Proxies for birth control. Comment on Kelly and Cutright, Sociological Focus, April 1983.

Abstract

Kelly and Cutright (1983), using regression techniques, conclude that birth control is among the more important determinants of Swedish illegitimacy. To derive this conclusion, they use changes in the marital fertility of wives aged 35-39 as a proxy for birth control. They maintain that annual change in the marital fertility rate of wives aged 35-39 is not likely to be greatly influenced by annual change in factors other than birth control. The "argument" appears to derive from the "desired family size" model of childbearing--a basic assumption of social demography. In it simplest form it states that most couples do not practice birth control until they reach a preconceived goal, or desired family size. It thus implies that a change in family size preferences will most affect the birth control practices of the oldest reproductive age groups. The simple form of the model has been questioned by the failure of Western couples to reproductively compensate for a major proportion of their child deaths, by the proportions of Western couples who say they would have preferred larger families than they actually had, by the predictive inadequacy of family size preferences, and by suggestions that age may be the more important determinant of reproduction. As a result some demographers now concede its inadequacy. Others are trying to relax its assumptions, with as yet problematic success. Essentially every Western fertility decline to date has been characterized by an increasing concentration of childbearing in the youngest age groups. In discussing this pattern social demographers have maintained that it could only have come about by a decline in family size preferences. This then is the standard argument supporting Kelly and Cutright's proxy for birth control. The authorities who offer it generally ignore the difficulties with the desired family size model and simply assert without justification that couples do in fact conform to it. Data on the age patterns of chronic disease and on the reproductive effects of environmental stressors suggest that the modern age pattern of fertility could also be produced by a deteriorating environment. Kelly and Cutright are incorrect in asserting that factors other than voluntary birth control could not be responsible for changes in fertility at ages 35-39. At best they may argue that their proxy is uniquely definitive provided that the desired family size model can be saved and provided the health of Western populations has not been compromised by technological change. At issue is a debate between what Dunlap calls the human exemptionalist and the ecological world views.