The hidden cost of bananas: pesticide effects on newborns’ health

We study the effects of aerial fumigation of banana plantations on newborns’ birth weight during the period 2015-2017 in Ecuador. We use mothers’ addresses and information on the perimeter of the plantations to create an individual measure of newborns’ exposure to pesticides. We use this measure to implement three independent identification strategies to address the endogeneity of exposure to aerial fumigations. First, we consider a difference-in-differences strategy that exploits seasonal variations in the use of pesticides across provinces. Second, we estimate a difference-in-differences model that considers geographical variations in the use of pesticides across comparable crops. Third, and finally, we estimate a maternal fixed effects model to examine the effect of pesticides on siblings who had a different residence during gestation and who were exposed to different levels of fumigations. Our first empirical model shows that newborns exposed to pesticides, when their first gestational trimester coincides with the periods of intensive fumigations of the plantations, have a birth weight reduction of between 38 and 89 grams. Moreover, exposure to pesticides increases the likelihood of low birth weight and low Apgar score at the first minute by around 0.35 and 0.33, respectively. The second model finds that newborns exposed to fumigated banana plantations have a birth weight deficit of between 29 and 76 grams, when compared to those exposed to other fumigated crops. Finally, the maternal fixed effect model show that girl newborns exposed to pesticides have a birth weight deficit of 346 grams when compared to non-exposed siblings.