Social Pressure in the Digital Age

Social Pressure in the Digital Age

By David Searle, Research Manager

People are more likely to obey social norms when they feel their behavior is being monitored. Social pressure includes both reminding voters of the social norm of voting and making them feel accountable by informing them that whether or not they vote is public information.

Over the last decade, social pressure mail has become one of the most consistently successful strategies for increasing voter turnout. Our most recent GOTV meta-analysis, covering over a hundred tests of social pressure mail, finds that this tactic has had large effects on turnout and has been highly cost-efficient. Social pressure mail can come in different forms, be it hard, gentle, or positive. Hard social pressure directly compares an individual’s vote history to their neighbors’, and the gentle version only mentions a voter’s own vote history. Positive social pressure stays positive regardless of an individual’s history by thanking them for being a voter in a previous election and mentions that their vote history is public.

Despite these differences, all social pressure forms have one key component in common: personalization. Each voter receives an individualized mailer, such as a voter report card or a personally-addressed letter with their vote history or voter grade. This level of personalization differentiates social pressure mail from many other GOTV tactics and provides concrete evidence that voters’ vote history is public information and being monitored. 

Translating social pressure to paid digital ads

Most social pressure experiments have occurred via mail, but how can we deliver social pressure online? While paid digital ads can be targeted to groups of voters, they cannot be tailored to each target in the same way that social pressure mail can. We’ve seen across several tests that social pressure can be difficult to translate into the digital realm. In a pair of 2014 tests, Planned Parenthood Votes and Progress Texas found modest success incorporating one social pressure component into their digital ads by mentioning that individuals’ vote history is public and searchable. However, the ads’ blanket statement did not include the key aspect of personalization that targeted mail social pressure delivers.

Michigan League of Conservation Voters (LCV) conducted a separate test in 2014 that found modest effects of paid digital ads on turnout. Michigan LCV incorporated individuals’ vote histories into the ad creatives, grouping targets by their vote history in the two previous elections. The creatives mentioned whether an individual voted in 2012, 2010, neither, or both years. One drawback of this approach is that the ads used the words “you voted” rather than addressing the individuals by name, making the ad content less personalized.

In 2018, Oklahoma Observer Democracy Fund, in collaboration with Voter Participation Center, Rising Tide Interactive, Prof. Katherine Haenschen, Hal Malchow, and AI, conducted a new test of digital social pressure, encouraging click-throughs to a website where an individual could search their own as well as their families’, friends’, neighbors’, and coworkers’ vote histories. Once individuals clicked through to BadVoter.org, they were identified as a “Great voter,” “Good voter,” “Decent voter,” “Bad voter,” “Not yet a voter,” or “Not really a voter?” depending on when they last voted. The test targeted young voters and voters of color across three states (NC, OK, OH), and the ads were delivered entirely on Facebook. Overall, the ads did not appear to increase turnout. There was a small boost to voter turnout in Oklahoma, a somewhat quieter electoral environment where the budget per target was larger than in the other two states and in which BadVoter had been present longer and received more media coverage.

In short, unlike mail where social pressure has outperformed other tactics, so far digital social pressure tests have found effects in line with other digital tests. 

Digital social pressure: A tough nut to crack

Digital social pressure messages are somewhat limited in the extent to which they can fully incorporate the key aspect of social pressure -- personalization. Michigan LCV customized messages based on individuals’ voting records, but without the full degree of personalization of social pressure mail. And while the Oklahoma Observer Democracy Fund 2018 test provided individualized voter grades, these were viewable only after clicking through to Badvoter.org, which few voters did. These limitations to personalization in paid digital ads may explain why paid digital approaches to delivering social pressure have so far been unable to replicate the success of mail-based social pressure.