![[Adobe Stock]](https://sp-ao.shortpixel.ai/client/to_webp,q_glossy,ret_img,w_750,h_375/https://www.psypost.org/wp-content/uploads/2025/01/tiktok-750x375.jpg)
A recent study evaluating political TikTok videos found that traditional, policy-focused campaign ads improved voter favorability toward Vice President Kamala Harris, while trendy, short-form clips had a more limited impact. The research, published in Political Research Quarterly, highlights how short-form video platforms influence political attitudes but struggle to overcome entrenched biases. These findings suggest that while social media can mobilize certain voters, it may not easily persuade those outside a candidate’s existing base.
In the months leading up to the 2024 presidential election, political campaigns heavily invested in digital spaces to reach younger audiences. The Harris campaign notably focused on TikTok, a platform where users share short-form videos that generally run around 60 to 90 seconds. Social media allows candidates to bypass traditional news outlets and speak directly to voters. However, political scientists are still working to understand exactly how these short, algorithmic video feeds affect voter opinions.
Past research has established that voters often evaluate Black women politicians differently than other candidates. Black women in public office frequently navigate overlapping layers of racial and gender biases. Some voters hold negative views rooted in racial resentment or modern sexism. These biases can shape how people interpret a candidate’s tone, sometimes causing them to view a Black woman as aggressive when she uses standard political attacks.
Other psychological and behavioral concepts also play a role in how voters consume political media. Negative partisanship occurs when a voter’s behavior is driven more by a strong dislike of the opposing political party than by an affection for their own party. Attack advertisements frequently rely on this dynamic by tearing down an opponent rather than offering positive solutions. Outrage discourse involves speech that is inflammatory or misleading, often aimed at stoking fear or anger in its audience.
Campaigns also sometimes use implicit or explicit racial language to mobilize specific voting blocs. While this can activate support among targeted groups, political scientists debate whether explicit appeals risk alienating moderate voters. The concept of intersectionality helps explain how overlapping identities, such as race and gender, create unique political environments for candidates.
To investigate how different types of short-form videos perform in this environment, political scientists Sydney Carr-Glenn of the College of the Holy Cross, Chaya Crowder of Loyola Marymount University, and Christine Slaughter of Boston University designed a targeted experiment. They aimed to test whether exposure to specific messaging frames on TikTok could alter a voter’s opinion of Harris. The researchers focused on the abbreviated timeline of the 2024 race to capture immediate digital reactions.
The researchers recruited approximately 1,200 participants through an online survey platform during the early voting period in November 2024. The sample was matched to national demographics, though the researchers included extra Black participants to ensure they could properly analyze responses within that demographic group. Each participant was randomly assigned to watch one of five short videos. They viewed the content outside of the actual TikTok application, simulating the experience of having a video texted to them by a friend.
One group served as a control and watched a 13-second, nonpolitical video of teenagers dancing to an Eminem song. The other participants viewed actual TikToks posted by the official Harris campaign account. These four political videos represented different messaging strategies.
The first campaign video focused on policy, running for over a minute. It featured Harris discussing freedom and her platform over a popular Beyoncé song, accompanied by a montage of American citizens. The second video relied on an identity appeal. This 11-second clip mixed audio of Michelle Obama discussing “Black jobs” over a montage of Harris’s career.
The third video featured a campaign rally where Harris challenged Donald Trump to a face-to-face debate. This 30-second clip represented a form of political incivility or outrage discourse. The final video highlighted negative partisanship by attacking J.D. Vance’s stance on a national abortion ban. It featured headlines about abortion restrictions layered over a popular pop song.
After viewing their assigned video, participants answered questions about their favorability toward Harris, their likelihood to share the video, and their voting intentions. The researchers utilized statistical models to isolate the effect of each video from other demographic variables. They found that the traditional, policy-oriented video was the only piece of content that meaningfully increased overall favorability toward Harris compared to the control group. The other three videos did not produce a similar boost in average favorability across the entire participant pool.
While the trendy videos did not elevate overall favorability, all four political videos increased the likelihood that participants would share the content with their own social media followers. This dynamic indicates that provocative content can still achieve viral spread even if it does not improve a politician’s overall image. The policy and debate videos also made participants more likely to believe that social media is an effective tool for generating enthusiasm about an election. None of the videos significantly shifted overall voting intentions for the entire participant group, which the researchers said provides limited evidence that the videos affected votes for Harris.
The researchers noted several nuances when breaking down the data by demographic groups. Black participants reported higher favorability toward Harris after watching the videos focused on identity and abortion rights compared to non-Black participants. Democratic voters naturally showed high levels of support across all categories. Female participants expressed greater favorability toward Harris than male participants after viewing the policy and debate videos.
Racial resentment and modern sexism are academic scales used to measure subtle or institutional prejudices rather than overt bigotry. Participants who scored high on these standard psychological measures showed lower favorability toward Harris after watching the videos. This was particularly evident when they viewed the debate clip, which featured direct confrontation. The authors noted that short-form social media content does not erase preexisting prejudicial attitudes, as these biases continue to act as strong deterrents to positive evaluations.
Age and social media habits also influenced the results in unexpected ways. Older voters sometimes displayed more favorable attitudes than younger participants after viewing specific clips. Participants who reported being frequent TikTok users actually exhibited lower favorability toward Harris in some conditions. The researchers suggested this may indicate that TikTok-using audiences are less engaged by traditional campaign or debate messaging.
The study involves specific design choices that provide context for the results. Participants watched a single isolated video during the survey rather than scrolling through an algorithmic feed. In a natural setting, TikTok users encounter multiple related videos tailored to their interests in a single session. Viewing a single clip may not fully capture the cumulative effect of sustained social media engagement.
Future research could explore how voters respond to seeing a volume of political content natively within an application’s interface. Political scientists could also investigate how user-generated internet memes compare to official campaign content. As campaigns increasingly rely on digital outreach, understanding the exact mechanisms of online persuasion will remain an important area of study.
The study, “Social Media and Candidate Favorability: TikTok and Support for Harris During the 2024 Presidential Election,” was authored by Sydney Carr-Glenn, Chaya Crowder, and Christine Slaughter.