Spatial theories of electoral competition

Abstract

Rational choice theory dominates the study of American electoral politics, but a profound disconnect exists between formal spatial modeling and empirical reality. Traditional spatial models, such as the median voter theorem, often fail to account for the multidimensionality of issues and the diverse motivations of candidates. Contemporary theoretical developments, including probabilistic voting and policy-seeking models, frequently function as post-hoc justifications for “stylized facts” rather than as predictive frameworks. The proliferation of unobservable variables within these models—such as candidate risk aversion and specific utility functions—limits their falsifiability and hinders systematic empirical application. Bridging this gap requires a transition toward comparative statics and research designs that analyze candidate behavior in response to exogenous environmental shifts. Furthermore, empirical inquiry must allow for the possibility that candidate strategies are shaped by factors other than optimal strategic play, such as organizational habits or cognitive biases. Restoring the scientific utility of spatial theories necessitates balancing theoretical nuance with testable propositions that distinguish rational action from alternative behavioral explanations. – AI-generated abstract.