Bell & Howell Information and Learning: Foreign text omitted ...
In this paper I examine the evolving association between educational attainment and the timing of births. In the late 1970s, women with four-year college degrees had lower first birth rates before age 30 than women with less education, but rates offirst births were similar for the two groups after age 30. From the 1970s to the 1990s, first birth rates decreased before age 30 for all women, but increased after age 30 only for women with four-year college degrees. Parity 2 birth rates also increased for college graduates with a first birth after age 30. These results document widening educational differences in fertility timing between 1975 and 1995, which may reflect period changes at later ages in women's work and family lives.
In this paper I examine trends in U.S. women's birth rates in their later reproductive years for the period 1975 to 1995. I pose several questions concerning this two-decade span. First, what are the chances that women who delay childbearing to age 30 will eventually have children? Second, how many children will they have? Third, what is the association between education and fertility during these ages?
For answers to these questions, I examine data from the June 1990 and June 1995 Current Population Surveys to estimate trends in birth rates from 1975 to 1995. Then I reconstitute the birth rates into synthetic estimates of total births for the late 1970s and early 1990s. Results are consistent with a standard interpretation: Conflicts between women's work and family lives reduce fertility in early adulthood for all women, and especially for college-educated women. Yet I find a compensating increase in family formation rates ofter age 30 only for women with four-year college degrees.
Theoretical interest in birth timing derives partly from the link between delayed fertility and recent transformations of women's work lives. Working women commonly postpone childbearing as a way to coordinate their work and domestic roles (see Bianchi and Spain 1986; Van Horn 1988); this would imply that successive cohorts of women increasingly will postpone childbearing as their labor force involvement increases. Cross-sectionally we would also expect women with high labor force involvement, such as college graduates, to be the most likely to postpone childbearing. Indeed, Rindfuss, Morgan, and Offutt (1996) have documented declining birth rates for women in their early-middle reproductive years, especially women with college degrees.
Although it is widely agreed that women's work leads to postponement of fertility, it is less clear whether postponement leads to later births or to childlessness. As a step toward resolving this issue, I make an explicit distinction between decreasing birth rates before age 30 and rising birth rates after age 30. In addition, I argue that parity-specific birth rates after age 30, although affected strongly by birth rates before age 30, are also subject to distinct social and economic influences on women's lives.
Decreasing birth rates before age 30 have been well documented in the United States. Because fewer women enter their later reproductive years with their childbearing already completed, recent increases in births at later ages may be considered a direct consequence of delayed family formation at younger ages. Many current arguments on fertility timing proceed as follows: The reason why women (especially college-educated women) are having births at later ages is that they are not having births at younger ages, and the reason why they are not having births at younger ages is that their career orientations and demands for high-quality (and costly) childcare do not permit them to start families early in their careers. Hence, especially for college-educated women, the competition between work and family roles in the early adult years causes births to be