Estimating Weather and Climate Impacts with Panel Models

This paper begins with an assumed nonlinear effect of climate on an economic outcome and then derives the resulting short run and long run behavior to changes in weather. The short run annual effects are a function of intertemporal weather deviations and the interaction between weather deviations and the long-run marginal effect of climate. One can identify the underlying nonlinear climate effects with these interaction terms. We illustrate these principles by showing how both weather and climate affect annual GDP per capita in a panel of countries. GDP is generally more sensitive to short run changes in temperature than long run changes. A long run 1 warming has a negligible effect on GDP per capita overall but it will cause losses in the hottest countries of -1% to -1.5% per C. Precipitation also has a very small effect on GDP per capita. These estimates just reflect the direct effect of temperature and precipitation on the economy and do not reflect nonmarket damages and damages caused by sea level rise and storms.