Heather Joshi and Patricia David. Demography: Analysis and Synthesis. Editor: Graziella Caselli et al., Volume 1, Elsevier, 2006.
Features of the world beyond the individual and the couple determine whether, why and how fertility is controlled. The economic and social systems provide the opportunities and constraints within which choices can be made. They also help determine whether people are aware that choice is possible, whether, as Coale (1973) put it, fertility is within the sphere of conscious control. The economic and social context may also help to set the standards of behavior between family members and the values to which they aspire. How far each sort of consideration applies to the explanation of fertility changes already observed, and how far social and economic variables can be used to predict, or manipulate, the future course of human reproduction remain controversial. This chapter is mainly concerned with the social and economic explanations of deliberate actions whose cumulated outcome is the rate of human reproduction. Leaving biologic factors aside, already discussed in Chapters 30 to 34, we treat fertility as a form of behavior rather than as a response to innate drive that has been bred in to survivors. We also leave aside, for the most part, the question of partnership formation and dissolution that are dealt with in Chapters 25 to 29.
Four topics will be considered. The first three sections elaborate some of the theoretical explanations of fertility offered in recent literature. First we describe a behavioral framework for the understanding of social reproduction (and make a passing reference to social biology). Next we review some approaches from just two of the social sciences, economics and sociology. Finally, we consider some contemporary issues, in both developing and industrial countries, on which empirical applications of these approaches have shed both heat and light. The questions selected are as follows:
- Why does fertility decline with development?
- Can fertility decline in poor countries?
- Women’s education: a materialist or ideational influence?
- Is childbearing for old age security?
- What makes fertility fluctuate in rich countries?
- Are state policies effective in accelerating fertility decline?
- Are state policies effective in preventing subreplacement fertility?
This is more an account of territory through which we have traveled than a complete Michelin Guide. We have not attempted a comprehensive coverage of topics or literature. We hope to spark enough interest to encourage the reader to discover the numerous authors we have neglected (particularly outside the Anglo-Saxon literature) and that what we do cover will show the way for further interdisciplinary endeavor.
Supply and Demand: A Framework of Factors Influencing Fertility
The level of fertility in a society or a group is the cumulated outcome of deliberate actions (behavior) and biologic functioning, social pressures and chance operating for each pair of parents, actual or potential. The constraints that condition behavior are largely determined by the social environment they inhabit. Coleman and Schofield (1986) observed that demography still has no central paradigm linking processes at the microlevel with those operating at the societal level. Microlevel behavior must be involved if there is any change in fertility in the aggregate, at the macrolevel; but microlevel variation can also be observed, and linked to variations within an economic and social system that is in aggregate unchanging. The primary focus of most (nonbiologic) fertility theories is therefore on the individual.
Contextual factors external to the individual or couple that affect behavior and/or decision-making regarding family formation must work through mechanisms that operate at the individual level. One general framework that has proven useful is set out by Easterlin, Pollack, and Wachter (1980). They classify these mechanisms into those affecting the biologic production of births, supply or natural fertility, and demand factors, which may reduce wanted fertility below the biologically feasible maximum. Their approach, sometimes identified as that of the Pennsylvania School, is a general formulation not only because it allows for fertility sometimes being constrained by biologic limits, but also because it allows for the preferences, which define demand, themselves to be determined within the system.
The Pennsylvania approach is contrasted with the Chicago-Columbia School whose exponents include Becker (1981), Schultz (1981), and Cigno (1991). This group of writers adopt a demand approach to fertility, seeing decisions about the number, timing (and quality) of children as the outcome of rational decision making, maximizing the utility to be gained from fertility decisions subject to the resources available and the rates (prices) at which they may be transformed into competing sources of utility. It is assumed that tastes or preferences remain fixed, and are unified within couples. The reactions of the rational decision-maker to changes in income and prices are predicted.
One variant of the supply-demand framework is to investigate the constraints around which decisions are made. For Hobcraft and Kiernan (1995) the constraints on parenthood are biology, money, security, ideas, and especially time. For Folbre (1994), constraints are socially structured—by membership of social groups, such as social class, religious, or ethnic groups and gender. Membership of such groups overlaps, and thus individuals are subject to diverse, possibly conflicting pressures. In different circumstances different types of group membership may be dominant. Her book contains a very extensive review of the evolving structures over several continents and centuries. This recognition that the demographic actor’s tastes and behavior are affected by group identity within broad societal influences helps to account for differential fertility within societies. It also suggests that group norms and interests may mediate the impact of social and economic considerations upon any given individual. The overlapping nature of group membership adds complexity to the notion of self-interest.
Another variant of the supply-demand framework, as adopted by Bulatao and Lee (1983) adds specific consideration of the reconciliation of supply and demand. The deliberate regulation of the supply of children is conditioned on both demand factors and the costs of, or obstacles to, fertility regulation. These include social influences on attitudes to, knowledge of, and access to contraception, abortion, or the practice of infanticide; the gender balance of power to make decisions regarding control of reproduction; and husband-wife communication regarding fertility regulation. Where births are unwanted, there is said to be unmet need for control of fertility.
The next two sections elaborate factors than can be thought of as influencing the supply-demand model of fertility decisions.
Supply factors affect the number of surviving children by regulating the length of the reproductive life span and the spacing of births within it. Beyond purely biologic factors, they include social constraints on the timing of entry into marriage or sexual union, and practices affecting the timing of parenthood. Constraints on behaviors directly affect the biology of childbearing, such as norms regarding breastfeeding and postpartum abstinence, and use of contraception or abortion. These regulate the pace of childbearing during the reproductive years. Supply factors also include features of the material environment (such as nutrition and hygiene) that may affect the chances of conception and spontaneous abortion.
Factors that affect survival of liveborn children, particularly where there are high rates of infant and child mortality, can be thought of as affecting the supply of surviving progeny. Infant mortality may indirectly affect subsequent fertility through shortened breastfeeding even if there is not any deliberate intention to replace a child who dies (Bongaarts, 1983). The end of childbearing is affected by social factors—also often group-specific—such as those affecting maternal survival and widowhood; norms governing separation, divorce, and remarriage; and those influencing deliberate means of terminating childbearing such as sterilization, terminal abstinence, or effective contraception.
Some supply determinants are more clearly the result of the behaviors of the individual, such as breastfeeding practices and the timing of marriage, variable across and often within societies around the world. Others, like child or maternal mortality levels, are not behavioral in nature, but are affected by the material conditions of life and available health technology. The social environment (a macro consideration) determines the resources available for the care of mother and child, and the knowledge (technology) of how this should be done, and sets norms for acceptable practices. The social environment also provides any knowledge about methods of birth control and affects their cost, social as well as monetary.
The technology of producing children of a given standard goes beyond biologic factors to the investment of resources made in children once born, building up their human capital of health and skills. Changes in any of this technology modify what reproductive outcomes are feasible and permit different outcomes to be chosen. The state of knowledge about producing everything else can also change, which will also affect the setting in which fertility decisions are made. Note, too, that in the production of successful surviving children, if nothing else, the venture is beset with risk and uncertainty. Outcomes could be quite different if it became possible to determine the sex of a child in advance or guarantee that a woman would still be able to conceive if she postponed motherhood until her 40s.
These factors, proximate determinants of childbearing, will be affected by the social and economic environment. Society, for example, can influence who enters sexual unions, at what age, and on what terms. In historic times in Europe, norms surrounding marriage acted as an important safety valve to control fertility (as well as containing sexuality), preventing the formation of new households when economic conditions were poor. Marriage is now less important with respect to childbearing in the United States and Europe, and other social norms regulate age at first birth. The first birth may precede formal marriage in many African societies, where proof of fecundity may be an important precondition to finalizing the marriage bond. Far more important to reducing the overall level of fertility is the deliberate regulation of fertility within marriage.
To economists at least, it seems natural to turn to the analogy of the buyer’s side of a market once fertility enters the realm of control, not merely limited by biosocial supply constraints. When families can do little but accept however many (or few) children God sends, the question of whether they have adapted their behavior to achieve desired fertility does not arise. Once it becomes technically feasible, and thinkable, to reduce childbearing below the natural maximum, it becomes relevant to ask what fertility is desired, and to seek explanations for reproductive behavior in its motivation as well as its means.
People may be able to specify the fertility they would like to achieve by stating a family size, perhaps depending on sex composition. They may be able to distinguish between an ideal family and the number that they would prefer given existing circumstances. Even parents who embark on family formation with no clear goal, may come to weigh up the advantages and disadvantages of adding another birth: the marginal child (Leibenstein, 1974). The death of an existing child may motivate another birth as a replacement (Vallin and Lery, 1978), but the question still remains of why a particular target family size was desired, and whether it is worth the costs of achieving it. Those who decide to remain voluntarily childless (or child-free to put it more positively) can be thought of as taking the decision that even the first child is not worth having, but as being generally influenced by the same considerations as those who choose to limit a positive number of children (Keilman 1996).
The demand approach to fertility applies microeconomic theory of rational decision making to the whether and when to control fertility. The decision-maker is aiming to attain maximum feasible utility from the sources of satisfaction available, which include children. How much these are valued (i.e., how much satisfaction they provide) will be affected by the parents’ preferences, or tastes. The parents’ commitment to bringing a child into the world and bringing it up to adulthood is likened to investment in a consumer durable. The benefits expected must be weighed against the cost of achieving them, which is the forgone return from an alternative investment of their resources.
This reduction of complex, private, and moral decisions to the self-interested calculus of the counting house is not universally palatable. The theories do recognize that children are not exactly like physical assets, as they cannot be sold. They can also allow for parents acting in their children’s, as well as their own, interests. Through the assumption of altruism, one decision-maker draws satisfaction from the well-being of another. Indeed Folbre (1994) caricatures the actors in neoclassical fertility theory as a family of Rational Economic Men, perfectly altruistic and harmonious at home, and single-mindedly self-seeking in the market place. Such agents need not be constantly adjusting their private lives to economic signals; it suffices that some economic changes induce some people to adjust their family formation behavior, for it to be worth thinking about what such adjustments might consist of. The assumption of a united household is being challenged by the application of game theory to bargaining models. This highly technical literature is not covered by the present review.
Social Biology: A Digression
Why is there a demand for children? Does biology help explain the demand as well as the supply side? The fundamental reasons people bear children are often assumed to lie in a biologic imperative to reproduce. Evolutionary biologists, working with animal models of reproductive behavior, have tried to extend these to the human organism. These theorists now propose that when it is in the individual’s interest to overexploit existing resources, populations will grow too large for the available pool of resources and the population will experience excess deaths as a result (a Malthusian check on the growth of the population).
Lately, some of these theorists have examined reproductive strategies as adaptive mechanisms that individuals used to maximize the fitness of their genetic inheritance. These life-history theorists are concerned with understanding how alternative reproductive strategies evolve to affect reproductive success. Use of the term strategy suggests conscious choice, but its use by biologists does not make such an assumption.
One school of thought suggests that such reproductive success may be achieved by having fewer offspring, and investing more parental effort in their quality. This may be the best strategy for ensuring the long-term reproduction of their genetic traits (Chisholm, 1993). Attention is focused on how physiologic factors that directly affect reproduction have evolved, and why norms that affect birth spacing patterns, such as prohibitions on intercourse or prescribed breast-feeding durations, are found to vary across different cultural contexts (Wood, 1994). Why have these evolved in this particular way?
According to some life-history theorists, mortality risk is what determines these differences in the reproductive strategy that is adopted (Promislow and Harvey, 1990, 1991; cited in Chisholm, 1993). When mortality rates are high, the most efficient strategy is to maximize the number of offspring: a short-term strategy to ensure the survival of their genetic traits, because at least some of their offspring will survive to reproduce. When mortality is low, so the theory goes, the best strategy is to invest more parenting effort in fewer offspring: a long-term strategy to maximize the number of descendants in the distant future by maximizing the reproductive potential of their offspring.
Because theorists who focus on human fertility must contend with biologic mechanisms and psychological forces that drive human sexuality and control of conception or the number of surviving children, attempts are now being made to synthesize work from the fields of evolutionary biology and developmental psychology (Chisholm, 1993).
Do parents, in response to an innate drive to reproduce their own traits, adjust their family-building patterns in response to an assessment of risk or a calculation of the marginal returns on their investment, (Kaplan, 1994)? Is this how cultural differences in the proximate determinants of fertility evolve? Hobcraft and Kiernan (1995) wonder whether it is biologic drives that keeps reproduction going at all. Evolutionary biology can portray regression to two-child norm as a trade-off between carrying on the lineage or gene pool and the costs of childbearing. We turn to the social sciences for other perspectives.
Economic Explanations of Fertility Behavior
One feature of the environment is the economy, the availability of resources through production and exchange. The demand approach is to consider fertility decisions as being made by rational agents, setting advantages against disadvantages, benefits against cost. An objective, reflecting personal tastes and norms or values of the peer group or society at large, is maximized subject to the resources available, their relative prices and the technology. A rational choice would imply that taking any other decision would have left the parents worse off. Changes in the economic environment, if big enough, could lead to adjustment of family formation.
The utility function sets out the valuation the decision-maker puts on various sources of satisfaction. The objectives are essentially the fruits of reproduction, the fruits of production and the enjoyment of leisure. In new home economics models (as those originating in the Chicago-Columbia School are also known), the satisfaction parents derive from children depends on both their number and their quality. This key construct denotes health and skills, which will depend on how much parental time and other resources are invested in them. Parents have other sources of satisfaction in their own consumption of other goods. If they are altruistic toward their children they will also draw utility indirectly from their children’s (and future descendants’) standard of living. Parents therefore gain from an increase in either their own consumption, or child services, if there is an increase in their resources (budget) meaning they do not have to sacrifice any of the other. Whether they will choose to trade one for the other when the budget is fixed, depends on whether their relative valuation of the outcome exceeds the cost of achieving it. The valuation reflects individual (or socially conditioned) preferences. The question of relative costs is central to the analysis, as being the most likely front on which the economy may send signals to formers of families, who may react by engaging in substitution, a reaction to changes in relative prices.
The level of resources, the budget, is given by the couple’s endowment of physical and human capital. The latter represents their earning potential that can be accumulated through past investments in their health and skills. The greater the resources at the couple’s disposal, the larger the range of possible combinations of consumption and children they may choose. Increases in the budget lead to income effects in contrast to substitution or price effects. The range of opportunities may also be widened by improvements in technology, the rate at which resources may be transformed into either of the desired outputs, or indeed in the safety, reliability, and cheapness of fertility control. The range of outcomes chosen may be constrained by peer pressure or social prescriptions. Hence the contrast between neoclassical economists’ interests in what choices people make with what Duesenbery (1960) described as sociologists’ interests in why people have little choice to make. Because we are interested in apparently deliberate behavior in fertility control, we elaborate the economic framework a little further.
Before examining what costs of childrearing might consist of, let us first list ways in which parents may benefit from having children:
- Consumption-like benefits: the intrinsic value of the child to the parents, for its own sake;
- Income earning: the value of the child’s labor to the parents;
- Security: especially in old age and in times of stress; Perpetuation of the lineage;
- Net cash benefits for children from the state
Blake (1972) adds to the customary economist’s catalogue of utilities an item recognizing the social roles of parenthood. It confers status within the kinship system and provides channels for mutual emotional support between generations. Every sort of benefit varies according to the number of children already in the family, and across social settings. It may vary between males and females and be subject to numerous sources of uncertainty. Benefits may or may not take into account the welfare of the next generation. They need to be evaluated as a present value, giving a smaller weight to benefits in the distant future than to those nearer the present. This weight (discount factor) represents the couples’ rate of time preference and any allowance they make to write down risky outcomes. A low discount rate and altruism toward future generations (which tend to go together) would increase the present value of any given stream of benefits expected, and tend to increase fertility. Conversely, a reduction in altruism (intergenerational solidarity), all else equal, would lead to lower fertility. This could be a special case of the sociological notion of individualization (Lesthaeghe and Surkyn 1988).
Among the costs of an additional child are as follows:
- Physiologic wear and tear of pregnancy, childbirth, and lactation (maternal depletion), risks to the mother, and costs of obstetric care;
- Value of goods and services spent on the child, and invested in quality;
- Value of market time devoted to child-rearing: earnings forgone, usually by the mother;
- Value of time diverted from nonmarket activities, such a leisure, sleep, care of existing children;
- Any planned bequest (transfers of nonhuman capital; e.g., dowries)
Once again the importance of each item depends on the context. Uncertainties affect how much any one birth project might cost. A lower risk of infant and child mortality can raise the rate of return on investments in children considerably. Improved child survival is often thought to reduce fertility, by cutting the number of births needed to produce a given surviving family size. But the demand approach to fertility points out that the rising rate of return could induce increased investment. This could involve a higher number of births but the investment might also be in the quality of existing children (Schultz 1981). It is sometimes also assumed that rational parents will assess the child’s own risk or future potential (a child’s endowment) as they make decisions to invest in or curtail the care they give—their investment in child quality.
Even given the context, the cost of any particular child is not fixed, different levels and timing of investment will produce different levels of quality, and different children will need different investments to produce the same result. There is enormous variation in the world as to how much is invested in children, and also enormous variation in the amount of investment that is paid for by the family and the state, depending for example on how medical and educational services are financed. For the moment, we assume that the private costs of children (i.e., those not born by the state) are borne by the family. Distribution of costs between mothers and fathers is considered in subsequent sections. Note that other family members, particularly siblings, may bear some of the cost of an additional child.
The item within costs referring to the diversion of the mother’s productive labor into reproduction has received massive attention. Childrearing is assumed to displace time from other work, and it is the woman whose time is assumed to be interchangeable between production and childrearing, rather than the man’s, because that is how parents generally divide up the work. Almost universally, the woman is the primary care giver, taking responsibility for childrearing and domestic work as well as her biologic role of carrying and nurturing the newborn. The father, assuming he is around, takes the primary breadwinning role in the economy beyond the family. The traditional division of labor leads to gains from specialization, realized by both parties if resources are pooled. It is normal, natural, and, according to Becker (1991) and Cigno (1991), an efficient way of maximizing gains from specialization without running into coordination and enforcement problems that might beset a larger commune. However, the domestic division of labor between men and women is often reinforced by wage discrimination against women in the labor market, which will tend to discourage women’s participation in the paid economy, and might not be economically efficient.
Whatever the merits or shortcomings of this interpretation of the sexual division of labor, wherever it exists, it implies a trade-off between women’s paid work and reproduction, which had come to play a central role in economists’ attempts to model fertility trends in developed countries. Because the family is assumed to be choosing both childbearing and the woman’s labor force participation at the same time, some further factor determining them both must be identified if the chicken and egg problem of reciprocal causation is to be disentangled. Easterlin (1980) suggests that in the (then) contemporary United States the primary factor would be the generation’s relative income and dismisses the liberation of the new woman, as a major explanatory force. This contrasts with the view of Davis (1984), but Easterlin’s dismissal of the importance of women’s social and economic status as an engine of demographic change in the U.S. is echoed by Joshi (1985,1990) also pointing to the very modest nature of women’s advance into the labor market. For new home economics, the independent factor is the woman’s wage if she’s a potential labor force participant. It is the relevant price of her time, reflecting income foregone for every hour devoted to unpaid maternal duties. The higher the woman’s wage, for example because of her education, the lower would fertility be predicted.
Preferences and Values
The set of preferences, which evaluates the objectives in the demand model can almost interchangeably be described as tastes, desires, wants, needs, aspirations, goals, values, interests, and ideologies (Pollak and Watkins, 1993, p. 481). One might add norms. The orthodox approach prefers to treat preferences as given outside the system (Stigler and Becker, 1977), as not being the source of change into which economists should inquire or hope to make predictions and that can either be safely ignored or left to other social sciences. Economists have therefore generally seen little point in subtle differentiation between influences on motivation.
The Pennsylvania School of economic demography is an exception. They declare a need for interdisciplinary dialogue to understand how preferences are formed and may themselves adapt to changes in the economic environment (endogenous preferences). They point out that the Chicago School’s definition of the maximand as a set of unmeasurable commodities generated by household production would enable them to describe the results of a change in tastes as if it had been a change in unobservable household technology. Although the two interpretations are observationally equivalent, the conclusions to be drawn about whether a change is for the better are different. If preferences do not change and a technical change increases household production possibilities, the family must be at least no worse off. If household production changes because preferences change, the utility of the two situations is not comparable (Pollak and Watkins, 1993).
What can be studied about changing preferences are social mechanisms of diffusing new values or norms, and the role that economic change may play in setting the goals to which people aspire. The most well known hypothesis of an economic influence on agents’ aspirations (endogenous preferences) is Easterlin’s relative income hypothesis, which states that young people planning to start family life will attempt to achieve or improve upon the standard of living their own parents managed to set for them. This is also an example of a theory formulated at a macro rather than a micro level.
So far the exposition has gone along with the assumption of home sweet home, the couple and the family acting in concert. This assumption deserves scrutiny. It would be reasonable if we knew that some mechanism was at work regulating the standards of private behavior, ensuring that family members treated one another fairly, if not lovingly, so that each party shared the gains of marriage for example. Religions have generally taken the lead at promulgating family values, and so, to a lesser extent have secular value systems. It was the early Catholic Church, for example, that insisted that each spouse should consent to a marriage. The state, in the form of social workers and family law, sets sanctions for those who break the rules, although it does not normally venture into private territory and it has changed the rule to allow more individual freedom. State regulation of the responsibilities of marriage partners and of parenthood is seen by Cigno (1991) as essential to ensure that contracts are enforced. If the sexual division of labor keeps a woman away from the labor market, she loses, forever, the opportunity to accumulate marketable skills, and is vulnerable should the marriage end without proper compensation.
As Folbre (1994) points out, parents do not necessarily share costs and benefits equally, particularly when the father parts company with the mother. If the father deserts without contributing to the child’s maintenance, the mother is left with double demands on her time and fewer resources to invest in the child’s quality. The absent father probably loses out on the joys of parenthood, but if he stays around he may get more than his fair share of the benefits. If he has patriarchal control over the family, he is likely to contribute less, at least of time and effort to domestic activities and be able to get the children, as well as their mother, to do his bidding.
Situations of marked difference between the economic and social status of men and women are likely to affect parents’ perceptions of the value of a child’s contribution—sons and daughters will be valued differently. Sons are also more valuable, provided parents have access to their earnings, wherever the earning opportunities for females are inferior to those for males. Daughters have an alternative value in principle as auxiliary household labor, but where marriage takes young women away to live with the bridegroom’s family, it is sons’ wives who make this contribution, and any dowry to be paid for daughters would be an example of a bequest adding to her cost.
Female participation in the labor force is often confused with the status of women, although they are not at all the same thing. Employment does not automatically entail autonomy, although it may help. Women’s wages may be a better approximation, as they show upon what terms women had paid jobs and, unless someone has control over their income, the purchasing power that it permits them.
Blank (1993) points out that individuals are differently empowered to take decisions and make choices affecting their lives, for social, psychological, and economic reasons. Relatively low empowerment of women relative to men is a characteristic of gender relations in many societies, but there are differential levels of autonomy across different groups. As the distribution of empowerment changes, one should expect shifts in decisions about the formation of families.
The distribution of costs and benefits depends not only on the division of labor between the sexes, but also on the family organization and availability of other care givers and the societal provision for bearing some of the costs of reproduction (Lloyd, 1994). The costs of fertility may be partly shared by the wider community, for example, in the form of free or subsidized health and education services or tax allowances and cash benefits for children in a modern welfare state. Similarly the collectivity may have an interest beyond the individual in the size of coming generations as contributors to the pension scheme or as home-grown members of the labor force or military recruits. As such, they bring benefits to other citizens besides their parents, particularly those with no children of their own, and particularly where immigrants would not be satisfactory substitutes. In countries where fertility is considered too high, this could be because the public sector is bearing more of the costs than the private decision-makers (e.g., in the Caribbean departments of France) and/or because people beyond the family have to bear adverse consequences of congestion, competition for scarce jobs or resources. Such spillover consequences for others than the decision-maker are technically known as externalities and constitute a reason why one might want public authorities to intervene in private decision making. The effectiveness of public policy to change population trends is discussed in subsequent sections of this chapter.
A Sociological View
A sociologic perspective considers the context of social organization, cultural norms and the sanctions by which societies have attempted to regulate reproductive behavior. Culture, religion, and ideational change mediate the economic influences on fertility, and offer normative constraints on behavior.
Cultural sanctions on reproductive behavior are the meanings organized and shared by groups that emphasize criteria of right and wrong (Kreager, quoted in Kertzer, 1995). Ideology, custom and culture also reinforce the authority structure within the family—between parents and children and between men and women. The importance given in most societies to the regulation of sexuality and marriage and to family and kinship systems, which determine and are the result of reproductive behavior, is evidence that fertility processes are a central concern of cultural systems and social organization (Kertzer, 1995).
Fertility behavior is influenced by both sacred and secular values. The first statement of demographic transition theory by Notestein (1945) suggests that religious values and the moral codes predominant among pre-transition societies are important in keeping fertility high. With the movement toward a secular society, brought about as by modernization, the taboos surrounding deliberate control of fertility break down, bringing on the transition to low fertility.
In this view, all aspects of life are integrated in traditional societies; family life, economics, and politics are all governed by religious norms. With the coming of modern industrial forms of economic organization, these aspects of social organization become differentiated, and individuals are freed from the all-encompassing constraints of the dominant value system. They gain more freedom to choose different values and beliefs, particularly those related to family life. The process of modernization and the development of capitalism in Europe were believed to be accompanied by a move toward secular individualism. Traditional religious ideas gave way to self-expression, in turn affecting the individual expression of sexuality and family values. Parenthood was no longer a dominant moral obligation dictated by religious or social values (Simons, 1986). Lesthaeghe (1983) has suggested that this emergence of individual autonomy in European cultures led to behavioral patterns oriented toward the welfare of nuclear households rather than toward larger kinship groups. As individuals perceived themselves to be more self-reliant, changes in family formation, dissolution, and reproduction took place. These changes, he suggests, involved an increased orientation toward child welfare, and this pattern contributed to reinforcing greater independence from the family itself, as economic production shifted from familial to nonfamilial modes.
The sociological view of fertility focuses attention on how particular forms of cultural and social organization lead to norms and values that influence when and how childbearing takes place. One well-known example of such social norms is those found in parts of West Africa, which regulate when sexual relations may begin following a birth. Such culturally prescribed practices may have arisen to protect the new infant from the competition of a sibling whose birth follows too closely and interrupts breast-feeding, which can lead to malnutrition and ill health or death. Although traditional values dictate abstinence until the child is weaned or walking, these norms may break down with the influx of new ideas or values, such as those accompanying the spread of Islam or the increasing influence of the West. Islam proscribed sexual relations for a 40-day period, much shorter than the traditional norms. With the advent of modern technologies, however, women are able to maintain the prescribed spacing of their births, while resuming sexual relations with their husbands, who may use the new norms to justify their own pressure to resume relations (Bledsoe et al., 1993).
Religious values may be upheld only among small groups, who are able to maintain fertility norms quite distinctive from the society in which they function. One such group is the Hutterites, an Anabaptist religious sect living on the American Great Plains, whose religious principles prohibit contraception and lead to a completed family size of almost 10, a much higher level of fertility than of the surrounding society. Secular society will tolerate nonconformist fertility behavior as long as it remains the choice of a minority and does not threaten to destroy the legitimacy of the prevailing norms and values (Simons, 1982).
Chamie (1981) studied religious differentials in fertility among Lebanese women, and found that non-Catholic Christians had higher fertility than Catholics and both groups had lower fertility than Muslims. However, he also found substantial differences among the different Muslim groups, fertility highest among Shia women, slightly lower among Sunni Muslim women, and among the Druze somewhere between that of Catholics and non-Catholics. It seems it is not religious norms and values, per se, that lead to these differences, but other norms shared by different religious groups. Such would fit with Folbre’s view that group membership helps to determine reproductive behavior. In some cases these pro-natal norms are strengthened by the minority status of the group, feeling the need to keep up their numbers in a somewhat hostile environment. Such an explanation has been offered for the relatively high fertility of Muslims in India and Walloons in Belgium.
There is also heterogeneity in the fertility of Israeli Jews, dependent not on ethnicity, but on the religiosity of certain groups (Friedlander and Feldmann, 1993). In Israel, a modern developed society by most standards, the fertility rate of around 2.8 births per woman is still well above most European countries. This national average disguises large differences in fertility between religious groups. The total fertility rate (TFR) is 4.5 among all religious Jews in Israel, while ultra-Orthodox Jews have very high fertility, a TFR of 7.5. This group of highly religious Jews shuns the values of the outside world, while sharing group norms that lead to early marriage and very high fertility. However, their high fertility does not entail extra cost to parents because family arrangements are such that younger children are looked after while women work and men spend their time in religious study.
Although rejecting the values of the outside world, this pattern of childbearing and rearing is supported by the wider society. The success of these pronatalist religious values is due in part to arrangements in the wider society, which provides economic supplements such as child allowances and housing assistance, and contributions from external religious groups, their own institutions and the government. In this case, it appears that the pressure on individuals to conform to religious sanctions and fertility norms is powerful, and as the size of this community increases, their family-building pattern could result in increased political power.
The historic evidence from the European Fertility Project (Coale and Watkins, 1986) exposed the shortcomings of Notestein’s classic theory of demographic transition. This relied heavily on modernization and changes in the mode of production to explain fertility change. This project suggested that fertility was most strongly related to cultural boundaries, defined by language, ethnicity, or geography. Cleland and Wilson (1987) synthesized this historic evidence and contemporary findings of the World Fertility Survey program, conducted in less-developed countries around 1975. Their findings counter the economic view of fertility behavior as the product of rational decision-making. They present evidence that fertility decline was related less to socioeconomic factors than to social variables and that the diffusion of new ideas about the acceptability of modern birth control was its main determinant. The formation of the nation state in the process of modernization creates another mechanism for direct and indirect influence on reproduction, at macro- and microlevels (see Casterline, 1989 and subsequent sections).
Many attempts to explain demographic processes, including fertility behavior, have been post-hoc explanations for observed changes, mostly driven by the types of data conventionally collected in surveys (McNicoll, 1980). Lacking a theoretical perspective, these explanations have yielded a puzzling array of possible factors and configurations that may lead to reproductive change. McNicoll elaborates what has been called an institutional theory of fertility change. Unlike ideational theories, this allows some role for the vastly differing contexts, both historic and current, in which communication about fertility takes place. He argues that institutions rooted in the history of each society give rise to local patterns of social organization: the family and local community; law, and the local dimension of public administration; stratification and opportunities for mobility; the labour market; and state-individual relations. These institutional endowments, as well as the transactions individuals make in adapting their hopes and expectations about the future to the institutional constraints, together trace the path a particular society takes to fertility decline (McNicoll, 1994). Individuals are not merely passive receptors of social norms, but actively try to re-negotiate the institutional constraints in their own interest. Such a theory attempts, as Greenhalgh (1995) puts it, to situate fertility in its cultural, political, and economic context. Such an approach provides the possibility of testing hypotheses about how particular aspects of social organization translate into microlevel reproductive behavior, providing a role for individual action in shaping reproductive outcomes. Folbre (1994) would add collective action by interest groups to the script. Writers concerned with the influence of social values on fertility include Lesthaeghe and Surkyn (1988), who cite the demonstration by Simons (1982) that the baby boom and bust of 1945 to 1985 in England and Wales was foreshadowed by an indicator of religious adherence (Easter Day Communicants). Preston (1986) also believes that value changes play an important role in fertility trends, although he feels they can ultimately be traced back to an economic rationality.
Even in societies that have not yet undergone a fertility decline, people are not merely passive acceptors of normative constraints. Recent evidence from West Africa suggests that individuals tinker with social norms regarding the spacing of births (Bledsoe et al., 1993). However, in West Africa cultural norms do not include the idea of stopping after a preferred family size has been reached. This suggests that although the means of fertility regulation are well known, it is changes in ideas regarding the acceptability or desirability of family limitation within marriage that leads to the stopping behavior characteristic of significant fertility declines.
The social construction of the family differs across cultures and may explain why fertility has not declined in some regions, even those where development is far advanced. Cleland and Wilson found that sub-Saharan African and Middle Eastern societies differ from other regions in fertility preferences. Socioeconomic differences in fertility preferences in these regions are more pronounced than divergences in fertility itself.
The social organization of the household—itself an outcome of cultural forces that lead to different configurations of household structure—influences marriage timing and the ability of the household to support dependents. Because these seem bound up with fertility differences, a key question, then, becomes, “What determines family systems?” In much of sub-Saharan Africa, the boundary of the family extends to a large kin or lineage group, which makes it more likely that the larger group to which individuals owe allegiance will retain much control over fertility behavior. Practices such as child fosterage, arrangements where the extended family provides child care, and state provision for education, “break[s] down household economic boundaries and spread[s] the impact of additional children on family resources across a wider kin network” (Lloyd, 1994, p. 11-12). Such institutional supports to high fertility may prevent the spread of the small family norm in these regions.
The family mode of organization framework hypothesizes that the extent to which the family organizes individual members’ activities plays a role in fertility change (Axinn et al., 1992). This can occur in two ways. Fertility change may be caused by a changing demand for children as a result of desiring to maximize the outcome of the new, nonfamily organization of activities (structural economic change) or may be due to ideational change, a result of the diffusion of new ideas and fertility goals through alteration of the group (nonfamily) with whom family members have contact. In this way, this approach attempts to connect macrolevel social change to individual-level fertility decisions and behavior. This seems a promising way to approach a test of the rival economic and sociologic theories of fertility change, but there is still scant evidence on which to judge these hypotheses.
Questions About Fertility and Society
This section looks at some of the answers, and controversies, that social scientists have offered to certain questions. Answers are summarized at the end of each section.
Why Does Fertility Decline with Development?
One of the greatest paradoxes facing the demand approach to fertility is to reconcile the historic (at least) association of fertility decline with economic development. In the economist’s view, the latter increases the resources available to potential parents. If their decision making resembles that of a consumer demanding a normal good (i.e., one for which demand increases as income increases), one might expect economic growth to raise the demand for progeny.
One interpretation is that positive income effects are outweighed by increasing costs of parenthood as economies develop. There are higher standards of consumption to be provided, fewer opportunities for child labor, and increasing investment in education and training required by an increasingly sophisticated economy.
Just as economic growth involves the accumulation of physical capital, it also requires increasing standards of skill and sophistication among the workers. New home economics appeals to its distinction between child quality and child quantity in this context. Economic development offers improved opportunities for investing in child quality (better survival, better schools, better employment prospects). This encourages parents to invest more resources per child than they would have done at a lower level of income and, given the right assumptions about relative income elasticities, to substitute quality for quantity. They cut family size to raise the quality of life for each child in the smaller family. This resembles propositions reached independently by evolutionary biologists. Although the detailed apparatus of this theory is difficult to prove or disprove, it is consistent with much historic experience. Fertility decline is associated with improvement of child health and the spread of schooling (investment in human resources) and with accounts of the prospective cost of education as being a motive for controlling fertility (Crook, 1997).
Other ways in which economic development may reduce the demand for children could be pressure of space in urbanization; competing employment opportunities for women, as the home and the workplace become separate; and the reduced risks in the economic environment against which a large family may be viewed as insurance. Women’s employment and insurance are discussed in subsequent paragraphs.
The classic demographic transition theory was based on the idea that differences in fertility are rooted in economic structures. Traditional peasant economies are organized along familial lines, whereas capitalist economies are not. With increasing industrialization comes disintegration in the familial mode of production and the pronatalist structures it engenders. As industrialization spreads, economic modernization is accompanied by the diffusion of the small nuclear family norm. As we have seen, the findings of the Princeton European Fertility Project did not support the classic view of fertility transition. Change spread to all social groups too quickly to reflect real changes in structural economic conditions. A threshold was apparent: when fertility declined by 10%, it tended to proceed quickly to low levels (Watkins, 1986). The timing of fertility change seemed to follow cultural boundaries and to be associated with indicators of social development, such as education or literacy, rather than with economic development (Cleland and Wilson, 1987; Brass and Kabir, 1982), and its pace was closely associated with a shift from fundamentalist to secular world views (Lesthaeghe and Wilson, 1986).
Cleland and Wilson (1987) argue that the fundamental force behind fertility decline is not the changing balance of costs and benefits of childbearing, but a change in ideas and norms, acting independently of the level of economic development. They suggest that “the fundamental forces of change operate at the societal level” (p. 24), not at the microlevel of individual decision making. Differences between high- and low-fertility groups in the World Fertility Survey are not to be found in their reported preferences for children, but in couples’ willingness or ability to regulate their fertility in accord with these preferences.
Fertility transitions, both historic and modern, are predominantly driven by stopping behavior. After a specified number of children, couples adopt birth control or otherwise limit the size of their families. One suggestion is that modernization (by this we mean the shift in economies driven by technological advances that result in increased output) brings about a spread of new ideas, a change in mentality, from a belief that family size is a result of chance to a clear notion of what family size ought to be, in van de Walle’s phrase numeracy about family size. Means of regulating fertility are known in most societies, but “any contraceptive technique is probably most effective after the couple has reached the number of children they do not want to exceed” (van de Walle, 1992, p. 501).
How these new ideas arise and why new norms develop are still unanswered questions. Cleland and Wilson (1987) argue that the growth in human mastery over nature and widespread mortality decline has facilitated the spread of fertility control by undermining the reproductive imperative. Once the latent desire for lower fertility appears, it is the spread of knowledge regarding effective means of birth control that eliminates excess or unwanted births.
Development is likely to facilitate the spread of knowledge. Development is also likely to induce the latent desire for lower fertility, particularly if it is of a broad-based, rather than trickle-down nature. But economic development cannot be claimed as a necessary and sufficient condition for fertility decline as implied in the Best Contraceptive slogan.
Finally, the explanations of the link from development to fertility decline are multiple and not mutually exclusive: reduced benefits; increased costs; demand for quality; improved contraceptive technology or delivery; changed perceptions of what behavior is advantageous, appropriate, or even thinkable. These could act singly or jointly. What operates in any particular situation does not necessarily apply in others. Just as economic development cannot be simply replicated from one country to another, the course of demographic development can follow diverse paths, reflecting local features of both social and economic structures, and drawing on the technology and experience of those who have gone before.
Can Fertility Decline in Poor Countries?
A closely related debate is whether fertility decline is possible without economic development. The view that it is not, or at least not likely, goes back at least as far as the Bucharest Population Conference of 1974, when the slogan was coined, “Development is the Best Contraceptive.” This echoes classic demographic transition theory that social and economic development is its prime cause. Development reduces dependence on children or raises the cost of rearing them. Thomas (1991) restated this slogan. He argues that programs offering the means of fertility control will find more willing takers if accompanied by measures to reduce the risks or provide alternative sources of insurance to poor parents.
In poor societies, the need for child labor, either for wages or in family enterprise, is one source of the demand for children. Insurance against risk in old age or in times of crisis, which children can provide, is another (Cain, 1982). One source of security, available to some, is land holding, but any impact this may have on the need for a large family through greater security may be offset by a demand for child labour. (For further discussion of old age security see subsequent sections.)
In the joint-family systems of much of Asia, old-age support is a family concern (Cain and McNicoll, 1987). Sons bear the primary responsibility for care of elderly parents, leading to a demand for sons, and a good probability that at least one will survive to support his parents in old age. Considerations of security, therefore, can outweigh the cost-benefit accounting, which might precede the decision not to have another child (Cassen, 1978).
Furthermore, the high mortality levels usually prevalent in the poorest societies may lead to more births in order to replace children who die to ensure that at least some children, especially sons, survive to adulthood. In poor societies, where education levels are low, the costs of raising children may be exceeded by the benefits those children will bring when they reach the age at which they are net contributors to the family income. For a number of reasons, therefore, when parental income is low, the emphasis is on child quantity rather than quality.
Nevertheless, traditions of filial support may be compatible with low fertility (Cleland, 1993; Weinstein et al., 1990). Examining fertility decline in Bangladesh, a range of factors thought to reduce demand for children were reviewed (Cleland et al., 1994). No significant changes in school enrollment, the value of children’s labour, social security benefits from the state, female participation in the labor force, or improvements in the educational or social status of women were found. Conversely, some argue that economic reversals can induce fertility decline. For example, increasing landlessness could drive families from agriculture to the service and sales sector, where fewer children are needed to supply labor and costs of raising them increase. The Bangladesh study detected no significant deterioration in the economy to an extent that might decrease demand for children. Despite the evidence that factors affecting demand for children had not changed, fertility began to fall in the late 1970s, accelerating in the mid 1980s, reaching the level of about 4.5 children per woman by 1990. The authors suggest that the long-term decline in child mortality in Bangladesh led to a greater willingness to adopt innovative behavior. All that was necessary was access to the means of fertility control, provided by a strong government family planning program, which effectively removed barriers to use of contraceptives by delivering advice and supplies at the doorstep. Despite the evidence that demand for family planning was low before the initiation of the national program, its comprehensiveness appears to have maximized the spread of new behavior and sustained it.
These conclusions may be compared with those of Dow et al. (1994) for Kenya. Here too, the resurgence of a family planning program accompanies fertility decline, but a deterioration of economic conditions is judged to have triggered the motivation to seek its services. However, identifiable transfers (that is, support) to the elder generation did not appear to have changed.
In yet another context, Martine (1996) examined the fertility decline over a period of intense social change and deep recession in Brazil. Total fertility fell dramatically and very quickly from an average of six births per woman in the early 1960s to 2.5 births per woman in the mid 1990s. This was achieved primarily through high rates of abortion and sterilization. Martine traces increased motivation for fertility regulation to institutional changes in health and social security. The expansion of health services served to legitimize the use of birth control, even in the face of strong church opposition, and increased the level of knowledge about, and access to, effective means of regulating fertility. Government-sponsored attempts to modernize also accelerated rural-urban migration, and increased interest in consumer goods, which in turn increased the costs of childrearing. Mass media are also given credit for promoting consumption and eroding traditional ideas about male authority. Coverage of the social security system was expanded and a guaranteed income in old age established. This shifted responsibility for the elderly from the family to the state, and despite deliberate pronatalist elements of the social security system, such as child allowances, the impact of the system was in fact antinatalist. Changes in government policies played an unintended role in Brazil’s fertility decline.
Martine (1996) places less weight than others on the argument that Brazil’s economic difficulties lessened demand for children (Carvalho et al., 1981). Rather, his evidence suggests that a constellation of factors—structural and ideational—are at the heart of the sustained transition. The aggregate analysis is persuasive, but does not directly connect these changes in the larger society to mechanisms at the individual level. Can diffusion of new ideas be linked directly to changes in individual behavior?
To put it briefly, fertility has declined in poor countries and poor communities. This is not necessarily as a result of poverty. Ideational and structural factors short circuit any iron link between development and fertility decline.
Women’s Education: Materialist or Ideational Influence?
One oft-cited benefit of development is the increased access to educational opportunities that may accompany changes in the demands of the labour market. How increased access to education affects the status of women in general, and their fertility, is a major concern of policy makers today.
It is well known that, for a number of reasons, educated women are more successful than their uneducated sisters in bearing and rearing healthy children. They are thus less likely to bear more children than their family size goals (see Cochrane, 1979 and Cleland, 1986, for evidence on maternal education and fertility; Cleland and van Ginneken, 1988 for a review of the evidence regarding child survival). The strong and nearly universal link between female education and fertility decline has provided further evidence on how and why reproductive change occurs. It suggests that the effects of education are both material and ideational in nature.
Education can indirectly influence fertility through several pathways: by affecting supply factors, the demand for children, and the costs of control. The mechanisms through which it works are unclear, partly because in some circumstances very small amounts of school attendance—just a few years—are sufficient to observe an effect. Given the unequal and very uncertain quality of education widely available in many societies, this effect can hardly be due to actual knowledge imparted and makes it unlikely that strikingly different attitudes toward the world are imparted and absorbed by girls in so short a time. Some have proposed that besides exerting a powerful independent effect on fertility, education (or schooling) may act to change reproductive behavior by increasing a woman’s autonomy (her ability or freedom to make decisions about her own fertility) (see Basu, 1996, for a detailed exposition of these issues). Moreover, the relationship may partly reflect selectivity especially where the general education level is low. That is, educated women are more likely to come from families of higher status or wealth, who want to educate their daughters and also may impart an orientation to lower fertility. This confounds the relationship between fertility and education (Sathar, 1996). We now look at the evidence for all of these relationships.
Jeejebhoy (1995) provides us with an extensive review of evidence on the matter from demographic surveys. Education can influence the supply factors by breaking down traditional norms about breast-feeding and postpartum abstinence. Even a little education shortens the duration of breast-feeding. Breast-feeding may be incompatible with the employment educated women seek. They are also more likely to be aware of and able to afford alternatives to breast milk. As education increases, postpartum abstinence decreases, and this is thought to be due to a greater intimacy between husbands and educated wives. These education effects raise fertility, unless countered by effective birth control.
On the other hand, education raises the age at which a woman marries (Appleton, 1996), reducing the length of her reproductive life. The reasons are many: Educated women have a greater role in decisions about whom and when they marry; they are likely to have greater control over resources due to employment before marriage; and, in some settings, they are less marriageable than uneducated women (Jeejebhoy, 1995). However, postponement of marriage may simply be followed by couples catching up to those marrying earlier by increasing the pace of subsequent childbearing (Sathar, 1996).
Women’s education, at least beyond a few years of schooling, can also reduce the number of desired children. Economic explanations emphasize the increased costs of an educated mother’s time and increased rationality of decisions about family size and desired sex composition. When the education of women becomes widespread, the value of their labor increases. They become more active in the labor force, and more economically independent. This in turn can reduce the felt need for male children, either as labor or as security in old age (Cochrane, 1979).
Women’s education can also reduce the obstacles to contraception, lessening the psychosocial costs. There is a direct relationship between education and the practice of contraception, due in part, perhaps, to more frequent interactions with the outside world (Basu, 1992). Thus, educated women are also more likely to know about birth control and to have better access to effective methods because they have more experience in dealing with the world beyond the household. In this sense, education may work by facilitating the spread of knowledge about reproduction.
Another mechanism proposed is that education increases a woman’s status and autonomy, giving her more control over her own fertility and affecting her motivation to bear children. There is a good deal of evidence that a higher threshold level of education must be reached in highly stratified societies than in egalitarian ones before fertility starts to decline (Jejeebhoy, 1995). This may be especially true where gender stratification is strong, because it limits the domains over which educated women can make decisions and determines the amount of education necessary before they are taken independently. The social context also conditions the relationship between education and a woman’s motives to bear children. Where her opportunities are greater, she may gain status in other ways than by bearing sons, and may need to depend on them less for security in old age. In gender-stratified societies, more education is necessary before women are confident of their own economic and social status to revise their fertility goals (Jeejebhoy, 1995). It may be that the education-fertility link is strongest where education actually can be translated into employment.
Education may also help to increase a woman’s autonomy over her own body, including her right to refuse sex (Orobuyloye, 1995). It may also lead to increased husband-wife communication regarding fertility decisions and use of contraception. Evidence from Africa suggests that educated women are more likely to be in agreement with their spouse’s fertility preferences than uneducated women (Bankole and Olalaye, 1995), but this does not always result in lower fertility. In Kenya and Ghana, for example, men and women share a common goal of high fertility.
Educated women are more likely to marry men who are also educated, and although women’s education is a consistently stronger predictor of fertility than their husband’s education, his role in shaping fertility behavior is clearly pertinent. Husbands’ roles in family-building strategies and decisions are important, but the meager evidence available does not reveal any consistent associations between husband and wife preferences and other factors (Mason and Taj, 1987). Basu points out that both men’s and women’s fertility preferences are likely to be affected by the position of women in society, so that for example, if women are denied a place in the labor market, both husband and wife should want more sons. This may extend to the other actors in the household who influence family-building strategies, such as mothers-in-law (Basu, 1992). Others have suggested that a husband’s non-family experiences (such as changes in residence, and work and leisure activities outside the family) have a positive effect on contraceptive use (Axinn, 1992). Opportunities for work and residence outside the family may depend to some extent on the husband’s educational attainment. If research designs exclude men, important mechanisms involved in behavior change may be missed.
Jeejebhoy’s extensive review of education-fertility linkages concludes that in societies with strong gender stratification, the most powerful effects of education are to reduce traditional inhibitions on the supply factors—to raise fertility. Only in more egalitarian settings do the fertility-inhibiting effects of education—on marriage age, family-size preferences and use of contraception—counterbalance or exceed these. This may explain why the relationship between education and fertility is less clearcut than the relationship between education and contraceptive practice. In some settings contraception may be initiated only after unwanted births have already occurred. Jeejebhoy (1995) argues that this may be the case where women achieve the power to make reproductive decisions rather late in their married lives, when desired family size has already been exceeded, rather than earlier in their life course.
In highly gender-stratified societies, religion and patriarchal kinship structures shape marriage practices and female economic roles and sometimes make them seem impervious to education, or at least resistant to its effects. Cultural systems and social organization that strongly influence reproductive behavior often have their roots in the predominant economic structures and modes of production, at least historically. For example, higher levels of literacy and employment among South Indian women do not explain observed differences in female autonomy—and fertility—between North and South India, which have deeper historic roots (Basu, 1992).
This may be one reason why community context—community-level effects—seems to operate when we look at the relationship between fertility change and education. Generally speaking, high societal levels of education are strongly correlated with low fertility at the aggregate level (Cleland and Jeejebhoy, 1996). However, the overall level of education in a society can exert a powerful effect on its own to influence the fertility behavior of individuals regardless of their educational attainment. Cleland and Jeejebhoy show that, in the Indian states of Kerala and Tamil Nadu, where literacy rates are high, fertility at the individual level is inversely related to years of schooling. Both states have much lower total fertility than the low-literacy states of Uttar Pradesh and Bihar. In the latter, however, there is no consistent relationship between a wife’s education and fertility at the individual level. The contextual effect of education exerts a strong influence on fertility behavior that is independent of its individual effect.
A striking example of how contextual factors may change is illustrated in a study examining education and fertility decline in two Chinese provinces (Freedman et al., 1988). Early in the shift from moderate to low levels of fertility in both provinces the educational level of the production brigade had an effect on fertility that was independent of the effect of an individual woman’s educational attainment. In rural areas, fertility of women at each level of education increased steadily with the proportion of the local population that was illiterate, and where the proportion illiterate was as high as 40%, the effect of individual educational level was inconsistent (as seen in Uttar Pradesh and Bihar).
Following large-scale campaigns and intensive family planning program effort, these local (contextual) differences were leveled, and the overall level of literacy in the area was no longer consistently related to the level of fertility. Moreover, the effect of education at the individual level also disappeared very shortly after inception of the government program. Where the government family planning program was strong, knowledge and supplies quickly became available to everyone in the area, and social differentials in fertility disappeared (Riley and Gardner, 1997). Government policies seem to have the power to diminish educational differences in contraceptive use and fertility.
LeVine et al. (1991), tried to investigate what processes at the individual level that lead to lower fertility might be affected by increased education. Their study in two Mexican communities found that mothers with more education used child health and family planning services more frequently than less-educated mothers. Both factors increased child survival and lowered fertility, but some intervening processes were found that provide clues to how increased schooling may translate into changes in fertility.
Using in-depth interviews and observation, they found that in contrast to less-educated women, more schooling led women to aspire to higher status for themselves. They were more likely to report making major household decisions jointly with their husbands. Educated women also spent more time talking to their infants and were more responsive to them than uneducated women. Such behavior has long-term effects on an infant’s cognitive development. They expressed higher occupational aspirations for their children, while reportedly expecting less help from their children in old age. These findings confirm that schooling influences the attitudes of individuals, and also points to the effect education appears to have in increasing parental investment in a child’s human capital or quality, directly raising the costs of children by demanding more of a mother’s time. Whether new fertility preferences and practices preceded and promoted these educational differences in attitudes and behavior or resulted from them cannot be deduced, but this study is evidence of schooling’s wide-ranging effects.
Finally, there are multiple mechanisms by which increased female education affects fertility behavior, and good evidence that its influence is both materialist and ideational in nature.
Is Childbearing for Old Age Security?
Caldwell’s theory of the fertility transition is an elegant synthesis of material and social forces (Caldwell, 1982). In the process of development, not only does fertility decline but there is also an associated reversal of intergenerational wealth flows. In the high-fertility society, the elders are in control and draw economic support from their adult offspring. This creates a motive for high fertility, or at least an obstacle to the adoption of family limitation. The more surviving children, the more secure will be those parents who survive into old age. Those with few or no surviving children will be endangered themselves. In the course of development, children become a less reliable source of support; the gerontocracy weakens, especially as the younger generation gains skills and becomes more mobile. Even without enforcement problems, alternative sources of security become more attractive: physical assets, purchased insurance, the capital market and related institutions like credit unions or friendly societies, and, eventually, social insurance pensions providing old age security through the state. Caldwell notes that the transition does not just involve the cessation of wealth flows from younger to older generations, but a reversal. Parents invest more resources in the human capital of their offspring, and bequeath to the next generation whatever has not already been used up.
There is conflicting evidence on whether the wealth flows are always in the directions Caldwell posits (e.g., Dow et al., 1994). In rich counties there are gross flows in both directions, and mutual dependence, in some cases of the elderly and the middle-aged (Thane, 1989). In some poor settings, fieldwork among primitive groups in South America suggests children do not contribute more than they consume and points to other motivations or drives to reproduce (Kaplan, 1994).
People may still want to call on family members to provide physical care, companionship, or emotional support in old age, and in some cultures it is important to know that a descendant will be on hand to lead one’s funeral rites. In India, it has traditionally been important for a son to survive to perform this function, but isolated, although significant, incidents have recently been reported of daughters taking this responsibility (P. Visaria, personal communication).
In more developed societies, with increased life expectancy, the issue of old age dependency gains salience over that of who takes care of the funeral. The establishment of collective institutions for the material support of the elderly in developed countries removes one reason for having children. Cigno (1991) points out that there must be other reasons keeping childbearing going at all. He also points to the irony that the pay-as-you-go pension systems, mostly established when the elderly were relatively less numerous than they are becoming, are facing severe difficulties because of the declining size of the contributing generations. If their existence is deterring childbearing, pension schemes can be said to contain the seeds of their own destruction. One may not need to have children to provide a private source of income in old age, but someone has to rear the next generation of contributors to the pension scheme. This externality has led to the suggestion that pension entitlements should be relinked, positively to the number of offspring produced (Demeny, 1987; Vallin, 1992).
At present, a large class of prolific pensioners are actually penalized in the cash support they can expect. Most state (and all private) pensions are positively related to the earnings. Women who have raised children often forgo at least some of their earnings during their labor force years, which tend to be reflected in lower pension (Joshi and Davies, 1994). Thus the opportunity costs of childbearing are carried on into old age, reinforcing low fertility and further undermining the long-term viability of the pension scheme. In the short term, elder dependents and child dependents appear to be in competition for the resources available. In the longer term, investment in the young is still essential to ensure resources for the old (Qvortrup, 1994).
Developed economies have many institutions other than parenthood to secure income and services for the aged. Less-developed societies have hitherto had relatively few aged people to support and these have not necessarily been dependent on their own offspring. The empirical importance of the old age security motive for childbearing has been inconclusively debated. Old age security is still a reason for reproduction in aging societies, but one that is more likely to be acted on collectively than individually.
What Makes Fertility Fluctuate in Rich Countries?
Has the fertility transition finished? Have the developed countries with their high-quality population, their highly developed state and market institutions, and their high levels of birth control technology reached a new low-fertility equilibrium? Will decline continue to be sustained below replacement and beyond or is it the nature of reproductive behavior in advanced societies to fluctuate?
Richard Easterlin’s famous hypothesis is that in mature and stable posttransitional society, fertility would have a tendency to fluctuate, sometimes below, sometimes above replacement. As spelled out in his popular book, Birth and Fortune, the central mechanism is relative generation size and its influence on aspirations relative to earnings (Easterlin, 1980). Members of small cohorts would reap advantages in education and the labour market; young adults would find life relatively easy, so that their aspirations for living standards would be attained sooner, and permit the forming of larger families than generations who were more numerous relative to their predecessors. A key link in the chain is the assumption that standards are set by those the cohort enjoyed during their upbringing. This relative income hypothesis is an example of how preferences are determined within the system (endogenous). Easterlin argues that relative generation size determines a number of other series, from wives’ labour force participation, to divorce, suicide, and crime. The evidence that relative generation size affects earnings is mixed (Wright, 1991). The evidence that a small generation became the parents of a baby boom is strongest for just one cycle in the United States and perhaps Australia, Canada (Wright, 1989), and France (Leridon, 1978; Chenais, 1983). Ermisch (1988) did find a modest contribution of relative generation size, among other variables, to his analysis of parity-specific birth rates in England and Wales between 1952 and 1985.
The prediction of self-perpetuating swings has not been borne out. By the 1990s, the children of the 1960 to 1970s baby bust should have been generating a new boom, but fertility in the United States has only moved from around 1.8 in the 1980s to around replacement (2.1) in 1990 to 1992 (Monnier and Guibert-Lantoine, 1995). Perhaps the conditions of economic stability on which the hypothesis was founded have not held out or perhaps the model never really worked.
A rival model, which also predicts fluctuating fertility in advanced economies, was proposed by Butz and Ward (1979) building on the microeconomic Chicago school model of Willis (1973). Their prediction was, contrary to Easterlin’s, that fluctuations in fertility would be counter-cyclical: more births in hard times than times of prosperity. Their reasoning was based on the growing proportion of wives in the labour force, for whom increased wages would mean substitution away from childbearing. As the proportion of two-earner couples increased, presumed positive income effects of men’s wages would be outweighed by the negative effect of rises in women’s wages. The upswing of the 1950s baby boom would fit with a regime where income effects dominated. The end of the boom would be triggered by sufficient two-earner couples confronting a continued wage growth. The model was fitted to data for 1948 to 1956 and successfully predicted 1957 to 1975, but its fit to the earlier part of the century or to subsequent years is questionable (Olsen, 1994). Given continued rise in women’s labour force participation, we should expect continued decline in fertility into the 1990s, but as we have seen, such a trend has reached a plateau (in most other advanced countries except Southern and Eastern Europe and Japan).
One reason for the failure of these models to find confirmation in statistical data may be the annoying habit of history not to stand still long enough for a cyclic process to display any regularity. Another view, perhaps another way of saying the same thing, is that there are other factors to be taken into account. These include the policies adopted by governments (described in subsequent sections), the growing compatibility of motherhood and employment, particularly via the purchase of child care, the growing instability of marriage, the dissociation of fertility from marriage, and a climate of uncertainty and global insecurity (Ermisch, 1990, Olsen, 1994). Most of these factors point to continued low fertility, if not continuing flight from parenthood. Hobcraft and Kiernan (1995) see “little prospect of a resurgence of parenthood” without the help of “biological and social pressure to reproduce.”
A word is necessary about the role of contraceptive innovation in accounting for fluctuations in fertility in industrial countries. In particular it is very widely believed the introduction of the contraceptive pill brought about the 1960s baby bust, and that periodic scares about the health hazard of the pill bring about blips in the birth rate as users rush away from that type of birth control without substituting another. Murphy (1993) argues that changing attitudes to the contraceptive pill are a major element in short-term fluctuations in British fertility. In the work by De Cooman et al. (1987), evidence of pill scares was not abundant, but one year did show a slight excess of births, otherwise unexplained. An alert about the health effects of some types of pill in October 1995 was followed by a modest rise in both abortions and maternities in England and Wales (Wood et al., 1997).
The fact that there is a range of modern and traditional means of contraception should mean that changes in the price, availability or quality of one product would result in people choosing another method. Some of the boom births were described as unwanted in demographic surveys (Westoff and Ryder, 1977). Using demographic surveys for France, Leridon (1985) estimates that 21% of births in 1965 were not desired, and an equal number were not planned to arrive at that particular time.
It is not only the absence of effective methods of birth control that made this situation possible, but also a general context in which the child that was one too many was not altogether so … Contraceptive practice may have slackened during the 1960s due to the fact that couples’ intentions had become less certain during this period.
Over the years 1965 to 1975, as contraception improved, unwanted fertility was halved, accounting for a drop of 0.32 on the period fertility rate. However, most of the births in both 1965 and 1975 were planned. The fall in planned births accounted for 0.28 and in mistimed births by 0.21. Even the unwanted births could not have been too seriously unwanted as the knowledge, in principle, existed of how to prevent them. Even if the methods were somewhat old-fashioned, inconvenient, and/or unreliable they served to keep fertility low in the 1930s. One therefore needs an explanation of the motivation, as well as the means, to end the baby boom. Certainly, the demand approach would admit that improved means of birth control would tip the balance in favor of its use at the margin, and could theoretically have a large impact if the conventional methods were insuperably distasteful. There is also the argument that the users of birth control became more effective, or careful, as well as the methods themselves.
Reliable birth control, accessible to all women has a number of other consequences, which could in turn help bring about fertility decline. It has permitted widespread sex outside marriage. This increased opportunities to form, and move between, partnerships. It has increased the freedom of both men and women, which has arguably led to the revision of norms about the formation of sexual unions. The easier sexual mores added to other factors weakening the institution of marriage (and ironically led to extramarital births). With partnerships less solid, people become more cautious about committing themselves to parenthood (Ermisch, 1990).
Posttransitional fertility is still relatively recent, even in industrial countries. It may still be too soon to generalize, but it does not seem to be characterized as either a regular cycle or an inexorable downward trend. No universally robust predictor has been found. Those that have been considered include the relative generation size of men, the relative wages of women, the reliability of marriage, the level of religiosity, and the effectiveness of contraception. The latter may be the outcome of choices rather than a determinant.
1 Ce n’est pas seulement l’absence de méthodes efficaces de regulation des naissances qui a rendu une telle situation possible, mais aussi un contexte général dans lequel l’enfant “en trop” ne l’était pas tout à fait…. Il a pu se produire dans les années 60 un relâchement de la practique contraceptive, parce que les intentions des couples étaient devenues plus incertaines à cette époque.
Are State Policies Effective in Accelerating Fertility Decline?
Few issues have raised such heated debate in international circles as the question of whether population growth is so rapid that its effects should be curbed through deliberate antinatalist government policies. The effectiveness of state policies in curbing the rate of growth through promotion of family planning is a major concern of demographers. There is still a good deal of skepticism that policies can be effective in accelerating fertility decline due to the widespread belief that only socioeconomic transformation can succeed in providing couples with the motivation to reduce the number of children they bear. Crook (1977) points out that there was no state encouragement of the fertility decline in England and Wales, in the 19th century, quite the contrary. Yet almost all third world fertility declines involve some degree of state involvement. Mauldin and Berelson (1978) and later Lapham and Mauldin (1984) were among the first to attempt a systematic analysis of the impact of family planning program. These studies pointed to the association between the level of socioeconomic development and a measure of program effort, but the major conclusion was that strong programs do reduce fertility, a finding that reassured governments and donor agencies (Lee et al., 1995).
The 1994 International Conference on Population and Development, held in Cairo, marked a growing consensus among developed and developing countries, and among women’s and environmentalist nongovernmental agencies, that they share an interest in advancing women’s access to family planning. National policies and plans of action were agreed in Cairo, which had the objective of “incorporating population concerns in all relevant national development strategies” (United Nations, 1995, p. 67). There remains a need, however, to examine the factors that promote the effectiveness of policy and related public family planning services (Lee et al., 1995). The debate still hinges on whether family planning programs can effectively reduce unwanted fertility, in the absence of other broader social change.
Pritchett (1994) argues that the challenge is to reduce people’s fertility desires, not to address the unmet need for family planning. He argues that the question is still whether family size preferences are determined by economic influences or social and cultural forces, concluding that family planning program effort, which provides knowledge and access to the means of control, is not a dominant factor determining differences in fertility. Rather, a low level of demand for children is both necessary and sufficient to reduce actual fertility. While pronatalist policies that reduce access to contraceptives will weaken the effect on fertility of any changes in preferences, the reverse may not work. Improvements in access to contraceptives are neither sufficient nor necessary for fertility reductions to occur, he contends, contrary to the study in Bangladesh. Some active family planning programs have gone on the offensive to recruit users using coercion. The Emergency in India in 1975 and some stages of the One Child Policy in China would be cases in point. Economic incentives (and sanctions) are then focused on the motivators and officials of the program (Crook, 1997). Curiously the much resented manner of the Emergency Family Planning Drive resulted in the overthrow of the Government of India, but not a sustained family-planning backlash.
New work to assess the demographic consequences of divergent population policies in a series of comparisons between pairs of countries similar in terms of crude socioeconomic indicators has recently been completed (Lee et al., 1995). This study’s main conclusions were that when policy elites, responsible for putting family planning on the policy agenda, did not sustain momentum by building broadly based political coalitions, the family planning agenda was less well formulated and implemented. The study supports the view that government policies and programs can make a major difference, at least to the timing of fertility decline. In three of the four pairs, in countries with stronger programs the fertility decline started earlier; in the remaining pair, the pace of decline was sharper in the country with the stronger program. Moreover, countries with weaker family planning programs displayed more pronounced social differences in contraceptive use. In those with stronger programs, the main function of the program was to accelerate the spread of fertility control from the urban and educated elite to the rural and less educated. National policy-making by elites, and strong institutional and financial support for family planning programs are unlikely to have had direct relevance to fertility desires of individuals in these countries. These conclusions are in direct contrast to Pritchett’s.
Both analyses appear influenced by ideology. Neither presents direct individual level evidence that might indicate which changes first, fertility preferences or access to the means of control. Thus, we return to the lament of Coleman and Schofield (1986) that demography suffers from lack of a theory that would integrate the influence of societal factors, such as social policies or economic development, with the individual-level behaviors of which national fertility levels are made up, a theory to integrate macro and micro.
In short, there are findings that government policies can make a difference to fertility, but also suggestions that they may not always be successful.
Are State Polices Effective in Preventing Fertility Falling Below Replacement?
Over the past century, there has been a gradual increase in political interest in population and family matters in developed countries. Common developments and diversity within them across countries are fully described by Gauthier (1996a) and in greater detail for Canada, the United States, Australia, and Great Britain by Heitlinger (1993). They include public regulation of abortion and contraceptives, cash benefits to families, policies facilitating the employment of women in general, policies facilitating the combination of paid work and parenthood, provisions for childcare and (although not specifically treated in these books) housing policy and immigration control. Few governments, other than the French, have explicitly stated pronatalist goals, although Nordic countries, especially Sweden, have a tradition of facilitating family life through liberal policy on voluntary parenthood, relatively generous social security, and a deliberate policy of promoting sex equality through recognizing a woman worker’s right to parenthood (rather than a mother’s right to employment). Since the 1970s the parental insurance scheme for the workforce has established right for fathers as well as mothers to have paid leave from employment in connection with their children, in an attempt to make gender equality and family life more compatible. The package of family friendly policies in Scandinavia also includes widespread public involvement in childcare. Elsewhere, policies affecting the family have been more likely to subsidize a traditional breadwinner-housewife division of labor between parents. The history of the development of this complex of policies suggests that they are in part the outcome of demographic change (Gauthier, 1996a). Their effectiveness at changing demographic trends is more debatable.
Gauthier (1996b) reviews evidence of effects on fertility (and family structure) of the impact of fiscal incentives on fertility. The evidence is for nothing more than a very modest effect (which was not altered when she included maternity provisions and child care provision in a study of twenty-two industrialized countries). The study by Ekert (1986) of eight European Community countries suggested that cash benefits of the order offered in France raised total fertility by 0.2 children. A similar result was confirmed by Blanchet and Ekert-Jaffé (1994) using data on 11 European countries, and by Ermisch (1988) on the basis of births of different order in England and Wales. His estimate was that a doubling of child-benefit would, eventually, raise completed family size by 0.15 children. This implies that should the government wish to be so un-British as to manipulate the birth rate, the costs of achieving significant change through cash benefits would be prohibitive.
Ermisch’s analysis (and that of the study that preceded it, De Cooman et al., 1987) suggests that most of the reaction of births in England and Wales to changes in economic signals was in timing, particularly of the first birth, rather than completed family size. Postwar swings in period fertility have generally had a larger component of tempo rather than quantum, timing and spacing rather than completed cohort fertility. (Olsen, 1994; Hopflinger, 1994; Walker, 1995) While swings in birth timing may not make any difference to the family’s ultimate size, they make a difference to the overall age structure, and make it difficult to plan age-specific services like schools, jobs for school leavers, and, eventually, pensions. It therefore remains important to understand and anticipate movements in births, even if there is little prospect of controlling them. Cigno (1991) suggests that governments are unintentionally fueling the deferment of childbearing through tax policies conducive to female labour force participation. This argument could be countered with the suggestion that female labor force participation need not be incompatible with childbearing, if policies are in place to facilitate their combination.
Such policies have been developed in Sweden. In the early 1990s, Sweden had one of the highest levels of fertility in Europe, and many commentators have been tempted to suggest that the package of family-friendly policies were responsible for raising the birth rate up to replacement levels. Two recent studies cast some doubt on such an interpretation. Both stress the importance of re-timing of Swedish births around historic and projected cohort fertility close to two children for all cohorts born this century. Hoem (1990) explains how a specific change in the parental leave regulations in 1986 created an inducement to space births more closely. There is evidence people did this, a response, perhaps unintended, to a policy change, but mainly one of tempo rather than quantum. Walker (1995) also concludes that there has been little effect of economic variables on life-cycle fertility between the 1950s and the 1990s, and that the impact of state policy was less important than broad trends of economic change in the labor market. He reaches this conclusion by estimating how the cost of childbearing (the shadow price) has changed over the years. He constructs series for the value of parental benefits, public child care, and child allowances that are compared with direct costs and opportunity costs in the labor market including foregone opportunities to accumulate human capital. The compression of wage differentials, another aspect of Swedish policy, during the 1970s and 1980s, is said to have reduced incentives for early childbearing and could also account for the shift in birth timing over the period (see also Cigno, 1991, Chapter 8), but pronatalist policies would have had to be even more generous than the standards set in Sweden to have much importance in the short or long term. The last resort for those who wish to defend Swedish family policy as a pronatalist instrument (rather than something that has merit in its own right) is that without parental leave and childcare apparatus the size of the Swedish family would have been smaller. This is an unfortunately unverifiable proposition. Drops in fertility in Sweden (and Eastern Germany) in the 1990s, may still prove to be temporary adjustments.
It remains to be seen whether pronatalist policies are likely to be effective at changing anything other than timing, and it is unlikely that we will see this put to the test. Hobcraft and Kiernan (1995) offer another reason to be skeptical of the state as engine of a reversal of the fertility decline—that it is unlikely that the public budget would ever devote substantial enough resources to make much difference. As Folbre (1994) would put it, the vested interests against investing in children as public goods, leave social reproduction increasingly the responsibility not just of private families, but the women left holding the baby as male responsibilities are eroded. She calls for a new definition of family policy enshrining an egalitarian restructuring of the genders.
Although state policies can certainly help to speed up the decline of high fertility, there is little evidence that they can successfully apply the brakes, should they want to, when the process falls below replacement level. Fertility and family life might be facilitated by a wide range of economic and social changes, but there is unlikely to be a substantial sustained response to measures aimed only at women and to the proximate determinants of their fertility.
The mastery of nature displayed in the control of fertility is itself one aspect of economic development. It is possible for this aspect to run ahead, behind, or alongside the mastery of other theaters of endeavor. When they run together, ideational and materialist explanations are compatible perspectives on the same process, the economist’s signal is the sociologist’s sanction. However, development is not always even, nor can one driving force always claim primacy. In some situations it may be possible to test the predictions of economic and social explanations against one another, but not all. Explanations that are observationally equivalent may have different implications for policy or prediction according to what the underlying mechanism is. It is thus difficult to reach simple conclusions about the social and economic determinants of fertility, other than to point to their complexity. Different forces on beliefs and behavior operate at individual, family, community and societal levels with different strengths, on different time scales and in different points in time and space. Furthermore there is feedback between the economic and the demographic, reflecting human kind’s capacity to adapt and negotiate. The data one would ideally need to test such a model quantitatively defy description, let alone practicality in poor countries (David, 1998).
The lack of suitable longitudinal, multilevel, data (on paid and unpaid work, family history and structure, attitudes in the home and outside, and so on) may be one reason why social scientific knowledge of the determinants of fertility is not a clear consensus. It is easier to draw some conclusions about the beliefs of social scientists than about the beliefs and reproductive behavior of real people. The economic approach sees rational agents deciding to invest in child quantity, child quality or birth control according to changing price signals in the economic environment. The sociological view is that changing norms regarding reproduction lie behind fertility decline, accompanying, preceding and perhaps even promoting economic development. The ideational theory of fertility change has been criticized because it appears to allow no role for the cultural context in which diffusion of ideas takes place. Institutional theories of fertility contend that particular historic and political configurations shape social institutions in any society, and give rise to particular patterns of social organization. The sociologist’s emphasis on the diffusion of ideas complements the economist’s model in which agents are assumed to be equipped at no cost with perfect information that is assumed to permeate society (for an exception see Pollak, 1985).
Economic and sociologic analysts of fertility are too often in competition, but there is common ground. Notable attempts to synthesize across disciplines are offered by Folbre (1994), Hobcraft and Kiernan (1995), Pollak and Watkins (1993), and Pritchett (1994). For Folbre, constraints are gendered (men and women do not always face the same choices about social reproduction), but the salience of gender differences has varied over time and space.
Cigno’s exposition of economic decision making in and about the family takes a more narrowly focused view. He confidently concludes that the decline in size of the family, and in its role as an organizer of transactions between generations are brought about as a response to changes in the structure of incentives and disincentives associated with economic development, rather than caused by changes in preferences or culture.
Indeed, some of what goes under the name of culture may be interpreted as the debris of past states of the economic environment—a consequence of the fact that, when the structure of incentives and disincentives changes, it takes time to realize it, time to decide how best to respond to it, and time to modify the pattern of one’s life (Cigno, 1991, p. 195).
Most other economists writing about the family have been less sanguine about the completeness of the explanations to be offered by models of constrained optimization. Montgomery and Trussell (1986, p. 266) point out that the family still has to be set aside from other subjects of economic research because of “the pervasive roles played by individual physiology, by uncertainty, by irreversibility and by social norms.” Irreversibility refers to the fact that responsible parents at least cannot trade in their children even if they turn out unsatisfactorily, like Gary Becker’s Rotten Kid.
Readers may be more comfortable with the line taken by Pollak and Watkins (1993). They suggest that different cultural settings mediate how preferences and opportunities operate within groups differentiated by “language, ethnicity, region, and dress” (p. 488). They also suggest that opportunities and preferences are inseparably bonded, as the warp and weft of a fabric, and warn against pulling it apart. In declaring the need for a multidisciplinary approach they remark:
The rational actor model is silent about the boundary between choice behavior and nonchoice behavior and about the determinants of nonchoice behavior. Perhaps ‘culture’ can fill this analytic vacuum by describing or explaining the location of the boundary separating nonchoice and choice behavior (p. 499).
The strength and diversity of social norms helps explain variation in behavior between and within societies. Secular and sacred values help to set targets and to sanction deviant behavior. But just as individuals may not be purely rational actors, neither are they merely passive agents of culture. They can shape and re-shape their culture, and may re-negotiate the norms, rules or values to their own benefit. The political and economic forces shape the cultural context of behavior, and individuals continually renegotiate their behaviors, rationalizing them within the context of social norms (Bledsoe et al., 1993).
Neither the view of individuals as passively adhering to cultural values, nor as rational utility maximizers is satisfactory. The notion that human motivation extends beyond economic incentives is useful not only for the understanding of fertility, but also relevant to other fields of activity, including the management of a paid workforce. No one set of explanations is sufficiently developed to provide an integrated view of why and how reproductive change occurs. There is a need for testable theories for how sociocultural and economic forces can be linked to variations in individual behavior. Both macro- and microlevel analysis is required to provide an understanding of how the political economy evolved and how the individual creates his definition of proper behavior (Greenhalgh, 1995; Kertzer, 1995).
The theories and evidence we have assembled paint the picture of the process of social reproduction as vitally connected with the process of production. They are not always harmoniously adjusted to one another. The gender division of labor and power sets the terms of cooperation and conflict between men and women in the reproductive and productive spheres. As their work so often spans reproductive and productive fields, it is obvious that women’s lives need more than economics to explain them, but so do men’s. Just as men and women have seldom been successful at cooperating on equal terms to achieve reproductive and productive goals, social scientists have not always found cooperative solutions to the challenge of understanding these important processes. We hope this chapter points toward a constructive renegotiation of the division of social scientific labor.