In the weeks following Trump’s election, as elites tried to make sense of what went so horribly wrong (as they saw it), many an accusing eye turned its gaze toward Facebook. There were now-familiar criticisms of the social network site — notably that it places us into echo-chambers, feeding us stories that reinforce our political views. But a new scourge also emerged: Facebook, according to reports from BuzzFeed and other sources, had become a breeding ground for misinformation and purposeful disinformation.
The rise of fake news on Facebook is worrisome, to be sure. And to their credit, Facebook executives took the issue seriously and introduced a system through which articles flagged by users will be vetted by fact-checkers — a welcome, if far from perfect, development. But there is an equally worrisome risk that as we jump to fix the fake news problem, we may end up destroying our best hope to break through our echo chambers.
That hope is, contrary to much of what you’ve read, Facebook itself. Yes, contrary to common belief, Facebook and other social media sites may not be the fomenters of hyper-polarization often thought; they could, in fact, be a powerful force that pushes us out of our ideological silos.
Some of the evidence that social media serves more as a bridge than a divider comes from studies that, as it happens, set out to make the case that social media split us into political bubbles. But those bubbles, it turns out, are far from airtight. Something from 15 percent to 25 percent of the people we connect to on social media, on average, come from across the political divide. Those numbers might seem small, but they are at least comparable to the level of diversity we come across in real life. (According to the 2008 American National Election Survey, 20 percent of the partisans Democrats spoke to about politics were Republican and 22 percent of Republicans’ political interlocutors were Democratic.)
We’re more likely to engage with “second-tier” friends online
And there are reasons to suspect that our online networks might be more productively diverse than our real-life circles.
When it comes to our daily lives, homophily — the tendency to surround ourselves with like-minded people — is a strong force. But homophily is weaker among our more distant connections, those people whom we see rarely. And this is where Facebook comes in: The social network has proven especially good at facilitating connections with our second- and third-tier friends. People we rarely, if ever, schedule social events with offline nonetheless hover perceptively in the margins of our online social networks.
And social media may not only increase the odds we connect to someone on the other side of political spectrum — extending a “friend” request, accepting one — but also that we share our views with them. Etiquette and our instinct to avoid conflict keep us from bringing up politics at the dinner table or around the water cooler. For better or worse, those rules of decorum get tossed out online.
Research confirms that that the diversity of our online networks, combined with our relative disinhibition online, indeed exposes us to views we wouldn’t otherwise come across — and may even lead us to moderate some of our more extreme views.
When Facebook researchers examined the diversity of their users’ networks in 2015, they found — echoing the studies mentioned above — that on average our bubbles are quite porous: 18 percent of self-described liberals’ connections were conservative and 20 percent of self-described conservatives’ friends were liberal. More surprisingly, those connections delivered to the social network’s users a diverse set of political news: 24 percent of the news items liberals saw were conservative-leaning and 38 percent of the news conservatives saw was liberal-leaning.
And we might not just be seeing those diverse stories on our feed; research by the computational statistician Seth Flaxman and colleagues suggests we may be clicking on them as well. Flaxman tracked the web-surfing behavior of 50,000 internet users and compared the diversity of news they read — taking note of whether these news consumers arrived at sites via a news aggregator, using a search tool, directly accessing a news site, or through social media. News aggregators and search sites had the best bubble-bursting record: They led internet users to the widest array of stories and to the most stories that challenged users’ political views. Social media trailed aggregators and search on this measure; still, people who relied on social media for news got more diverse and challenging information than did the people who navigated directly to their favorite news sites.
Exposed to different points of view, people changed their news-consumption habits
That exposure to diverse news might even turn out to be a moderating force. Pablo Barbera looked at the political leanings and networks of Twitter users from 2013 to 2014. He not only found — again, echoing other studies — that users had remarkable diversity in their friends: For half of all users, at least 30 percent of friends did not share their ideology. He also found that users whose online friend circles included at least 20 percent who disagreed with them became more centrist over time, as measured by the news sources they followed.
To be sure, neither Facebook nor Twitter represents the ideal of a “public sphere.” As anyone who has de-friended a particularly offensive acquaintance, or simply taken a Facebook mental health break, will attest, Facebook exchanges can often feel more like cage-fighting than Socratic dialogue. Not all dialogue is productive dialogue, researchers tell us: Without a foundation of respect, for example, discussion with a political opponent may make us dig our heels in — and convince us that our political foes are as ignorant and evil as we always knew them to be.
That, indeed, may be the true issue with social media — not that it seals us off from opposing views, but that it exposes us to the worst in political dialogue: to heated ad-hominem attacks, cherry-picked evidence and, as we’ve seen in recent weeks, false information.
Those polarizing forces need to be tempered, but as we confront that problem we should be careful not to accidentally dismantle the bridge social media offers. Consider the recent offensive against fake news. As others have noted, the primary risk of “fake story” fixes and other attempts to regulate Facebook is that they will be perceived as biased (or, worse, be biased) toward one side of the political divide. Such a perception could have the backfiring effect of making users on the extremes flee the network — perhaps to read alternative media, effectively sealing themselves off from opposing viewpoints. Then we will, indeed, have airtight echo chambers on social media.
One way to avoid the appearance of bias is to make any fake-news detector (or other de-polarizing tool) optional — preferably “opt-in” rather than “opt-out,” so users feel truly empowered in creating their online environment. To help encourage the adoption of those tools, Facebook could provide nudges by posting notices when one’s friends turn on a fake-news-evaluation app.
This bruising election season has left some Americans understandably doubtful that social media can be force for democratic good, but we can’t give up hope. That’s not just because social media is likely to increasingly be our gateway to political information, but because, contrary to conventional wisdom, it still holds considerable potential to bridge our divides.
Julia Kamin is a PhD candidate in the department of political science at the University of Michigan.
The Big Idea is Vox’s home for smart, often scholarly excursions into the most important issues and ideas in politics, science, and culture — typically written by outside contributors. If you have an idea for a piece, pitch us at firstname.lastname@example.org.