Economics Department†

This paper develops a general model of elections in which two candidates poll voters prior to taking policy positions. Private polling provides candidates private information about the voters’ preferred policies. Unlike electoral competition models without private information, platforms diverge in equilibrium. Candidates who receive moderate signals adopt more extreme platforms than their information suggests, while candidates who receive more extreme signals moderate their platforms relative to their pollsters ’ advice. Contrary to previous models of elections, we show that some platform differentiation increases voters ’ welfare. Although candidates’ platforms diverge in equilibrium, they do not do so by as much as voters would like. As a result, the electorate always prefers private polling to public polling, as public polling reduces the choices that candidates provide. For similar reasons, some noise in the polling technolog