Markus Eberhardt

 Democracy does cause growth: Comment

This comment revisits the causal relationship between democracy and growth as recently studied in Acemoglu, Naidu, Restrepo, and Robinson (2018, ANRR). My analysis provides a number of extensions which significantly qualify the findings of these authors: I first demonstrate the fragility of ANRR’s empirical results — loss of statistical significance and collapsing long-run democracy coefficients — to the omission a fraction of the sample’s observation. I suggest that this fragility is due to two implicit assumptions in their empirical analysis: first, a common equilibrium democracy-growth nexus across countries; and second, the absence of strong cross-section dependence, arising from common shocks with heterogeneous impact and/or productivity spillovers, among other potential sources. Using novel methods for policy evaluation with interactive fixed effects I then demonstrate that in the long-run democracy does indeed have a positive average effect on per capita income, and that this finding is robust to significant sample reductions. However, at 10% the average effect is between one-half and one-third the magnitude suggested by the results in ANRR.

Lunch from 12.30pm