Completed Fertility and its Timing: An Economic Analysis of U.S. Experience Since World War II.

Abstract

Presents an economic model of fertility behavior designed to explain and predict both current fertility rates and expected family size emphasizing the distinction between male and female earnings in affecting fertility. It is hypothesized that an increase in husbands earnings raises family income and leads to an increased demand for children. An increase in the wifes income however raises the cost of children due to the rise in the opportunity cost of childbearing and rearing. The role of both current and future income and costs is emphasized particularly the role of substitution of births between time periods. A temporary increase in the wifes wage provides an incentive to postpone births while if the wifes wages were expected to increase in some future period current births would be less costly relative to future births. The option to substitute births between time periods perhaps leaving completed fertility unaffected means that the simple relationships between current fertility and current economic variables may be an inappropriate guide to the relationship between economic variables and ultimate family size. The effects of economic variables on the timing of births are distinguished from the effects on the level of completed fertility by estimating an auxiliary model that predicts future male and female income based on age and year specific data from 1957 to 1975. A fertility rate model showing the effect of both current and future economic variables on current fertility is subsequently estimated and the effects of female and male income changes on both current and completed fertility are simulated. Male income has a positive effect and female income a negative effect on both current and completed fertility. More than 1/2 of the observed effect is due to indirect influence of these economic variables on expectations concerning future income. The strong tendency to substitute births between time periods in response to female wage rates suggests that most of the year-to-year change in period fertility rates is due to altered timing of births. The regression model allows forecasting of the life cycle of couples fertility at different starting dates with the level of expected completed fertility computed as the couple ages. Cohorts have substantially revised their target fertility in response to changes in economic variables. Total fertility rate is decomposed into timing and completed fertility using these estimates. The paths of these components over time show that the 1950s baby boom resulted from gradually increasing expected family size and from a compression of births into those years. The subsequent decline in fertility was primarily the result of downward revisions of completed fertility expectations not due to deferred births. These results indicate that the current low level of fertility should continue apart from temporary increases when the time cost of children falls. (Authors modified)