By UNC Kenan-Flagler Business School Professor of Finance and National Bureau of Economic Research (NBER) Research Associate Ric Colacito
Climatologists project that global temperatures may rise by up to four degrees Celsius over the next century. This projection raises a natural question: “Can we assess the impact that this temperature increase will have on the U.S. economy?” In the paper, “Temperature and Growth: A Panel Analysis of the United States” which I coauthored with Bridget Hoffmann of the Inter-American Development Bank and Toan Phan of the Federal Reserve Bank of Richmond, we answer this question by employing a rich historical dataset of U.S. temperatures, which allows us to exploit random fluctuations in seasonal temperatures across years and states in a panel regression framework. In this empirical setting, we investigate the effect of a state’s average seasonal temperatures on the growth rate of gross state product (GSP).
Our analysis documents that rising summer temperatures negatively affect the growth rate of GSP and, to a smaller extent, an increase in the fall temperature positively affects this growth rate. Furthermore, the negative economic impact of rising summer temperatures has accelerated substantially during the last 20 years or so. Our research also documents that summer temperature has a particularly strong effect in states with higher average temperatures, most of which are in the South.
We also find that an increase in the average summer temperature has a pervasive effect throughout an entire range of industries, not just on those sectors naturally exposed to weather or changing climatic conditions, such as agriculture and construction. Indeed, our research indicates that an increase in the average summer temperature negatively affects the growth rate of economic output of many industries, including finance, real estate, insurance, services, retail and wholesale. Together, the affected industries account for more than a third of national gross domestic product (GDP).
Finally, we combine our estimated impact coefficients with climatologists’ projections of the expected temperature increases under various emissions scenarios to get an assessment of the possible long-term repercussions of rising temperatures on U.S. economic activity. Specifically, we look at the projected seasonal temperature increases for the U.S. for the period 2070-2099 from 16 general circulation models under three different greenhouse gas emissions scenarios: high, medium and low emissions. We show that under the most conservative emissions scenario, the projected trend in rising temperatures is expected to reduce the growth rate of U.S. output by 0.2 to 0.4 percentage points by the end of the century. These figures are not negligible: at a historical growth rate of U.S. GDP of 4% per year, this would correspond to a reduction by up to 10%. The results are even more dramatic when considering the high emissions scenario. Here, the reduction of economic growth could reach up to 1.2 percentage points, corresponding to roughly one-third of the historical annual growth rate of the U.S. economy.
It is important to note that these estimates likely represent upper bounds because they do not consider any adaptation that might occur in the longer term or technological progress that could dampen the effect of temperature on economic growth.
Read the full paper, here.
Based on the paper: “Temperature and Growth: A Panel Analysis of the United States” by R. Colacito, Bridget Hoffmann, and Toan Phan, published in the “Journal of Money, Credit and Banking”, Volume 51, Issue 2-3, March-April 2019, pages 313-368.