Eyal Frank, UChicago

Abstract:
New technologies that deliver large private benefits see widespread adoption, even if little is known at the time regarding potential adverse side effects. However, in the presence of negative externalities, the large-scale adoption of new technologies can lead to an unequal distribution of gains and losses. An example of such a technology is the rapid and widespread adoption of DDT for civilian use after World War II, which ushered in an era of high synthetic pesticide use. While DDT was banned in the US in 1972, it is currently still used in 24 countries, and there are periodic calls to reinstate its use. In this paper, we use two natural experiments that created variation in DDT use: The agricultural use, namely in cotton cultivation, and the use of DDT in forest pest control operations. We leverage baseline agricultural cultivation intensities, as well as detailed forest spraying data, and test DDT’s historical impact on health: We find strong evidence that in the US South, where DDT was used extensively in cotton production, infant mortality rates (IMR) increased by 11% relative to the baseline in 1945. This increase in IMR is mostly driven by non-white IMR. Looking at DDT use for the control of forest insects, we find that counties that were downstream from the spraying locations experienced meaningfully higher IMR, but counties upstream did not. In ongoing work, we are testing for the impacts on agriculture. Because DDT was effective against target and non-target insects, we examine how pollination-dependent crops experienced differential changes in yield growth rates. We find that while corn, cotton, and wheat saw large gains from the adoption of DDT, apple yields suffered large declines. These results highlight DDT’s overall negative effect on short-term health, which might have exacerbated existing health inequalities, and potentially had differential agricultural productivity gains.