Theoretical Issues

We begin by distinguishing between analytic frameworks and causal/behavioral theories. Analytic frameworks are useful ways to organize data, and they capture structural aspects of the process. Fertility research has produced widely accepted and very useful analytic frameworks. However, these analytic frameworks are largely silent regarding the more distal social causes of fertility trends and group differences. There is much greater disagreement regarding the relative value of these more distal causal theories. We address analytic frameworks and causal theories in turn.

Analytic Frameworks

Two mutually informing analytic frameworks have been central to much recent fertility research: the life course and the proximate determinants frameworks. The biological nature of fertility determines the structure of each framework. In fact, both frameworks rest on very straightforward observations. The life course perspective adopts a sequential model because children tend to be born one at a time, not in lots (Namboodiri 1972: 198). Moreover, because women are biologically restricted to having children only between menarche and menopause, fertility may be considered as an irreversible, time-limited sequence.

This sequential structure can be used to decompose overall change into age and birth order (or parity) components (see Morgan 1996). Or it can be adopted to compare the fertility regimes of different groups. For instance, when do two groups behave differently and when do they behave similarly? This structure also makes explicit the fundamental life course principle that events and their circumstances at time t can influence behavior at time t + 1. Most researchers now view fertility outcomes as resulting from a series of sequential decisions. For instance, permanent childlessness results most often from a series of decisions to postpone childbearing and not from firm decisions made early in life to remain childless (see Rindfuss, Morgan, and Swicegood 1988).

The proximate determinants paradigm provides a second organizing framework. It rests on the observation that the sequential biological process is influenced through only a few mechanisms, specifically, variables that influence sexual activity, the likelihood of conception, and the likelihood that conceptions result in live births (see Davis and Blake 1956). Bongaarts and Potter's (1978) operationalization of the proximate determinants demonstrates that most fertility variability between populations and over time can be accounted for by the following four determinants: (1) marriage and marital disruption (as indicators of the segments of the life cycle when women are sexually active), (2) postpartum infecundability (the period after a birth without ovulation; its length is determined primarily by the duration and intensity of breastfeeding), (3) use and effectiveness of contraception, and (4) induced abortion. Three other determinants are occasionally or potentially important: (5) the onset of permanent sterility, (6) natural fecundability (frequency of intercourse), and (7) spontaneous intrauterine mortality. However, they do not vary as much among populations as the first four.

The life course and proximate determinant frameworks together provide a crucial foundation for understanding the mechanisms that influence individual and aggregate fertility. Descriptive work using these frameworks identifies precisely "what needs to be explained.'' As an example, the most dynamic fertility component in the U.S. baby boom and bust was the timing of the first birth (Ryder 1980). This observation begs the question: What accounts for this changing timing of family formation? Likewise, if an observed fertility decline can be attributed, within the proximate determinants framework, to changes in marriage, then a very different explanation is required other than that the change is due to increasing contraceptive use. Overall, then, theories of fertility change and variability are incomplete if they do not specify where in the life course and through which proximate determinants the social, economic, and cultural factors operate.

Causal/Behavioral Theories

Fertility transitions are complete in many developed countries and are in progress in much of the rest of the world. The transition model has three stages: relatively high and stable fertility, followed by a period of fertility decline, and then followed by relatively low and stable fertility. This fertility transition is part of the demographic transition model that includes similar changes/stages in mortality (see chapter 10 in this Handbook). The demographic transition theory, based heavily on observed, historical changes in the West, linked fertility and mortality changes to social, economic, and family changes caused by industrialization and urbanization (Notestein 1953). Given a very long time frame, all economic transitions (from rural/agrarian to urban/industrial) have been accompanied by fertility declines. But demographic transition theory has not performed as well in accounting for the timing of fertility decline. This poor fit of data to theory has led to a number of revisions, extensions, and elaborations of demographic transition theory.

Specifically, substantive behavioral explanations for fertility transition focus on at least one of three elements: the nature of fertility decisions, the information and knowledge available to decision makers, and the institutional context for decision making. We first consider high fertility in pre industrial settings. Widespread evidence indicates that the high fertility of many populations coincided with a "natural fertility'' regime, one in which potential decision makers did not limit their number of births via changed behavior at higher parities (Henry 1961). There are three possible reasons: (1) decision makers were motivated to have as many children as possible, (2) decision makers did not know how to limit fertility, or (3) fertility control was not licit. Important behavioral theories focus on each alternative.

The first alternative is consistent with micro economic models of choice that stressed the economic value of children. Children provide substantial labor in preindus-trial settings and have relatively low direct and opportunity costs. One version of the argument for the rationality of high fertility is found in Caldwell's (1982) work. Caldwell argued that preindustrial patriarchal family structure (an institutional context) allowed older individuals and males to appropriate wealth from younger and female family members. This positive "wealth flow'' from children toward the senior generation motivated higher fertility. High fertility, in turn, produced greater wealth, power, and prestige for patriarchs. According to this and other arguments stressing the economic value of children, fertility was high in preindustrial contexts because children were net assets.

Empirical evidence regarding the economic value of children in preindustrial contexts is mixed. Children clearly performed substantial work, but their rearing and support also required substantial investment. The current consensus is that children were not universally perceived as net economic assets in preindustrial settings.2 These mixed results on the economic cost of children shifted attention to children's roles as adults in supporting their parents. In many contexts older persons were dependent upon their children for support in old age. Prior to old age dependency, many relied upon children for support in case of crisis. A large literature has focused on the import of children in providing old age security and risk insurance (e.g., Cain 1983; Nugent 1985).

Finally, the cost-benefit analysis of high fertility focuses on noneconomic values of children. Across a range of institutional contexts, parents give reasons for having children, such as having a child to love and care for, having a sibling for a previously born child, having a son or a daughter, or bringing the family closer together (see Bulatao 1981).

Others have argued against individual cost-benefit calculations in many preindus-trial contexts. van de Walle (1992), for instance, has claimed that many persons did not conceptualize family size as a key decision variable because factors determining it seemed beyond their ability or willingness to control. The vagaries of mortality or the social construction of morality pushed effective family size control beyond the reach of most.

Note that this argument does not mean that people had no knowledge of birth control. Rather it implies that they were not willing to use these mechanisms. In some settings knowledge of techniques may have been the limiting factor, while in others, limiting family size was beyond the "calculus of conscious choice.'' Proponents of these two positions have waged an intellectually engaging debate for over two decades. Evidence for both positions exists, and there is no reason to assume that a single answer is appropriate for all settings. For instance, evidence of infanticide in preindustrial China shows that under given institutional arrangements controlling family size and composition was of extreme importance. In addition, the extended family arranged marriage to control family size. Abstinence and abortion also have a long history in China (Lee and Feng 1999). In contrast, in Western Europe there was little evidence of family size control before the onset of fertility decline. Knodel and van de Walle (1979) have discussed the evidence in detail. For instance, they point out that nonmarital and marital childbearing declined in tandem. While one might argue that the desire to limit marital fertility was absent prior to the onset of decline, the European historical context included strong negative sanctions for nonmarital childbearing. Thus, incentives to avoid nonmarital childbearing clearly existed prior to the onset of fertility decline. Indeed, the nearly simultaneous decline of marital and nonmarital fertility suggests that knowledge of birth control means, not motivation, was the missing ingredient. As

2 Evolutionary arguments point out that other mechanisms generally guaranteed a substantial supply of children (Potts 1997) and that exploitation of the younger generation is not a viable evolutionary strategy (Kaplan 1994).

further evidence, women in developing countries frequently report having more children than desired. This is prima facie evidence against the claim that families in preindustrial settings desired as many children as they could have (Shorter et al. 1971; Knodel and van de Walle 1979).

The fertility transition's second stage views family size declines as resulting from conscious actions by decision makers, specifically decisions to curtail childbearing at lower parities. Depending on their explanation for the preceding natural fertility regime, theorists point toward institutional change that transforms children from net assets to financial liabilities, or ones that provide new knowledge or changed norms that allow for family size control.

Caldwell (1982), for instance, points to schooling and nonfamilial employment opportunities that undermined the power and control of patriarchs and shifted the net flow of wealth toward children. Others have incorporated the importance of mortality decline. For instance, Easterlin and Crimmins's (1985) framework explicitly includes the possibility that declining infant and child mortality produced an increase in surviving children. For some decision makers, surviving children (the supply of children) now exceeded the desired number (the demand for children), producing a motivation for fertility control. In sum, a large body of work argues that rational decisions in changed contexts tended to lead to a new decision—to stop childbearing after a desired number was achieved.

For those arguments that stressed the lack of contraceptive knowledge or the presence of normative injunctions against using them, key factors in the decline were likely the spread of knowledge or new ideas legitimizing contraceptives or the small family size ideal. These ideas could have included broad, increasingly popular, anti-natalist ideologies of individualism and self-actualization. Such ideologies justified nonfamilial activities and aspirations (Lesthaeghe 1983). But the diffusing knowledge may have also included new information about techniques or the diffusion of the technologies themselves. The evidence suggesting a role for diffusion processes is powerful. For instance, the European Fertility Project (see Coale and Watkins 1986) characterized the geographical pattern of European fertility decline as a ''contagion process.'' Geographical proximity and measures of interaction (e.g., shared language) were strong predictors of the timing of fertility decline. Likewise, reviews of evidence from developing country fertility surveys have shown patterns of change far too rapid to be attributed solely to decision-maker adjustment to changing objective, socioeconomic circumstances (Cleland and Wilson 1987).

Consistent with earlier adjudication between these positions, one need not choose one or the other as a universal answer (see Mason 1997). In an analysis that we believe best captures the contemporary demographic consensus, Bongaarts and Watkins' (1996) review of postwar fertility declines finds evidence that both structural change and diffusion processes are at work.

The third and final stage of the demographic transition is low fertility that approximates very low mortality. Such a balance is logically necessary; over the long run neither positive nor negative growth rates can continue. An emerging issue of both practical and intellectual import is the question of whether economically advanced societies will have fertility levels that even approximate replacement levels.

One position is nicely characterized by Bumpass (1990), who argues that the long-term factors that have reduced fertility ''have not run their course.'' These factors include ''structural'' changes in the way we live and work that make children costly

(in economic terms and in terms of foregone opportunities). Secular forces also include ideologies of self-actualization and individualism that could become even more powerful and pervasive antinatalist ideologies (see Lesthaeghe 1983; van de Kaa 2001).

As an example, many see increasing female labor force participation as a key structural, secular, antinatalist factor. The standard microlevel home economics approach posits that declining gender discrimination and greater access to the labor market increases the cost of having children. This increased cost results from women's exit from the labor force to bear and raise the children and the costs of labor force exit on career trajectories. This argument has led to the very widely held view that increased labor force participation by women would depress fertility rates. In the past, evidence for such an association was common at both the individual and aggregate levels.

More recent arguments, however, stress the importance of institutions in conditioning the effects of secular structural change (Rindfuss 1991, Rindfuss, Guzzo, and Morgan 2003). For instance, if one assumes that female labor force participation increases will continue, the question then becomes what societies using which institutions can make accommodations that allow women to more easily work and have children?

Perhaps the best evidence that societies vary on this dimension is the changing aggregate-level association between female labor force participation rates and the total fertility rate. Traditionally and according to most theories, this association should be negative; higher labor force participation is associated with lower fertility. But in low-fertility contexts, the opposite appears to be true (see Rindfuss and Brewster 1996; Rindfuss, Guzzo, and Morgan 2003). This cross-sectional association occurs because, in the past two decades, the association between labor force participation and fertility has varied dramatically by country. In the U.S., for instance, increases in labor force participation have not been accompanied by decreases in fertility. In Italy over the past two decades, for contrast, the association between labor force participation and fertility has been strongly negative (Rindfuss, Guzzo, and Morgan 2003). What aspects of context weaken the incompatibility of work and family obligations? Many point to available, high-quality day care, flexible work environments (flex time and parental leaves, for example), and more egalitarian gender roles that provide women with a domestic helpmate and a reduced "second shift.'' In sum, this second view holds that fertility levels are determined by adjustments in the institutions of family, economy, and public policy (for different conceptualizations of these adjustments see Esping-Ander-son 1999; McDonald 2000). Thus, the future of fertility depends upon societal adjustments that ease work-family conflicts. Some countries will make or have made such adjustments, while others have not and may not (see Morgan and King 2001).

The above review underemphasizes several important issues. First, our review has largely ignored the larger social context in which the debate about fertility occurred. The politics of the Cold War period, concerns about a global "population explosion,'' and the feminist movement all provided an urgency and brought resources and attention to the study of fertility transitions. This political context helped define high fertility as a social problem of the highest order and thus motivated action at many levels (see Hodgson 1988; Hodgson and Watkins 1997). Two levels of action were by governments and by nongovernmental organizations, both of which organized and funded family planning programs. These programs clearly sped fertility declines in many countries. A second theme underemphasized here is linked to theories drawn from closely observed local experiences (e.g., Watkins 2000; Kertzer and Fricke 1997). Known as "thick description,'' these theories provide details of change specific to certain locales and resonate more closely with the experiences of the respondents. Such a focus is missed by the broad review just presented.

Pregnancy Guide

Pregnancy Guide

A Beginner's Guide to Healthy Pregnancy. If you suspect, or know, that you are pregnant, we ho pe you have already visited your doctor. Presuming that you have confirmed your suspicions and that this is your first child, or that you wish to take better care of yourself d uring pregnancy than you did during your other pregnancies; you have come to the right place.

Get My Free Ebook

Post a comment