1985 Multilevel effects of socioeconomic development and family planning programs on children ever born. American Journal of Sociology 91(3):616-649.
Jeffery, R., and A. Basu 1996 Schooling as contraception? Pp. 15-47 in R. Jeffery and A. Basu, eds., Girls' Schooling, Women's Autonomy and Fertility Change in South Asia. New Delhi: Sage Publications.
Jejeebhoy, S.J. 1995 Women's Education, Autonomy and Reproductive Behaviour: Experience from Developing Countries. Oxford: Clarendon Press.
Kabir, M., M. Moslehuddin, and A.A. Howlader 1988 Husband-wife communication and status of women as a determinant of contraceptive use in rural Bangladesh. Bangladesh Development Studies 16(1):85-97.
Kiernan, K., and I. Diamond 1983 The age at first birth in the United Kingdom. Population Studies 34:363-380.
King, E.M., and M.A. Hill, eds. 1993 Women's Education in Developing Countries: Barriers, Benefits and Policies. Baltimore, Md.: The John Hopkins University Press.
Lee, Y.J. 1994 Education, employment and residential independence among young adults in Korea 1970-1990. Paper presented at the Conference on Women's Status in Korea. Vancouver, Canada. July 7-10.
Lesetedi, L.T., G.D. Mompati, P. Khulumani, G.N. Lesetedi, and N. Rutenberg 1989 Botswana Family Health Survey 11 1988. Gaborone, Botswana: Central Statistics Office and Ministry of Health; and Columbia, Md.: IRD/Macro Systems, Inc.
LeVine, R.A., S.E. LeVine, A. Richman, F.M. Tapia Uribe, C. Sunderland Correa, and P.M. Miller 1991 Women's schooling and child care in the demographic transition: A Mexican case study. Population and Development Review 17:459-496.
LeVine, R.A., E. Dexter, P. Velasco, S. LeVine, A.R. Joshi, K.W. Stuebing, and F.M. Tapia Uribe 1994 Maternal literacy and health care in three countries: A preliminary report. Health Transition Review 4:186-191.
Mahmud, S. 1995 The Context of Women's Work in Urban Bangladesh: The Need to Move Beyond Purdah. Bangladesh: Bangladesh Institute of Development Studies (mimeographed).
Mason, K.O. 1984 The Status of Women: A Review of Its Relationship to Fertility and Mortality. New York: The Rockefeller Foundation.
Martin, T.C., and F. Juarez 1995 The impact of women's education on fertility in Latin America: Searching for explanations. International Family Planning Perspectives 21:52-57.
Mauldin, W.P., and J.A. Ross 1991 Family planning programs: Efforts and results, 1982-1989. Studies in Family Planning 22:350-367.
Mitra, S.N., M.N. Ali, S. Islam, A.R. Cross, and T. Saha 1994 Bangladesh Demographic and Health Suney 1993-4. Dhaka, Bangladesh: National Institute of Population Research and Training.
Muhuri, P.K., A. Blanc, and S. Rutstein 1994 Socioeconomic Differentials in Fertility. Demographic and Health Surveys Comparative Studies 13. Calverton, Md.: Macro International Inc.
National Council for Population and Development and IRD/Macro Systems Inc. 1993 Kenya Demographic and Health Survey 1993. Columbia, Md.: National Council for Population and Development and IRD/Macro Systems Inc.
Preston, S., ed. 1978 The Effects of Infant and Child Mortality on Fertility. New York: Academic Press.
Rodriguez, G., and R. Aravena 1991 Socioeconomic factors and the transition to low fertility in less developed countries: A comparative analysis. Paper presented at the Demographic and Health Surveys World Conference: Washington D.C. August 5-7.
Rodriguez, G., and J. Cleland 1980 Socioeconomic determinants of marital fertility in twenty countries: A multivariate analysis. WFS Conference Proceedings 2:325-414.
Rosa, K. 1989 Export-oriented industries and women workers in Sri Lanka. Pp. 196-211 in H. Afshar and B. Agarwal, eds., Women, Poverty and Ideology in Asia. Basingstoke, England: Macmillan.
Salaff, J.W. 1981 Working Daughters of Hong Kong: Filial Power or Power in the Family? Asa Rose Monograph Series. Cambridge, England: Cambridge University Press.
Sathar, Z. 1996 Women's schooling and autonomy as factors in fertility change in Pakistan: Some empirical evidence. Pp. 133-149 in R. Jeffery and A. Basu, eds., Girls' Schooling, Women's Autonomy and Fertility Change in South Asia. New Delhi: Sage Publications.
Shah, N.M. 1986 Pakistani Women: A Socioeconomic and Demographic Profile. Honolulu, Hawaii: East-West Center.
Singh, S., and J. Casterline 1985 Socioeconomic determinants. Pp. 199-222 in J. Cleland and J. Hobcraft, eds., Reproductive Change in Developing Countries. Oxford: Oxford University Press.
Steele, F., I. Diamond, and S. Amin 1996 Immunisation in rural Bangladesh: A multilevel analysis. Journal of the Royal Statistical Society Series A 159:289-299.
Streatfield, K. 1989 Fertility Decline in a Traditional Society: The Case of Bali. Canberra, Australia: Australian National University Press.
Stichter, S., and J.L. Parpart, eds. 1990 Women, Employment and the Family in the International Division of Labour. Macmillan International Political Economy Series. Basingstoke, England: Macmillan.
United Nations 1995 Women's Education and Fertility Behavior: Recent Evidence from the Demographic and Health Surveys. New York: United Nations.
1987 Fertility Behaviour in the Context of Development: Evidence from the World Fertility Survey. New York: United Nations.
Westoff, C.F., A.K. Blanc, and L. Nyblade 1994 Marriage and Entry into Parenthood. Demographic and Health Surveys Comparative Studies No. 10. Calverton, Md.: Macro International.