Democracy, regime duration, and growth

Abstract

Findings of regularities in statistical relationships between regime type and economic growth are debatable, because specifications are variable. We argue that the effect of regime type on growth is contingent on the duration of the regime. Further, the effect of duration depends on regime type: older autocracies grow at a different rate than older democracies. The reason is that the causes for decelerating growth are fundamentally different in democracies and autocracies. Our estimations are run on a large (unbalanced) annual panel of 155 countries from 1950 – 2006. In this sample, autocracies grew around 0.75 percentage points more slowly than non-autocracies, holding constant the regime-specific effect of regime duration. Autocratic growth rises for the first 30 years, after which growth rates fall rapidly. Non-autocracies always grow faster than equally old autocracies, but the differences are especially pronounced when young and old regimes are compared. Some implications of this result for the political system of Brazil are discussed.