This rejoinder responds to Kenneth Roberts’s criticisms of my review published in these pages in August 2016. Roberts’s book, Changing Course, is a major contribution to scholarship on Latin American party systems, but Roberts and I disagree about three key issues. Firstly, because Roberts measured the economic effects of different kinds of party systems beginning in 1980, his coding of party systems must be immediately prior to that. Secondly, Roberts and I agree that many Latin American party systems exhibited deep change during the long era of important substitution industrialization (roughly 1940–80). In my judgement, the coding of party systems should reflect these changes more than Changing Course does. Thirdly, following principles for understanding causation, I argue that Changing Course does not conclusively show that labour-mobilizing party systems caused deeper economic crises and greater party system instability than elitist party systems.
