Partisan Memes: A Content and Engagement Analysis across 3 U.S. Election Cycles

• In Event: Popular Culture: Communicating and Shaping Political Attitudes

Abstract

Political memes play an important communicative role in American politics. In online spaces where attention is limited and political interest is low, partisan memes can be a key component of a person’s information diet (Denisova, 2019). Memes are known to be a means for people to contribute to or participate in ongoing political and social debates (Milner, 2013). Importantly, political memes are created and shared by users, as opposed to political discourse communicated through a media outlet or politician; therefore, a meme’s content and engagement metrics are reflective of how Americans think and/or feel about certain issues or people. Scholars have pointed to memes’ inability to meaningfully gauge public opinion (Denisova, 2019), yet they also note that memes contribute to polarization (Kristensen & Mortensen, 2021, Paz et al., 2021). Considerable research regarding political memes has emphasized their importance during specific contexts or case studies such as the COVID-19 pandemic (Kristensen & Mortensen, 2021) or as visual responses to events such as #MeToo (Andreasen, 2020) or #proudboys (DeCook, 2018). Thus far, work has produced mixed evidence about the effects of memes in relatively isolated, individual contexts.

The present study does not speak to the effects of memes, but instead to paint a bigger picture of what content comprises the political meme space. We take a broader look at memetic content by conducting a content analysis of a representative sample of political memes on Facebook shared by popular, hyper-partisan Facebook pages to answer the following research questions:
RQ1: What political issues, politicians, and strategies (e.g. personal attack, call to action, patriotic symbolism) are most prevalent in political memes?
RQ2: Is engagement greater for memes made of likeminded content (e.g., likeminded politicians, Dem/Rep-owned issues) or non-likeminded content (e.g., outgroup politicians/issues)?
RQ3: What types of political memes garner the most engagement (emotional reactions, comments, shares)?
RQ4: How do memes differ during a midterm election compared to presidential elections?

Using CrowdTangle, we retrieved 2,986 memes published to the top 5 most-followed Republican hyper-partisan Facebook pages and the top 5 most-followed Democrat hyper-partisan Facebook pages during the 2016 (n = 986), 2018 (n = 1,000), and 2020 (n = 1,000) U.S. election cycles. Three independent coders analyzed the politicians/parties/media/celebrities featured in each meme, the issue(s) mentioned in each meme, and a variety of strategies that could be used to generate engagement with the meme such as personal attack, emphasis on American identity, symbols of patriotism, or mention of a political outgroup. At this stage of the research, we have completed the content analysis and are working to analyze the data. To do so, we combine engagement data with the content analysis to observe not only types of memes created by Facebook pages but also how Facebook users respond to different types of memes.

Ultimately, this study aims to contribute to the growing body of research on political memes by presenting characterizations of political memes that are representative across Republican and Democrat hyper-partisan Facebook pages and three U.S election cycles. These data aim to present trends with regard to who is featured in political memes, how politics are discussed within memes, and how peoples' engagement with memes might have evolved over time. The results of this study will lay out what constitutes hyper-partisan political memes with the goal of providing a better foundation for scholars interested in studying why this type of content affects, or fails to affect, political attitudes and behaviors.

Authors

• Jessica Renee Collier, Mississippi State University

• Yujin Kim, The University of Texas at Austin

• Natalie Jomini Stroud, University of Texas, Austin