Here’s something everyone knows: Social media is driving American politics into a ditch of partisanship. Political junkies log on and cocoon themselves in a bubble of friendly punditry, appealing fake news, and outrageous acts from the other side. Every retweet and every like is another moment of identity confirmation, another high five to our friends, another reminder that we’re right and they’re wrong.
The result is, well, this ugly mess — President Donald Trump, red and blue Americas, polls showing we fear and hate the other party more than ever before, conspiracy theories growing like weeds, a polity where agreement is impossible and everyone is angry. Damn you, Facebook! Curse you, Twitter! (Instagram, you’re cool.)
But what if this obvious analysis is wrong? What if social media isn’t driving rising polarization in American politics?
That’s the conclusion of a new paper by Levi Boxell, Matthew Gentzkow, and Jesse Shapiro. Their study, released recently through the National Bureau of Economic Research, tests the conventional wisdom about polarization on social media nine ways from Sunday and finds that it’s wrong, or at least badly incomplete.
Their approach is simple. Using data from the American National Election Survey, they compare the most web-savvy voters (the young, where 80 percent used social media in 2012) and the least web-savvy voters (the old, where fewer than 20 percent used social media in 2012) on nine different tests of political polarization. The measures cover everything from feelings about political parties to ideological consistency to straight-ticket voting, and the data shows how polarization changed among these groups between 1996 and 2012.
The results? On fully eight of the nine measures, “polarization increases more for the old than the young.” If Facebook is the problem, then how come the problem is worst among those who don’t use Facebook?
To be thorough, Boxell, Gentzkow, and Shapiro also construct panels based on internet access and find much the same thing — polarization is accelerating fastest among those using the internet the least:
Does this mean the internet isn’t making us more polarized? Not necessarily. The young are becoming more polarized, and it’s possible social media is part of the reason. But given that older Americans who don’t use social media are polarizing faster than younger Americans who do, it’s clear that this is about more than whom you follow on Twitter.
“Something has to explain the rising polarization of older Americans,” says Gentzkow, an economist at Stanford. “We don’t argue against the view that social media is important. It’s just not the whole picture.”
I asked Gentzkow what he thinks might be part of the fuller picture. “I have two main hypotheses,” he replied. “One is stuff that has nothing to do with media at all but is structural, like increasing income inequality. The second is non-digital media, and cable TV and talk radio in particular.”
The latter piece makes particular sense if you think about the fact that older Americans make up the base of both the cable and talk radio audiences. More than a third of talk radio listeners are over age 65, and half of Fox News’s audience is over age 68. As bad as getting your news from Facebook can be, it’s often far better than relying on Fox News or Rush Limbaugh.
The authors’ data only goes until 2012, so it can’t tell us much about the 2016 election. But even though Trump’s use of Twitter was remarkable, my guess is the main way it mattered was by setting the agenda for more traditional news outlets, particularly cable news and talk radio. Remember, it was seniors — only 6 percent of whom use Twitter — who pushed Trump to the White House. He won 53 percent of voters ages 65 and over, but only 37 percent of voters ages 29 and younger. Trump is the Twitter-using president, not the president chosen by Twitter’s users.
Social media is new, it is transformative, and it is certainly changing American politics. But it’s not the only force at play, nor even the main one. And while it’s hard for news junkies (myself included) to remember, most people’s media feeds tilt more toward baby pictures, wedding announcements, and funny videos than political punditry. Those of us who follow lots of politicians and politicized news sources are outliers, and we shouldn’t extrapolate too much from our weird experience.
Whatever is tearing our politics apart is deeper and more universal than the digital filter bubbles that get so much attention — and it seems to be most powerful among the people least likely to get their news from social media.