Making sense of The Anxious Generation

Jonathan Haidt’s latest book The Anxious Generation came out on March 26th. It has since reignited the debate about social media, young people, and mental health. I have now read the book. I’ve read the reviews and critiques. I’ve listened to the podcasts. And I’m finally ready to offer my “hot take”—which is actually now a very cold take I’ve been thinking through over the past month (plus 12+ prior years of research).

Note: this is my current thinking on these issues, which I’m sure will continue to evolve as new research comes out, and as we work toward finding the best possible solutions.

So, here goes nothin’!

What’s the book about?

The central premise is that rates of mental illness have been rising in young people since the early 2010s, and smartphones and social media are largely to blame (see the accompany Atlantic piece). Haidt argues that childhood independence and play have been declining since the 1980s, and that the transition to a “phone-based childhood” between 2010 to 2015 accelerated this problematic shift.

He also proposes solutions for governments, tech companies, schools, and parents to “roll back” this phone-based childhood, like enacting age verification and waiting to give kids smartphones until high school.

Where do people disagree?

A widely-cited review of the book in Nature argues that there is “no evidence that using these platforms is rewiring children’s brains or driving an epidemic of mental illness.” Many academics similarly argue for a lack of evidence for Haidt’s claims. Some argue the evidence is correlational, rather than causal (i.e., trends in social media use and mental health problems are rising together, but one is not necessarily causing the other). Others argue that the effect sizes are too small to be meaningful, or that the majority of evidence simply does not support his argument.

Critiques of Haidt’s proposed solutions include: concerns about government overreach (or under-reach), belief that they are unrealistic or impractical, and flat-out conviction that they will not improve kids’ mental health.

What do you think? Did social media cause the youth mental health crisis?

In general, I think the decline of the “play-based childhood” and rise of the “phone-based childhood” likely played some role in the current mental health crisis (see more here). But, at the end of the day, this is my opinion. I do not think we have enough high-quality scientific evidence to prove this theory right or wrong.

I agree with many of Haidt’s conclusions. Where we differ, I think, is in our confidence in the evidence behind them. To me, this theory remains just that: a theory.

Why don’t we have good evidence one way or the other?

This question that we’re trying to answer—Did the introduction of social media and smartphones cause the increase in mental illness among young people in recent years?—is simply a very hard question to answer with data. Population-level trends are difficult to explain. Most likely, there are many causes.1 And the research methods we have available to us do not do a good job of getting us to an answer.

Many of the studies are correlational, as we know. Many do a bad job of measuring “social media” or “mental health.” But even a good experimental study cannot give us proof one way or the other. For example, an experiment might tell us whether limiting an adolescent’s social media use right now, in 2024, improves their mental health symptoms. But a teen avoiding social media in 2024 is a very different situation than all teens (hypothetically) having never used social media in the first place. It can certainly inform the answer to the population-level question, but it won’t get us all the way there.

The truth is, it may be a very long time before we have definitive proof one way or the other.

So, then, what do we do? Do we dismiss this theory? Or do we dip our toes into the treacherous waters of common sense? Relying on common sense in the sciences is, generally, a very bad idea. Especially when it comes to evaluating the effects of new technologies (see: moral panics about bicycles, the printing press, cars, telephones…). But I do think a common sense argument is warranted here. I have to believe that the drastic shift in the early 2010s to a world of smartphones and social media played some role in the youth mental health trends that started at that time.

What do you think we should do about it?

In general, I agree with some of the solutions Haidt proposes. But, again, I do not think we have strong evidence that they will work. In fact, we will likely never have enough evidence to say that, for example, age 14 is safer than age 13.75 for introducing smartphones, or 16 is better than 15 for social media.

So, the question becomes: how much evidence do we need to act? As a parent, I’ll likely try to delay giving my child a smartphone. I do not need a randomized controlled trial to tell me to do this. But what about schools? What about governments? How much evidence do we need to make legislative decisions? When the science can only take us so far, how do we make policy that does the most good, and the least harm, for our children? We do it by trying our best. Using the evidence where possible. And trying to avoid making things worse.2

Research can only get us so far here. Science can take us to the edge of the riverbank, but eventually, we need to forge across. The sooner we recognize the limitations of the evidence, the sooner we can find a path to the other side.

What do you think of Haidt’s “four reforms”?

Haidt proposes “four foundational reforms” that communities can use to roll back the phone-based childhood. Here they are, along with what I think about them:

1. No smartphones before high school?

In general, I think it makes sense to delay smartphones. The start of high school seems as good a time as any, particularly given the “collective action” problem Haidt describes (and that anyone who’s tried to delay giving their child a phone knows well). However, I maintain that every family is different, and some might have very good reasons for getting their child a smartphone earlier. Note: a “dumbphone” can be a good first phone option. For more on this, check out my discussion with Dr. Emily Weinstein about parent smartphone “pledges” and relevant considerations.

2. No social media before age 16?

Again, I think it makes sense to delay social media when possible. Is 16 a magic number? No. Currently, the age of “Internet adulthood,” based on the Children’s Online Privacy Protection Rule (COPPA) is 13. We have to set it somewhere. Should that “somewhere” be 14? 15? 16?3 Science cannot answer this, beyond telling us that, on average, older teens generally have developed more skills (impulse control, lower sensitivity to peer feedback) that might help them navigate social media [see this recent APA report], and that younger teens, on average, may struggle more with social media.

Many argue that this age limit is unrealistic. I understand this, but I also think this is a separate question. Let’s first decide whether it’s a good idea in theory before we decide whether it will work.

I’ve also heard disagreement over whether we should: (a) limit social media access below a certain age, or (b) make changes to the platforms to make them safer and healthier for kids. This is a false dichotomy. We can, and should, do both. We already essentially do the former (at age 13), though it may make sense to raise this age. We have not yet done the latter, but we certainly need to.4

3. Phone-free schools?

Yes, especially for middle and elementary school. Schools have many different considerations to weigh in how they might implement this, but I think we have some evidence (and common sense) that phone use during the school day can cause distraction and lower-quality in-person socialization.

4. More childhood play and independence?

Yes. Better and longer recess, play clubs, time outside, greater independence and responsibility, and more opportunities to engage in the “real world”—all good ideas for kids. This is, of course, a systemic issue, too—making sure that all kids have access to these opportunities is critical.

Are there any benefits to smartphones and social media?

Yes! Most children (and their parents) will agree that many good things happen on their devices. For example, opportunities for creativity and self-expression, staying in touch with friends and family, learning and discovering new things, entertainment and convenience. Not to mention, opportunities for teens to connect with others who are similar to them, particularly for those who may feel marginalized in their “offline” lives, like LGBTQ+ teens (note, though, that these teens often also encounter more online risks).

These benefits are important, and any solutions we implement should aim to avoid unintended consequences that diminish these opportunities.

In my mind, this becomes a “line-drawing” problem. Where do we draw the line on what counts as “social media?” Can kids still get these benefits from YouTube, or text messaging, or some other platform that doesn’t involve an algorithmically-curated feed designed to maximize engagement? And where do we draw the line on age? We’ve already effectively decided (with COPPA) that these benefits are not so great as to require 12-year-olds to need access to social media. What about 13-year-olds? 14? 15?

When we talk about restricting access to social media, as parents, or as a society more broadly, we also need to think about what we’re replacing it with. How can we ensure that kids have access to the things they need—a supportive community, opportunities to learn—whether that’s online or offline?

Is there anything we can all agree on?

In my view, we all—Haidt, his critics, and everyone else—agree on many of the things that matter most.

Most of us recognize there are problems with the current design of smartphones and social media platforms for kids. If you were to design a platform from scratch to provide the healthiest, most enriching environment for children, would you come up with TikTok? No. I don’t think anyone is arguing that.

We agree that something should be done about this—though exactly what that should be, and who should do it, is less clear.

We agree that many non-screen things matter for mental health, too: safe and supportive school and home environments; access to basic needs like food, shelter, and healthcare; and positive family and peer relationships, to name a few. In fact, for me, some of the most resonant sections of The Anxious Generation were those about children’s lives off-screen—about the need for play and independence, the importance of community, and the value of practices like quiet reflection and shared rituals.5

Most importantly, I believe we all agree on the ultimate goal: to do right by our kids.

And that gives me hope we’re heading in the right direction.