The link between regional temperature and regional income: Econometric evidence with sub-national data

We study the effect of temperature on economic development on the sub-national level, employing cross-sectional data for up to 15,533 sub-national units from two distinct sources. In contrast to the existing cross-country literature on the temperature-income relationship, our setting allows us to exploit sub-national heterogeneity through the inclusion of country fixed effects and mitigate omitted variable bias. We find no negative relationship between regional temperature and four different measures of economic development (per capita GDP, growth of per capita GDP, nightlights and gross cell production). We also test whether temperature is non-linearly related to income (with hotter regions being potentially particularly prone to adverse effects of temperature on income) but find no evidence in favor of such a relationship. Finally, we examine whether the effect of temperature on economic development is especially pronounced in poorer regions (e.g., due to weaker adaptation), but find no robust evidence for this proposition. In sum, our findings suggest that adaptation to temperature differences could be feasible and relevant.