Unlock Editor’s Digest Lock for Free
FT editor Roula Khalaf will select your favorite stories in this weekly newsletter.
In economics, moving from a concentrated market with a small number of dominant players to a highly competitive market with small players is usually considered a good thing. However, it is less clear that society will benefit if the products that these players are targeting with a variety of consumer sets are, as we know, alternative versions of reality.
Some content could not be loaded. Please check your internet connection or browser settings.
Due to all the attention given to misinformation, the narrow focus on objective falsehood deflects us from much more fundamental change. New democratic risks are less than what people believe in wrong. They have always. They are no longer wrong or otherwise do not believe in the same thing with each other.
If this sounds exaggerated, try a recent example. In January, after Elon Musk attracted new attention to the UK’s grooming gang scandal via social media platform X, nearly half of British reform voters said they had heard about recent news stories compared to one in ten workers or LIB DEM voters.
This is not a question of whether a particular information is true or false. It seems like if different people encounter the same information at all.
In the past, established media organizations have followed the same news agenda, primarily within the borders. However, older gatekeepers and norms are increasingly bypassed in an increasingly unbounded, divided information environment.
This causes the continuous divergence of online public platforms, including social media, to be forked into an obviously right-leaning space with plenty of different agendas. As a dual citizen of X and Bluesky, there is a clear difference in the topics found on the two platforms.
Another weakness of misinformation discourse is: This means that it is a unique problem of one “side”. Research shows that on average, American conservatives are more likely to believe false statements about climate change, while liberals are more likely to believe false statements about nuclear power. Other US studies have found that people who attended university are not judges with better news truth than those who only have high school education.
I will not emphasize this to criticize a particular group. It’s exactly the opposite. I would like to do so, most people are left, right, uneducated – do not simply interrogate every claim they encounter.
Humans are efficient maximis players and are looking for shortcuts on every occasion. The truth is that the majority of us never invest the time to fact-check or evaluate all the information we consume. If it appears to be plausible and comes from sources that we are not actively distrustful of, then that is enough.
Combining this heuristic with the explosion of start-up information providers that tend to work in pockets of unprecedented fragmented media landscapes, we get interesting results.
Evidence regarding echo chambers and filter bubbles has been mixed up to date, but most studies predate a very recent shift to recent real-time recommendation algorithms.
Analyzing data from the UK election survey, we found that those who received the news from Tiktok were more likely to become supporters of reform between 2021 and 2024, even after controlling for age, gender and education levels.
In particular, this pattern is much stronger among men than women, and is consistent with the idea that different groups currently live in political environments online, even within the same platform.
Some content could not be loaded. Please check your internet connection or browser settings.
Similarly, in Germany, the strong Tiktok presence of both AFD and Die Linke is believed to have boosted their support among young men and women, respectively, contributing to the harsh gender gap between the voting patterns of young Germans in last month’s election.
The link between this fragmentation and democratic dysfunction appears everywhere you look. The American polarized media ecosystem is coincidental by political divisions that are far broader than countries with a more cohesive media landscape. Longitudinal studies have found that divided media produce polarized politics.
And beyond age range, sources of information differ most significantly from previous generations (in terms of young men and young women) and show the broadest ideological disparities on many scales.
Discourses of misinformation will undoubtedly ring out, but as teens glued to screens, they miss the broader context.
John.burn-murdoch@ft.com, @jburnmurdoch
Data Sources and Methodology
An analysis of the association between the use of Tiktok in the UK’s general elections and support for reform was conducted using waves 21 and 28 of British election surveys, which have been following the same British adults over several years, measuring how their political views and other attitudes change over time.
We used logistic regression models to predict the likelihood that someone would vote for reform in the 2024 election. This is subject to the political left and right scale (measured three years ago), the use of various social media platforms for news, and the use of different social media platforms at gender, age and education levels. Interaction terms were included, assuming that due to the interaction between social media platforms and sex, different people could see different content on each platform.
You can find the R code that replicates the analysis here.
