Abstract
Robust estimation of the impact of political institutions on economic development requires the identification of valid instruments for institutional quality. Acemoglu et al. [2001] introduced the use of colonial settler mortality rates as such an instrument. Our paper develops a more eclectic theory of colonial development, and compares the performance of the settler mortality model to alternatives incorporating instruments reflecting the production structure of colonial economies. Ceteris paribus, colonies with a natural comparative advantage in pastoral agriculture were more likely to experience European settlement that led to non-extractive institutions. Some – but not all – of Acemoglu et al.’s conclusions are robust to the use of a wider set of instruments.