TikTok and US Public Opinion

We provide some basic facts from survey data about TikTok usage and make sound estimates of its influence on US public opinion and voting intention. The data are from nationwide mTurk samples, weighted to reflect the population’s age, education, and religious upbringing. Role-players are omitted. These 2020-2021 surveys contain a wide range of important control variables. TikTok usage spanned the full spectrum from never-users (62%) to avid users, using TikTok multiple times per day (8%). TikTok users tend to hold stronger religious beliefs, to have been reared in upper working class or middle class homes, to be men, and to be Black. All those differences are mild tendencies rather than yawning gulfs. TikTok users tend also to use other social media (correlations), most especially SnapChat. These pre-existing differences demand multivariate analysis to isolate the unique influence of TickTok on public opinion. To this end, we use regression analyses controlling for an extensive panoply of influences which could be confounded with TikTok usage. The more people used TikTok the higher their evaluations of BOTH the Democrats and Republicans handling of the Covid epidemic (a key issue at the time). That holds after regression filters out the effect of the others influences, including a general propensity to use social media. Note that the measure of TikTok usage is general (not focused on news or politics). Nearly every other variable that influences public opinion on party-political issues, lifts evaluations of one party and depresses evaluations of the other. Moreover, TikTok users are more confident and trusting of government and major health-related institutions. For reasons not yet clear, TikTok usage warmed users feelings toward Joe Biden, but did not affect their feelings about Donald Trump one way or the other.