The Real Problem with Facebook Virality

Why objective journalism loses against partisan content

Jul 19

How much did Facebook's algorithms contribute to the political polarization which resulted in Donald Trump's narrow 2016 victory?

One of the more popular explanations for the polarization is that Facebook is an echo chamber or a bubble for like-minded people. Liberals who had posted selfies of themselves voting for Hillary Clinton were stunned by the Trump victory, and likely had no Facebook friends who (openly) voted for Trump. The New York Times' Kevin Roose wrote an influential column during the 2020 election posing the question of whether the "right-wing Facebook bubble" would lead to a second surprise Trump win. Facebook CEO Mark Zuckerberg denied that Facebook was an "echo chamber" in an interview with Axios' Mike Allen, saying that the characterization was "just wrong." However, he admitted, ""It's true that partisan content often has kind of a higher percent of people...engaging with it, commenting on it, liking it." Roose, in a follow-up column last week, wrote, "Facebook is not a giant right-wing echo chamber. But it does contain a giant right-wing echo chamber."

Recent research shows Facebook's problem isn't so much that it is an echo chamber, but that posts mocking an out-group go viral more frequently than posts praising an in-group. In a recent study, psychology researchers Steve Rathje, Sander van der Linden, and Jay Van Bavel, of Cambridge and New York University analyzed 2.7 million Facebook and Twitter posts from news media and congressional members in 2020. They found that the effect of posts using out-group language was "4.8 times as big as that of negative affect language and 6.7 times as big as that of moral-emotional language." In other words, posts mocking a politician of an opposing party were more likely to be shared than simply negative posts or ones that appealed to emotions or morals.

The researchers specifically dismissed the echo chamber hypothesis. They argued that even when users saw posts that periodically challenged their views, they were unlikely to change their views because the opposing view was also partisan. In other words, if a Trump supporter saw a post mocking Trump rise to the top of his feed, it would not be likely to change his opinion. "Even if people are exposed to more cross-partisan content than expected, our findings suggest that opposing views on social media may be excessively negative about one’s own side," they wrote.

Journalism is one of the biggest losers in the propensity for out-group content to go viral on Facebook. Since research shows that objective news tends to go less viral, news organizations stand at a disadvantage against partisan outlets without the rules of objectivity. According to the Pew Research Center, 53 percent of Americans often or sometimes get their news from social media.

Last Wednesday morning, the story of the Texas Democrats fleeing the state to try to block a Republican voter ID law dominated the top news and politics posts with the most Facebook engagement. Indeed, it was a big story. However, the top four linked posts were all from conservative outlets using out-group animosity:

I had to scroll to the 35th most interacted post to find an objective story by the Austin American-Statesman. It had 83 shares, versus the most interacted post by ForAmerica, which had over 27,000. (ForAmerica describes its mission on its website as “Making freedom go viral, one click at a time.”)

According to the Wall Street Journal, Facebook in 2018 shut down research efforts to make its platform less divisive. A few years later, the platform favors content that is as divisive as ever -- to the detriment of journalism.

Share Public Sphere