Policy Persuasion and Electoral Competition

The paper studies the impact of biased media news on policy outcomes in the classical Hotelling-Downs setup of winner-take-all elections with state-dependent payoffs. When voters and the two parties do not observe this state but an ideologically biased media does, it can indulge in two forms of cheap talk messaging: policy advice (that comes before parties commit to platforms) and party endorsements (that come once parties can be readily identified by their electoral platforms). We show that while the standard result of policy-convergence remains robust, the Median Voter Theorem is violated in a number of important ways. In particular, when there is too much uncertainty in the environment we show that parties converge to media pandering platforms in ‘fear’ of influential media endorsements at a later stage of the electoral process. We call this media apprehension that yields party leaderships to be ‘media darlings’. We then prove a staggering result in which we show that the welfare of voters exhibit a nonmonotonic relation with the size of the media bias: starting from no bias, voter’s welfare falls as this bias rises up to a point after which a higher bias increases this welfare. This implies that while media skepticism is widespread and driven by perceived ideological bias of the media, extremist media presence is better for the majority of uninformed voters than moderate ones. We then show that while on one hand if party endorsements were all that the media could provide then a complete ban on the media would unambiguously benefit voters, the presence of media that only provides policy advice on the other hand always benefits them. In spite of that our second surprising result is that only if the media bias is significantly large a full media ban is better than having a media which indulges in both policy advice and party endorsements. The paper then studies other issues involving the voters’ choice between indirect democracy (where the media informs the society, parties choose platforms and voters elect parties) and the dictatorship of the media (where voters delegate decision rights to the media elite) and a party’s choice between candidates and platforms. We also look at issues concerning editorial delegation and provide a theory of equilibrium editorial bias. Some of our results have direct implications for the Cheap Talk literature. JEL Classification: C72, D72, D82.