Estimating climate damage functions for firm value added using microdata

This study estimates climate damage functions for productivity using firm-level microdata while incorporating climate extremes such as precipitation variability, flood exposure, and drought intensity. Motivated by recent methodological debate regarding spatial correlation and limited independent variation in aggregate climate–growth estimates, the analysis adopts a long-difference framework that isolates persistent climate effects rather than short-run shocks. Using harmonized data from the World Bank Enterprise Surveys covering 125 countries and high-resolution climate data from ERA5 reanalysis, the results reveal a robust downward-sloping and convex temperature damage function, indicating that warming imposes increasingly large productivity growth losses in hotter baseline climates. While global-average effects are statistically weak, region-specific estimates uncover substantial heterogeneity. Extensions incorporating precipitation and climate extremes preserve the qualitative damage pattern and generally imply smaller projected losses, suggesting that part of the temperature effect may operate through additional climate channels. Leave-one-region-out tests confirm that the results are stable and not driven by any single region.