Theme Panel: Politics in 60 Seconds: Short-Form Video, TikTok, and Political Communication

August 28, 2025 APSA Annual Meeting, Theme Panels Comments Off on Theme Panel: Politics in 60 Seconds: Short-Form Video, TikTok, and Political Communication

In-Person Full Paper Panel

Participants:

  • (Chair) Eunji Kim, Columbia University
  • (Discussant) Margaret E Roberts, University of California, San Diego

Session Description:
Short-form video platforms have transformed the political communication landscape. TikTok surged in popularity in 2018, particularly among younger audiences, prompting copycat products like Instagram Reels and YouTube Shorts, both released in 2020. The affordances of these platforms—an immersive, vertical viewing format, a low barrier to content creation, and an attention-driven, hyper-personalized feed—have generated tremendous demand for this new medium, giving rise to a newly emergent class of content creators (and therefore political actors). The rise of TikTok has reshaped political communication on social media, from elite discourse and information flows to grassroots organizing and traditional campaigning.

Despite their growing cultural and political significance, short-form video platforms remain understudied within political science research. Part of the challenge lies in the scarcity of publicly available, granular data from TikTok, which dominates the short-form video space. The lack of transparency around its recommendation systems and the anonymized nature of user behavior have hampered systematic investigation. Moreover, TikTok’s rapid adoption worldwide raises pressing questions around algorithmic bias, misinformation, echo chambers, and foreign influence.

Indeed at the time of this writing, TikTok faces a potential ban in the U.S., stemming from concerns about data privacy and potential Chinese state influence on the platform’s inner workings. The panel’s participants have contingency plans in place to ensure relevant findings come out of data collected irrespective of any disruption to TikTok service (with the exception of the activation study) and in some cases to leverage any potential ban in quasi-experimental designs.

This panel brings together scholars who, despite these data challenges, are pioneering innovative methodologies and drawing on novel datasets to deepen our understanding of short-form video’s role in contemporary politics. They explore a range of pivotal themes: How is political information produced and distributed on short form video platforms? How can we best use video, audio, and text together to better undesrtand short form video platforms? Which types of political messages or narratives tend to go viral, and what factors drive their amplification? How much hate speech is observable on the platform, and how well does the platform moderate it? What is the causal effect of starting to use TikTok? And does TikTok’s algorithm favor messaging that aligns with China’s geopolitical interests?

Donnay et al. provide a comprehensive analysis of hate speech on TikTok, combining descriptive statistics with experimental methods. They map the prevalence of hate speech across regions, time, and topic, while evaluating TikTok’s responsiveness in terms of content moderation.

Guinaudeau et al. examine how political information is created and disseminated on TikTok. Drawing on a large, representative sample of TikTok videos published in the United States with a focus on the months preceding and following the U.S. 2024 election, their initial findings indicate that political content spreads faster than entertainment, and that there’s notable concentration of influence among a small group of creators, contradicting the platform’s democratic image. The study highlights the risk that politically extreme discourse could dominate the platform, and calls for further scrutiny of TikTok’s algorithms and societal impact.

Guess et al introduce a new method for analyzing TikTok style video content using audio, video, and text. They apply this strategy to an exploratory analysis of the political content on TikTok posted around the 2024 U.S. presidential election. The topic model style strategy will be applicable to other forms of diverse short-form video content.

Barnehl et al. investigate whether TikTok artificially amplifies pro-China content, addressing concerns about potential state media influence on the platform. The paper proposes a method of assessment that centers on using an LLM-classifier to identify content aligned with Chinese state media objectives and examines algorithmic distribution via longitudinal analyses of views, likes, comments, and moderation. Preliminary initial analyses have yet to yield clear evidence of preferential treatment or asymmetric moderation of pro-China content.

Rutherford et al. conduct an “activation study,” to investigate the political and social consequences of starting or increasing TikTok use. The research provides a clean, causally identified method of assessing the impact of TikTok use on political attitudes, knowledge, and social wellbeing, providing a more holistic understanding of the platform’s influence compared to prior deactivation studies. It will also be one of – if not the – first social media “activation” experiment, complementing a series of prior “deactivation” experiments on other platforms, but allowing for an even more direct measure of the quantity of interest: the impact of platform usage.