The muddled science on teens and social media

Extensive interviews with Jon Haidt and Candice Odgers

Last week, I had an editorial on the explosive success of Jon Haidt’s blockbuster book, The Anxious Generation, which details his claims that the phone-based childhood created by social media is “the major cause” (his words) of the mental health crisis among young people. My column was not a litigation of whether his claims are correct (I’m not an expert on that), but rather whether he is out ahead of the science (he acknowledged to me that he is) and whether he has more responsibility to use his platform to explain that the science is unsettled (I concluded he does).

The main critic who has been willing to be on the record on this is Candice Odgers, a psychologist at UC-Irvine who wrote a now-pretty-famous brutal review in Nature of Haidt’s book. This is the most devastating quote in the review:

On the first day of the graduate statistics class I teach, I draw similar lines on a board that seem to connect two disparate phenomena, and ask the students what they think is happening. Within minutes, the students usually begin telling elaborate stories about how the two phenomena are related, even describing how one could cause the other. The plots presented throughout this book will be useful in teaching my students the fundamentals of causal inference, and how to avoid making up stories by simply looking at trend lines.

Oof. Probably not on Haidt’s virtual refrigerator.

Odgers was willing to talk to me, as she has done with lots of reporters. A few of the quotes ended up in the editorial, but there’s lots of good stuff in the interview that I’m recounting here. I asked her about her decision to review the book. She said that she was invited by both the New York Times and Nature, but decided to do the review in Nature. “I said to Nature that I couldn’t write anything positive,” she said. “After Nature sent me an advance copy of the book to review I wrote back to say, ‘I have been struggling in my review of this book as it is certain to be a best seller, given the fear narrative and scare tactics that are taken throughout, but the science is…absolutely awful.’” Good for Nature for continuing with the review and running it.

When I asked her why she went forward, she said, “this is a pretty damaging story for young people.  If someone was going around telling a story with science about the causes of childhood cancer, we would correct the record.  People are not standing up to it because the moral panic and fear around this issue has reached a fevered pitch making it a widely unpopular opinion, therefore it takes so much time to come forward.” She’s right about that: Odgers has been getting so many requests that she sends out a pre-written piece (“Social Media Fact Sheet”) before the interview. Also, I contacted a number of other prominent psychologists who also disagree with Haidt’s analysis but who wouldn’t go on the record because of the time sink that it would be to jump into the fray.

“The most shocking thing,” Odgers said, “is the distance between what people believe and what you can detect in the data.  What is the mechanism that causes these things?” She says there is “more evidence in the solution space that technology can be used to deploy therapies on the positive side than evidence for bad effects on the negative side.”

To Haidt’s credit, he also engaged in an extensive exchange with me by email. Around the same time, there was a story on PBS News Hour that discussed the controversy where Haidt did not respond to interview requests. There was another in the New York Times where his collaborator, Jean Twenge, was quoted. So, I was very grateful he engaged. Probably the most important important thing Haidt told me was that “It is true that I am promoting a social change program…and I am doing this before the scientific community has reached full agreement.” But Haidt’s contention, and Twenge said the same thing to the Times, is that there is no harm in what they are proposing, so waiting for scientific consensus to form when — in their view — significant harms are occurring is not responsible.

I asked Odgers if she agreed that there was no harm in implementing Haidt’s recommendations. She disagreed strongly and had two reasons. The first is “the opportunity cost for sending everyone out chasing the data to back up these stories, instead of keeping their focus on the main issues that are threatening the mental health of young people and figuring out how to better support them in the online spaces where they spend most of their time.” We see a lot of papers on the effects of social media on various behaviors, and it is absolutely the case that it is almost always very hard to get a signal. So the search for the answer could take a very long time even if there is one.

Further, Odgers said - and as a long-time college professor who usually teaches first-year students who are still teenagers, I strongly agree — “it sends a message that this is a shameful activity.  When adults think about social media, they think about the awful way we behave on Twitter, not the many positive things that young people do.  Large-scale surveys find that parents cite arguments over digital technology use as a leading cause of conflict in their families.”

My anecdotal experience is that adults have at least as much trouble getting off their phones as teens do. We had a great paper about this a few years ago in Science Advances called “Kids these days: Why the youth of today seem lacking.” This paper has strong evidence that every generation says this about youth because we old folks always map our current traits on to them. In this case, we’re doing that without acknowledging our own addictions to our phones. Haidt, I’m sure, blames the social media companies and the parents for enabling the behavior, but I think the shaming element is unavoidable: after all, it’s the kids on their phones that is apparently the problem. If you read my graduation speeches, you’ll see that I disagree strongly with blaming students for societal ills. (Insert your favorite line from The Who’s rock opera, Tommy.)

There’s a third harm, in my view, which is teaching the public bad science. We know that once incorrect ideas are lodged in the public consciousness — such as vaccines causing autism or hydroxychloroquine curing COVID-19 — they are almost impossible to dislodge completely. So, if Haidt has to walk any of this back, it’s going to be really difficult.

Finally, there’s the question of whether a scientist who can reach millions — as Haidt is clearly doing — has an obligation to do more for critics who don’t have the same platform. Odgers has (for better or worse) certainly become better known as a result of all of this, but she is not on every talk show, and her ideas are not being lodged in the wider conversation. When I asked Haidt if he should have been more forthcoming about the disagreements in the book, he said, “I don’t think so.”

I asked Odgers about the fact that certainly plenty of parents sense that the phones might be bad for their children and that is why this resonates. She agreed. “A lot of people say, ‘I don’t need a study to tell me what is bad for my kids,’” she said.  “That’s a value-based argument and part of parenting. Caregivers make all kinds of decisions about how they want their children to spend their time.  But you cannot then say that science supports the idea that social media causes things like suicide or is causing an epidemic of mental illness.”

I decided on the title “Unsettled science on social media” for my editorial. That title is likely to hold up for some time. Too bad so many people think we’re done.

_________________

For more reading, here’s a section of Odgers pre-interview handout that has some more things to read about this.

Hancock et al. 2022. Psychological Wellbeing and Social Media Use: A Meta-Analysis of Associations between Social Media Use and Depression, Anxiety, Loneliness, Eudaimonic, Hedonic and Social Wellbeing (2022) who analyzed 226 studies and concluded that across 275,728 participants indicated that social media use was not associated with overall wellbeing, with an effect size equal to approximately zero.

Valkenburg et al. 2022. Social Media Use and its Impact on Adolescent Mental Health: An Umbrella Review of the Evidence (2022) who reported weak and mixed associations between the use of social networking sites and wellbeing.

Orben. 2020. Teenagers, Screens, and Social Media: A Narrative Review of Reviews and Key Studies. Reviews 80 systematic reviews and meta-analyses and again concluded that while a small negative correlation between digital technology and adolescent wellbeing can be located, that it is not clear whether the association is driven by other factors.

Ivie et al. 2020.A Meta-Analysis of the Association Between Adolescent Social Media Use and Depressive Symptoms. This study performed a meta-analysis of 11 studies (10 cross-sectional, 1 longitudinal) from 2012-2020 with a total of 92,371 and cautioned interpretation of an association due to the small effect size and high variabiltiy acorss studies. The concluded that …. “… prevention programs and public policy would be better served focusing on these well-established risk factors with larger effect sizes than contributing to a moral panic about the effect of technology use, especially given the lack of supporting data.”