Get @ the vote : using Facebook and email to increase voter turnout

Abstract

This dissertation investigates the effects of the two most commonly used forms of digital media – email and Facebook – as mechanisms of voter mobilization. The widespread adoption of digital media in America makes it an idea conduit for voter mobilization, but to date there is minimal research that attempts to use Facebook to increase turnout, and few studies that use email to successfully boost participation. These studies leverage unique affordances of both mediums to increase voter turnout: Facebook increases the visibility of users' behaviors on the platform, and email messages and Facebook advertisements are inexpensive and easy to send to mass audiences. The results engage with existing literature on the power of social norms and how they can be used to drive behavior changes. To explore this topic, I conducted four field experiments designed to leverage Facebook and email messages to increase voter participation during the 2014 general election in Texas. These experiments adapt social pressure messaging, which emphasizes the public nature of voting records and attempts to increase the visibility of voting behaviors, for digital communication platforms. In implementing two of the studies, I developed a new method of conducting field experiments on Facebook randomized at the level of the individual that are implemented with the help of confederates. Results were analyzed using logistic regression. The most effective method – directly shaming people for failing to vote in the ongoing election – produced a statistically significant increase in turnout of 22%, which is much higher than what has been obtained through traditional analog methods. Directly praising friends for past participation was also able to raise turnout by 9%. Additionally, seeing others be praised for voting was able to increase turnout amongst new and infrequent voters. The second two studies build on past research by combining the email messages with Facebook advertisements and sending multiple waves of reminders. They show that multiple rounds of email and social pressure messaging can generate small increases in turnout. The findings demonstrate that Facebook and email can be used to increase voter turnout, and that the effects of mobilization within peer-to-peer networks are much larger than those obtained from unsolicited mass-email messages. This work contributes to existing theory by demonstrating that voting behavior circulates and can be induced through networks. Furthermore, the heightened visibility of user behaviors within online social networks was able to amplify the effects of the treatments beyond what has been produced in an offline context. Overall, the results show that digital media can be used to increase voter turnout, and offer reasons to be optimistic about the future of democracy in our increasingly digital society.