Mandate and paternalism: A theory of large elections

Abstract

We propose a game theoretic costly voting model of large elections that incorporates the assumption that mandate matters. This innovation is motivated by empirical evidence that US Representatives with larger victory margins on average vote in a more partisan manner. If voters are paternalistic, this new model predicts strictly positive limiting turnout rates as the population grows arbitrarily large. The model also preserves stylized comparative statics results of costly voting models, including the underdog effect and the competition effect. Finally, we develop an innovative computational strategy to solve the model for large, finite, electorates and show that our results are not only qualitatively, but also quantitatively relevant.