Think back to last year: Where did you most often go to get election news?
Pew Research recently asked a bunch of Americans about this, and this past week wrote about the results. They found who you voted for had a huge impact on where you got your news.
For Donald Trump voters, 40 percent said Fox News was their "main source." Every other option was under 10 percent.
Among Hillary Clinton voters, 18 percent said CNN was their main source, the highest share for any news outlet on the Democratic side. MSNBC was next in line for Clinton voters at 9 percent.
Were Clinton voters going to Fox News at all? Not really.
Only 3 percent said they regularly went to that source. And it goes both ways – Trump voters didn't mention MSNBC enough to make the list.
NPR and the New York Times also made the Clinton voters list, and didn't appear for Trump voters – those two sources tend to have more liberal audiences.
This pattern carries over to online: 11 percent of Trump voters said they regularly got news from Breitbart or Drudge Report, which skew conservative.
And 24 percent of Clinton voters looked to Huffington Post frequently, while 10 percent named Buzzfeed. Both tend to lean liberal.
You're picking things you already agree with
So why does it matter? Well, some people argue looking at things that just repeat what you already believe isn't very helpful.
Selective exposure is one of the sciency names for this. It's the theory that people choose to take in some types of media, while avoiding others. Going a step further, it's often applied to beliefs – people will seek out and take in things that affirm what they already believe, and ignore things that challenge those beliefs. (Here are a couplestudies that back this up.)
There are other names (you might hear "confirmation bias" for example), but "filter bubble" seems to be the most common when talking about news and the internet.
The idea is that this selective exposure is exacerbated because websites and online services are relying more and more on tailoring user experience.
Even Facebook has an opinion on your political opinions, and uses it to serve up things it thinks you'll like.
This might be a bad thing ...
Filter bubbles limit what viewpoints we see, and instead of having our beliefs challenged or broadened, we get trapped, author Eli Pariser has argued.
There were a whole bunch of internet think pieces on this after the 2016 election, with the fake news narrative and questions about its effect on the results. These professors argue the echo chamber gets bigger, and fake news is more likely to spread.
WIRED said the filter bubbles are "destroying democracy" because we "miserably fail to penetrate other social bubbles."
"This is especially damaging because peer views and referrals are the strongest, most convincing form of marketing," the author wrote, later adding: "Until the election results, a little more than half of us didn’t realize that the other half of the country was frustrated enough to elect Trump."
Or maybe it's not filter bubbles' fault
Now here's the but.
Internet Policy Review published a study in March of 2016 that found "little empirical evidence that warrants any worries about filter bubbles." Part of the reason is because there are always holes – you'll probably still see conflicting content around, even if you've tailored everything.
That said, the authors note it's very early, and say it's worth more study as things progress.
And Scientific American argued filter bubbles and content personalization aren't really the problem at all – it's more that people are stubborn as heck when it comes to political beliefs, and will interpret things differently (or just ignore them) based on their principles.