Unbundling Ex-Colonies: A Comment on Acemoglu, Johnson, and Robinson 2001

Abstract

In this comment on AJR (2001), we argue that a bundling of all former colonies into one ‘colonial’ theory of comparative development is problematic for several reasons. During the mercantilist wave of mainly Latin American colonization between 1500-1830, strong capitalist institutions were largely non-existent in the Western world. During the scrable for Africa, starting in 1885, good institutions were an option, but the disease environment had then drastically changed. We show that when AJR’s sample of 64 former colonies is disaggregated into a Latin American, an African, and an Asian/Neo-European subsample, the proposed relationship between settler mortality and institutions is weak or rejected for Latin America and Africa.