9
Women's Education, Marriage, and Fertility in South Asia: Do Men Really Not Matter?

Alaka Malwade Basu

Introduction

The quest for pathways in the relationship between female education and fertility in developing countries has run the gamut of the proximate determinants proposed by Davis and Blake (1956). Some of these proximate determinants, such as the length of breastfeeding or of postpartum abstinence, have been found to be related to education in the direction of potentially increased fertility. These findings have led to much speculation in the literature on the surprising magnitude of those variables which act in the direction of reducing fertility, since the net impact of education on fertility is generally negative.1 Thus, for example, given that birth intervals are often shorter for educated women, we are impressed by the role of modern contraception, which seems not only to compensate for the resultant increase in fecundity, but also to reduce fertility to levels lower than those seen with longer birth intervals.

Now that the negative relationship between female education and fertility has been clearly established, further research on this issue, at least in demography, seems to consist of doing more of the same, albeit with bigger and better data sets. The uniformity is particularly striking in the way some of the proximate determinants of fertility are conceived and interpreted. While such uniformity of definition and analysis is understandable in the case of single-component

1  

The finding that educated women want lower family sizes and seem to be better able to achieve such lower fertility is remarkably universal in a range of fertility surveys and smaller studies across the world (for a review, see Jejeebhoy, 1995; see also Diamond et al., this volume).



The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement



Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.

OCR for page 267
--> 9 Women's Education, Marriage, and Fertility in South Asia: Do Men Really Not Matter? Alaka Malwade Basu Introduction The quest for pathways in the relationship between female education and fertility in developing countries has run the gamut of the proximate determinants proposed by Davis and Blake (1956). Some of these proximate determinants, such as the length of breastfeeding or of postpartum abstinence, have been found to be related to education in the direction of potentially increased fertility. These findings have led to much speculation in the literature on the surprising magnitude of those variables which act in the direction of reducing fertility, since the net impact of education on fertility is generally negative.1 Thus, for example, given that birth intervals are often shorter for educated women, we are impressed by the role of modern contraception, which seems not only to compensate for the resultant increase in fecundity, but also to reduce fertility to levels lower than those seen with longer birth intervals. Now that the negative relationship between female education and fertility has been clearly established, further research on this issue, at least in demography, seems to consist of doing more of the same, albeit with bigger and better data sets. The uniformity is particularly striking in the way some of the proximate determinants of fertility are conceived and interpreted. While such uniformity of definition and analysis is understandable in the case of single-component 1   The finding that educated women want lower family sizes and seem to be better able to achieve such lower fertility is remarkably universal in a range of fertility surveys and smaller studies across the world (for a review, see Jejeebhoy, 1995; see also Diamond et al., this volume).

OCR for page 267
--> variables such as intercourse frequency,2 it seems unnecessary to reduce other variables to such mechanical numbers.3 In particular, a narrow perspective continues to characterize the demographer's treatment of marriage as an intermediate variable in the relationship between socioeconomic factors and fertility. The timing of marriage (and in more sophisticated versions, the start of a sexually active life) is taken as the only way the fact of marriage affects fertility, whether natural or controlled. This is especially the case for studies on Asia.4 That is, most analyses that use marriage as an intermediate variable to study fertility are concerned with the question of differentials in when women marry, and not at all with the question of whom they marry.5 In addition, because female education continues to have a strong relationship with fertility even when the standard household variables (of income, occupation, and so on) are controlled, the explicit conclusion in the literature is that educated women have a single-handed role in their lower fertility. This conclusion in turn is taken to suggest that women's autonomy is important to fertility decline (a proposition first put forth by Dyson and Moore, 1983). This latter conclusion is 2   Even here, however, the imaginative researcher can be concerned with more aspects of this variable than the simple presence or absence of a particular act of sexual intercourse. 3   This is not to say that demographic research is unaware of the many aspects of a single variable. Breastfeeding in particular has now undergone detailed scrutiny, and we know it cannot be related to fertility in terms of simple measures such as its duration; we must also note whether it is total or partial, disciplined or available on demand, and so on. 4   Mainstream demography has long been interested in the interrelationships among different forms of marriage (polygyny in particular) in sub-Saharan Africa and a few local communities elsewhere; in addition, there is a small literature, focused on the developed world, on the implications of marriage dissolution and reformation for fertility. But for Asia, where marriage is relatively stable and universal (and perhaps because it is seen to be stable and universal), the only measure of variation deemed interesting is that of its timing. 5   There are too many studies with this focus on the relationship between female education and age at marriage to make it possible to provide a few representative references. See, for example, the scores of references to the relation between education and the age at marriage in a recent comprehensive review by Jejeebhoy (1995). This book itself also addresses the intermediate role of marriage almost entirely in the context of its timing. There is, admittedly, increasing awareness of one other dimension of marriage in semi-anthropological demographic research specific to Asia: the question of whether modernization or education or other factors associated with lower fertility are also associated with an increase in "love" versus "arranged" marriages (see, for example, Jeffery and Jeffery, 1993; Caldwell, 1996). But this notion of love marriages has yet to catch on in the standard demographic literature, as the brief reference to it in Jejeebhoy's review suggests. In the literature on sub-Saharan Africa, there is also an acknowledgment of the impact of education on the kind of marriage women enter into—in particular the impact of female education on polygynous marriages. In addition, there is some interest in the effect of education on interspousal factors, such as the age gap between husbands and wives. Such analyses, if extended to look more carefully at the husbands of educated women, and especially in regions outside Africa as well, should move the debate in demography on the relation between female education and marriage much further.

OCR for page 267
--> buttressed by the several studies that do indeed find educated women have more freedom in decision making and action on a range of domestic and extradomestic matters (see, for example, Basu, 1992; Vlassoff, 1996; Morgan and Niraula, 1995; and the reviews in Jejeebhoy, 1995, and Sathar, 1996). But from this conclusion (that educated women have greater autonomy as well as lower fertility), an implicit conclusion that has much less empirical basis is drawn: that educated women have greater reproductive autonomy than uneducated women. And a further implicit conclusion follows: that there is somehow an intrahousehold conflict in reproductive preferences that is resolved in the woman's favor when she is educated and therefore has greater autonomy in reproductive decision making. Marriage is brought into this framework in a somewhat confusing way. There is a direct effect of later marriage on natural fertility, of course, simply because of the reduced period of exposure to the risk of pregnancy and childbearing. But later marriage is also believed to reduce the demand for children through the woman's increased premarital exposure to the kind of culture and ideas that encourage lower fertility. In addition, later marriage is believed to make it easier to achieve these lower fertility goals through increased access to and autonomy in using birth control information and services. At the same time, these are all things that education is believed to do directly as well, so that the educational effect through later marriage is primarily in reducing the period of exposure to potential childbearing. Jejeebhoy (1995) concludes that in this sense, the delayed marriage associated with women's education has an "unintended" impact on fertility in that educated women do not delay marriage as a means of reducing their fertility. This chapter attempts to isolate only one strand of the female education-marriage-fertility relationship. It argues that aspects of marriage other than its timing are relevant to fertility. Quite apart from whether educated women marry early or late is the question of the kind of men educated women marry. Are these the kind of men whose reproductive goals need to be changed or overruled by their educated wives? That is, is reproductive autonomy an essential ingredient of the education-fertility relation? The chapter suggests that it is not. The marriage market ensures that educated women will find the kind of husbands who share their reproductive preferences and that intrahousehold gender inequalities in daily life need not imply gender inequalities in reproductive goals. Analyses of the female education-fertility relation that focus only on the timing of marriage and therefore its effects on the attitudes and abilities of the woman assume that while there is something special about the educated woman, there is nothing special about the educated man. This chapter argues that there is something special about the educated man who marries an educated woman, even if there is much less to distinguish the educated man in general than the educated woman. Merely by marrying an educated woman, this man is saying something

OCR for page 267
--> about himself; he is not just a random educated man who then has his attitudes and preferences molded by his educated spouse. This hypothesis is not inconsistent with the empirical finding that the wife's education has a much clearer relation to fertility than does the husband's education. The special feature of the husband of the educated woman is not his education, but rather other characteristics that standard surveys have not attempted to capture.6 Women's Education And The Marriage Market Assume four categories of marriageable individuals: less-educated men, more-educated men, less-educated women, and more-educated women, represented, respectively, by A, B, C and D. At a very broad level of generality, A and B can marry from among C and D. But in reality, given their relative numbers and norms about marriage in South Asia, groups A and D are much more restricted in their spousal options than are groups B and C. What are these relative numbers? Assume for the moment that A + B = C + D. That is, assume that there are roughly equal numbers of marriageable men and women. Given the difference in age at marriage in countries such as India, this assumption is not strictly true—there should be more women than men. But given the sex differential in mortality that also exists in this region, it may be fairly correct to assume that the total number of women of marriageable age is about the same as the total number of men that group of women can potentially marry, that is, men aged about 5 years older than these women. It requires a much greater stretching of the facts, however, to assume that the numbers of men and women in the two educational categories are also about equal. Given the differential emphasis placed on the schooling of boys and girls in many parts of the developing world, it is more likely by far that B is a much larger group than D, and A is probably a much smaller group than C. Any table on the sex differential in educational attainment of the population in South Asia should make this quite obvious. For example, Table 9-1 is based on the sample households in the National Family Health Survey in India. The gender differences in education are obvious from this table. For example, while 22.5 percent of men aged 20-25 are illiterate, and 11.7 percent have studied beyond high school, the corresponding figures for women aged 15-19 (these men's potential spouses) are 43.8 percent and 1.7 percent. Men aged 25-29 show a similar dissimilarity to women aged 20-24. 6   Perhaps it is the kind of education that matters more for men than for women. A general measure such as "years of schooling" may be able to discriminate much better among women than among men.

OCR for page 267
--> TABLE 9-1 Educational Levels of Marriageable Men and Women: India       Literate     Sex Age Illiterate Up to primary school (%) Middle and high school (%) Above high school (%) Men 20-24 22.5 21.7 44.0 11.7 Women 15-19 43.8 21.4 33.0 1.7 Men 25-29 28.8 23.4 35.3 12.4 Women 20-24 52.3 18.3 22.8 6.6   SOURCE: Internal Institute of Population Sciences, 1995. Thus, in every population of equal numbers of marriageable men and women, there are significantly fewer less-educated men than less-educated women, and a larger number of more-educated men than more-educated women. Why do these relative numbers matter? They matter because they severely restrict the marriage market for two categories of individuals—less-educated men and more-educated women. They do so because of the norm of hypergamy in much of South Asian society (see Basu, 1998). In a general way, the system of hypergamy refers to the practice of women marrying men of higher ritual and social status than themselves. This is not simply a matter of preference. It is an injunction that is set forth in the scriptures themselves and that dictates marriage rules on the subcontinent even today. For example, the ancient Hindu lawgiver Manu states, ''If a virgin makes love with a man of a superior caste, the king should not make her pay any fine at all, but if she makes love with a man of the rear castes, he should have her live at home in confinement" (Doniger, 1991:191). And again, "According to tradition, only a servant woman can be the wife of a servant; she and one of his own class can be the wife of a commoner; these two and one of his own class for a king; and these three and one of his own caste for a priest" (Doniger, 1991:44). In the present context, marrying "up" by women requires that they marry men at least as educated and preferably more so. The education of girls may in fact increase the chances of their finding such high-status grooms. Indeed, on this aspect of hypergamy, South Asia is perhaps not unique at all. Around the world it seems to be as common for women to marry men more educated than themselves as it is for them to marry men older than themselves. Yet this phenomenon is not exactly the same as the positive assortative mating with respect to education hypothesized by Becker (1981) and characteristic of most societies in the developed world (see, for example, Epstein and Guttman, 1984; Warren, 1966; Layard and Zabalza, 1979; Schirm, 1986). In the South Asian case, it applies to women, but may not apply to men. In turn, this

OCR for page 267
--> difference implies that men in group B can choose their partners from among women in group C or D (since they are barred only from marrying women more educated than themselves, not those less educated), while men in group A are restricted to women in group C. Similarly, women in group C can marry men in group A or B, but women in group D are restricted to men from group B.7 These restrictions are fortunate in at least one respect because they mean that in principle everyone can find a potential spouse: the larger numbers of less-educated women can turn to more-educated men to make up for the shortage of less-educated men, and the larger numbers of more-educated men can turn to less-educated women to make up for the shortage of more-educated women. Does this phenomenon happen in practice at more than an anecdotal level? That is, does it happen in statistically significant numbers? Tables 9-2 and 9-3 (again from the Indian National Family Health Survey) certainly suggest that it does. The data in Table 9-2 show that virtually none of the sampled women with more than a high school education have married men with no education,8 whereas the data in Table 9-3 show that a substantial number of the sampled men (though fewer in the southern states, where the gender gap in education is less pronounced) with more than a high school education have married illiterate women. Moreover, the vast majority of sampled women with more than a high school education have married men who also have more than a high school education,9 while far fewer than half the sampled men who have gone beyond high school have wives who have done the same. Looking only at more-educated brides and grooms, that is, at members of groups B and D, it is thus obvious that men have a much wider field than women. The main thesis of this chapter is that the way they play this field has a bearing on the relationship between female education and fertility. While the majority of more-educated women marry more-educated men, there are two kinds of more- 7   Indeed, the anthropological literature from South Asia is rich in descriptions of the marriage squeeze that occurs when marriages are hypergamous for any characteristic, not just education. In general, this hypergamy squeezes the marriage market for higher-status women (as measured by income, education, caste, or any other feature) and lower-status men, explaining to some extent the prevalence of dowry in the upper-status groups and of bride-price in the lower-status groups (see, for example, some of the readings in Uberoi, 1993). 8   The disaggregated National Family Health Survey data indicate that the few educated women who do marry uneducated men are almost all in the southern part of India. In addition, they are all in rural areas, that is, areas where there is a lower premium on education, and other socioeconomic factors, such as income, employment, caste, or landholding, substitute for education among men in the marriage market. 9   Kerala, both urban and rural, is a little different in this regard. Relatively larger numbers of women with more than a high school education have married men less educated than themselves. This new phenomenon has been attributed to the rising unemployment of educated men in the state, so that employment rather than education has become the preferred feature of hypergamy (Rajan et al., 1996).

OCR for page 267
--> TABLE 9-2 Educational Level of Husbands Relative to Wives: India, 1992-1993   Husband's Education Level (%)   Education of Wife Illiterate Up to middle school Above middle school Illiterate 28.4 56.9 14.7 Up to middle school 4.0 59.0 37.0 Above middle school 0.3 10.0 89.7   SOURCE: International Institute of Population Sciences, 1995. TABLE 9-3 Educational Level of Wives Relative to Husbands: India, 1992-1993   Wife's Education Level (%)   Education of Husband Illiterate Up to middle school Above middle school Illiterate 92.5 7.2 0.3 Up to middle school 64.3 33.7 3.0 Above middle school 24.0 33.0 43.0   SOURCE: International Institute of Population Sciences, 1995. educated men—those that marry more-educated women and those that marry less-educated women. The hypothesis here is that those who do the former are a select group who themselves have low reproductive goals, so that controlling family fertility requires no extraordinary effort on the part of their more-educated wives. There is no conflict of wills involved, and there is little spousal difference in the returns to childbearing. Likewise, those more-educated men who marry less-educated women are themselves more likely to have the same high-fertility goals as their wives and as

OCR for page 267
--> their less-educated male counterparts. Thus again higher fertility results from the spousal partnership, rather than women's low-fertility intentions being crushed by overbearing husbands with high-fertility desires. The Possibility That There Are Two Kinds Of Educated Husbands How do we know that there are two kinds of more-educated men? That is, how can we assume that educated men who marry educated women are not a random sample of all educated men, but a group selected in some way? If they were a random sample, we could indeed infer that a strong relation between female education and fertility and a relatively weak one between male education and fertility reflects the primary or even sole impact of the woman's education on reproductive behavior, with men's reproductive goals being immaterial or even contrary to women's, as is often implied in the literature. The notion that educated men who marry educated women are in some way different from educated men who marry uneducated women makes intuitive sense; marriage is too important a life event in most societies to be left to random circumstances. Conscious decision making is involved, and both the individuals concerned (or, more commonly, their families) have a list of preferred characteristics, whether these preferences are dictated by norms, ambitions, or tastes. The nature of these preferred characteristics in turn tells us much about the individual or family that holds these preferences. Thus the educated man (or his family) who marries an educated woman is saying something about his preferences, given that he has a pool of educated and uneducated women from which to choose. We may note that the preferences of the educated woman in the marriage market are less transparent in this regard, given that she is effectively barred from choosing her partner from the larger pool of uneducated as well as educated men. It thus follows that the educated man who marries an educated woman is displaying a different set of preferences from those of the educated man who marries an uneducated woman, a difference that cannot be accounted for by their relative levels of education. A field inquiry seeking to discriminate between these two sets of men would therefore not get very far by focusing on their education; information about other aspects of their lives and personalities is necessary. Since intuition is not an acceptable form of evidence in academic (and especially demographic) research, more empirical and measurable means of demonstrating that there are indeed two kinds of educated husbands are needed. This chapter proposes two possible lines of evidence. The first focuses on identifying some of the specific ways in which educated men who marry educated women differ from those that marry less-educated women, even before they get married. That is, what relevant background characteristics could distinguish these two sets of men? The second focuses on demonstrating a similarity between educated

OCR for page 267
--> women and their husbands, assuming it is possible that the lower fertility of educated women is not an outcome of their ability to reject the different reproductive goals of their husbands, but an outcome of having husbands who think like them. To begin with the first line of evidence, the fact that male education itself has only a weak relationship with fertility is not surprising. In most parts of the developing world, and especially in contemporary South Asia, families are well aware of the returns to schooling and increasingly are even embarrassed about illiteracy, especially in males (see, for example, Jeffery and Jeffery, 1993). In addition, the education of sons faces no cultural or attitudinal inhibitions. I have yet to come across a survey in which the bulk of respondents did not reply "as much as possible" when asked to what extent they would like to educate their sons. The bounds of this possibility are set primarily by resource availability, so that male education levels are a proxy for the schooling facilities and financial resources available to households, rather than for the household attitudes and values that female education more often reflects.10 Educational differentials in men thus capture little more than their resource differentials. Moreover, we know from a mass of research on historical and contemporary populations that economic factors can explain fertility declines only imperfectly. Thus if the male is the unit of analysis and we want to relate male characteristics to fertility, we need markers other than education to identify the bundle of values and attitudes that make up the "modern" world view.11 In principle, it should be quite straightforward to collect empirical information on differentials among educated men. The trouble is that most data sets have not gone beyond searching for other proxies for education to capture such differentials. Thus income, occupation, and some background family characteristics (such as caste in the Indian context) are the kinds of things measured by most large surveys, whereas what we need is some information about differences in the world views of different kinds of educated men. Do some educated men (and, by extension, their families) live by a different set of values and have a different set of attitudes toward life from those of other educated men? Is this what makes them gravitate toward educated women when the time comes for them to marry? As already mentioned, if a family can afford to educate its sons, it does so; 10   Female education does also partly reflect resources and services. But the use of these resources and services is much more conditioned by norms and values than is the case for males. 11   That some form of "modernization" or, more accurately, "westernization" inspires the transition to lower fertility is now acknowledged in the literature to be as important as, and often more important than, changes in affordability occasioned simply by changing incomes. Indeed, even those changes in the costs and benefits of childbearing caused by changed incomes are usually mediated by changes in attitudes and aspirations. If this were not the case, increases in income would lead to a rise in the demand for children.

OCR for page 267
--> there is no larger principle involved. But in marriage, men are expressing more than their financial constraints; they are expressing a preference. Therefore, one way of differentiating among educated men is by looking at their marriage partners. But then, of course, one must ask whether the husbands of educated women are different because they were always different, or have become different through spousal influence. Fortunately, information on premarital differences in world views requires no specialized survey instruments. One does not need to devise psychological tests of modernization to identify the educated man who will prefer an educated wife. One can of course ask young unmarried men about their preferred spousal characteristics. But such a measure would suffer from the same kinds of potential biases as any attitudinal questions asked in the impersonal style of a survey. This chapter suggests a much simpler measure of modernization. If one assumes (quite reasonably) that female education is as much an outcome as a determinant of modern attitudes and values, one potential way of classifying educated men (and their families) according to these attributes would be to look at the investment made in the education of their sisters.12 The latter cannot logically be the result of the educated women these men marry, unless their sisters are very much younger than they. If one can thus demonstrate that educated men who marry educated women come from backgrounds that are relatively modern to begin with, one must give less credence to theories that explain the lower fertility of educated women in terms of their own characteristics alone. Most data sets in demography already routinely collect some background information on all household members. From this information it should be possible to measure the educational levels of the girls and women in the husbands' natal families and to relate these levels to those of the wives of these men.13 If such a relationship does indeed exist, it will provide strong empirical support for the thesis that there are two kinds of educated men. The hypothesis is that households in which both sons and daughters receive an education are qualitatively different from those in which only the sons are sent to school (or remain in school for any length of time), and that the men who marry educated women are more likely to belong to the former group. 12   Alternatively, one could look at the education of their mothers. But in a society in which female education is a recent event, this indicator might not be discriminatory enough. Moreover, the education of mothers may reflect the attitudes of the grandparents of the husbands more accurately than those of their parents. 13   One can foresee several problems with getting these tabulations, of course. For example, given the nature of existing data sets, one would be able to look at the educational levels of sisters only when the families are joint. In addition, one could look only at the education of the unmarried sisters of the husband, since his married sisters would be living elsewhere. But these problems may not exist in all data sets, and in any case, future research on this issue would merely need to add some questions to the standard household survey instrument to eliminate these potential biases.

OCR for page 267
--> TABLE 9-4 Educational Level of Husbands' Sisters According to the Education of Their Wives: India, 1992-1993   Education of Sisters of Husbandsb (%)   Education of Wivesa Illiterate 1-6 years of school 7+ years of school Illiterate 71.4 17.6 11.0 1-6 years of school 35.0 35.0 30.0 7+ years of school 6.4 19.1 74.5 NOTES: Sample Size = 186. Includes only those households that recorded the presence of wives as well as sisters in the household survey. a Includes women aged 20-29 who are the wives, daughters-in-law, or sistersin-law of the household head. b Includes women aged 15-24 who are the sisters and daughters of the household head. SOURCE: Computed from Indian National Family Health Survey data. Table 9-4 lets us look for such a relationship in the results of the Indian National Family Health Survey. The table plots the educational distribution of the available daughters and sisters (that is, the agnates of the respondents' husbands) against the education of the wives, sisters-in-law, and daughters-in-law of the household (that is, the in-marrying women) to see whether educated women marry into homes in which female education is already valued.14 Table 9-4 is primarily indicative since the data have many limitations (see note 13); nevertheless, the results are strongly indicative. Educated women are much more likely to marry into homes in which the daughters are also educated. Compare, for example, Tables 9-2 and 9-4. While only 28 percent of the husbands of illiterate women in Table 9-2 are illiterate themselves, as many as 71 percent of the sisters-in-law of the illiterate women are illiterate (Table 9-4). On the other hand, while 90 percent of the husbands of women with 7 or more years of schooling have 7 or more years of education themselves, 74 percent of the 14   By restricting the age ranges of the two categories of women-those marrying into the family and those born in it-Table 9-4 manages to exclude the daughters of the married women from the set of women born into the family before the woman entered it; a woman aged 20-29 is extremely unlikely to have a daughter aged 15-24.

OCR for page 267
--> sisters-in-law of these educated women are also as educated as the women. That is, sex differentials in education are much higher in the families of the men who marry less-educated women. This finding in turn suggests that the families of the educated men that choose to marry educated women are qualitatively different from the families of the educated men that choose less-educated or illiterate wives. This finding is not simply one more demonstration of the existence of positive assortative mating for education. For one thing, while it is true that the more educated the woman, the more educated her spouse, it is not similarly true that the more educated the man, the more educated his spouse. Instead, from the husband's perspective, the assortative mating occurs for his sisters' and his wife's education. That is, the more educated are his sisters, the more educated is his wife. The resultant implications for fertility could receive further support from analyses that would try to estimate the relationship between the woman's fertility and the education of her sisters-in-law, controlling for the woman's own education. According to the present hypothesis, this relationship would be stronger than that between her husband's education and her fertility. In other words, what we see here is a case of educational hypergamy for women (that is, women tending to marry men more educated than themselves), but a continuing social homogamy for spouses (that is, a tendency for spouses to belong to similar social backgrounds). This combination of educational hypergamy and social homogamy is not unique to the Indian case used to illustrate the present discussion; instead it seems to be a feature of most societies in most parts of the world (see, for example, Bozon, 1991, for France).15 When the data needed to study husbands' background are not available (and even when they are), the second possible line of evidence mentioned above—the similarities between educated women and their husbands (and by extension, the similarities between less-educated women and their husbands, whether educated or not)—can strengthen the case for stating that the husband is an important intermediate variable in the relationship between female education and fertility. This is a difficult research undertaking given the trend in academic and policy circles to focus on intrahousehold conflict, treating any convergence of interests as incidental at best and a result of bulldozing at worst. Even on a narrowly demographic matter such as reproductive goals, the assumption in the literature is that there must be an interspousal difference in interests. 15   The implications of social homogamy for fertility may explain some findings from unrelated parts of the world. For example, Mascie-Taylor and social class than it is in South Asia, this finding may reflect (1986) reported for a British sample that as educational heterogamy increased, so did fertility. Given that the woman's education in Britain would be more straightforwardly a function of her economic the socioeconomic heterogamy of the high-fertility couples. Their analogue in the South Asian case would be educated women married to educated men with uneducated or poorly educated sisters.

OCR for page 267
--> Two examples will illustrate the point. The first is a stylized representation of the situation in the developing world from an influential World Bank publication on the virtues of female education (King and Hill, 1993:vi): A poor family has six children. The mother never attended school, was married at age 15 and remains illiterate. Her husband earns most of the family's meager income and decides how it is spent. Since his own economic security depends on his sons' ability to support him in his old age, he insists that the boys go to school while the girls remain at home to do chores.... Summers (1993:vi) provides a second example: A poor family has three children. The mother went to school for five years and is able to read and do arithmetic well enough to teach school in the village. As her last birth was extremely difficult, she and her husband adopted family planning. She now has more time and resources to spend on her family. Hoping for a better future for her children, she insists that they all go to school and practice their reading every night.... How real is this characterization? In the first family, is it only the man that wants many children and especially many sons? If the wife had her way, would fertility necessarily decline? And in the second family, are there fewer children, and do they all go to school, only because the wife wishes this? Despite the hypothesizing and the rhetoric, there is really very little evidence in the literature to support these contentions. The trouble is that while the literature posits that education leads women to have a smaller need or desire for children or a better ability to achieve their low-fertility desires, it is generally silent on male preferences, except by implication. When we say that female education is necessary for fertility decline, are we saying that: In the case of low female education, women want fewer children than men, but cannot articulate or execute their wants? or Women with low education want more children than men, while those with high education want as few children as men, and such convergence of wants is necessary for fertility to decline? or In the case of low female education, both men and women want more children than they do in the case of high female education? (This is not to imply that husbands and wives have similar fertility goals altruistically, although that is not as impossible as the literature on intrahousehold conflict would have us believe; it could just be that their fertility goals are similar because their socioeconomic and external circumstances are similar.) What about the empirical evidence on spousal differences in reproductive

OCR for page 267
--> goals? Given the amount of hypothesizing on this matter that is implicit in the literature on intrahousehold disparities, it is surprising that there is so little actual information available on this question from the field. Most of the few surveys that have concerned themselves with differences in reproductive preferences between spouses have had to rely on the reporting of only one spouse, usually the wife. This is, of course, partly a matter of convenience given the tradition in demographic research to focus on the woman in data collection and analysis. But the result is that we have no sound basis for assuming that husbands think differently from wives in the area of reproductive goals. Indeed, in an extensive review of the literature, Mason and Taj (1987) found little evidence of gender differences in fertility goals. A search of the more recent literature on this subject, especially the few studies in which both husbands and wives were interviewed, finds little reason to alter this conclusion (see, for example, Ezeh, 1993; Bankole, 1992; Jeffery and Jeffery, 1993; Stycos, 1996). One can of course think of several reasons for this finding of a similarity in husband-wife reproductive preferences that are compatible with couples having different goals in reality. First, as Mason and Taj (1987) discuss, most of the available literature may refer to populations that are already undergoing a fertility transition, and it is in pretransition societies that one would expect to find the greatest gender differentials. Second, since the review measured net differences between the average family-size goals of husbands and wives, as opposed to the proportion of individual couples in which husbands and wives had differing goals, there may have been a tendency for subgroups with a difference in one direction to be balanced out by those that differed in the other. If that is the case, it is not clear how important differences in fertility goals are in aggregate reproductive behavior. Third, family-size ideals may be a reflection of behavior, rather than predating and/or determining behavior. But this supposition does not apply to younger cohorts who have yet to attain their desired family sizes. Fourth, family-size goals may be similar between spouses because both spouses have accepted and are stating the preferences of the more powerful spouse. This kind of dominance is difficult to detect by the ordinary survey method and is understandably inferred from other kinds of evidence and theories about intrahousehold relationships. Fifth, and this is a more charitable variant of the fourth interpretation, there may be a convergence of fertility goals between husband and wife as the two influence and are influenced by each other, not necessarily through pressure, but through persuasion and constant exposure to one another's views and beliefs. There is also a sixth and most plausible possibility: that husbands and wives do indeed enter a marriage with similar attitudes toward reproduction and family size. This possibility is not at all incompatible with great intrahousehold gender inequality. That is, it is quite plausible that husbands and wives do have similar, independently generated fertility goals because their motivations and background circumstances are similar.

OCR for page 267
--> In any case, it is not at all clear theoretically why male and female reproductive goals should differ. Moreover, if they should, there are as many theories that would lead one to expect higher fertility goals in women as there are theories that would imply lower fertility goals in women. For example, the literature on the status of women and fertility that rationalizes the high fertility of women with low autonomy in terms of their need for sons to provide risk insurance and old-age security does not suggest that these women are so fertile only because their husbands and mothers-in-law override their wishes. At the same time, when this literature addresses the limited access of women to contraception and their limited ability to be heard in the matter of reproductive decision making, the inference is that they would have fewer children if only their families were more receptive to the idea. In theory, however, it makes much more sense for the husband of the low-autonomy (or, in this context, poorly educated) woman himself to want many children because the conditions under which women are limited in their life choices are the very ones that should increase men's own need for sons to provide risk insurance and old-age security. Similarly, when the wife is in the position of supporting herself or being supported by her daughters in an emergency, she (or her daughters) is also in a position to support the husband when the need arises, so that the interests of both spouses are well served by fewer children. That is, all theories about the low-status woman's need for several children are consistent with her husband's and household's need for several children as well. I would therefore rewrite the earlier quotations from King and Hill (1993) and Summers (1993) as follows: A poor family has six children. Their mother never attended school, was married at age 15, and remains illiterate. Her husband earns most of the family's meager income and decides how it is spent. Since their economic security depends on their sons' ability to support them in their old age, both parents insist that the boys go to school, while the girls remain at home to do chores. and A poor family has three children. The mother went to school for five years and may or may not teach in the village school (the link between female education and employment being much more complex than that implied in the original quote). Hoping for a better future for their children and themselves, both parents insist that they all go to school and practice their reading each night. Female Education, Marriage, And Reproductive Motivation If it is true that the educated men who marry educated women are qualitatively different from the educated or uneducated men who marry uneducated women, if an important part of this difference lies in the lower fertility goals of

OCR for page 267
--> the former, and if these lower-fertility goals are shared by the men's educated wives, the question of the reasons for these lower goals still remains. Why do educated women and their husbands want fewer children? These common motivations are not the focus here, but some possibilities are nevertheless worth reviewing briefly. If it is true that educated wives and their husbands both desire fewer children than do uneducated wives and their husbands, it cannot also be true that the woman's control over reproduction or over her husband (or mother-in-law) is the critical issue. It is the couple's control over life in general that makes both husbands and wives interested in controlling fertility. Indeed, contrary to the common wisdom, the woman's control over any aspect of life—sexual, reproductive, or otherwise—need not follow from her education even if her education is associated with low fertility (see, for example, several of the papers in Jeffery and Basu, 1996). Continued intrahousehold hierarchies are quite compatible with fertility decline even in the face of rising female education, and in fact seem to be characterizing recent fertility declines in much of Asia today (see, for example, Greenhalgh, 1985; Basu, 1995). I would suggest that an important determinant of intentional fertility decline associated with female education is an increase in the family's or, more narrowly, the husband-wife team's united ability to manipulate the environment, rather than an increase in such ability in the woman alone. Female education is partly a proxy for this joint ability, since it is a proxy for the husband's characteristics that cannot be measured by his education. In addition, the wife's education is a facilitator of this ability because there are now two individuals who can appreciate and reinforce each other's understanding that low fertility is one way to satisfy the new aspirations and wants resulting from their education (as well as their exposure to the mass media, of course). As Freedman (1979) and others have pointed out, one should not underestimate the value of such changing aspirations in general and increased material aspirations in particular, independently of change in the objective conditions of life, in motivating fertility control. In addition, female education allows households as a whole to better exploit the opportunities that open up because of education, an exploitation that often hinges on smaller family size. For example, although an educated woman may not have much direct control over whether she grinds her spices on the grinding stone or in an electric grinder, her education and her exposure to the mass media tell her clearly that the electric grinder is more convenient (and more fun), and that one way of affording an electric grinder is to have one less child for whom school fees must be paid. If, in addition, the woman has fewer but more-educated children, they can provide her with a television to watch during the time she saves by having an electric grinder. In India (and in South Asia in general), the woman's education has also reinforced her awareness that to effect such transfers from children to parents, one needs sons, not daughters, and that with daughters, in fact, the transfers are all in

OCR for page 267
--> the opposite direction. This kind of reasoning ability may explain why educated women want fewer children than do uneducated women, but are not much more indifferent than are uneducated women to the sex composition of these fewer children. For example, in the all-India survey of the Operations Research Group (1991), only 1.5 percent of educated women were indifferent to the sex composition of their children, although the minimum number of sons wanted was, at 1.6, lower than it was for illiterate women at 2.0. The husband of this educated woman will, as is argued in the last section, already have similar low-fertility desires (except perhaps that the electric grinder is replaced in his imagination by a motorcycle). Another important determinant is likely to be the hypothesized relation between parents' and children's education. If the mother's education increases the possibility of her daughters' education, the overall proportion of children being educated is higher in families with educated wives, thereby increasing the costs and decreasing the immediate benefits of children—another rationale for reduced fertility. In addition, if the literature on the impact of mass schooling on fertility has any basis (see, in particular, Caldwell, 1980, and Bongaarts, 1996), then at the macro level, too, there should be a fertility impact as larger proportions of children are sent to school—even if these larger proportions are merely a mechanical outcome of the greater probability that the daughters of educated mothers will be sent to school. Yet rising material aspirations and greater investments in the education of daughters are but two of the ways the more ambitious partnership that characterizes the couple in which the wife is educated can provoke a fertility decline. Other ways include the differential impact of changes in the broader economy (couples with some education may be more likely to benefit from such changes if they have fewer children, or may even have more to lose with high fertility than do uneducated couples); access to modern views about family size in general; changing aspirations for one's children; the prestige of education being able to compensate for the loss of status associated with low fertility in uneducated families; the higher incomes that reduce the need for children as security; and the reduced fatalism about life in general and fertility control in particular that brings conscious birth control within the calculus of human choice. The important point is that none of these changes hinge on the ability of the educated wife to override the wishes of an ignorant or conservative husband; merely by marrying her, this husband has already demonstrated that he can be as modern as she is. Some Implications The policy implications of this chapter do not tally exactly with those derived from a framework in which women's education is connected to fertility decline purely or even largely through its effects on women's control over their lives and their reproductive performance. Control over reproduction is not im-

OCR for page 267
--> plied if the men who marry educated women also have low fertility goals themselves. This is not to make the case that women's control over reproduction is not a desirable goal—it is—but to clarify that such control is not a necessary condition for fertility decline because there is often no conflict of reproductive goals to begin with, even given the lower-fertility preferences of educated women. Such women are likely to be married to men who share their fertility preferences even if they do not otherwise discard the trappings of patriarchy. That is, patriarchal family structures are not necessarily inconsistent with fertility decline (on this, see also Weinstein et al., 1994). Given the impact of education on knowledge about and aspirations for a better (or at least different) material life, even in the face of unchanged autonomy in domestic or reproductive decision making,16 female education remains a major goal of antinatal policy. Whatever their origins, educated women do clearly have lower fertility preferences, and a convergence of reproductive preferences between spouses should undoubtedly make it easier for fertility to decline than if one partner alone must impose fertility control. In addition to conscious policy, however, the goal of female education may also be easier to attain as female educational levels rise further and increase the demand for educated wives from the brothers or other relatives of these women, and as uneducated women now find themselves increasingly excluded from marriage with desirable grooms. Acknowledgments I am grateful for many suggestions for improving the paper from participants, from Kaushik Basu, and from two anonymous referees. I am also grateful to Annabel Perkins for research assistance. References Bankole, S.A. 1992 Marital partners' reproductive attitudes and fertility among the Yoruba of Nigeria. Ann Arbor, Mich.: University of Michigan. (mimeographed) Basu, A.M. 1992 Culture, the Status of Women and Demographic Behavior: Illustrated with the Case of India. Oxford: Clarendon Press. 1995 The many routes to a fertility transition: Fertility decline and increasing gender imbalances in Tamil Nadu, India. Ithaca, N.Y.: Cornell University. (mimeographed) 1998 Anthropological insights into the links between women's status and demographic behavior: The notions of hypergamy and territorial exogamy. In A.M. Basu and P. Aaby, eds., The Methods and Uses of Anthropological Demography. Oxford: Clarendon Press. 16   But. as stated at the beginning of this chapter, education may also actually increase female autonomy and undoubtedly frequently does; the main point is that this is not necessary for a negative education-fertility relationship to exist.

OCR for page 267
--> Becker, G. 1981 A Treatise on the Family. Cambridge, Mass.: Harvard University Press. Bongaarts, J. 1996 Remarks made at National Research Council Workshop on Female Education and Fertility in Developing Countries. Washington, D.C. February 29-March 1, 1996. Bozon, M. 1991 Mariage et mobilite sociale en France. European Journal of Population 7(2): 171-190. Caldwell, J.C. 1980 Mass education as a determinant of the timing of fertility decline. Population and Development Review 6(2):225-255. Caldwell, B. 1996 Female education, automony, and fertility in Sri Lanka. Pp. 288-321 in R. Jeffery and A.M. Basu, eds., Girls' Schooling, Women's Automony, and Fertility Change in South Asia. New Delhi, India: Sage Publications. Davis, K., and J. Blake 1956 Social structure and fertility: An analytic framework. Economic Development and Cultural Change 4:211-235. Doniger, W. 1991 The Laws of Manu. New Delhi: Penguin. Dyson, T., and M. Moore 1983 On kinship structure, female autonomy and demographic behavior in India. Population and Development Review 9(1):399-440. Epstein, E., and R. Guttman 1984 Mate selection in man: Evidence, theory and outcome. Social Biology 31(3-4):243-278. Ezeh, A.C. 1993 The influence of spouses over each other's contraceptive attitudes in Ghana. Studies in Family Planning 24(3): 163-174. Freedman, R. 1979 Theories of fertility decline: A reappraisal. Social Forces 58(1):1-17. Greenhalgh, S. 1985 Sexual stratification: The other side of ''growth with equity." Population and Development Review 11(2):265-314. International Institute for Population Sciences 1995 National Family Health Survey (MCH and Family Planning): India, 1992-93. Bombay: International Institute for Population Sciences. Jeffery, R., and A.M. Basu, eds. 1996 Girls' Schooling, Women's Autonomy, and Fertility Change in South Asia. New Delhi, India: Sage Publications. Jeffery, P., and R. Jeffery 1993 Killing my heart's desire: Education and female autonomy in rural North India. In N. Kumar, ed., Woman as Subject. Calcutta: Bhatkal and Sen. Jejeebhoy, S.J. 1995 Women's Education, Autonomy and Reproductive Behaviour: Experience from Developing Countries. Oxford: Clarendon Press. King, E., and M.A. Hill, eds. 1993 Women's Education in Developing Countries: Barriers, Benefits, and Policies. Baltimore, Md.: The Johns Hopkins University Press. Layard, R., and A. Zabalza 1979 Family income distribution: Explanation and policy evaluation. Journal of Political Economy 87(5, part 2):S133-S161.

OCR for page 267
--> Mascie-Taylor, C.G. 1986 Assortative mating and differential fertility. Biology and Society 3(4): 167-170. Mason, K.O., and A.M. Taj 1987 Difference between women's and men's reproductive goals in developing countries. Population and Development Review 13(4):611-638. Morgan, S.P., and B.B. Niraula 1995 Gender inequality and fertility in two Nepali villages. Population and Development Review 21(3):541-561. Operations Research Group 1991 Family Planning Practices in India: Third All-India Survey. Baroda: Operations Research Group. Rajan, S.I., M. Ramanathan, and U.S. Mishra 1996 Female automony and reproductive behaviour in Kerala: New evidence from the recent Kerala Fertility Survey. Pp. 269-287 in R. Jeffery and A.M. Basu, eds., Girls' Schooling, Women's Automony, and Fertility Change in South Asia. New Delhi, India: Sage Publications. Sathar, Z.A. 1996 Women's schooling and autonomy as factors in fertility change in Pakistan: Some empirical evidence. Pp. 133-149 in R. Jeffery and A.M. Basu, eds., Girls' Schooling, Women's Automony, and Fertility Change in South Asia. New Delhi, India: Sage Publications. Schirm, A. 1986 Assortative mating when individuals differ by many traits. Paper presented at the annual meeting of the Population Association of America, San Francisco. Stycos, J.M. 1996 Men, Couples and Family Planning: A Retrospective Look. Population and Development Program Working Paper. Ithaca, N.Y.: Cornell University. Summers, L.H. 1993 Foreword. Pp. v-vii 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. Uberoi, P., ed. 1993 Family, Kinship and Marriage in India. Delhi: Oxford University Press. Vlassoff, C. 1996 Against the odds: The changing impact of schooling on female autonomy and fertility in an Indian village. Pp. 218-234 in R. Jeffery and A.M. Basu, eds., Girls' Schooling, Women's Automony, and Fertility Change in South Asia. New Delhi, India: Sage Publications. Warren, B.D. 1966 A multiple variable approach to the assortative mating phenomenon. Eugenics Quarterly 13(4):285-290. Weinstein, M., T.H. Sun, M.C. Chang, and R. Thornton 1994 Co-residence and other ties linking couples and their parents. In A. Thornton and H. Lin, eds., Social Change and the Family in Taiwan. Chicago, Ill.: University of Chicago Press.