An Informational Rationale for Political Campaigns

Abstract

Existing theoretical and empirical research has a dim view of campaigns: they are distractions at best and exercises in vote buying at worst. We study a setting in which citizens are rational, but imperfectly informed, and where elections serve three goals: aggregating preferences, aggregating information and selecting good politicians. We highlight two essential roles of campaigning. First, campaigning allows candidates to tell voters what they intend to do while elected. Second, during the campaign, candidates present themselves to citizens, so citizens can learn about the capabilities of politicians that are relevant to their functioning. Without campaigns, voters learn preciously little about politicians or platforms and elections cannot achieve the aforementioned goals. We then extend our model to explore (1) when politicians commit to their platform, when they remain ambiguous and how this affects issue salience, (2) how campaigns determine when elections revolve around personality and when around policy and (3) when politicians can claim an electoral mandate for their platform. Joint work with Otto Swank.