No Better Than Soup? Comparing Null Experimental Effects of Political Facebook Ads Across Persuasive and Instrumental Measures of Effectiveness

Abstract

Studies on digital advertising effects claim that the primary purposes of online ads are persuasive: They seek to change vote choice or voters’ attitudes toward candidates. But recent scholarship has noted that social media’s unique affordances encourage electoral campaigns to use them in specific ways, such as using Facebook’s ads for email list-building. We conceptualize such strategic campaign goals as instrumental purposes of advertisements. We develop novel measures to test these instrumental effects. In an online survey experiment using Facebook ads from the 2020 Biden and Trump campaigns, we test our theory with list-building, fundraising, and persuasion ads. In a factor analysis, we find that instrumental and persuasive effects are related, but distinct, aspects of candidate support. We also test the effects of these ads on persuasive and instrumental outcomes. Our control treatment was an ad for a can of Progresso soup. We found no main persuasive or instrumental effects of any advertisement type. These ads performed no better than soup.