A Psychologist Explains Four Reasons the Internet Feels So Broken

Jay Van Bavel is a professor of psychology and neural science at New York University. His lab has published papers on how the internet became a fun-house mirror of extreme political opinions, why the news media has a strong negativity bias, why certain emotions go viral online, why tribalism is inflamed by online activity, and how the internet can make us seem like the worst versions of ourselves. At the same time, Van Bavel emphasizes that many of the group psychology dynamics that can make social media seem like a dumpster fire are also core to what makes humankind such a special and ingenious species. We discuss the four dark laws of online engagement and the basics of group psychology.

If you have questions, observations, or ideas for future episodes, email us at [email protected].

In the following excerpt, Derek and Jay Van Bavel dive into the basics of group psychology and discuss how that plays a part in social media engagement.

Derek Thompson: Jay, you and I are going to spend most of this episode talking about your empirical research on how the internet, and social media in particular, holds a fun-house mirror to the human condition and why I think the incentives and engagement structures of social media are truly driving many people slowly insane. But first, how would you describe your work? What’s the big question that you and your lab are trying to answer?

Jay Van Bavel: Yeah, I think the big question that’s animated me for the last 20 years in my research is about the nature of groups and identity and how they work. I started originally looking at the brain, how our brain responds to being in a group or seeing members of other groups. And then we’ve been—our brains, which are very evolutionarily ancient, have been dropped in this modern technological environment with social media and artificial intelligence. And now, I’m fascinated by studying how that all plays out: how is that manipulated, how do we get triggered in various ways, and so forth.

Thompson: So obviously, the brain is a pretty ancient piece of machinery. Group psychology has to be a relatively ancient concept. We’ve been living in groups for hundreds of thousands of years before the internet came along. Is there a general Psych 101 principle of, like, “This is your brain on groups”? If I went to Group Psychology 101, what would I learn on day one about how groups change my own identity and my sense of what is true and false in the world?

Van Bavel: Great question. So maybe I’ll give you a couple principles. The first is: All it takes is a flip of a coin for us to feel part of a group and start to identify with it. And so we study this in the lab. It’s called a minimal group, where we just literally flip coins and put people on the blue team or the red team. But if anybody has ever grown up playing recreational sports, you go to the gym and you randomly get picked onto a side to play a sport for an hour or two, and that automatically triggers a different psychological mindset. And even people who were strangers a few minutes before are people we suddenly like more, pay more attention to, have an affinity for, and we’re able and willing to cooperate with them in ways we wouldn’t have if they were strangers.

And this happens to us also if you go to a professional sporting event. You’re giving a high five to someone sitting next to you who has a Yankees jersey if you’re a fan, even though you might [have never known] them before, might never see them again in your life. And so that’s the psychology that all of us walk around in.

The other part of it that a lot of people don’t get, but it’s part of the examples I give, is that it shifts from minute to minute. So again, I can create groups in my lab. You walk into a stadium with a jersey on and see everybody else wearing those colors. You have a different identity. You turn on the TV, and it’s election night, and you’re cheering for your party to win. Your country goes to war. You’re under attack. An even broader, more inclusive identity gets triggered. Or let’s say you turn on a movie and you’re watching an alien invasion. In that case, it’s just all of humanity against this alien invasion. So we can get triggered to think of ourselves as different types of members of groups or as members of different types of groups depending on the situation we’re in.

Thompson: It’s almost as if the desire to belong to any group is pre-downloaded in us, and therefore thinking outside of groups seems to be the really key challenge. If it is so easy for us to adopt a group psychology, then being online, where it’s all just one big mass of group psychology overlapping group psychology overlapping group psychology, the hardest thing to do is to try to “think for yourself” when we so want to think as if we belong to any of these groups.

So having spent several hours reading your papers from the last few years this morning and yesterday and thinking about them as a collective, I came up with, what I am calling for the moment, “the four bad laws of internet and social media engagement.” Because I feel like you and your lab have been—in researching how group psychology and dynamics and social media and the internet affect us, you found your way to these four laws of what it is that drives bad engagement on the internet. And I’m just going to read them out now, and then we’re going to walk through them. I’ll repeat them at the end, and then we can analyze them. But here are the four laws that I’ve taken from your research.

Number one is that negativity drives engagement. Number two, extremism drives engagement online. Number three, out-group animosity drives engagement online. And number four, moral, emotional language drives engagement online. And for listeners at home, if you didn’t understand what some of those terms meant, we will absolutely define them in just a few minutes. But what I want to do with you with our time together is to explain where these ideas come from—because they’re all derived from your work in the last few years—and what they mean.

I want to start with a paper you published last year on negativity bias in the news. In a randomized study of 105,000 headlines and 370 million impressions from a data set of articles published by the online news dispensary Upworthy, you found that so-called “negative” words increased the click-through rate by more than 2 percent, and the prevalence of positive words in a news headline decreased the likelihood of any headline being clicked on. In your own words, what did this paper show that was so important?

Van Bavel: Yeah, I think one of the things that psychologists are very familiar with is the idea that bad is stronger than good, that people pay more attention to or are more motivated by bad things than we are good things. And I’ll go back to a bit of an evolutionary explanation for this again: You can imagine your ancestors walking around the African savanna, and they’re looking for food to eat. And so there’s a motivation to try to find food. But if they potentially see something and it could kill them, that’s dramatically more significant to their survival. And so they have to err on the side of avoiding things that are negative and risky. And our ancestors who did that over generation after generation after generation were more likely to survive. And so we have brains that are wired to detect threats and super-tuned to them more than they are to detect rewards.

And so that gets manipulated, classically, by the news. The famous quote from journalism is, “If it bleeds, it leads.” But of course social media and internet websites have hijacked that as well. And the data we had, I’ll just say briefly, came from the classic website that engineered virality online. It was from Upworthy. So we got access to the Upworthy data archives. And what Upworthy was famous for in the early days of the internet was doing A/B testing. So they would take a single news story, and they would try one headline on their website and see how many people clicked and then randomly assign half of the other people who came to their website to see a different headline. And so that way, they would test different headlines, look at their data, and then pick the one that was most popular and keep it up. And they were the pioneers in this.

In fact, they were so viral at one point, they had more traffic than The New York Times. They were so viral that Facebook had to change its news algorithm because people were getting really frustrated by all the clickbait from Upworthy headlines and leaving Facebook to go get it. And so they really engineered this through data. So when we got that data set, we thought, “Well, we can see not what their theory is about what works; we can see with their experimental A/B testing what actually gets people to click through an article.” And, of course, they’re more likely to click to negative content, negative headlines for the exact same story.

This excerpt was edited for clarity. Listen to the rest of the episode here and follow the Plain English feed on Spotify.

Host: Derek Thompson
Guest: Jay Van Bavel
Producer: Devon Baroldi

Subscribe: Spotify