Endogenous Democratization

A rich dictatorships more likely than poor dictatorships to collapse and be replaced by democracies? Consider, for example, Chile, which in 1985 had a per capita gross domestic product (GDP) of $3,400 and was under dictatorship, and Benin, which in the same year had a per capita income of about one-third of Chile’s, $1,108, and was also under dictatorship. Setting aside other differences between these countries, did their relative levels of development make a transition to democracy more likely in Chile than in Benin? Anyone who has followed recent scholarship would be likely to answer no. This is the answer that would follow from Adam Przeworski and Fernando Limongi’s “Modernization: Theories and Facts,” a study that hit the field of political development like a bolt of lightning and immediately changed the landscape.1 In it they reconsider the classic proposition that economic development favors democracy, identifying an ambiguity in this proposition. Why do we observe a higher proportion of democracies among rich countries than among poor countries? Is it because development increases the likelihood that poor countries will undergo a transition to democracy? They call this an “endogenous” theory. Or is it because development makes democracies, once established, less likely to fall to dictatorships? They call this an “exogenous” theory. The conceptual distinction is brilliant, and Przeworski and Limongi offer evidence that the exogenous theory holds and the endogenous one fails. Their conclusion has been deeply influential among social scientists and policy analysts. In a review of Przeworski and Limongi’s later book (coauthored with Michael Alvarez and José Antonio Cheibub), which restates the rejection of endogenous democratization, David Brown writes: “In a convincing fashion, the authors argue that modernization