The growth effect of democracy: Is it heterogenous and how can it be estimated?

We estimate the effect of political regime transitions on growth with semi-parametric methods, combining difference in differences with
matching, that have not been used in macroeconomic settings. Our semi-parametric estimates suggest that previous parametric estimates
may have seriously underestimated the growth effects of democracy. In particular, we find an average negative effect on growth of leav-
ing democracy on the order of -2 percentage points implying effects on income per capita as large as 45 percent over the 1960-2000 panel.
Heterogenous characteristics of reforming and non-reforming countries appear to play an important role in driving these results.