We explore whether fetal and postnatal exposure to tropical cyclones affects education and income in adulthood by using World War I draft records linked to census data. Difference-in-differences estimates indicate that white males born in hurricane-prone US states who experienced a hurricane in utero or as infants had 5% lower income. Labor force participation was unaffected, while education and migration account for a small portion of the effects on income. Empirical tests suggest the persistent impact of damage is an unlikely channel. Thus, we attribute the findings to lower health capital stemming from temporary disruption in the aftermath of storms.

We make use of a new data resource--merged birth and school records for all children born in Florida from 1992 to 2002--to study the relationship between birth weight and cognitive development. Using singletons as well as twin and sibling fixed effects models, we find that the effects of early health on cognitive development are essentially constant through the school career; that these effects are similar across a wide range of family backgrounds; and that they are invariant to measures of school quality. We conclude that the effects of early health on adult outcomes are therefore set very early.

Little is known about the role birth order plays in delinquency and adult crime outcomes that carry significant externalities. We use rich data sets from Denmark and Florida to examine these outcomes and explore potential mechanisms. Despite large environmental differences between the areas, we find remarkably consistent results: in families with two or more children, secondborn boys are 20%–40% more likely to be disciplined in school and enter the criminal justice system than are their firstborn male siblings. We rule out health at birth and school quality as mechanisms but find evidence for the role of parental time investment.

Recent evidence indicates that boys and girls are differently affected by the quantity and quality of family inputs received in childhood. We assess whether this is also true for schooling inputs. Using matched Florida birth and school administrative records, we estimate the causal effect of school quality on the gender gap in educational outcomes by contrasting opposite-sex siblings who attend the same sets of schools--thereby purging family heterogeneity--and leveraging within-family variation in school quality arising from family moves. Investigating middle school test scores, absences and suspensions, we find that boys benefit more than girls from cumulative exposure to higher quality schools.

We explore how access to modern hospitals and medicine affects mortality by leveraging efforts of the Duke Endowment to modernize hospitals in the early twentieth century. The Endowment helped communities build and expand hospitals, obtain state-of-the-art medical technology, attract qualified medical personnel, and refine management practices. We find that Duke support increased the size and quality of the medical sector, fostering growth in not-for-profit hospitals and high-quality physicians. Duke funding reduced both infant mortality—with larger effects for Black infants than White infants—and long-run mortality. Finally, we find that communities aided by Duke benefited more from medical innovations.