I haven’t read the book or the linked articles. But I also see why the piece was killed; it doesn’t make a convincing argument.
The primary direct evidence you cite (the Yondr data) is pretty tangential to the question of whether the smartphone is to blame for the drop in teenaged mental health.
As you describe it, the data only addresses smartphone bans in schools. That’s maybe 6 out of 24 hours each day. It’s 6 hours when there’s a lot of other things going on. There’s plenty of room for smartphone harm to be happening the other 18 hours a day.
We’ve banned smoking in schools forever. And back in my day plenty of teenagers smoked and it was provably bad for them.
A more relevant question for school smartphone bans is whether they improve academic performance other than teenagers self-reporting on their attention levels (likely to be wildly biased). You don’t provide evidence about that. But it (if supported by data) would be a much more reasonable justification for school phone bans.
You also conflate “phone use” with “social media”. Children use phones for games and media consumption in significant amounts as well. All of the other references to studies you cite are specifically about social media. Could the psychological harm come not specifically from social media but more generally from phone use displacing important developmental activities?
This would be a far stronger article if you had a more compelling facts supporting your argument within it.
I’m not arguing that the book doesn’t provide academically appropriate levels of evidence to support its claims, that public policy should be based on flimsy evidence or that you’re wrong in any way. Maybe if I clicked on the links you provided I would find your argument well-supported. But as written, this article falls short in some of the same things you accuse the book of.
Robert, you make some fair points, and I love that you're engaging with the substance rather than just picking a team. You're right that the Yondr data covers only the school day, and right that phone use and social media get conflated too often in this whole debate. Those are real limitations.
But I think the central charge, that the piece commits the same sin as the book, doesn't hold. The book makes a strong causal claim. Phones are the major cause of the teen mental health crisis. A claim that strong carries a heavy burden of proof. The piece makes a much more modest claim. That the strong version isn't supported by the evidence yet. Those are not symmetric. Asserting causation requires robust data. Pointing out that the data isn't there requires far less. So they are not falling short in the same way.
Your smoking example is interesting, because it actually shows the gap. Smoking has a clear biological mechanism and decades of overwhelming causal evidence. That is precisely what we don't have for phones. If anything, reaching for smoking underlines how thin the causal story for phones still is.
And on the social media versus general phone use distinction, I'd argue that cuts against the book rather than the article. If we can't even isolate what's doing the harm, that's another reason to be cautious about a confident single-cause narrative.
Worth clicking the links. I think you'll find the argument better supported than it reads at first.
Just to make it super clear, I am NOT defending Haidt. The picture is clearly more complicated and there are so many more factors that could be at play including helicopter parenting / the end of free-range childhoods, the loss of regular exposure to danger, the ever looming and unaddressed threat of climate apocalypse… I agree that smartphones are a too-convenient boogeyman.
When I see documentation of small AVERAGE effects I am always led to wonder about the extent to which children are differentially susceptible to environmental influences due to variation in developmental plasticity in general or even to just social media exposure, presuming, of course, that selection effects have been discounted. If this proves to be true then the question becomes for which kids is social media problematic—and even for which may be beneficial? Failing to distinguish positive, negative and no effects across individuals could account for both small or null ones.
Spot on Jay! It’s been far too long since our MVR days! As Urie B might say “for whom, under which circumstances and with which effects over what period of time”. Be well!
I echo what many others have said, in that this is an incredibly challenging question to answer because the data are inherently messy.
Debate on this topic often conflates: 1) social media; 2) smart phone use; and 3) screen time. These are all, related, but different issues. Obviously one cannot use social media without screens, but one can have screen time without social media.
But, it goes beyond that. How exactly is one using social media? Is it just to see family photos? To watch funny cat videos? Or to interact with one's school age friends, as an adjunct or substitute to other "IRL" interactions?
Likewise, removing access to smartphones in the classroom does not get at the larger question of their impact. Like others, I will use a COVID analogy.
We've heard people proclaim that "science shows masks don't work in preventing COVID." But, again, that depends on mask use. Imagine there is a group of 20 people, and one of them has COVID. They all wear a mask for 8 hours of the workday. Except for during lunch. (This is essentially the policy we had with air travel - you must wear a mask, except for when eating). If you remove the barrier temporarily, it's going to increase the risk of transmission. (The same could be said of condom use).
If children have access to smartphones and are using them to engage on social media before school, after school, and on weekends, we can't expect that removing those smartphones during the school day will automatically translate to better classroom behavior and better learning. Perhaps a student was being cyberbullied by one of their classmates all weekend long over social media - they aren't going to magically forget that when they are in the classroom with them, just because they don't have access to their phone.
Similarly, If years of smartphone use have altered habits, sleep, social interactions, or attentional patterns, we should not necessarily expect those effects to disappear simply because access is restricted during the school day.
So, I think it is fair to question the causal evidence behind Haidt's claims. But I also think it is fair to question whether studies restricting smartphone access during school hours are actually testing the same hypothesis. Evaluating adolescents with years of prior exposure who continue to use smartphones and social media outside of school may tell us something about the effects of partial restriction, but it may tell us much less about the broader effects of growing up in a smartphone-mediated environment.
Did any of these studies account for the effects that repeat covid infections might be having on these kids, since mental illness, brain fog, etc are all things long covid can trigger or make worse?
Seems like any study done since 2020 should account for covid's impact and try to rule that out whether it's concerning mental or physical health.
Thank you for writing this. Having previously worked on child online protection initiatives with the ITU and various stakeholder groups across Southeast Asia, including governments, educators, industry, and civil society, I think this is a very important point.
I increasingly worry that many governments are overselling bans as standalone solutions to much deeper social, educational, and infrastructural problems. If expectations are framed around dramatic improvements in mental health or academic performance, the reality will likely prove far messier and more modest.
Young people do not simply disappear from digital life when platforms are restricted. They migrate elsewhere, often into less visible and less moderated spaces. They are so creative!
What still seems missing from many policy discussions is equal investment in alternatives: healthier digital ecosystems, youth-centered public-interest platforms, digital literacy programs for both parents and youth, community initiatives, and meaningful offline social infrastructure.
Otherwise, we risk creating a policy cycle driven more by symbolic reassurance than sustainable long-term support for young people.
All the phone ban studies aren’t analysing true phone bans - all the pupils still own smartphones. We need a study on schools who have a rigorous brick phone policy - not smartphones in pouches or lockers which are misused (or they put a burner phone in and keep their smartphone). So there’s no data proving a true phone ban doesn’t work.
It’s quite surprising in this article you can’t see that pouches are because of the mere presence effect. They essentially say: own a smartphone that is harmful to you but here’s a special package to put it in. Yonder are a company who jumped onto the phone ban idea as a way of making money for themselves.
What we want and have always wanted is a genuine phone ban and restrictions on ownership of smartphones and social media age limits.
I recommend you read the work of Jared Cooney Horvarth if you haven’t.
From someone on the front line, true phone bans are working. By true phone ban I mean a brick phone policy, discouragement from school about purchasing a smartphone and rigorous bag checks to enforce it.
Thank you for highlighting this: “and that the decline in adolescent mental health is more likely caused by a lack of parental support, economic distress, a decline in parental mental health, or many other possibilities.”
Those working on mobile phone bans know that pouches don't make a difference as kids still get access outside school hours (which is when they would mostly use social media anyway). The only time you see an impact, and loads of head teachers agree on this, is if you fully ban phones so that they cannot even be bought onto the school site. Then that changes culture and you see an impact.
I see why there are reasons to be skeptical of phone bans in schools, but the recent NBER paper did find that after the first year, subjective wellbeing went up for the next 2 years. So, even if the phone restrictions didn't affect test scores or other measures, isn't it still a good thing if kids are happier in general?
I haven’t read the book or the linked articles. But I also see why the piece was killed; it doesn’t make a convincing argument.
The primary direct evidence you cite (the Yondr data) is pretty tangential to the question of whether the smartphone is to blame for the drop in teenaged mental health.
As you describe it, the data only addresses smartphone bans in schools. That’s maybe 6 out of 24 hours each day. It’s 6 hours when there’s a lot of other things going on. There’s plenty of room for smartphone harm to be happening the other 18 hours a day.
We’ve banned smoking in schools forever. And back in my day plenty of teenagers smoked and it was provably bad for them.
A more relevant question for school smartphone bans is whether they improve academic performance other than teenagers self-reporting on their attention levels (likely to be wildly biased). You don’t provide evidence about that. But it (if supported by data) would be a much more reasonable justification for school phone bans.
You also conflate “phone use” with “social media”. Children use phones for games and media consumption in significant amounts as well. All of the other references to studies you cite are specifically about social media. Could the psychological harm come not specifically from social media but more generally from phone use displacing important developmental activities?
This would be a far stronger article if you had a more compelling facts supporting your argument within it.
I’m not arguing that the book doesn’t provide academically appropriate levels of evidence to support its claims, that public policy should be based on flimsy evidence or that you’re wrong in any way. Maybe if I clicked on the links you provided I would find your argument well-supported. But as written, this article falls short in some of the same things you accuse the book of.
Robert, you make some fair points, and I love that you're engaging with the substance rather than just picking a team. You're right that the Yondr data covers only the school day, and right that phone use and social media get conflated too often in this whole debate. Those are real limitations.
But I think the central charge, that the piece commits the same sin as the book, doesn't hold. The book makes a strong causal claim. Phones are the major cause of the teen mental health crisis. A claim that strong carries a heavy burden of proof. The piece makes a much more modest claim. That the strong version isn't supported by the evidence yet. Those are not symmetric. Asserting causation requires robust data. Pointing out that the data isn't there requires far less. So they are not falling short in the same way.
Your smoking example is interesting, because it actually shows the gap. Smoking has a clear biological mechanism and decades of overwhelming causal evidence. That is precisely what we don't have for phones. If anything, reaching for smoking underlines how thin the causal story for phones still is.
And on the social media versus general phone use distinction, I'd argue that cuts against the book rather than the article. If we can't even isolate what's doing the harm, that's another reason to be cautious about a confident single-cause narrative.
Worth clicking the links. I think you'll find the argument better supported than it reads at first.
Just to make it super clear, I am NOT defending Haidt. The picture is clearly more complicated and there are so many more factors that could be at play including helicopter parenting / the end of free-range childhoods, the loss of regular exposure to danger, the ever looming and unaddressed threat of climate apocalypse… I agree that smartphones are a too-convenient boogeyman.
When I see documentation of small AVERAGE effects I am always led to wonder about the extent to which children are differentially susceptible to environmental influences due to variation in developmental plasticity in general or even to just social media exposure, presuming, of course, that selection effects have been discounted. If this proves to be true then the question becomes for which kids is social media problematic—and even for which may be beneficial? Failing to distinguish positive, negative and no effects across individuals could account for both small or null ones.
Jay Belsky
Exactly. I think you'll dig this paper:
https://psycnet.apa.org/fulltext/2026-84042-001.html
Spot on Jay! It’s been far too long since our MVR days! As Urie B might say “for whom, under which circumstances and with which effects over what period of time”. Be well!
I echo what many others have said, in that this is an incredibly challenging question to answer because the data are inherently messy.
Debate on this topic often conflates: 1) social media; 2) smart phone use; and 3) screen time. These are all, related, but different issues. Obviously one cannot use social media without screens, but one can have screen time without social media.
But, it goes beyond that. How exactly is one using social media? Is it just to see family photos? To watch funny cat videos? Or to interact with one's school age friends, as an adjunct or substitute to other "IRL" interactions?
Likewise, removing access to smartphones in the classroom does not get at the larger question of their impact. Like others, I will use a COVID analogy.
We've heard people proclaim that "science shows masks don't work in preventing COVID." But, again, that depends on mask use. Imagine there is a group of 20 people, and one of them has COVID. They all wear a mask for 8 hours of the workday. Except for during lunch. (This is essentially the policy we had with air travel - you must wear a mask, except for when eating). If you remove the barrier temporarily, it's going to increase the risk of transmission. (The same could be said of condom use).
If children have access to smartphones and are using them to engage on social media before school, after school, and on weekends, we can't expect that removing those smartphones during the school day will automatically translate to better classroom behavior and better learning. Perhaps a student was being cyberbullied by one of their classmates all weekend long over social media - they aren't going to magically forget that when they are in the classroom with them, just because they don't have access to their phone.
Similarly, If years of smartphone use have altered habits, sleep, social interactions, or attentional patterns, we should not necessarily expect those effects to disappear simply because access is restricted during the school day.
So, I think it is fair to question the causal evidence behind Haidt's claims. But I also think it is fair to question whether studies restricting smartphone access during school hours are actually testing the same hypothesis. Evaluating adolescents with years of prior exposure who continue to use smartphones and social media outside of school may tell us something about the effects of partial restriction, but it may tell us much less about the broader effects of growing up in a smartphone-mediated environment.
Did any of these studies account for the effects that repeat covid infections might be having on these kids, since mental illness, brain fog, etc are all things long covid can trigger or make worse?
Seems like any study done since 2020 should account for covid's impact and try to rule that out whether it's concerning mental or physical health.
Thank you for writing this. Having previously worked on child online protection initiatives with the ITU and various stakeholder groups across Southeast Asia, including governments, educators, industry, and civil society, I think this is a very important point.
I increasingly worry that many governments are overselling bans as standalone solutions to much deeper social, educational, and infrastructural problems. If expectations are framed around dramatic improvements in mental health or academic performance, the reality will likely prove far messier and more modest.
Young people do not simply disappear from digital life when platforms are restricted. They migrate elsewhere, often into less visible and less moderated spaces. They are so creative!
What still seems missing from many policy discussions is equal investment in alternatives: healthier digital ecosystems, youth-centered public-interest platforms, digital literacy programs for both parents and youth, community initiatives, and meaningful offline social infrastructure.
Otherwise, we risk creating a policy cycle driven more by symbolic reassurance than sustainable long-term support for young people.
I recently reflected further on this question here: https://digitalserendipities.substack.com/p/if-we-ban-social-media-for-children
All the phone ban studies aren’t analysing true phone bans - all the pupils still own smartphones. We need a study on schools who have a rigorous brick phone policy - not smartphones in pouches or lockers which are misused (or they put a burner phone in and keep their smartphone). So there’s no data proving a true phone ban doesn’t work.
It’s quite surprising in this article you can’t see that pouches are because of the mere presence effect. They essentially say: own a smartphone that is harmful to you but here’s a special package to put it in. Yonder are a company who jumped onto the phone ban idea as a way of making money for themselves.
What we want and have always wanted is a genuine phone ban and restrictions on ownership of smartphones and social media age limits.
I recommend you read the work of Jared Cooney Horvarth if you haven’t.
From someone on the front line, true phone bans are working. By true phone ban I mean a brick phone policy, discouragement from school about purchasing a smartphone and rigorous bag checks to enforce it.
Thank you for highlighting this: “and that the decline in adolescent mental health is more likely caused by a lack of parental support, economic distress, a decline in parental mental health, or many other possibilities.”
This is never discussed enough. If parents aren’t doing well and there are financial issues in the home, this can have negative effects on adolescents. This article could be useful, https://journals.sagepub.com/doi/abs/10.1177/09637214251410195
Those working on mobile phone bans know that pouches don't make a difference as kids still get access outside school hours (which is when they would mostly use social media anyway). The only time you see an impact, and loads of head teachers agree on this, is if you fully ban phones so that they cannot even be bought onto the school site. Then that changes culture and you see an impact.
I see why there are reasons to be skeptical of phone bans in schools, but the recent NBER paper did find that after the first year, subjective wellbeing went up for the next 2 years. So, even if the phone restrictions didn't affect test scores or other measures, isn't it still a good thing if kids are happier in general?