An Empirical Model of Agricultural Subsidies with Environmental Externalities

Résumé

Agriculture is widely recognized as one of the sectors most vulnerable to extreme temperatures. Yet, crop losses are estimated to form only a modest share of aggregate macroeconomic damages from climate change, since agriculture accounts for a small share of global GDP. These estimates, however, arise from analyses that largely ignore the critical role of agriculture as an upstream sector in global production networks, including the sectoral and spatial linkages connecting local agricultural output to other sectors and regions. In this paper, I develop a a novel reduced form method to incorporate input linkages between sectors and countries that I use to estimate the aggregate impacts of extreme heat in agriculture. A multi-region multi-sector production network model illustrates how extreme heat in agriculture can propagate to downstream sectors across countries by reducing supply availability and increasing intermediate input prices. Exploiting the geographic distribution and temperature sensitivity of 118 crops across the world, I construct a measure of exposure to extreme heat in agriculture and show that it induces substantial losses to downstream sectors, across national borders, and beyond first degree linkages. Counterfactual exercises reveal that input linkages are responsible for approximately 70% of the total value added losses induced by extreme heat in agriculture. The analysis demonstrates that adaptation in agriculture to heat in regions that are central to global production networks can have substantial co-benefits downstream and in other locations.