Extreme weather events, climate expectations, and agricultural export dynamics

Rising temperatures and shifting precipitation patterns threaten agricultural yields in many key global production regions. This paper assesses the impact of growing-season extreme weather events on agricultural export outcomes in the short run, as well as the association between agricultural exports and long-run climate expectations and variance. Our analysis matches information on bilateral trade flows with high-resolution, geospatial data on growing area, planting and harvest dates, and weather for three highly traded staple crops-maize, soybeans, and rice-which together account for almost half of global calorie consumption. We use an econometric gravity model to estimate the short-run effects of weather volatility and a nonparametric series regression to infer long-run climate-export associations. We then use our estimates to simulate the effects of various climate and weather counterfactuals on the agricultural export landscape. We find that 2-standard-deviation extreme weather events (measured using the water balance deficit) reduce maize, rice, and soybean bilateral export values by 48.2%, 53.4%, and 21.7%, respectively. Our long-run results imply that increases in the standard deviation of weather are associated with lower export values across all three crops. An increase in the frequency of extreme events has the potential to greatly shift current commodity export patterns. Understanding these shifting patterns of trade is necessary to implement trade policy that enables countries to leverage their evolving comparative advantages and ensure the effectiveness of trade as a tool mitigating the negative production effect of climate change.