Empirical Findings

The fertility literature is huge. We will introduce it by identifying 10 major "social facts,'' i.e., empirical regularities from this literature.

1. Fertility in populations not using contraception and abortion varies substantially.

2. The timing of the onset of fertility transition (vis-a-vis objective socioeconomic conditions) is highly variable.

3. Existing institutions influence the fertility transition, thus the process of change varies from place to place and has historical continuity.

4. The fertility transition involves a collective evaluative assessment of social conditions and possible responses.

5. Once the fertility transition begins, it does not stop until fertility reaches levels of approximately two children or lower.

6. Fertility change is a period, not a cohort, phenomenon.

7. Fertility delay is fundamentally antinatalist.

8. Valid and reliable retrospective fertility histories can be collected from women in a broad range of settings. Reliable information on abortion cannot generally be obtained from respondents.

9. Long-range fertility intentions (individual or aggregate) have low predictive validity (at both the individual and aggregate level).

10. The fertility desires/intentions of men and women (and husbands and wives) are similar; and the impacts of their intentions/desires on subsequent fertility are similar.

We now discuss these in greater detail. These statements are not immutable "facts," but rather they reflect current, accepted wisdom.

1. In populations not using contraception and abortion, fertility varies substantially.

As noted earlier, natural fertility results if there are no attempts to control family size. In practice, however, natural fertility is frequently operationalized as involving no contraception or abortion (Henry 1961). Fertility is high in natural fertility populations—but how high? Females can have children as early as the midteens and can continue until the late 40s. Theoretically, women could have nearly one birth per year. Thus the theoretical maximum fertility, in the absence of all behavioral constraints, could be as high as 35 births!

In fact, no population has averaged anywhere close to this theoretical maximum level of fertility. Instead the classic example of a high-fertility population, the Hutterites, has fertility one-third this high. From 1880 to 1950, the U.S./-Canadian Hutterite population increased from 443 to 8,542 persons (Eaton and Mayer 1953). This is the world's fastest known natural growth rate (4.21% annually), with families averaging around 10 to 12 children (Ingoldsby 2001).

On the other end of the natural fertility spectrum lie the Dobe !Kung hunter-gathers, residents of the Kalahari Desert in Africa prior to 1975 (Howell 1979, 2000). The reported TFR for this natural fertility population was about 4.5 births per woman. Thus, the question, "How can natural fertility populations be so different from each other, and why are even the highest observed rates much lower than the theoretical maximum?''

The answer to both questions relies heavily on the proximate determinants framework described earlier. All known societies have encouraged practices that, through biological mechanisms, reduce fertility well below maximum levels. Key features are norms about union formation and dissolution (specifically, marriage) that impact coital frequency and the risk of pregnancy. Late marriage (indicating the postponement of sexual intercourse) reduces the years available for childbearing and thus the number of births.

The second important determinant of these differences in fertility is breast-feeding and postpartum amenorrhea (Bongaarts and Potter 1983). It is now well established that breast-feeding leads to a substantially longer postpartum period without ovulation than the typical 1.5- to 2-month interval that is experienced by women who do not breast-feed (Leridon 1977). Also, the intensity of breast-feeding affects the likelihood of ovulation. Women who exclusively breast-feed their children have a significantly lower chance of ovulating than do women who supplement breast-feeding with other food.

The !Kung typically breast-feed for three years and Hutterite women, for less than half this period.

In short, the Hutterite-Dobe !Kung natural fertility differential can be traced to greater time spent out of sexual unions (especially due to separation and union dissolution) and especially a much longer and more intense period of breast-feeding among the !Kung. In general, differences in natural fertility can be accounted for by these same two proximate determinants (see Bongaarts and Potter 1978, 1983).

2. The timing of the onset of the fertility transition (vis-a-vis objective socioeconomic conditions) is highly variable.

Demographic transition theory attributes fertility (and mortality) change to the process of economic development, especially the transition from a rural agrarian society to an urban industrial one. This leaves unanswered the question of what part of this process was most crucial for fertility decline. Was it changed occupations, urban living, or increased educational attainment that produced fertility decline? Further, what level of change in these aspects of economic development or its correlates was necessary to initiate a fertility decline?

The current consensus is that this view is overly mechanistic. There are no "threshold levels'' of these macroeconomic indicators that consistently predict the onset of the transition. Likewise, there are no identifiable macrolevel changes that consistently predict the speed of the transition. Some argue that these findings must be interpreted cautiously, and one should not imply that economic development plays no causal role. Specifically, if multiple causes of decline are acknowledged, and if one views industrialization and urbanization as fundamental but distal causes (that need not produce synchronous change), then the role of economic development would receive greater support (Mason 1997).

3. Existing institutions influence the fertility transition, thus the process of change varies from place to place and has historical continuity.

Some of the reasons for the "loose" connections between socioeconomic change and fertility lie in preexisting differences in cultures and social institutions. For example, Greenhalgh (1988) argues that Chinese populations were among the first to experience fertility decline compared to others at similar levels of development. She attributes this to a historical and institutional context that made the number and sex composition of children a focal point of family strategy. In short, the Chinese populations began with a historical legacy that legitimated family size control and linked mobility strategies to number of children. Chinese groups quickly adopted modern contraception as a modern technology consistent with more costly traditional ones (including infanticide). In the Chinese context, the adoption of contraception was for limiting family size (specifically adopted by older women at higher parities).

In contrast, traditional African fertility regimes have been more concerned with a wide spacing of births as opposed to their number (Caldwell, Orubuloye, and Caldwell 1992). The link between limiting the number of children and upward social mobility was less apparent in these contexts. Institutions such as child fosterage may have played a role by spreading the costs of children across families, reducing the immediate impacts of rising child costs. Thus, the adoption of contraception was attractive as a substitute for postpartum abstinence and with the ideas that healthy children were produced by wide spacing (that could be aided by contraceptive use). As a result, the initial adoption of contraception in Africa tended to be simultaneous across ages and parities.

In short, Chinese and African family traditions influenced the speed and nature of their fertility transition. Chinese institutions hastened the transition (by its traditional emphasis on the size and composition of families and its use of postnatal control, explicitly, infanticide). African extended family and lineage institutions retarded change. The nature of the transition was also influenced. In Chinese populations fertility decline fell almost entirely due to contraceptive use after the desired number and composition of children were born. In Africa, fertility fell because of the wider spacing of births and birth limitation.

4. The fertility transition involves a collective evaluative assessment of social conditions and possible responses.

In a recent attempt to explain contemporary fertility transitions, Bongaarts and Watkins (1996) replicated the claim (discussed in 2, above) of a modest relationship between development indicators and changes in fertility. However, they argue that the diffusion of information about birth control techniques and ideas that legitimate small family size are important determinants of the timing of fertility change. This idea was central in the reports from the European Fertility Project (see Coale and Watkins 1986). Once a region of a country began a fertility transition, neighboring regions that shared a common language experienced a fertility decline shortly thereafter, regardless of the region's level of development. In this spirit, Bongaarts and Watkins (1996) conclude that social interaction in the form of exchanging information and ideas, evaluating their meaning in a given context, and social influences that encourage or discourage certain behaviors are significant factors in the transition from high to low fertility. Their measures of societal contact added significant explanatory power to their model of fertility transition. Watkins' work in contemporary African settings (e.g., Watkins 2000) describes at a microlevel how women's conversations helped to construct an understanding that fertility control was safe, appropriate, and advantageous.

5. Once the fertility transition begins, it does not stop until fertility reaches levels of approximately two children.

A well-known finding from the European Fertility Project is that once a 10% decline in fertility occurred (for any province), an irreversible transition was underway (Coale and Watkins 1986). Data in the Bongaats and Watkins study (1986) also show remarkably steady tendencies toward decline once the process is underway. Table 8.2 shows a cross-sectional, global view of the transition as of 2000. Of the 187 countries for which the UN provides data, only 19 have not yet shown evidence of fertility decline. These countries are primarily in Africa and include over 20% of the population of Africa. Thirty-two countries, again primarily in Africa, have recently begun a decline and another 73 are well into the transition. For practically all these countries, the lowest fertility observed is the most recent estimate, clearly indicating a steady march toward replacement level (or lower) fertility. Twenty-three countries, over half of the Asian countries, have TFRs that have fallen below 2.1 in the last decade or so. Coupled with the 39 countries that reached low fertility several decades ago (and experienced a post-World War II increase and then decline), these data indicate that by 2000 over 45% of the world's population lived in a country with replacement level fertility or below. Note that only two countries have experienced the transition to low fertility and have stabilized at levels above replacement (Argentina and Uruguay have levels that have remained close to 3.0 for the 1950 to 1980 period).

6. Fertility change is a period, not a cohort, phenomenon.

Earlier we noted that changing fertility rates can be described as occurring by cohort replacement or by pervasive period change. Ryder (1965) developed a paradigm of social change based on the concept of cohort replacement. The cohort perspective posits that trajectories of experience are frequently set by events early in life and are resistant to change subsequently. Cohort explanations stress the unique experience of a specified birth cohort (see Ryder 1965; Cherlin 1992). Change by cohort replacement comes slowly and steadily over time as new cohorts, in an orderly way, replace older ones.

Period explanations, on the other hand, emphasize the idea that shifts in fertility seem to affect all age groups at the same time. For example, shifts and changes in family attitudes and values may broadly impact nearly everyone's lives at once. Thus, the effects of these shifts are not unique to any one age group of people.

Twentieth-century U.S. fertility changes bear the unmistakable impact of period factors, including the Great Depression, wars, and economic cycles. Evidence from other developed countries is similar: changes in fertility are period driven, and cohort factors are weak or nonexistent (Ni Bhrolchain 1992).

7. Fertility delay is fundamentally antinatalist.

Although not invariant in magnitude, the timing of fertility is linked consistently to the number (or quantum) of births. This timing-number link can be seen for individual

Table 8.2. Stages of the Transition to Low Fertility in the Major Areas of the World by 2000.
100 Pregnancy Tips

100 Pregnancy Tips

Prior to planning pregnancy, you should learn more about the things involved in getting pregnant. It involves carrying a baby inside you for nine months, caring for a child for a number of years, and many more. Consider these things, so that you can properly assess if you are ready for pregnancy. Get all these very important tips about pregnancy that you need to know.

Get My Free Ebook

Post a comment