In the United States, like in other countries, the agrifood supply chain faces challenges from a growing population and less predictable weather conditions. Extreme weather events such as droughts decrease agricultural yield and harvested areas, impact the domestic trade of agricultural products and, in turn, food manufacturing. We investigate this relationship at the state level by estimating the food manufacturing production function in a two-stage instrumental variable estimation process. We first assess how drought affects trade in animals and fish, cereal grains, and all other crop products. Next, we estimate a nested production function for processed food. Our findings indicate that the impact of a drought is far from being confined to the area where it happens. At the national level, we find that a 1% increase in drought in the states producing agricultural commodities reduces their exports to other states by 0.5% to 0.7% which, in turn, reduces food manufacturing production by an average of 0.04%. The capacity to shift the origin of import flows, adjust their volume, and substitute agricultural inputs supports the resilience of the food manufacturing sector. We further estimate the 48 × 48 pairwise dependence across states and by commodity group. While cereal grain production is more spatially concentrated than other crops, the agrifood supply chain can enhance its resilience by sourcing from geographically diverse counties within key supplier states and improving multistate coordination. These findings provide important insights for policymakers and industry stakeholders willing to reduce the food system vulnerability to extreme weather events.
Yim et al. (2025) studied this question.