Caroline H. Bledsoe, Jennifer A. Johnson-Kuhn, and John G. Haaga
Over the last century and a half, organized mass education in the Western style has spread across the world. The second half of the twentieth century in particular has seen immense transformations in both the educational and reproductive expectations of young people in the developing world. As of the end of World War II, women in Asia, Latin America, and Africa had few opportunities to attend school and little access to what are commonly referred to as modern methods of fertility control. Today, after two decades of dramatic rises in literacy and gross enrollment ratios (Craig, 1981; Hill and King, 1993; United Nations, 1995:15), formal Western education enjoys a position of global prestige. Throughout virtually every country in the world, Western-style schooling has become the most coveted type of preparation for adult female life. Whether she aspires to life as a secretary, a doctor, or the wife of an urban man with rising prospects, almost any young woman in the world, if she had the choice, would probably go to school.
During the same decades in which this revolution in mass education was creating a new social climate of career possibilities for young women, a reproductive revolution was occurring in many parts of the world: early and rapid childbearing were no longer assumed to be the inevitable or enviable patterns of life. Almost half the married couples in developing nations now use some form of modern contraception. The desired family sizes reported by both women and men in survey responses are notably smaller almost everywhere outside Africa than was the case two decades ago. It is unlikely that these two virtually simultaneous changes in educational opportunities and fertility have no causal links. Yet under what circumstances can a causal relationship in the conventionally
expected direction-—education influencing fertility—actually be identified? Through what mechanisms does it operate?
In cross-country analyses, a generally consistent inverse relationship appears: women with more schooling have lower fertility than those with less (Adamchack and Ntseane, 1992; Ainsworth et al., 1996; Castro Martín, 1995; United Nations, 1995; Jejeebhoy, 1995).1 The assumption of a causal flow from education to fertility has led many governments to support women's education not only to foster human rights and national development, but also to reduce fertility rates, promote the use of modern contraception, and improve child health. Indeed, a pressing need to educate girls for precisely these reasons was one of the strongest messages emerging from the 1994 International Conference on Population and Development in Cairo:
The increase in the education of women and girls contributes to greater empowerment of women, to a postponement of the age of marriage, and to a reduction in the size of families (United Nations, 1994:Ch. XI, Para. 11.3).
This statement is appropriately concise. A Programme of Action agreed upon by more than 180 nations and other signatories is not the right place for nuance, qualification, or counterexample. However, the empirical record does not support the idea that such a simple causal process operates everywhere. Understanding the nature and strength of the relationship between education and fertility remains a central challenge both for scholars seeking to explain demographic and social change and for policy makers who must decide on the allocation of scarce public resources.2 This volume takes up that challenge by bringing together analyses from several research perspectives to reexamine the education-fertility relationship and to rethink conventional lines of logic in the education-fertility paradigm. Although the geographical focus of most of the case studies is Africa, the papers, as well as this introduction, draw from a wider world literature.
Empirical Association Between Education And Fertility
The effects of education on fertility have held enduring interest in demography. Aside from a woman's age and marital status, educational attainment is
probably the variable most frequently included in fertility analyses in developing countries. If the relationship between an individual's education and fertility were consistent across time and place, and if its temporal trajectory at the societal level proceeded in the generally assumed direction—changes in schooling precede changes in fertility (or perhaps, the expectations of reproductive behavior—the matter would appear to be straightforward: because an individual's education is almost always finished by the time childbearing begins, education must lower fertility (or perhaps the expectations of reproductive behaviors) and not vice versa. Of course, the question of who receives schooling would remain pressing, as would the question of why education seems to have this effect on fertility. As noted above, however, the assumption of a simple causal relationship between higher levels of female schooling and lower fertility is not universally supported by the empirical record.
Those who see girls' education as a route to lower fertility can point to countries on both ends of the education-fertility continuum as evidence. Sri Lanka has unusually low fertility rates for a low-income country (total fertility rate [TFR] of 2.3 children per woman) and an unusually high female literacy rate (87 percent of women over age 15). Pakistan, with a much lower literacy rate (24 percent of adult women) has a much higher fertility rate (5.6 children per woman). But there are also anomalous populations in which high female literacy rates coexist with high fertility rates. Jordan, with 86 percent of women literate, has a TFR of 5.6, while in Bangladesh, a relatively low TFR of 3.3 coexists with a low female literacy rate of 26 percent.
In Jejeebhoy's (1995) analyses, the inverse association was found to be monotonic and significant in fewer than half (26) of the countries examined, including only 1 of 21 countries in sub-Saharan Africa. In those countries where the expected negative association did prevail, fertility differences were often small; the negative relationship appeared in regions such as Latin America that have more-developed economies and well-established schooling systems (see also Cochrane, 1979; Cochrane and Farid, 1989; Castro Martin, 1995). The expected pattern held least often where levels of per capita income and literacy were low and where there was gender imbalance in literacy levels (see also Diamond et al., this volume). Cross-sectional analyses of 6 of 26 countries in which Demographic and Health Surveys (DHS) were conducted (Burundi, Kenya, Liberia, Indonesia, Sri Lanka, and Bolivia) even show a reversal of the overall negative association between educational levels and fertility at lower levels of education: TFRs were higher among women with 1 to 3 years of schooling than among those with no schooling (United Nations, 1995:29). In 9 countries, almost all of them in sub-Saharan Africa, women with some education had higher completed fertility than did those with no schooling. Comparing the relationships between education and both fertility and child mortality, Cleland and Kaufmann (1998) report a highly consistent, almost straight dose-response relationship with child mortal-
ity.3 With respect to female fertility, however, they report considerably more variation, and they argue that the fertility-education relationship appears far more context-dependent than is the case for child mortality.
Various explanations for such results have been advanced. In high-fertility contexts, for example, small amounts of education have been hypothesized to undermine lengthy postpartum breastfeeding and sexual abstinence that in previous times would have delayed a new pregnancy (see, e.g., Lesthaeghe et al., 1981:16-21). Alternatively, these unexpected findings may reflect the ability of educated women to achieve their numerical reproductive goals, whether these are goals of high or low fertility (Cleland and Kaufmann, 1998).4
As Diamond et al. (this volume) point out, the most recent DHS surveys show that the inverse relationship between education and fertility is emerging more strongly than ever; most of the contrary cases have begun to fall away. However, the pattern that has emerged has not always followed the temporal order that a straightforward thesis of education causing fertility decline would imply. Rodriguez and Aravena (1991) show, for example, that fertility differentials by education were low in Latin American countries early in the transition, then opened up, then narrowed again during the 1960s and 1970s. In Indonesia, education increased and fertility declined simultaneously (Oey-Gardiner and Sejahtera, 1996), and most of the Asian fertility decline has occurred among populations that are not literate or, for that matter, wealthy or urban (Caldwell, 1993; see also Freedman 1995, on the case of Bangladesh). These findings suggest either that education may be only one of many possible precipitates of fertility decline or that it may co-occur with fertility decline and other societal changes.
As with the effects of formal education, or training, on cognition or fertility,
assertions about the effect of education in conveying accurate information about the biology of fertility must be made cautiously.5 It has been shown (United Nations, 1995:71) that larger proportions of women correctly identify the most fertile period in the ovulatory cycle in countries where average levels of schooling are higher. Yet across countries, the correspondence between knowledge of the fertile period and fertility rates is quite low. In Thailand, with a TFR of 2.2, 13 percent of women identified the middle of the cycle as the most fertile, compared with 44 percent in Sri Lanka, with a TFR of 2.7. In sub-Saharan Africa, where fewer than 10 percent of women in 5 of 10 DHS countries analyzed by the United Nations (1995) identified the middle of the cycle as the most fertile, two of those countries with very low knowledge, Botswana and Zimbabwe, were two of the first three countries in all of Africa to undergo a fertility transition. In Botswana, with the lowest TFR of all the United Nations' 10 African countries, knowledge of the ovulatory cycle was the lowest among all 26 countries analyzed: 3.3 percent.
A number of researchers are careful to point out that education need not necessarily cause a fertility decline (e.g., Cochrane, 1979:7,31-32; Mason, 1993; United Nations, 1995:103; Castro Martín, 1995:194; Ahlburg et al., 1996) or produce behaviors considered conducive to lowering fertility. Roudi (1995:16) observes that in Jordan, which was found in the 1990 DHS to have the most closely spaced births in the Middle Eastern region, women with at least a secondary education, despite their later age at first marriage and higher levels of modern contraception, were more likely to have closely spaced births than those with less education. Analogous patterns have been in observed as well in Taiwan. Freedman (1995:22), pointing to a marked increase in births among educated Taiwanese women who delay marriage and childbearing, argues that ''period measures such as the TFR are likely to exaggerate the fertility decline, which is made up when postponed births occur at a later age." Clearly, Western education, whether or not it played a key role in lowering fertility initially, may now be exerting pressures or offering inducements to keep fertility low. Yet given such variation in observed empirical results, education cannot be a necessary cause for fertility decline at either the individual or the societal level. The question is under what conditions education or some aspect of it might be a sufficient cause.
The following discussion is organized under two rubrics: cognition and social influence. As for the first, there is a long tradition of work behind the demographic view that schooling transforms the way individuals think or gives them access to new information; as educated people, they will more likely want fewer
children and have the personal knowledge or skills needed to be able to implement their desires. Accordingly, education is often hypothesized to have its greatest influence over young women's future reproductive lives by making them significantly different from uneducated women. What evidence is there of such cognitive effects that might bear on fertility?
With regard to social influence, behind many discussions of education as a key variable in fertility research is the Parsonian view of the individual as a socialized actor (Parsons, 1964:207-226). In this view, individuals' actions stem from the values they are imbued with by institutions in the external world. As individuals internalize the values of new social institutions and cultural patterns, they gain skills and behavioral norms that allow them to adopt new practices differing from those of previous generations who were not imbued with these values. This view of the socialized actor pervades descriptions of modernization and its impacts on individuals (see especially Inkeles, 1969). Quoting from Inkeles, Easterlin (1983:563) describes the personality changes that modernization appears to entail as "an increased openness to new experience, increased independence from parental authority, belief in the efficacy of science, and greater ambition for oneself and one's children." Applied to fertility, education is said to break down traditional beliefs and customs, inducing couples to make conscious efforts to limit family size. It may also alter cultural norms that oppose the use of birth control, shift consumer tastes to goods rather than children, present new lifestyle possibilities that compete with children, raise healthcare standards for those children one does bear, and (in general) encourage a problem-solving approach to life (Easterlin, 1983:63).
These speculations about the conventionally expected impact of modernization on individuals make intuitive sense. The problem is that the literature provides little direct evidence for these speculations or, indeed, for the socialized-actor explanation of innovations in fertility behavior. More serious is a logical conundrum concerning the perceptions of those who send children to school: Why would parents enroll their children in school if they know it is a forum where children will be encouraged to reject parental values? The issue is particularly acute in the case of girls: Why would parents enroll their daughters in school knowing that doing so may very well result in rejection of the society's high-fertility values? A number of answers might be advanced. For example: (1) parents are aware of the likely transformation to come and are willing to accept it; (2) whatever changes in behavior schooling entails are not as drastic as outsiders perceive; and (3) parents are willing for some of their daughters, though not necessarily all, to become transformed in the ways they expect schooling to effect.
While individual women are usually the exclusive focus of studies of the education-fertility relationship, the issue of the perceptions not only of the educated individual, but also the wider body of consumers of education raises a set of questions that have received far less attention in fertility studies. These questions
relate to how the groups in which a woman lives and acts affect the relationship between her education and childbearing. Of particular interest are relationships involving generational and conjugal links. To address these questions, we pay particular attention in what follows to the notion of "selection," a term generally used to describe sources of error in demographic statistical calculations. Here we put the notion to a highly sociological purpose in order to understand how young women may be recruited into or withdrawn from groups whose members are more likely to have lower or higher fertility.
Cognition And Social Influence And The Education-Fertility Relationship
Fertility and the Cognitive Effects of Schooling
While education may indeed impart knowledge and skills that bear directly on fertility, research on this specific point is surprisingly sparse. As a result, LeVine et al. (1994:304) point out, the processes by which learning experiences may produce demographic behaviors remain mysterious:
Among the least understood processes are those that mediate the effects of school experience during childhood on the behavior of adult women as mothers and reproductive decision makers in Third World countries. This is the "black box" in survey data linking years spent in school with demographic and health outcomes; it is often covered by speculative assumptions and interpretations. but rarely investigated.
The question of how Western-type schooling may transform thinking is a subject that sparks great controversy among those who study the sociology of education and literacy. Yet it seems quite paradoxical that almost no one in the field of education has displayed any interest in the demographic implications of a phenomenon of such historic importance as the global spread of mass education (cf. LeVine et al., 1991). Indeed, the presumed capability of education to transform individuals in ways that seem so intuitively compelling in the domain of population research are seen in more problematic terms by most of those in the field of international education and literacy (see, e.g., Graff, 1979, 1981, 1987; Heath, 1982; Street, 1984, 1993; Willis, 1977; Wagner, 1983; Meyer et al., 1992; and Daniels and Bright, 1996).
In the domain of studies that see literacy as transformative of society, comparative and historical psychological studies have explored what is regarded as a cognitive divide between literate and nonliterate people whereby the schemata people use to evaluate the world are transformed by the process of education (Goody and Watt, 1963; Ong, 1967, 1982; Clammer, 1976; see Collins, 1995:77-78, for a review of such work). In this framework, writing, because of its clarity, endurance, and precision—attributes that advocates of the position see
as a logical consequence of the medium itself—is described as superior to traditional oral modes of communication for transmitting ideas across space and time (e.g., Goody and Watt, 1963; Goody, 1968, 1977; Goody et al., 1977; Olson et al., 1985). Text and the written word can therefore replace reliance on memory and oral communication because their meaning can be understood independently of context.
In this view, education and its effects on social institutions and cognitive processes are taken as central objects of study: literacy is seen as increasing individualism, writing is seen as creating detachment because it is not context dependent, and literacy is seen as enhancing communication in extending it beyond face-to-face interaction and reducing dependence on memory for storing and retrieving information. Variants of this theory of cognitive modernization regard education, especially in its function of teaching literacy, as fundamentally altering basic forms of thought. These changes are said to allow people to calculate along longer time frames; to distinguish between opinion and truth on the basis of critical, externally informed evaluation; and to pursue autonomous courses of action instead of accepting received tradition or adhering to authority. LeVine and White (1986), for example, suggest that schools provide new decontextualized modes of communication that can be useful in confronting complex bureaucracies, as well as didactic models for childrearing practices. This formulation of language use is not meant to imply that nonliterate cultures are deficient or that people living in them give up traditional aspirations upon adopting Western institutions. Rather, local institutions of socialization create new combinations of indigenous and imported meanings, and schooling allows new modes of communication, new concepts of self-perception, and new kinds of thought.
In recent years, however, numerous critiques of the thesis of the primacy of literacy have emerged. One of the most visible critics, Street (1984:2-3), ". . . treats skeptically claims by western liberal educators for the 'openness,' 'rationality' and critical awareness of what they teach . . . [concentrating] on the overlap and interaction of oral and literate modes rather than stressing a 'great divide."' According to such critics, to believe literacy has qualities that surpass those of oral communication is to neglect the equally logical potential of writing to provide "cultural capital" in the creation of elite status (Bourdieu, 1984) or to obfuscate (Bledsoe and Robey, 1986).
Against the idea that literacy both marks and creates an intellectual or evolutionary divide between those with and without the ability to read (Levy-Bruhl, 1926), most recent studies argue that Western education has no particular monopoly on efficacy or application (e.g., Halverson, 1992; Street, 1993). As Collins (1995:80) notes, Goody (1968) has modified his original stance to distinguish between full and restricted literacy. Goody has also ". . . steadily weakened his claims about literacy and logic, turning instead to arguments about literacy and social organization; Olson's  latest treatment of the subject concedes that
alphabetic literacy is not inherently superior to other scripts, and literacy does not by itself cause cognitive or cultural development. .. ."
In the framework in which the meaning and impact of education are seen as deriving more from context than from inherent qualities, education is seen as imparting special skills, such as reading, that a child may be unable to acquire elsewhere. Yet neither the forms of logical thinking nor the kinds of knowledge that are imparted have an intrinsic advantage over those children might acquire through other modes of training. As McDermott and Peiu (n.d.) observe, most scholarship based on work conducted in the field instead of in the laboratory has by now concluded that it is difficult, if not impossible, to separate the effects of literacy from the social contexts in which literacy arises (see, for example, Lave and Wenger, 1991; Scribner and Cole 1981; Street, 1984; and McDermott and Peiu, n.d; cf. Havelock, 1986). It is instead the political and economic setting, including its existing societal hierarchies, that determines how skills will be evaluated and how they will be put to use. With respect to issues of fertility, these observations imply that students who learn their national language may well become better equipped to acquire family planning/economic resources by employing their language skills or gaining bureaucratic connections. Yet information on issues such as fertility regulation can travel independently of education and even of the national language, if the context demands.6
An example of how the local context shapes the meaning and fertility impact of education (or, more broadly, training) is that of trade apprenticeships in West Africa. Through the work of Guyer (n.d.), the National Research Council (1993) found that although secondary education is generally associated with later first birth for young women in African countries, very different levels of secondary education can be associated with similar patterns of entry into childbearing if other forms of training have delaying effects. Young women in Ghana, a country with vast trade industries and high levels of informal-sector female employment, including an established tradition of female apprenticeship, showed almost the same age distribution at first birth as did young women in Botswana, a country with a strong industrial base, despite very large differences in levels of secondary education (7 percent in Ghana versus 38 percent in Botswana). Such findings on the varied pathways into the reproductive years suggest that the structure of training or its content per se may be less important in delaying the initiation of a sexual and reproductive career than a more general extension of supervised training into young adulthood, combined with certain relationships with a mindful mentor, in ways that discourage early fertility.
Situational factors may even effectively reverse the usual assumptions of
causality implied in the education-fertility relationship. Research on the effect of education on fertility frequently takes as the object of study one person's life course, with a first period characterized by preparation for adulthood and a second by roles that are seen as adult in nature, particularly childbearing. While this ordered division allows formal modeling of schooling as an investment in human capital, it can be misleading. Schoolgirls who get pregnant may be forced by social pressures to leave school, an effect that some might call reverse causation (for the case of Africa, see National Research Council, 1993). In such instances, strong social opinions about the proper timing of childbearing may cause childbearing to influence education, rather than the reverse.
The possibility that context fundamentally determines the meaning and impact of schooling is also seen in power dynamics. One of the most prominent aspects of theory that seeks to explain the inverse relationship between education and fertility is the possibility that education may bestow power on women, who generally express a desire to stop childbearing earlier than do their husbands (e.g., Mason and Taj, 1987). While individuals may acquire special knowledge or skills in school, the mere fact of having attended school may evoke deference from others, a phenomenon analogous to the employment theory notion of credentialism (e.g., Knight and Sabot, 1990). (For discussion of more substantial questions of whether schooling inevitably confers power on women, see Jeffery and Basu  and Shah and Bulatao .) In fact, education may even have negative consequences in the education-fertility relationship: in a society with highly unequal access to literacy, those who lack literacy may experience discrimination. As McDermott and Varenne (1995:341) have noted in a sobering description of the effects of adult literacy programs in the United States, "the more people believe that literacy is cognitively and culturally transformative, the more they can find reasons to degrade those without such powers...."
Finally, a fundamental problem with assuming that education produces modern behavior is the question of what, precisely, constitutes modern behavior. It has been noted repeatedly, for example, that education is strongly related to the use of modern contraceptives, itself one of the strongest correlates of reduced fertility rates (for some recent figures, see United Nations, 1995:69). What has gone virtually unnoticed in such analyses is a clear increase as well with education in the use of "traditional" forms of contraception. Although traditional methods are typically considered less effective than "modern" ones and are thus seldom included in contraceptive analyses, virtually all 26 countries included in the U.N. analysis showed a monotonic increase with education in the use of traditional contraception. With the exception of Zimbabwe, there is more use of traditional contraception among the most educated women (10 or more years) than among women as a whole (United Nations, 1995:74). Education may indeed allow women to adopt rational behaviors that break the bonds of custom and superstition. Yet if modern contraceptive use is a key indicator of rational action, it is not clear why educated women across the developing world should be using
traditional contraceptives more often than women with less education unless "traditional" methods are more "rational" than we believed.
The mystery can be solved in part simply by acknowledging the enduring semantic problems involved in coding and interpreting survey responses. The U.N. figures combine the DHS category of "traditional" methods (abstinence and withdrawal) with that of "folkloric" methods, which includes everything from imbibing herbs to wearing Islamic scriptures around the waist in an amulet pouch. Hence, use of the word "traditional" to describe what is probably largely abstinence and, to a lesser extent, withdrawal with respect to the behaviors of educated women is hiding more than it reveals (see Curtis and Neitzel, 1996, for a similar observation; see especially Tables D.9 through D.11, pp. 86-88). This stark example implies that traditional methods should be taken seriously in analyses of fertility control (see also Santow, 1995, on the importance of withdrawal as a contraceptive method in the European fertility decline). Furthermore, while it is tempting to see as rational the kinds of behaviors that Western schooling is said to induce, few analysts would argue that educated women, though more frequent users of traditional contraception than uneducated women, are more ''irrational" in their reproductive decision making than are their uneducated counterparts. The bottom line, to which everyone would likely ascribe, is that women from all walks of life take steps to stop childbearing or control its timing when they need to do so. How they can do so is determined less by custom than by circumstance and resources.
Social Influences on the Education-Fertility Relationship: Who Is Sent to School and by Whom?
Seen as an instrument of modernity, education is most commonly assumed to bring about lower fertility by inculcating those who acquire it with new ways of thinking and perceiving. Yet social forces intrude upon this relationship at every point. Among the most powerful are those operating across generations. Young children seldom make choices about their own education. The implication is that if educational experiences are important in shaping fertility careers, the educational choices adults make for their daughters are as important as the choices the children eventually make for themselves.7
Among the most subtle yet pervasive domains of social influence is mass
education, whereby all members of the population, including those who have not attended school, may be affected by the changes that a new cultural climate of an educated populace entails (e.g., Caldwell, 1980). Mass schooling clearly changes parents' expectations about the circumstances their children and future grandchildren will confront. Prior to this societal change, having little or no education might suffice to establish one's progeny in a respectable economic situation and secure them advantageous marriages. In the new situation created by mass schooling, this is no longer the case. The paradoxical result is often that young couples in rapidly developing economies perceive a need to educate all their children. Though clearly better off in material terms than their parents, these couples decide they cannot "afford" as many children as their parents somehow afforded a generation earlier (see the discussions of Thai villagers reported by Knodel et al., 1987: Ch. 7). In such situations, education may affect fertility, but hardly by imparting new cognitive skills. The implication is that the education-fertility association may sometimes arise less from a process by which modernity is imprinted on individuals whose student years precede their reproductive careers than from pressures on the preceding generation, for whom the anticipated costs of educating any future children may act to discourage high fertility (see, e.g., Caldwell, 1980, and Axinn, 1993).
For most wealthy countries in the 1990s, parents who do not provide formal schooling (or make equivalent arrangements) for all their children well into the teenage years are considered negligent and are liable for legal prosecution. In less fortunate countries, parents who manage to educate some of their children into the teenage years are either rich or unusually determined. In some circumstances, the children themselves, by showing exceptional talent, may go further in school than their parents had planned. This means the characteristics of the situation that result in certain children attaining more schooling and others less may well lead to lower fertility, independently of what actually happens in school.
Situations such as this, in which levels of two variables are affected by unobserved characteristics of individuals or groups that are known to those involved but not to the researcher, are termed "selection" effects. Invisible selection effects can render groups noncomparable and lead to spurious correlations. Prominent examples come from the realms of employment and marriage (Heckman, 1976; Smith, 1980; Axinn and Thornton, 1992). Yet beyond its application to the question of sampling errors, the notion of selection offers a sociological handle for addressing the question of education and fertility in the developing world, insofar as it may reflect the outcome of intentional social actions to recruit certain people into, or filter others out of, one pool or another. Such processes lie at the heart of Guyer's (1996) work on life pathways in the development of "singularity" in Equatorial Africa, in which youthful abilities are constantly being cultivated for different career path potentials in a volatile economy (see also Guyer, 1993, and Guyer and Eno Belinga, 1995). Parents may choose formal schooling for children with certain attributes, such as intellectual
aptitude or a willingness to follow instructions. Other children may be seen as suited for local alternatives to schooling, such as trade apprenticeships.
Across the world, unequal access to higher levels of education, whether by gender, by perceived ability, or by class (e.g., Knodel and Jones, 1996), is the rule rather than the exception. In national education systems that offer limited numbers of places on a competitive basis, decisions must be made to invest in the education of certain children and not others. Such pressures strongly shape the outcomes of competition for places in secondary schools, a process Bourdieu and Passeron (1971) vividly document for the selection and cultivation of members of the French elite. Individuals may or may not be transformed by processes of pedagogy and communication in education, but they clearly are measured at a point in time after their different unobservable characteristics affected their selection into or out of the pool of continuing students.
Whether this broadened view of selection refers to a passive process of being filtered out of a particular group by forces beyond one's control or to a process of entering another group by individual choice (or by some combination of the two), it captures the situation in many contemporary developing countries in which the number of women who can complete a certain level of education is sharply reduced at each successive level. Pressures exerted by social class, notions of propriety, or academic ability that influence which individual women will continue in school and which will be withdrawn for early marriage and motherhood may mean that reproduction effectively influences education. Such forces of selection are particularly important in light of the vast variation in the quality and cost of schooling both within and across countries. A family may send an academically gifted daughter to an expensive school with high academic standards: she is thus selected to attend a school whose students are likely to delay fertility because they are firmly established in a professional career track. A girl whose character or talents are not judged worthy, however, may not be enrolled in school at all or may be withdrawn as soon as marriage becomes possible. Her parents may even seek to avoid the very outcomes that schools are said to promote—female autonomy, shifts of loyalty from the family toward the outside world, late marriage, and low fertility.
The possibility that observed sortings may represent outcomes of intentional social actions implies that the removal by families of certain women from the elite pool of continuing students may even occur with the intent that their future life will be channeled toward high fertility. Similar forces of selection arise with respect to admission standards for students seeking transfer to a new school or reentry after a hiatus, issues of singular importance to women who seek to resume their schooling after childbirth. They may also apply to situations involving girls from low-status families, children from large families, or children whose births were unwanted.
Besides the complications posed for the education-fertility relationship by selection and associated generational forces are those involving the dynamics of
marriage. Women's choices about schooling and childbearing are inevitably shaped by the kinds of support available to them from men who may provide them with economic assistance, father their children, or protect them from other men. A variety of studies have shown that educated African women marry later; marry more educated men; and, at high levels of education, are less likely to marry (Gage-Brandon, 1993; Stambach, 1996). Yet powerful selection factors may affect fertility. The husbands of educated women, as Basu (this volume) points out for India, differ from those of uneducated women; as such they represent a distinct selected subset of men. This means that a certain fertility dynamic has been set up well before the women themselves take any reproductive decisions.
The forces by which young women are drawn into groups characterized by certain patterns of reproduction have been thinly documented and conceptualized. All deserve greater attention in both empirical research and theory building.
Overview Of This Volume
The chapters that follow contribute to our understanding of the above associations between education and fertility. Chapter 2, by Ian Diamond, Margaret Newby, and Sarah Varle, summarizes some recent empirical evidence, primarily from the DHS on the education-fertility relationship. The authors review a broad array of studies, positing that while primary education may affect fertility indirectly, by mediating the effect of various factors, secondary and higher education may influence fertility more directly by making people "more able to make independent decisions based on an assessment of the likely costs and benefits of different actions." They also ask, however, how the same measure of schooling can lead to very different fertility outcomes depending on the social, economic, and political circumstances. They identify multiple possible pathways to explain the influence of education on fertility: employment/opportunity costs, the nature of marriage, familiarity with bureaucratic institutions, and reference communities.
Most studies of education and fertility see schooling as imbuing students not with unthinking adherence to what they are taught, but with the ability to evaluate information and problems for themselves, and in particular to break loose from traditional beliefs. In Chapter 3, however, Anthony Carter shows that such models see students as passive recipients of educational messages, with education imprinting itself upon them in some way that will induce them to bear fewer children. If this is so, he shows, the students themselves need no discussion; they are the bearers of family planning messages and followers of new reproductive expectations. Drawing on theory from linguistic anthropology, Carter describes a model of schooling as social practice in which education and childbearing represent transformative moments in a life course shaped by the communities of practice in which the individual, by virtue of his or her educational experience,
becomes a participant. In this model, schools do not simply transmit ideas or cause women to bear fewer children through some series of prior or external forces. Rather, students actively select from among disparate pieces of knowledge from many different sources, including but not exclusive to the classroom, for use in different contexts.
Although education may in some way affect fertility, direct evidence on this question remains thin. Four of the chapters, those by Cynthia Lloyd and Barbara Mensch, Paul Glewwe, Duncan Thomas, and Bruce Fuller and Xiaoyan Liang, make significant contributions to the question of what specific forces may bring about such results and through what channels their effects may be felt. In Chapter 4, Lloyd and Mensch, asking what young women learn or experience in school that may bear on their entry into parenthood, review a wide range of literature on schooling and adolescence from the developing world, including their own recent study in Kenya. They examine the transition to adulthood for young women in terms of the possible links between schooling and low fertility, and shed light on the potentially fertility-relevant content of schooling. Their chapter brings together work on a variety of factors that may facilitate or constrain continued schooling for girls. Among these factors are school policies regarding marriage and pregnancy, school quality and academic achievement, gender-biased differences in study time, punishments, material investments, curriculum, and teacher attitudes.
In Chapter 5, Glewwe addresses the issue of comparability among schools. He suggests that if schools have a significant impact on demographic change, the specific cause of that impact is far less identifiably one of cognitive development among students than a simple measure of years of school attended or highest grade completed would imply. The use of such measures assumes that schools provide universally some comparable "dose" of learning in a year; school quality is treated as a constant and as having no bearing on parental choices regarding the length of their children's schooling. Using a neoclassical economic model, Glewwe demonstrates that a simplistic model of investment fails to explain parents' educational choices. He finds that parents are likely to invest in more schooling for their children if they perceive the available school to be of high quality. Because students will probably attend longer the kinds of schools likely to provide them with more cognitive skills, this observation confounds a presumed relationship between years of schooling and the amount of knowledge students acquire. Glewwe's work makes clear the need to ask why parents educate their children for certain numbers of years. His work also suggests that school quality cannot be taken as given; rather, empirical research is needed on what occurs in schools.
In Chapter 6, Thomas examines important empirical evidence concerning a possible mechanism through which the negative association between education and fertility might develop. He finds that among women in South Africa, reading comprehension skills are strongly associated with lower fertility, quite indepen-
dently of earnings (see also Oliver, in press). Thomas is thus able to argue that what makes a difference for fertility is not schooling in general, or even its duration, but specific comprehension skills learned in school. He argues that these results lend support to the thesis that women with stronger comprehension skills may be able to acquire and incorporate family planning information, although the specific mechanisms may vary across contexts. At the same time, Thomas' data reveal that cognitive skills alone cannot explain the variations in fertility. Women who leave school at natural breaks, such as those between primary and secondary school, have lower fertility than would be predicted if the schooling-fertility relationship followed a linear dose-response pattern. Thomas suggests that the women who leave school at natural breaks differ in systematic ways from women who leave at other times, meaning that processes of selection out of school may constitute a significant factor in the education-fertility relationship.
In Chapter 7, another study of South Africa, Fuller and Liang address the key question of who receives schooling. Using data from three South African provinces, they find that personal, family, and community characteristics all influence the likelihood that girls will stay in school. They also find support for a family model of household economics in which short-term economic resources and social obligations, rather than long-term investment strategies, predict the school attainment of eldest daughters. Their findings suggest that the community of educated women is both shaped and differentiated from that of uneducated women in ways that predate their schooling.
As for the possible effects of generational ties on the education and fertility relationship, Chapter 8, by Mark Montgomery and Cynthia Lloyd, offers an intriguing analysis of what might be labeled reverse causation through the notion of the "wantedness" of a child or a birth. Using DHS data, they find instances in which educational outcomes may be shaped by "exogenous" fertility at least as heavily as fertility is affected by education. They argue that in some instances, unintended or excess fertility may lead parents to withdraw children from school or to enroll them late. Their analysis also underscores the value of a situational view of childbearing and education choices, in which parents constantly revise their choices on the basis of changing circumstances. By focusing on births that women report they did not want, Montgomery and Lloyd transcend the simplistic assumption that any quantity-quality trade-off made by parents in childbearing decisions is a decision "made once-and-for-all, generally at the beginning of the reproductive lifespan" (Greenhalgh, 1995:22). Their findings suggest that the immediate constellation of the family is critical in this process.
Much research on education and fertility focuses on how schooling affects women's participation in labor markets. Yet marriage markets also figure critically in a couple's fertility. Schooling transforms or bolsters social positions in ways that affect profoundly who ends up marrying whom; thus the most important fertility decisions may be settled for all intents and purposes before the
couple actually meets. Certainly in the case of South Asia, where so many marriages continue to be arranged, the logic of assuming that schooling gives women power that might allow them to counter men's high-fertility demands must be weighed against the realization that most women in developing countries have partners who are more educated than they are. Thus, even if a woman has power conferred on her by education, her husband is likely to have more education than she does, meaning that the most powerful partner in a union is very likely to remain the man. The power of marriage markets in affecting subsequent fertility outcomes is central to Alaka Basu's analysis in Chapter 9. Countering the idea that educated women choose to reduce their fertility independently of their husbands, she argues that in India, men's education is nearly universally desired, so it cannot serve as a marker of particular values or preferences. However, men who choose to marry educated women differ systematically from other men. Coming from families that value women's education, these men were raised to seek the same in a spouse. Men who marry educated women, therefore, are largely in agreement with the low fertility aspirations of their educated wives well before the marriage.
In a concluding review in Chapter 10, Parfait Eloundou-Enyegue notes the parallels and contrasts in some of the themes of this volume with regard to questions raised 20 years ago in Cochrane's (1979) seminal book, Fertility and Education: What Do We Really Know?. In contrast to Cochrane's work, he attributes the cautions expressed here about how education may affect fertility to several factors. These factors include changes in demography as a discipline, its broadened geographical coverage from which generalizations must be drawn, its increasing sophistication as a social science, and the expansion of theoretical and statistical demographic models to grapple with the challenges posed by increasingly rich and complex data sources. As his observations reveal, the seemingly simple question of Cochrane's subtitle has grown into a set of deeper challenges to the field.
Adamchak, D., and P. Gabo Ntseane 1992 Gender, education and fertility: A cross-national analysis of Sub-Saharan nations. Sociological Spectrum 12:167-182.
Ahlburg, D.A., A.C. Kelley, and K.O. Mason, eds. 1996 The Impact of Population Growth on Well-being in Developing Countries. New York/ Berlin, Germany: Springer-Verlag.
Ainsworth, M., K. Beegle, and A. Nyamete 1996 The impact of women's schooling on fertility and contraceptive use: A study of 14 Sub-Saharan African countries. World Bank Economic Review 10(1):85-122.
Axinn, W.G. 1993 The effects of children's schooling on fertility limitation. Population Studies 47(3):481-493.
Axinn, W.G., and A. Thorton 1992 The relationship between cohabitation and divorce: Selectivity or causal influence? Demography 29(3):357-374.
Blake, J. 1981 Family size and the quality of children. Demography 18(4):421-442.
Blanc, A., and C. Lloyd 1990 Women's Childrearing Strategies in Relation to Fertility and Employment in Ghana. Policy Research Division Working Paper No. 16. New York: The Population Council.
Bledsoe, C., and K. Robey 1986 Arabic literacy and secrecy among the Mende of Sierra Leone. Man 21:202-226.
Bledsoe, C., A. G. Hill, U. d'Alessandro, and P. Langerock 1994 Constructing natural fertility: The use of Western contraceptive technologies in rural Gambia. Population and Development Review 20(1):81-113.
Bourdieu, P. 1984 Distinction: A Social Critique of the Judgment of Taste. Cambridge, Mass.: Harvard University Press.
Bourdieu, P., and J-C. Passeron 1971 Reproduction in Education, Society and Culture. London: Sage Publications.
Caldwell, J.C. 1980 Mass education as a determinant of the timing of fertility decline. Population and Development Review 6(2):225-255.
1993 The Asian fertility revolution: its implications for transition theories. Pp. 299-316 in R. Leete and I. Alam, eds., The Revolution in Asian Fertility: Dimensions, Causes, and Implications. Oxford: Clarendon Press.
Castro Martin, T. 1995 Women's education and fertility: Results from 26 Demographic and Health Surveys. Studies in Family Planning 26(4): 187-202.
Clammer, J.R. 1976 Literacy and Social Change: A Case Study of Fiji. Leiden: Brill.
Cleland, J., and G. Kaufmann 1998 Education, fertility and child survival: Unraveling the links. In A. Basu and P. Aaby, eds., The Methods and Uses of Anthropological Demography. Oxford: Clarendon.
Cochrane, S.H. 1979 Fertility and Education: What Do We Really Know? Baltimore, Md.: Johns Hopkins University Press.
Cochrane, S.H., and S.M. Farid 1989 Fertility in Sub-Saharan Africa: Analysis and Explanation. World Bank Discussion Paper, No. 43. Washington, D.C.: The World Bank.
Collins, J. 1995 Literacy and literacies. Annual Review of Anthropology 24:75-93.
Craig, J. 1981 The expansion of education. Review of Research in Education 8:151-210.
Curtis, S.L., and K. Neitzel 1996 Contraceptive Knowledge, Use, and Sources. Demographic and Health Surveys Comparative Studies No. 19. Calverton, Md.: Macro International, Inc.
Daniels, P., and W. Bright, eds. 1996 The World's Writing Systems. Oxford: Oxford University Press.
Delancy, V.H. 1980 The Relationship Between Female Wage Employment and Fertility in Africa: An Example from Cameroon. Unpublished Doctoral Dissertation. Department of Economics, University of South Carolina.
Desai, S., and S. Alva 1998 Maternal education and child health: Is there a strong causal relationship? Demography 35(l):71-81.
Easterlin, R.A. 1983 Modernization and fertility: A critical essay. Pp. 562-586 in R.A. Bulatao and R.D. Lee, eds., Determinants of Fertility in Developing Countries. Volume 2. New York: Academic Press.
Eloundou-Enyegue, P. 1994 Why Trade Quantity for Child Quality? A "Family Mobility" Thesis. Working Paper 94-15. University Park, Pa.: Pennsylvania State University Population Research Institute.
Freedman, R. 1995 Asia's recent fertility decline and prospects for future demographic change. Asia-Pacific Population Research Reports, No. 1.
Gage-Brandon, A.J. 1993 The formation and stability of informal unions in Côte d'Ivoire. Journal of Comparative Family Studies 24(2):219-233.
Goody, J. 1968 Introduction. Pp. 1-26 in J. Goody, ed., Literacy in Traditional Societies. Cambridge: Cambridge University Press.
1977 The Domestication of the Savage Mind. Cambridge: Cambridge University Press.
Goody, J., and I. Watt 1963 The consequences of literacy. Comparative Studies in Society and History 5:304-345.
Goody, J., M. Cole, and S. Scribner 1977 Writing and formal operations: A case study among the Vai. Africa 47:289-304.
Graff, H.J. 1979 Literacy, education and fertility, past and present: A critical review. Population and Development Review 5(1): 105-140.
1981 Literacy and Social Development in the West. Cambridge: Cambridge University Press.
1987 The Legacies of Literacy. Bloomington, Ind.: Indiana University Press.
Greenhalgh, S. 1995 Anthropology theorizes reproduction: Integrating practice, political economic, and feminist perspectives. Pp. 3-28 in Situating Fertility: Anthropology and Demographic Inquiry. Cambridge: Cambridge University Press.
Guyer, J.I. n.d. The economics of adolescent fertility in Africa. Background paper for the National Research Council Report on the Social Dynamics of African Population. Working Group on the Social Dynamics of Adolescent Fertility in Sub-Saharan Africa, Panel on the Population Dynamics of Africa , National Academy of Sciences. Washington, D.C.
1993 Wealth in people and self-realization in Equatorial Africa. Man 28:243-265.
1996 Traditions of invention in Equatorial Africa. African Studies Review 39(3): 1-28.
Guyer, J., and S.M. Eno Belinga 1995 Wealth in people as wealth in knowledge: Accumulation and composition in Equatorial Africa. Journal of African History 36:91-120.
Halverson, J. 1992 Goody and the implosion of the literacy thesis. Man (NS) 27:301-317.
Havelock, E.A. 1986 The Muse Learns To Write: Reflections on Orality and Literacy from Antiquity to the Present. New Haven, Conn.: Yale University Press.
Heath, S. 1982 Ways with Words: Language, Life, and Work in Communities and Classrooms. New York: Cambridge University Press.
Heckman, J. 1976 The common structure of statistical models of truncation, sample selection and limited dependent variables and a simple estimator for such models. Annals of Economic and Social Measurement 5(4):475-492.
Hill, M.A., and E.M. King 1993 Women's education in developing countries: An overview. Pp. 1-50 in E.M. King and M.A. Hill, eds., Women's Education in Developing Countries: Barriers, Benefits, and Policies. Baltimore, Md.: The Johns Hopkins University Press.
Inkeles, A. 1969 Making men modern: On the causes and consequences of individual change in six developing countries. American Journal of Sociology 75:208-225.
Jeffery, R., and A.M. Basu 1996 Schooling as contraception? Pp. 15-47 in R. Jeffery and A.M. 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 Behavior: Experience from Developing Countries. Oxford: Clarendon Press.
Knight, J.B., and R.H. Sabot 1990 Education, Productivity, and Inequality: The East African Natural Experiment. Oxford: Oxford University Press.
Knodel, J., and G.W. Jones 1996 Post-Cairo population policy: Does promoting girls' schooling miss the mark? Population and Development Review 22(4):683-702.
Knodel, J., A. Chamratrithirong, and N. Debavalya 1987 Thailand's Reproductive Revolution: Rapid Fertility Decline in a Third-World Setting . Madison, Wisc.: University of Wisconsin Press.
Lave, J., and E. Wenger 1991 Situated Learning: Legitimate Peripheral Participation. Cambridge: Cambridge University Press.
Lesthaeghe, R., P.O. Ohadike, J. Kocher, and H.J. Page 1981 Child-spacing and fertility in sub-Saharan Africa: An overview of the issues. Pp. 3-23 in H.J. Page and R. Lesthaeghe, eds., Child-Spacing in Tropical Africa: Traditions and Change. London: Academic Press.
Levy-Bruhl, L. 1926 How Natives Think. Princeton, N.J.: Princeton University Press.
LeVine, R.A., S.E. LeVine, A. Richman, F.M. Tapia Uribe, C.S. 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(3):459-496.
LeVine, R.A., S.E. LeVine, A. Richman, F.M. Tapia Uribe, and C.S. Correa 1994 Schooling and survival: The impact of maternal education on health and reproduction in the Third World. Pp. 303-338 in L. Chen, A. Kleinman, and N.C. Ware, eds., Health and Social Change in International Perspective. Boston, Mass.: Harvard University Press.
LeVine, R., and M. White 1986 Human Conditions: The Cultural Basis of Educational Developments. New York: Kegan Paul.
Mason, K.O. 1993 How female education affects reproductive behavior in urban Pakistan. Asian and Pacific Population Forum 6(4):93-103.
Mason, K.O., and A.M. Taj 1987 Differences between women's and men's reproductive goals in developing countries. Population and Development Review 13(4):611-638.
McDermott, R. and H. Varenne 1995 Culture as disability. Anthropology & Education Quarterly 26(3):324-348.
McDermott, R., and A. Peiu n.d. The cognitive consequences of Greek literacy. Unpublished manuscript.
Meyer, J.W., D.H. Kamens, and A. Benavot 1992 School Knowledge for the Masses: World Models and National Primary Curricular Categories in the Twentieth Century. Studies in Curriculum History Series: 19. Washington, D.C.: The Falmer Press.
Mueller, E., and K. Short 1983 Effects of income and wealth on demand for children. Pp. 590-642 in R.A. Butalao and R.D. Lee, eds., Determinants of Fertility in Developing Countries. New York: Academic Press.
National Research Council 1993 The Social Dynamics of Adolescent Fertility in Sub-Saharan Africa. Washington, D.C.: National Academy Press.
1998 From Death to Birth: Mortality Decline and Reproductive Change. Washington, D.C.: National Academy Press.
Oey-Gardiner, M., and I.H. Sejahtera 1996 Indonesian girls, too, get at least some education. Paper presented at National Research Council workshop on Education and Fertility in the Developing World. Washington D.C., February 29 - March 1, 1996.
Oliver, R. In press Fertility and women's schooling in Ghana. In P. Glewwe, ed., The Economics of School Quality Investments in Developing Countries: An Empirical Study of Ghana. United Kingdom: Macmillan.
Olson, D.R. 1994 The World on Paper: The Conceptual and Cognitive Implications of Writing and Reading. Cambridge: Cambridge University Press.
Olson, D.R., N. Torrance, and A. Hildyard 1985 Literacy, Language, and Learning: The Nature and Consequences of Reading and Writing. Cambridge: Cambridge University Press.
Ong, W. 1967 The Presence of the Word. New Haven, Conn.: Yale University Press.
1982 Orality and Literacy: The Technologization of the Word. London: Methuen Press.
Parsons, T. 1964 The Social System. New York: The Free Press.
Pitt, M. 1995 Women's Schooling, the Selectivity of Fertility, and Child Mortality in Sub-Saharan Africa. Washington, D.C.: The World Bank.
Rindfuss, R.R., and K.L. Brewster 1996 Childrearing and fertility. Population and Development Review 22(Suppl.):297-324.
Rodriguez, G., and R. Aravena 1991 Socioeconomic factors and the transition to low fertility in less developed countries: A comparative analysis. Pp. 39-72 in Demographic and Health Surveys World Conference Proceedings, Volume 1. Columbia, Md.: Institute for Resource Development/Macro International.
Roudi, F. 1995 An Analysis of Birth Spacing in the Near East. Prepared for the Asia/Near East Bureau of USAID by the OPTIONS Project. Washington, D.C.: Population Reference Bureau.
Santow, G. 1995 Coitus interruptus and the control of natural fertility. Population Studies 49(l):19-43.
Schultz, T.P. 1993a Investments in the schooling and health of women and men: Quantities and returns. Journal of Human Resources 28(4):694-725.
1993b Returns to women's education. Pp. 51-99 in E.M. King and M.A. Hill, eds., Women's Education in Developing Countries: Barriers, Benefits, and Policies. Baltimore, Md.: The Johns Hopkins University Press.
Scribner, S., and M. Cole 1981 The Psychology of Literacy. Cambridge, Mass.: Harvard University Press.
Shah, N., and R. Bulatao 1981 Purdah and family planning in Pakistan. International Family Planning Perspectives 7:32-36.
Smith, J.P., ed. 1980 Female Labor Supply: Theory and Estimation. Princeton, N.J.: Princeton University Press.
Stambach, A. 1996 'Seeded' in the market economy: Schooling and social transformations on Mount Kilimanjaro. Anthropology and Education Quarterly 28(4):545-567.
Street, B. 1984 Literacy and Theory in Practice. Cambridge: Cambridge University Press.
1993 Cross-Cultural Perspectives on Literacy. Cambridge: Cambridge University Press.
United Nations 1994 International Conference on Population and Development (ICPD) (New York). Report of the International Conference on Population and Development (Cairo, 5-13 September 1994). No. A/Conf.171/13, October 18, 1994. New York: United Nations.
1995 Women's Education and Fertility Behaviour: Recent Evidence from the Demographic and Health Surveys. New York: United Nations.
Wagner, D.A., ed. 1983 Child Development and International Development: Research-Policy Interfaces. San Francisco, Calif.: Jossey-Bass.
Willis, P. 1977 Learning to Labour: How Working-Class Kids Get Working-Class Jobs. London: Routledge & Kegan Paul.
Willis, R.J. 1973 A new approach to the economic theory of fertility behavior. Journal of Political Economy 81(2, pt. 2):S14-S64.