Maybe you call it a bubble. Maybe you call it a silo. Maybe you just call it an echo chamber. But whatever metaphorical, narrow and enclosed space you prefer, there’s a good chance you’ve been told that one of the great social problems of our time is Americans getting their political news from biased sources. Conservatives watch Fox News. Liberals watch MSNBC. The news tells us what we already believe and distorts reality around partisan talking points.
But here’s the thing: That’s not how it works. Turns out, news of our bloated, biased media diets has been greatly exaggerated. It could even be said that the “media bubble” narrative is … wait for it … fake news.
Last week, I wrote an article about how partisanship affects the way we interpret facts. Two people might see the same facts about the current impeachment investigation but interpret that news in wildly different ways. After that story ran, I got a lot of letters from folks who wanted to know how much of that effect was due to media bubbles. Sure, we interpret facts differently. But are we even getting the same facts?
Well, yeah, actually. Mostly, we are. That’s according to Brendan Nyhan, government professor at Dartmouth. “People have a notion from hearing about [information echo chambers] that most Americans are getting news and information from a very slanted media diet,” he told me. “Empirical evidence suggests that’s not true.”
Yes, seriously. Consider, for instance, the simple math of TV ratings. There are about 122 million Americans who told the Census Bureau that they voted in 2018. The vast majority of those voters don’t watch partisan cable news. FOX News and MSNBC pull in around 3 million viewers when their top hosts are on air. In contrast, around 5 million people tune in to each of the network nightly news shows. More Americans have a centrist media diet than a slanted one. And most Americans are basically fasting.
Even social media, which has provided new opportunities for echo chambers to form, doesn’t seem to be all that successful at politically isolating most of us. In 2018, Nyhan and his colleagues published a paper that found that, while Facebook really was a hive of fake news scum and villainy, the audience for those biased and often fabricated stories was relatively small. In a national sample of about 2,500 Americans, taken during the final weeks of the contentious 2016 presidential campaign, nearly 60 percent of all fake news visits came from the 10 percent of respondents with the most conservative media diets. Our national crisis is actually a niche issue.
Sure, we all do like to hear perspectives that verify what we already think we know. Who doesn’t enjoy being told that we are smarter than people we don’t like? But while lab experiments suggest that could cause us to silo ourselves, reality looks a little more nuanced. In a 2016 paper, researchers tracked a sample of 50,000 Americans who heavily consumed online news and found that their reading was overwhelmingly self-directed (rather than algorithm-selected) and mainstream centrist in orientation. And, while the use of social networks and search engines were associated with a larger partisan divide — they were also associated with people being more frequently exposed to more opinions they disagreed with.
That’s not to say that the media polarization that exists has no effect at all on our national politics. It might not be shaping the individual opinions of the vast majority of Americans, but the small percentage of highly polarized people are also the loudest, politically speaking, Nyhan said. “They consume a disproportionate amount of the news and the parties are most responsive to them because they’re the most politically engaged,” he said.
And that can present some problems of its own. Take, for instance, the influence that a vocal minority can have on the people closest to them. There’s some evidence that in-person conversations are actually more powerful polarizers than any TV station or website. In one study from 2018, the polarizing effects of talking to people who share your partisan political bent were more than twice as large as the effects of consuming politically biased media. Our coffee dates may be more effective conveyors of fake news than our Facebooks.
CORRECTION (Oct. 8, 2019, 5:15 p.m.): A previous version of this article misstated where Brendan Nyhan works. He is a professor at Dartmouth College, not the University of Michigan.
Maggie Koerth-Baker is a senior science writer for FiveThirtyEight. @maggiekb1