Abstract: This paper explores the causal pathway by which poor fetal health translates into reducing educational attainment and earnings as an adult. Using insights from the medical literature, I decompose low birth weight infants into two distinct subtypes: a symmetric type, which is characterized by cognitive deficits, and an asymmetric type, which exhibits little to no cognitive problems. Using data from a longitudinal survey of newborns, I establish three results: First, there is empirical evidence of brain sparing in the asymmetric subtype, but not in the symmetric subtype. Second, despite differences in cognitive impairment, both subtypes exhibit similar impairment to physical health. And finally, there is evidence that the causes and timing of onset during pregnancy are different for asymmetric and symmetric growth restriction. The results indicate that differentiating between these subtypes may offer new opportunities to identify the underlying casual relationships between health and human capital development, as well as uncovering the ''black box'' mechanism behind the fetal origins hypothesis. These results also have broad implications for the timing of policy interventions aimed at pregnant women.

Abstract: The fetal origins hypothesis has become increasingly popular in the economics literature. Despite convincing evidence that exogenous changes in the fetal environment negatively affect educational attainment and earnings, it is yet unknown whether this phenomenon is the result of a decrease in physical health or cognitive ability. This paper seeks to answer this question by recognizing heterogeneity in growth restricted infants. I split the cohort that experience the 1918 influenza pandemic in utero into two subtypes: an asymmetric type for which physical health is impaired but brain growth and cognitive function is spared, and a symmetric type for which both physical health and cognitive function is compromised. I find that for the symmetric type, which experienced the pandemic in early gestation, human capital measures are more significantly compromised than the previous literature suggests. On the other hand, for the asymmetric subtype, which experience the pandemic late in gestation, I find evidence of higher educational attainment and earning than the surrounding cohorts.