Childbearing and Maternal Employment
In this chapter, we describe trends in educational differences for several dimensions of childbearing including early childbearing, nonmarital childbearing, and bridal pregnancy. We also examine maternal employment. These analyses are based on the data described above in Sect. 3.2 and employ descriptive regression methods similar to those used in Chap. 3. This chapter concludes with a discussion of the results in both Chap. 3 and this chapter and suggestions for future research in this area.
Age at first birth is calculated from the fertility history modules in the NFS. Because detailed questions about childbearing refer only to women’s current marriage (prior to the 14th NFS), these analyses do not include births in previous marriages for married women or any births for currently unmarried women. Figure 4.1 presents trends in the predicted probability of having a child by age 22. Early childbearing, defined in this way, has declined over time for all groups except college graduates. Because the probability of early motherhood for college graduates has remained constant at almost zero, the decline in early birth among high school graduates implies a narrowing of the gap between these two groups. However, women in the intermediate educational groups of vocational and junior college graduates have diverged from high school graduates. The difference between vocational school graduates and high school graduates is significantly larger in the 1970s birth cohort (relative to the 1940s cohort). For junior college graduates, the gap is larger from the 1950s cohort on and grew significantly in the 1960s birth cohort. In contrast to the U.S., where those with some college increasingly look like those with a high
© The Author(s) 2017 43
J.M. Raymo and M. Iwasawa, Diverging Destinies,
Population Studies of Japan, DOI 10.1007/978-981-10-0185-7_4
Fig. 4.1 Predicted probability of having a child by age 22, by education and birth cohort
school degree or less, Japanese women with two-year degrees increasingly resemble their more educated counterparts—at least with respect to the likelihood of early childbirth (and early marriage as shown above in Sect. 3.3). The odds ratios for early childbearing among women with two-year degrees (relative to high school graduates) fell from 0.60 (vocational school) and 0.45 (junior college) in the 1940s birth cohort to 0.37 and 0.20, respectively, in the 1970s birth cohort. The odds ratio for university graduates relative to high school graduates remained at 0.05 across cohorts.
As in our analysis of early marriage, the 22-year-old threshold for defining an early birth is arbitrary and it is important to keep in mind that the meaning of this threshold is not constant over time. Because the average age of first birth has increased substantially, rising from 26.1 in 1980 to 29.3 in 2010 (NIPSSR 2016), motherhood by age 22 is much more uncommon for more recent cohorts. The cumulative fertility rate at age 22 declined from 0.20 for the 1940 birth cohort to 0.08 for the 1985 birth cohort (NIPSSR 2016). Unlike the U.S., where teen childbearing is of considerable policy and scientific interest, very few Japanese women become parents as teenagers. We therefore present trends in the predicted age of first birth (among women who have had a child) in Fig. 4.2. Here, we see that age of motherhood has increased for all groups except women who did not
Fig. 4.2 Predicted age at first birth, by education and birth cohort
complete high school. It is also clear that the delay in childbearing has been more pronounced for women with more than a high school education. The difference between high school graduates and the three groups with more education is significantly larger in the 1950s, 1960s, and 1970s cohorts relative to the 1940s cohort. This gap also grew significantly in the 1960s for women with a junior college or university education, again suggesting that family bifurcation was particularly pronounced in the 1960s birth cohort. To the extent that age of motherhood per se, rather than the experience of early childbearing, is related to children’s resources (McLanahan and Jacobsen 2015), these results are also consistent with a pattern of diverging destinies.