Tropical cyclones and fertility: New evidence from developing countries

Abstract

Does exposure to tropical cyclones affect fertility? This paper addresses this question by constructing a panel dataset from geolocated micro-data about the fertility history of mothers along with their local exposure to tropical cyclones for a sample of six developing countries for the 1985–2015 period. We then estimate the causal effect of tropical cyclone shocks on women’s likelihood of giving birth. We find evidence that tropical cyclone exposure has a significantly negative effect on motherhood. A cyclonic wind exposure between 60 and 117 km/h (resp. at least of 118 km/h) decreases the probability of giving birth by 7.8 (resp. 7.0) points a year after exposure. We also observed that the magnitude of the effect varies with the degree of cyclonic exposure associated with the mothers’ living environment and the number of children ever born. In particular, the fall in the likelihood of giving birth is lower for mothers living in cyclone-prone areas and for those who already have children. Alternative specifications of our baseline model provide further insights: (i) recent past exposure to cyclones was associated with a lower decrease in fertility when exposed once again; and (ii) no evidence of non-linearities was observed in the effect.