Aggregate Effects of Large-Scale Campaigns on Voter Turnout

Despite the central role of campaigns in electoral politics and despite many experiments on campaigning, we know little about the aggregate effects of an entire campaign on voter participation. Drawing upon inside information from a U.S. presidential campaign and utilizing a geographic research design that exploits media markets spanning state boundaries, we estimate that the 2012 presidential campaign increased turnout in highly targeted states by 7-8 percentage points. Further evidence suggests that the predominant mechanism behind this effect is traditional ground campaigning, which has increased dramatically in recent years. Additionally, we find no evidence of diminishing marginal returns to ground campaigning, meaning that voter contacts can have large aggregate effects over the course of a campaign. stract in PSRM publicatio