Competing demand-side explanations and populism: cross-national variation in a recursive ideational system

This article models the rise of populism in a country as a sequential recursive system driven by the interaction of voters’ contextual demands and the supply of populist political responses. We introduce to this literature high-frequency media event data extracted from a corpus of two terabytes of articles drawn from 72 countries over six years using natural language processing. Further, we introduce vector autoregression to the study of populism, which allows us to model the way voters’ grievances and politicians’ rhetoric and policy responses influence each other bidirectionally over time. Our analysis reveals that voter demands and political responses are best modelled as a recursive system. Given the positive feedback loop, efforts to curb the growth of populism and limit democratic backsliding will require both addressing underlying grievances (i.e. demand) and weakening or substituting for the efficacy of existing populist rhetoric (i.e. supply).