In April, social psychologist Jonathan Haidt published an essay in Atlantic in which he sought to explain, as the title of the play had it, “Why the last 10 years of American life have been especially stupid.” Anyone familiar with Haidt’s work over the past half-decade might have anticipated his answer: social media. Although Haidt admits that political polarization and factional enmity long predate the rise of platforms, and that there are many other factors involved, he believes the tools of virality – the Like and Facebook’s Share, Twitter’s Retweet function – have algorithmically and irrevocably corroded public life. He determined that a large historical discontinuity can be dated with some accuracy to the period between 2010 and 2014, when these features became widely available on phones.
“What Changed in the 2010s?” Haidt asks, reminding his audience that a former Twitter developer once likened the Retweet button to giving a four-year-old a loaded gun. “A mean tweet doesn’t kill anyone; it is an attempt to publicly shame or punish someone while broadcasting one’s own virtue, brilliance, or tribal loyalty. It’s more of a dart than a bullet, causing pain but no death. Even so, from 2009 to 2012, Facebook and Twitter distributed around one billion darts worldwide. Since then, we have been shooting each other. While the right thrived on conspiracy and misinformation, the left became punitive: “When everyone got a dart gun in the early 2010s, many left-wing institutions started shooting themselves the head. And, unfortunately, it is the brains that inform, educate and entertain most of the country. Haidt’s overriding metaphor for total fragmentation is the story of the Tower of Babel: the rise of social media has “unwittingly dissolved the mortarboard of trust, belief in institutions and shared histories that had held together a secular democracy vast and diverse.
These are, of course, common concerns. Haidt’s main concern is that social media use has made us particularly vulnerable to confirmation bias, or the propensity to fixate on evidence that backs up our prior beliefs. Haidt acknowledges that the existing literature on the effects of social media is vast and complex, and that there is something for everyone. On January 6, 2021, he was on the phone with Chris Bail, a sociologist at Duke and author of the recent book “Breaking through the social media prismwhen Bail urged him to turn on the television. Two weeks later, Haidt wrote to Bail, expressing frustration at how Facebook officials consistently cite the same handful of studies in their defense. He suggested the two collaborate on a comprehensive literature review that they could share, in Google Doc form, with other researchers. (Haidt had experimented with such a model before.) Bail was cautious. He said to me, ‘What I said to him was, ‘Well, you know, I’m not sure the research is going to confirm your side of the story,’ and he said, ‘Why don’t we see?’ ”
Bail emphasized that he was not a “platform basher”. He added: “In my book, my main point is: Yes, platforms do play a role, but we greatly exaggerate what it is possible for them to do – how much they could change things, regardless who is running these companies – and we deeply underestimate the human element, the motivation of the users. He found Haidt’s idea of a Google Doc appealing, in that it would produce a sort of living document that existed “somewhere between scholarship and public writing.” Haidt was eager to have a forum to test his ideas. “I decided that if I was going to write about this – what changed in the universe, circa 2014, when things got weird on campus and elsewhere – again, I better be sure to be right,” he said. “I can’t be satisfied with my feelings and my biased readings of literature. We all suffer from confirmation bias, and the only cure is other people who don’t share yours.
Haidt and Bail, along with a research assistant, completed the document over several weeks last year, and in November invited about two dozen researchers to contribute. Haidt told me of the difficulties of social science methodology: “When you first approach a question, you don’t even know what it is about. “Does social media destroy democracy, yes or no? That’s not a good question. You cannot answer this question. So what box you ask and answer? As the document took on a life of its own, actionable topics emerged: Does social media make people more angry or more emotionally polarized? Does it create political echo chambers? Does it increase the likelihood of violence? Does it allow foreign governments to increase political dysfunction in the United States and other democracies? Haidt continued, “It’s only after breaking it down into lots of questions that you can answer that you see where the complexity lies.”
Haidt came away feeling, overall, that social media was actually pretty bad. He was disappointed, but not surprised, that Facebook’s response to his article was based on the same three studies they’ve been reciting for years. “It’s something you see with breakfast cereals,” he said, noting that a cereal company “might say, ‘Did you know we have twenty-five percent riboflavin in more than the leading brand?’ They’ll tell you about traits where the evidence is in their favor, distracting you from the general fact that your cereal tastes worse and is less healthy.
After Haidt’s article was published, the Google Doc—”Social Media and Political Dysfunction: A Collaborative Review”—was made available to the public. Comments piled up and a new section was added, at the end, to include a mix of Twitter threads and Substack essays that emerged in response to Haidt’s interpretation of the evidence. Some colleagues and kibbitzers agreed with Haidt. But others, although they may have shared his fundamental intuition that Something in our social media experience was wrong, relied on the same set of data to come to less definitive, if not slightly contradictory, conclusions. Even after the initial flurry of responses to Haidt’s article faded into social media memory, the document, insofar as it captured the state of debate on social media, remained a living artifact.
Toward the end of the collaborative’s introduction, the authors warn, “We caution readers not to simply add up the number of studies on each side and declare one side the winner.” The document is over one hundred and fifty pages, and for each question there are affirmative and dissenting studies, as well as others that indicate mixed results. According to one article, “political expressions on social media and the online forum have been shown to (a) reinforce the partisan thought process of the expressers and (b) harden their pre-existing political preferences,” but, according to another, who has used data collected during the 2016 election, “During the campaign, we found that media usage and attitudes remained relatively stable. Our results also showed that Facebook information use was linked to a modest spiral of depolarization over time. Additionally, we found that people who use Facebook for news were more likely to see both supportive and counter-attitude news in each wave. Our results indicated that counter-attitudinal exposure increased over time, leading to depolarization. If results like these seem inconsistent, a perplexed reader turns to a study that says, “Our results indicate that political polarization on social media cannot be conceptualized as a unified phenomenon, as there are important differences between platforms”.
Interested in echo chambers? “Our results show that aggregation of users in homophilic clusters dominates online interactions on Facebook and Twitter,” which sounds compelling, except, as another team puts it, “We find no evidence supporting a strong characterization of “echo chambers” in which the majority of people’s sources of information are mutually exclusive and come from opposite poles. At the end of the record, the vaguely condescending recommendation against mere summons begins to make more sense. A document that was originally a bulwark against confirmation bias could, in the end, just as easily function as some kind of generative device to support anyone’s favorite belief. he, was simply throwing his hands in the air.
When I spoke to some of the researchers whose work had been included, I found a combination of general and visceral unease with the current situation – with the scourge of harassment and trolling; with the opacity of the platforms; with, well, the widespread hunch that of course social media is bad in many ways – and a contrasting sense that it might not be catastrophically bad in some of the specific ways that many of us have come to believe are true. It was no mere riptide, and there was no hint of joyful mythos; the problem was big enough to fix. When I told Bail that the result seemed to me to be that exactly nothing was unambiguously clear, he suggested that there was at least a solid basis. He looked a little less apocalyptic than Haidt.
“A lot of stories are just plain wrong,” he told me. “The political echo chamber has been massively overrated. That’s maybe three to five percent of people who are correctly in an echo chamber. Echo chambers, as hotbeds of confirmation bias, are counterproductive to democracy. But research indicates that most of us are actually exposed to a wider range of viewpoints on social media than we are in real life, where our social networks – in the original use of the term – are rarely heterogeneous. (Haidt told me this was a question on which Google Doc had changed its mind; he had become convinced that echo chambers were probably not as widespread a problem as he had imagined.) The echo chamber effect of social media could obscure the relevant counterfactual: a conservative might abandon Twitter just to watch more Fox News. “Coming out of your echo chamber is supposed to make you moderate, but maybe it makes you more extreme,” Bail said. The research is incomplete and ongoing, and it’s hard to say anything on the subject with absolute certainty. But that was, in part, Bail’s point: we should be less sure about the particular impacts of social media.