Ambiguous Platforms and Correlated Preferences: Experimental Evidence

This paper studies, theoretically and experimentally, a model of electoral competition that allows for platforms where candidates may be ambiguous about which policy they will implement if elected. We argue that uncertainty about the policy preferences of the electorate, combined with perceived similarity of voters and candidates, can lead to the latter running on these ambiguous platforms. By appealing to voters from both ends of the spectrum, such platforms can ensure electoral success for noncentrist candidates in a sufficiently polarized society. Ambiguous platforms pose a threat to democratic representation because winning noncentrists always implement policies in favor of a minority and against the preferences of the majority. In our laboratory experiment, ambiguous platforms are chosen frequently by candidates and gain notable support from voters. Our main treatment variation provides causal evidence that ambiguous platforms are more popular among noncentrist voters if one of the candidates is a known centrist.