Is temperature adversely related to economic growth? Evidence on the short-run and the long-run links from sub-national data

We investigate the effect of rising temperatures on economic development, using sub-national data for approximately 1,500 sub-national regions in 81 countries from the 1950s to the 2010s. Accounting for region- and time-fixed factors by means of a two-way fixed effects panel approach, we find no evidence that rising temperatures are adversely related to regional growth measured as changes in regional per capita gross domestic product (GDP). In addition to a panel setting, we also consider the long-run analogue of the panel model, exploring the relationship between regional temperature and growth over longer time periods. Applying this long-difference approach, we find evidence of a statistically significant negative association between temperature and regional economic activity. This suggests that intensification effects matter, meaning that the adverse relationship between temperature increases and growth may compound and materialize only in the longer run. What is more, we find that these adverse long-run effects of regional warming matter only to regions located in countries with relatively unfavorable economic and institutional conditions, that is, in countries with high levels of poverty, a lack of democracy, and a weak rule of law. This strongly points to the role of sound (country-specific) economic and institutional conditions in reducing vulnerability to higher temperatures. In line with this interpretation, we find no evidence for an adverse long-run relationship between temperature and growth for regions located in richer and democratic countries or those with an established rule of law.