The ‘Filter Bubble’ Explains Why Trump Won and You Didn’t See It Coming

Photo: Mark Wilson/Getty Images

Donald Trump’s victory is blindsiding, like stepping into a crosswalk and getting slammed into by a delivery guy cycling the wrong way down a one-way street. This is because, as media scholars understand it, we increasingly live in a “filter bubble”: The information we take in is so personalized that we’re blind to other perspectives. It simultaneously explains why Trumpism has flourished and why so many of us are insulated from it.

To George Washington University media-studies assistant professor Nikki Usher Layser, the filter-bubbledness of the Trump victory speaks to how, more than ever before, people now have the ability for “mass self-communication,” where you can share your perspectives, with friends you don’t see in person everyday, with the push of a button. Not just the people you see in real life, but that girl from high school you never really got along with but who agrees with your opinions today.

“We have always surrounded ourselves with people who agree with us [and] sought information we agree with, but there was at least a chance for serendipity,” she says, the chance that you’d discover something outside what you’d ordinarily choose to read yourself, even if you spend all day reading content. The big difference now, she told Science of Us, is the “autonomous decision-making” governing what stories you see. While newsrooms aren’t perfect, they can at least have the contextual awareness to offer contrasting narratives of what’s going in the news. “We can’t break out of patterns if we only consume information online, through our social feeds,” she says. It’s like explaining water to fish, this invisible, enveloping filter bubble.

The phrase got coined by MoveOn and Upworthy activist Eli Pariser, with his 2011 best seller, The Filter Bubble: What the Internet Is Hiding from You. His thesis was that while we might think of the internet as an impartial, universal library with Google serving as a superhuman Dewey decimal system, it’s remarkably, and perhaps pathologically, individualized. “Your filter bubble is this unique, personal universe of information created just for you by this array of personalizing filters,” he said in an interview with Amazon. “It’s invisible and it’s becoming more and more difficult to escape.” Back then, Pariser liked to say that Google used 57 signals to tailor its search results to you; today, the search giant says it’s over 200.

Given the events of last night — and the last year — this passage is strikingly prescient:

“Ultimately, democracy works only if we citizens are capable of thinking beyond our narrow self-interest. But to do so, we need a shared view of the world we cohabit. We need to come into contact with other people’s lives and needs and desires. The filter bubble pushes us in the opposite direction – it creates the impression that our narrow self-interest is all that exists. And while this is great for getting people to shop online, it’s not great for getting people to make better decisions together.

In a review of the book that same year, Slate head Jacob Weisberg argued that Pariser’s worries were of a piece with other web skeptics, a class of critics who worried that the web would turn “into everybody’s narcissistic ‘Daily Me’ feed,” though he thought that the web’s development in the previous 15 years didn’t attest to that. He quipped that Watson wasn’t going to beat Jill Abramson in news judgment any time soon, but here’s the thing: These algorithms aren’t optimizing for journalism, they’re optimizing for engagement.

Consider Facebook. The social network reaches 67 percent of American adults, Pew reports, and over 40 percent get news from the platform. It has “centralized online news consumption in an unprecedented way,” Jon Herrman wrote in the New York Times this August, and as such, it’s “hosting a huge portion of the political conversation in America.” It’s incredibly potent, and not just in the sense that manipulating people’s news feeds makes them more likely to vote.

In a 2015 study run by Facebook data scientists and published in Nature, researchers set out to test the filter-bubble hypothesis by looking at ten million de-identified Facebook users who self-reported their ideological affiliation over a six-month period. They found that users only clicked on 7 percent of “hard” content (politics, national news) in their feeds, as opposed to “soft” content like entertainment, sports, or travel. The researchers found that conservatives see about 5 percent less ideologically diverse content than their more moderate friends, with liberals at 8 percent. The Facebook algorithm, they concluded, makes it 1 percent less likely that people are exposed to cross-cutting content. More than anything, it’s the friends you have: “We show that the composition of our social networks is the most important factor limiting the mix of content encountered in social media.” While “news feed” is clever, sticky branding, it’s more “’my friend’s opinions’ feed.” Notably, the study with the largest data set on Facebook virality, out earlier this year, found that feelings of dominance predicted sharing, while arousal — getting angry or upset — predicted commenting.

As patterns of media consumption change, so do media structures — creating the space for extreme, identity-validating sites like Breitbart to flourish, as well as news and advocacy pages built specifically for Facebook, which Herrman, the Times media reporter, so precisely identified, and basically just exist on the platform, with names like “Occupy Democrats; The Angry Patriot; US Chronicle; Addicting Info; RightAlerts; Being Liberal; Opposing Views; Fed-Up Americans; American News; and hundreds more,” he writes. This also allows for the proliferation of disinformation sites run by Balkan teens. It’s “ideological media,” says Usher Layser, and it’s booming.

“Ideological media has grown to significant degrees, it can be automated, done by people who can put up decent websites and content that looks real, and people turn to ‘information’ that isn’t information,” she says. “What we see now is the birth of highly profitable small-scale ideological media that scales big on social because only one viral hit on one of these is enough to keep it afloat, and no original reporting is required. It’s genius. No infrastructure required to put these up.” Tiny sites like that shouldn’t be able to thrive, she says, since usually small sites are crowded out of the currents of virality. But today, there’s a “there’s a greater connection between the ideological media (even batshit sites) and elites,” like a Drudge or a Trump, she says. “Basically, the tie between right-wing crap sites and elites is stronger than ever before, and add Facebook + filter bubbles = viral splash.” The dark side of viral news isn’t cat listicles. It’s ideological scaling.

The ‘Filter Bubble’ Explains Why Trump Won