Twitter is not real life
Voters who get their news from Twitter are markedly different in our polls than the median voter
Whether you’re getting your information from TikTok, Instagram, Reddit, Twitter, cable news, or elsewhere, platforms are shaping your information diet in ways you may not even notice. Content is inseparable from the vehicle within which it arrives.
At The Argument, we ask respondents which platforms they regularly get news from every month,1 and our results revealed striking differences in opinion among the audiences of different news platforms. As an example, even though ICE’s net favorability rating is at -26 percentage points with all voters, it’s almost break-even with people who get their news from Twitter. Compare that to other social media platforms like Reddit and TikTok, where over 70% of voters viewed the agency unfavorably.
A large part of this divergence is down to selection, because people increasingly gravitate toward news sources that fit their worldview. (As one example, you won’t find too many liberals watching Fox News!) But in some cases, like with Twitter under Elon Musk, I think we have good reason to suspect that the platform is making people more right-wing.
For example, a study led by researchers at Northeastern University suggested that social media algorithms can amplify certain sentiments in ways that meaningfully influence respondent preference. And the numbers we have do lend some support to that theory; for instance, consider that Trump has lost the least support on Twitter.2
In fact, in our January survey, there was only one group that Trump remained popular with: Twitter users. With every other cohort, his numbers are charitably described as catastrophic.
That probably isn’t a coincidence. Musk’s interference with the algorithm has been well-documented, and the algorithm’s pro-Republican bias is evident to anyone who spends even a couple of minutes on the platform.
If you’re largely getting your news from Twitter, you might not even know that Trump is unpopular, because you wouldn’t even see a lot of the backlash. Instead, you might be under the mistaken impression that he has retained much of his support, and that his presidency is still treading water with the public instead of being near-radioactive.
Given how many politicians, activists, and staffers spend hours each day on the platform formerly known as Twitter, it’s fair to wonder whether they’re fully grasping where the country is at right now on a host of issues. Because if you ever wonder why the “vibes” feel like they do, the answer probably has a lot to do with your media diet.
With that said, algorithmic bias isn’t everything. Trump is still losing support among voters on Twitter. And among voters who get their news on TikTok, another platform that has faced its own accusations of right-wing content amplification, Trump has lost the most support, with approval ratings sitting at a staggering 25 points below his vote share in the last presidential election.
Where Americans get their news
Until recently, the tracking data needed to analyze news audiences was extremely hard to find in the public domain. But our survey project has tracked this since its inception, simply by asking respondents where they get their news from. With over 9,000 individual survey responses in our dataset, we can track and correlate political and social attitudes across media platforms in great detail — which has allowed us to build a surprisingly rare dataset and get hard, data-backed answers to our questions about news consumption in today’s world.
Our data suggests that there isn’t one single, shared platform that a majority of voters use for their news. Local television and Facebook have the largest news audiences in our survey tracking data, but the share of voters regularly using them for news still comes in at under 50%. And even platforms that feel ubiquitous in the discourse — like TikTok — are used by a relatively small share of consumers.
Unsurprisingly, when you look under the hood, there are vast compositional differences in platform audiences. For instance, the news audiences on social media platforms like TikTok, Reddit, Twitter, and Instagram all skew disproportionately younger, with Millennial and Gen Z voters comprising an outright majority of each. Meanwhile, the audiences for legacy media are much older; in fact, young voters made up less than 10% of cable television’s news audience, while voters over 65 accounted for nearly 40%.
It might not surprise you, then, to learn that there are massive differences in the political leanings of the audiences of different news platforms. The twist is that they don’t split as cleanly as “legacy media versus social media,” no matter what Elon Musk would have you believe.
For instance, Twitter is a pretty young platform; in fact, it’s the fourth-youngest one we tested. It’s also the most conservative platform among voters; 55% of voters who regularly use Twitter to get news backed Donald Trump in 2024, while just 35% backed Kamala Harris.3
Meanwhile, the people who watch broadcast television (like ABC or CBS) are a very boomer-heavy group, but they still come in as one of the most Democratic cohorts in our entire survey4 — 51% supported Kamala Harris in 2024, and just 39% voted for Donald Trump.5
Our findings also add some color to a larger debate: the firestorm over where conservatives and liberals get their news. It’s clear that print journalism leans disproportionately liberal (a finding substantiated by Pew Research Center’s polling).
The reasons for this are actually quite simple: Liberals read more and go to college at higher rates than conservatives do. Because of this, newspapers and news websites draw a significantly higher percentage of liberals than conservatives.
The danger of all of this sorting and slicing of people into separate news algorithms isn’t just that people aren’t getting the whole truth — it’s that they think they’re getting the whole truth. When someone on Twitter feels a vibe shift benefiting Republicans and someone on Instagram feels the opposite, neither of them are in the position to determine what’s actually happening in the country at large.
Here’s how we do it: In each of our surveys, we presented users with a list of news platforms and asked them to select all the ones that they regularly consumed political content on.
When comparing his current approval numbers to his 2024 margins.
For those of you who haven’t blocked 2020 from your memories, “Twitter is not real life” remains an extremely good maxim to operate under, but now for the opposite reason!
If you’re wondering why Bari Weiss’ CBS tenure has been received so poorly, this is part of the reason. The primary audience for CBS News is actually older liberals, and they’re not taking very kindly to Weiss ham-handedly overhauling one of the oldest institutions in American news media.
Valuable work; thank you. It should remind Argument readers that we are an analytical elite.
Surely we could compare republicans on reddit to republicans on twitter? Is the effect still there?