Mapping and Quantifying the Externalities of Factory Farms

Résumé

Factory farms account for an ever-growing share of animal agriculture worldwide, marked by intensive production practices and heavy geographic concentration. These operations generate large quantities of waste and have long raised serious concerns for environmental quality and public health, including rising threats such as antimicrobial resistance. Yet the environmental economics literature offers almost no robust, large-scale causal evidence on their impacts because systematic data on their location and evolution have been lacking. We assemble the first panel database of U.S. Animal Feeding Operations (AFOs), combining multi-state administrative records with nationwide coverage for 2011–2021 detected in remote sensing imagery, using a suite of machine learning approaches, including frontier geofoundational models. This resource makes it possible to precisely quantify the externalities of factory farming across multiple dimensions of environmental quality and health. The presentation reports preliminary results from two avenues that quantify and value such externalities. The first analyzes swine operations in Iowa and North Carolina, leveraging identification strategies based on spatio-temporal variation in animal production and extreme rainfall events. Results show significant increases in downstream fecal coliforms and nutrient concentrations, and reductions in dissolved oxygen, including from facilities below current regulatory thresholds. The second applies a hedonic framework to property markets, showing how proximity to AFOs is capitalized into housing prices and providing a revealed-preference lower bound on the costs these operations impose on surrounding communities. Together, the database and analyses provide new evidence on the magnitude and valuation of the environmental and economic externalities of factory farming, and contribute to a broader quantification of their social costs.