Welfare Impacts of Climate Change on Agriculture: Evidence from Over 1,000 Yield Studies

There is now a large body of scientific evidence based on experiments, process-based crop models, and econometric studies, documenting the expected impact of climate change on crop productivity. However, the implications of these changes for more salient economic outcomes such as production, prices, consumption, and welfare are poorly understood. In particular, recent scientific findings are not reflected in the calibration of damage functions in Integrated Assessment Models (IAMs), used to calculate the social cost of carbon (SCC), which are instead based on a small number of studies from the early-to-mid 1990s. In this paper we perform the first end-to-end analysis directly linking the scientific literature on biophysical climate impacts to the SCC. We do this by connecting a comprehensive meta-analysis of crop yield response to climate change, a computable general equilibrium model (GTAP), and the FUND IAM. We find negative effects of warming on most crops in most places and very limited potential for adaptation to offset declines. These yield impacts cause prices to increase between 24% (maize) and 1% (wheat). The welfare effects of these changes are mediated by terms-of-trade effects that tend to moderate negative impacts in net exporters (Brazil, Canada, United States) and exacerbate them in net importers (Middle East, Japan). Overall, damages from warming are negative in most regions and increase approximately linearly with temperature. Incorporating these new damage functions into FUND more than triples the SCC from $7 per ton to $23 per ton. This is due to impacts in the agricultural sector changing from benefits of $7 per ton to costs of $9 per ton. This has direct policy implications given FUND is one of three models used by the US government to calculate the SCC applied to cost-benefit analysis of climate-relevant regulations.