A1987
Title: Temperature and growth: A panel mixed frequency VAR analysis using NUTS2 data
Authors: Fabio Parla - University of Palermo (Italy) [presenting]
Andrea Cipollini - University of Palermo (Italy)
Abstract: The effects of an increase in temperature levels on regional economic activity are studied using data observed for 225 EU NUTS2 regions over the period 1981-2019. The regional economic growth is proxied by the real GVA growth rate (available at an annual frequency), while we use the level of temperature recorded during four seasons as a meteorological variable. The joint interaction between economic growth and seasonal temperature is modelled through a Panel VAR fitted to data sampled at a different frequency. The structural form of temperature shocks is identified by imposing a recursive ordering of the variables, with the temperature ordered before the proxy of economic activity. The empirical results show that an increase of 1 degree Celsius in the level of annual temperature is associated with a reduction in regional economic growth (-0.3 percent is the cumulative response of GVA after a two-year horizon). Moreover, we find that on impact, summer and, to a lesser extent, winter contribute the most to the decrease in economic activity. Results from a subsample analysis (i.e., 2000-2019) reveal a more detrimental effect of an increase in temperature on economic activity. Finally, the heterogeneity in the response across NUTS2 is investigated by splitting the sample into climate and income groups.
