It’s a sunny Wednesday in Brno, and we’re starting the ECREA 2026 conference proper now. I’m in a session on the political uses of TikTok, which begins with Eduard-Claudiu Gross, on the situation in Romania. Romania annulled the first round of its presidential election in December 2024, due to a massive Russian election interference campaign, and TikTok played a role in this; the re-run in May 2025 took place in this complicated and polarised political and media environment.
This study, then, gathered some 50,000 TikTok videos from June 2024 to March 2025, from 1,000 authors, and from these examined the 40 most-engaged videos. These break down into pro-European and sovereigntist, nationalist videos, and there are only few shared hashtags between these; they also have distinct discursive architectures, with pro-European videos emphasising themes of democracy, competence, empathy, normalcy, partnership, and truth (an institutional ist register), and the sovereigntist videos emphasising themes of crisis, tradition, sovereignty, and corruption (a populist register).
The same platform affordances are turned to opposite ends by these videos. Populist videos sought engagements through user polls, for instance, while institutionalist videos encouraged users to report disinformation. Engagement was somewhat greater for the populist than the pro-European side; populist videos also employed popular and patriotic songs, often against the express wishes of the artists.
There are no isolated echo chambers here, but some notable polarisation and segregation between the two sides. TikTok does not just reflect political polarisation, but also actively produces it; this also suggests that platform architecture should be better addressed by public policy (and not just the content).
