Robustness of the Impact of Climate Change on U.S. Corn Yields

One puzzling aspect of climate change literature is that regressions involving yield give much larger estimates of the impact of climate change than other methods. The model suggested by Schlenker and Roberts (2009) is replicated and then assumptions are modified. The model is robust to most changes in assumptions. Com yield change predictions are most sensitive to using long-term average data instead of yearly data. Using long-term average data and changing six other assumptions, the average predicted com yield decrease is 19.3% under RCP4.5 and 32.6% under RCP8.5. These results are closer to other research methods such as Ricardian models and plant growth simulations. This predicted rate of decrease is less than the recent increase in yield due to technological change. So, if the technological change continues, com yields would still go up over time.