A time to throw stones, a time to reap: how long does it take for democratic transitions to improve institutional outcomes?

We study the impact of democratic transitions on institutional outcomes. Using an event study method and a sample of 135 countries over the period 1984-2016, we observe that democratic transitions improve institutional outcomes. The effect appears within 3 years after the transition year. The results are robust to alternative definitions of transitions, alternative codings of pre- and post-transition years, and changing the set of control variables. We also find that both full and partial democratizations improve institutional outcomes. Transitions out of military regimes or communist autocracies do not. The effect of democratization depends on GDP per capita, education, and the regularity of the transition. Finally, the evidence suggests that the effect is particularly clear on the corruption, law and order, and military in politics dimensions of the index.