³ÉÈË´óƬ

Does development lead to the establishment of more democratic institutions? The key to the puzzle, we argue, is the previously unrecognized fact that based on quantitative regime scores, countries over the past 50 years have clustered into two separate, very distinct, yet equally-common stages of political development - authoritarian states with low levels of freedom on one side and democracies with liberal institutions on the other side of a bimodal distribution of political regimes. We develop a new empirical strategy - exploiting exogenous world economic factors and introducing new panel data estimators - that allows for the first time to estimate the effects of development as well as changing unobserved country effects in driving democracy at these different stages of political development. We find that income and education have the least effect on democracy when authoritarian regimes are consolidated and that only changing country effects, possibly accounting for institutional legacies, can lead to political development. Ironically, it is in highly democratic and wealthiest of nations that income and education start to play a role; however greater wealth and better educated citizenry can both help and hurt democracy depending again on what the country's institutional legacies are. Far from accepting the notion that much of the developing world is cursed by unchanging and poor long-run institutions, policy-makers should take note that with democratization we also see changing country-specific factors that in turn condition the difference income and education make for democracy.