Hollingshead, W., Gruzd, A., & Mai, P. (2026). Same platform, different stories: TikTok and the battle over immigration narratives. Media and Communication, 14. https://doi.org/10.17645/mac.11409
Summary
This paper offers one of the first systematic content analyses of how immigration is framed on TikTok, focusing on the Canadian context and the platform’s distinctive affordances of mimesis and interactivity. Through a mixed-methods content analysis of 344 high-reach English-language videos, the authors find that pro-immigration content substantially outweighs anti-immigration content — a reversal of the hostility documented in prior X/Twitter research. They argue that TikTok’s affordances act as a “double-edged resource,” simultaneously enabling digital inclusion and amplifying xenophobia. Crucially, users across ideological divides deploy the platform’s features in largely symmetrical ways, suggesting a shared memetic vernacular that transcends stance, while the primary axis of contestation in Canada is economic rather than identity-based.
Key Contributions
- One of the first content analyses of immigration framing on TikTok, extending a literature dominated by X/Twitter and Facebook.
- A replicable sampling methodology for studying geographically-bounded TikTok content in jurisdictions without Research API access (via Zeeschuimer scraping and time-varied searches).
- Operationalizes the Memetic Interactivity Codebook (MIC) for rule-based analysis of TikTok metadata.
- Bridges immigration framing scholarship and platform affordance theory, showing the same features support both inclusion and exclusion.
- Empirical insight into Canadian digital discourse during a period of shifting public sentiment and affordability anxiety.
Methods
Mixed-methods content analysis of TikTok videos collected in July 2025 using the Zeeschuimer Firefox extension (chosen because TikTok’s Research API is unavailable to Canadian institutions). Value-neutral English keywords and hashtags (e.g., “Canada immigration”, canadaimmigration) were searched twice daily over seven days to mitigate algorithmic personalization. Purposive filtering retained 2025 videos with over 100,000 plays, non-US location, and one video per user, yielding 344 videos after intercoder reliability checks. Immigration stance was manually coded against a 13-frame typology (Krippendorff’s alpha > 0.8), and affordances were coded with the MIC. Analysis used chi-square and Fisher-Freeman-Halton exact tests with post-hoc odds ratios and Bonferroni adjustment.
Findings
- Stance distribution: 41% pro-immigration, 13% anti-immigration, 8% neutral, 38% unrelated.
- Pro-immigration content was dominated by an “other” frame (62%) — consultant advice, peer support, citizenship celebrations — with economic benefits second (14%).
- Anti-immigration content was led by economic costs (47%), then cultural threats (18%) and nationalism (16%); identity frames were roughly one-third.
- Only one affordance showed a significant association with stance: anti-immigration videos used non-original audio far less (16%) than pro-immigration (44%) and unrelated (57%) videos.
- No Duet or Stitch use; community Toks appeared similarly across stances (~21%), often signaling Canadian patriotism.
- Humor served as a shared discursive register, enabling both solidarity narratives and racist anti-South-Asian tropes.
- Generative AI appeared in a minority of anti-immigration videos to construct dystopian “speculative futures” of an Indian “takeover” of Canada.
Connections
The paper’s observation that generative AI is emerging as a tool for racist “speculative worldbuilding” connects it to broader work on generative-AI-fueled disinformation such as Hackenburg2025-dj and DeVerna2025-dl. Its platform-vernacular and affordance-centric framing of ideological discourse resonates with TikTok and memetic-culture scholarship including Baym2026-tr and Copland2025-em. The focus on framing and cross-ideological narrative contestation also relates to information-disorder research like Hameleers2026-mc and Cazzamatta2026-lo.
Podcast
A research-radio episode discusses this paper: 🎧 MP3 · Spotify · Apple Podcasts