Acemoglu, Johnson, Robinson, and Yared (2008) document that the cross-country correlation between income per capita and democracy disappears once including country fixed effects. This paper tests the hypothesis that the effect of income on democracy might differ systematically across countries. A replication of the estimation in a less restrictive empirical framework provides evidence for significant but heterogeneous effects of income on democracy for former colonies and non-colonies, as well as within the sample of former colonies. These heterogeneous effects are related to colonial history and early institutions, and are robust to the use of alternative data and estimation techniques.