(Mis)Information and the Electorate: The Dynamics and Individual Level Effects of Political Advertising Claims

Abstract

This paper provides the first systematic examination of the relationship between the accuracy of the claims made in political advertising and political knowledge and turnout. Political advertising, especially negative political advertising, has been the focus of a lot of attention in political psychology and political communication over the last two decades: the 8-year period between Lau et al.’s (1999, 2007) two meta-analyses saw the number of studies more than double. In addition, Lau et al. note that the methodological rigor of this research has improved considerably. Conventional wisdom has evolved from Ansolabehere and Iyengar’s (1995) claim that negative advertising depresses turnout at elections, particularly among Independents, to arguments that negative advertising either has no unique effects on the electorate (Krasno and Green 2008), or that it often boosts turnout because of its denser content, greater need to back up its claims, and dramatic elevation of the stakes at elections (Franz et al. 2008).This research has, however, steered clear of the question of what such putative gains in knowledge or turnout represent, e.g., whether an enhanced ability to list likes and dislikes of a candidate and vote is based on accurate or inaccurate information, a more important issue to the quality of American democracy than whether turnout goes up or down. In this paper, which is the first of its kind, we endeavour to answer this question. We examine two relationships: between the accuracy of ad information and 1) political knowledge and 2) turnout. We focus on presidential advertising nationwide in the 2008 election, using three sources of data: our analysis of the accuracy of the claims made in all the presidential ads aired (which we purchased through VMS); turnout figures by DMA, aggregated from county level turnout data; and data from the Annenberg Survey that allow us to look at individual-level relationships with turnout and political knowledge over time, and at the psychological mediators of those relationships such as political efficacy. We coded more than 300 presidential ads and the accuracy of almost 1600 issue and trait claims.Beyond its implications for understanding of the impact of political advertising, our paper is also relevant to the theme of the conference in that the digital revolution has spawned a growth in “fact-checking” of ads, both through dedicated websites such as factcheck.org and through online media outlets such as the Washington Post. Our coding of the accuracy of the claims made in ads drew on these fact-checks where possible and our analysis of the relationship between the accuracy of ad claims and political knowledge over time will indicate whether misperceptions among the electorate have a tendency to persist in the face of enhanced fact-checking in the digital age.