The American Journal of Agricultural Economics just released my co-authored paper with Ahmad Wahdat on food system vulnerabilities (this an updated and revised version of the working paper I blogged about in November). The paper uses input-output tables to calculate how, for a given food processing sector in a given state, how diversified input purchases are across input categories.
Here is one paragraph from the conclusions:
“In this article, we identify the U.S. food industries’ vulnerability to upstream industries and labor occupations. We do this by calculating the exposure of food industries to upstream industries and labor occupations in terms of input purchases. Upstream industries are relevant for the purchase of intermediate inputs, and labor occupations are relevant for specialized labor. To complement our analysis of food industries’ vulnerabilities, we calculate the Simpson’s Diversity Index (SDI) and also estimate the gross output elasticity of inputs. We use data on input purchases, gross output, and labor earnings from Economic Modeling Specialists International (EMSI) and the U.S. Census Bureau from 2019. In terms of our main findings, we show that the animal slaughtering and processing industry has the lowest SDI score (0.66), which means the industry makes most of its purchases from a few upstream industries. Out of all upstream industries, food industries’ intermediate input purchases are exposed to a total of 22 upstream industries that can be classified under four groups: (a) agriculture, fishing, and hunting (RawAg); (b) food and beverage (FoodBev); (c) transportation and packaging (Transp); and (d) miscellaneous (Misc). Food industries rely heavily on (a) production labor and (b) transportation and material moving labor. Food industries’ gross output is the most elastic with regard to intermediate inputs from FoodBev industries (0.23) and RawAg industries (0.22). Among labor occupations, gross output elasticity is the highest with regard to production labor (0.1).”
You can read the whole thing here.