Fetal mortality. While my results demonstrate that in utero exposure to tropical cyclones can affect mortal- ity post-birth, the relationship between exposure and fetal losses is more difficult to identify because I do not have information regarding fetal deaths as only live births are reported. To investigate the possibility of tropical cyclones causing increased loss during pregnancy, I adopt the methodology used by Xxxxxxx and Xia (2018) by examining the effect of tropical cyclone exposure on the fraction of males in a given birth cohort. I use the model outlined in equation (5) and results are presented in Table 9. Though not precisely estimated, these negative effects are consistent with the fragile male hypothesis, which argues that the biolog- ical fragility of the male fetus makes it such that males are more susceptible to negative early life shocks than females in the womb (Xxxxxxx, 2000). Thus, any unobserved effects on fetal mortality would be disproportionately present in male fetuses, thus reducing the fraction of males in the birth cohort. Xxxxxxx and Xia (2018) also find that in utero exposure to typhoons in the Philippines reduces the fraction of males in the birth cohort, though their estimated effect sizes are larger.
