Closing the Turnout Gap: Can Planning Drive Young Americans to the Polls?

• In Event: Interventions and Innovations That Impact Turnout and Voting

Abstract

Voting rates among young Americans are consistently low - Holbein and Hillygus (2020) note that, based on data from the 2018 ANES, only 33% of young (18-29) potential voters turned out to vote, as compared to 66% of older (60+) potential voters. Moreover, this trend persists among young Americans who report intending to vote - according to longitudinal ANES data, the turnout gap between vote intention and validated voter turnout is on average around 17% higher for younger potential voters versus older (Holbein & Hillygus, 2020).
Even well-funded “get out the vote” (GOTV) campaigns have shown minimal effects on turnout among youth voters. One school of thought proposes that young potential voters lack sufficient motivation to turnout. More recently, Holbein and Hillygus (e.g., 2020) proposed an alternative explanation - perhaps young potential voters hold sufficient motivation to vote, but lack the ability to overcome obstacles and follow through on their voting intentions. This study used an online survey methodology to build on prior research by testing the effectiveness of two novel, theory-based GOTV interventions to increase turnout among young potential voters.
We recruited ~330 young Americans (ages 18-29) who were eligible to vote in the 2022 midterm elections through Dynata, a reputable survey panel provision company, in the 11 days preceding Election Day. Participation was restricted to only those intending to (e.g., didn’t already) and eligible to vote from one of five U.S. states with competitive 2022 Senate races - GA, NC, PA, WA, or WI. As voter eligibility might be particularly difficult for young potential voters to ascertain (e.g., in light of multiple part-time residences), study eligibility was assessed by screening potential participants based on the specific voter eligibility restrictions in the state in which they would vote.
All participants were randomly assigned between one of two GOTV interventions, or a filler-task control group. A “Why to Vote” (WTV) intervention aimed to increase turnout by strengthening existing sociopolitical beliefs and emphasizing the efficacy of voting as an avenue to achieve desired outcomes related to these beliefs. Five versions of the WTV were deployed (with random assignment within the condition) focusing on different issues salient among young voters (e.g., abortion access, immigration reform). This intervention was based in the work of Roseman et al. (2019), who describe a set of components that comprise the “narrative” of a strongly held belief. This intervention integrated voting within a narrative of action, aiming to build agreement to 24 target-belief-consistent statements culminating in a “behavioral” item highlighting voting as a means to achieve a belief consistent-outcome.
A “How to Vote” (HTV) intervention, based in the theory of Holbein and Hillygus (2020), tested whether low youth voter turnout results from failure to overcome obstacles to following through on vote intentions. This intervention walked participants through (1) practice navigating online state online election tools (e.g., checking one’s voter registration and polling place online); (2) making a plan to vote (e.g., whether one will vote in person or by mail); and (3) making back-up plans against obstacles often faced by young voters (e.g., being turned away due to lack of sufficient voter identification). Using Qualtrics logic, each participant was able to build their own plan for voting, choosing their intended vote modality (respective of relevant state deadlines) and then planning their vote (e.g., when they would vote, what ID they would use, etc.).
We collected participants’ full names, dates of birth, and addresses in order to manually validate voter turnout (against publicly available online state voting records) as a rigorous dependent variable in order to observe whether, as hypothesized, turnout in each of the two intervention groups was higher than in a filler-task control group. Validated voting is a state of the art method for assessing turnout, as it avoids the biases of self-reported turnout measures. Voting records were manually searched by three independent coders and matches were compared for consistency, resulting in a three-way IRR rate of 85%.
Results indicated that participants in the HTV intervention group were 2.16x more likely to turnout versus the control group, a significant difference, while participants in the WTV intervention group were not. While a large number of alternative explanations were considered via controlled, mediation, and moderation models, the success of the HTV intervention over the control remained robust. These findings suggest that future research should focus on testing and refining instructional voting interventions in diverse samples, and that policy changes to simplify or eliminate barriers to turnout (e.g., restrictive voter ID laws) may increase turnout among young Americans.

Authors

• Grace Evangeline Grady, Rutgers University

• Ira Roseman, Rutgers University, CCAS