Draft—Do not cite without permission The Marginal Child throughout the Life Cycle: Evidence from Early Law Variation

This paper tests whether the fetal origins hypothesis (Barker 1992), which posits that disparities in the preand peri-natal environment can account for long-term disparities in life expectancy, is applicable to the case of variation in early circumstances due to wantedness. To identify the effects of wantedness on fertility rates and on life expectancy, we exploit over-time variation in the legal restrictions on birth control and abortion within U.S. states from 1850 to 1920; we demonstrate that the adoption of these legal restrictions cannot be predicted by other changes in state circumstances, and that legal restrictions that do not directly impact fertility control do not affect our outcomes of interest. We find that about 7 percent more children were born in times and places when fertility control technologies were inaccessible than in states and times when such technologies were freely available. Members of these larger cohorts were roughly 5 percent less likely than those in other cohorts to survive their sixties or seventies. We impute that the typical unintended (or “marginal,” in the terminology of Gruber et al. 1999) child born solely due to legal restrictions on access to fertility control was almost 70 percent less likely to live to old age than the average child; this magnitude is similar to estimated effects of recent variation in fertility control on offspring’s early life outcomes such as receiving welfare or attending college. We conclude that wantedness, like other aspects of a child’s early life circumstances, has important effects on life expectancy.