Jonathan Haidt's 2024 book The Anxious Generation: How the Great Rewiring of Childhood is Causing an Epidemic of Mental Illness became an instant bestseller and a significant force in global policy debates. Haidt argues that a shift from a "play-based childhood" to a "phone-based childhood" — occurring roughly 2010–2015 — is the primary cause of a "crisis" of anxiety, depression, self-harm, and suicide among adolescents, particularly girls.
The book catalyzed legislative action across multiple countries, including Australia's ban on social media for under-16s and a wave of US state-level restrictions. Politicians cite it. Parents share it. School boards act on it.
But there is a serious problem. Researchers who have spent their entire careers studying adolescent mental health, children's digital media, developmental science, and media psychology — the people who actually built the evidence base Haidt draws on — have raised sustained, substantive objections to his core claims. Relying on those claims is NOT evidence-based policymaking and ignores what the science says (and doesn't say) about kids and social media.
This report presents those critiques. It begins by examining Haidt's own credentials and what they reveal about how his book was constructed. It then features the arguments of six world renowned and peer-reviewed domain experts whose work directly challenges his thesis. A Reading List and full Bibliography follow.
Jonathan Haidt is a social psychologist at NYU Stern whose academic record is built in moral psychology, moral emotions, and political psychology. His previous books — The Righteous Mind (2012) and The Coddling of the American Mind (2018) — were also published by non-academic presses, and are examinations of moral and social philosophy, not developmental science or media studies. Like The Anxious Generation, his published articles on social media, teens and mental health consist of reviews of existing research.
His arguments rely heavily on Jean Twenge's controversial work and on a set of rolling literature reviews compiled with his lead researcher Zach Rausch. His critics characterize this method of tallying studies regardless of quality as "vote counting" — an approach widely regarded as producing unreliable conclusions.
"It is true that I am promoting a social change program…and I am doing this before the scientific community has reached full agreement."
"It's a collision between a knowledge-discovery process, which is necessarily messy, and a narrative-generation process, which is unencumbered by critique."
✦ Moral psychology, intuition and emotions
✦ Political psychology & polarization
✦ Business ethics
✦ Not: adolescent development
✦ Not: media effects or screen time
✦ Not: child psychology or pediatrics
Holden Thorp, editor-in-chief of Science, questioned publicly whether a scholar with Haidt's public influence had a special obligation to communicate that the science he's describing is in fact inconclusive. When he asked Haidt if he wished he'd done more to highlight the disagreements, he said, "I don't think so."
Australia has banned under-16s from social media. Dozens of US states have passed restrictions, e.g., Florida and Utah. The UK and Canada are debating similar moves. Much of this legislation cites or is influenced by a popular non-fiction book about a hot button topic, whose core claims are widely disputed by the actual experts conducting the research on social media's impacts.
How about those bans? A University of Chicago working paper (Bursztyn et al., April 2026) surveyed 746 Australian teenagers four months after the ban — and found that only 1 in 4 teens aged 14–15 comply. Most believe their peers are still using social media platforms and report that a large majority (roughly two-thirds of peers) would need to quit before they would quit themselves. Compliant teens are seen as less popular. The researchers conclude compliance is more likely to diminish than rise over time.
Haidt's previous books were about exactly this kind of situation. In The Righteous Mind (2012), he showed how people use evidence to justify intuitions and strong emotions that they already have. In The Coddling of the American Mind (2018), he warned that protecting children by removing everything potentially risky or scary backfires because it produces more anxiety, not less.
His critics argue he is now doing both.
Haidt's evidence base is housed in a self-curated Google Doc. The document itself states plainly: "We have not done an exhaustive search of citation databases. We begin instead with articles published in or after 2014 that are being cited by scholars on either side of the debate."
The book's policy prescriptions target children and teenagers. Yet going through every experimental study in the Google Doc (as reviewed 27 April 2026), almost none involve actual children or teenagers. The studies Haidt counts as his experimental evidence are overwhelmingly conducted on college students and adults aged 18–35.
Several of the studies cited as providing conclusive evidence of "causality" involve very small samples:
A 2025 study in the Journal of Children and Media analysed 136 online news articles about screen time research to understand the types of stories social media users are most likely to share/repost. The findings describe why and how content like The Anxious Generation is much more likely to go viral:
Neuhaus & O'Connor (2025), Journal of Children and Media, 19:3, 619–633
Haidt's book combines alarming language with an overt scientific framing, focuses on children and teens, emphasizes negative effects, and limits its analysis to studies "that are being cited by scholars on either side of the debate." It fits the formula that Neuhaus & O'Connor find maximises social media reach, independent of underlying scientific merit. It also applies the theory of moral persuasion advanced in Haidt's previous books.
Candice Odgers has spent decades studying child and adolescent mental health, risk, and resilience. She is a Canadian researcher trained at Simon Fraser, leading the Child & Brain Development Program at CIFAR, and the Connecting the EdTech Research Ecosystem (CERES) network for the Jacobs Foundation. She is, unambiguously, an expert on children's brains and mental health. In a landmark review in Nature (2024), she delivered the most devastating single scholarly critique of The Anxious Generation.
Her central argument: the book's core claim is not supported by the science. The overwhelming majority of evidence Haidt draws on is correlational (i.e., when two things happen at the same time). It cannot establish causation (i.e., when one thing causes the other). Critically, correlations are sometimes coincidences. In this case, the causal arrow may in face run in the opposite direction, meaning that vulnerable young people may seek out social media more (vs. social media causing the vulnerability).
"Hundreds of researchers, myself included, have searched for the kind of large effects suggested by Haidt. Our efforts have produced a mix of no, small, and mixed associations. Most data are correlative. When associations over time are found, they suggest not that social-media use predicts or causes depression, but that young people who already have mental-health problems use such platforms more often or in different ways from their healthy peers."
"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."
"Haidt...is a gifted storyteller, but his tale is currently one searching for evidence."
Odgers cites the Adolescent Brain Cognitive Development (ABCD) study — the largest long-term study of adolescent brain development in the US — which found no evidence of drastic changes linked to digital-technology use. She also cited a 72-country analysis (Vuorre & Przybylski, 2023) showing no consistent global association between social media rollout and well-being changes.
Smartphones weren't the only thing that changed in kids' lives between 2010 and 2024. Odgers' point is precisely this: when multiple major changes happen simultaneously, picking one and calling it the cause requires evidence, not just a timeline. Here are some of the other things that tracked the same period — all also correlated with youth distress in the research literature:
Amy Orben (MRC Cognition and Brain Sciences Unit, Cambridge) and Andrew Przybylski (Oxford Internet Institute) are specialists in computational approaches to studying human behaviour and technology effects. Together they have produced some of the most technically rigorous large-scale studies of screen time and adolescent well-being research ever conducted, analysing data from 355,000 participants.
In their landmark 2019 paper in Nature Human Behaviour, they applied specification curve analysis — a technique that systematically tests all defensible analytical approaches to a dataset — across three large-scale datasets. The result: digital technology use explained at most 0.4% of the variation in adolescent well-being. For context: wearing glasses and eating potatoes showed similarly sized associations in these same datasets. It could even be the result of a rounding error.
"Extraordinary claims require extraordinary evidence. Right now, I'd argue [Haidt] doesn't have that."
"It doesn't matter if there's 19 studies in one direction and three in another, if those three studies are done well and have a large number of participants."
Przybylski has described Haidt's approach as focused on quantity over quality, which looks a lot like "vote counting." This is when a review of the literature tallies the number of studies pointing one direction against another, regardless of the methods used, the age or number of people involved, or the other crucial things one should consider when assessing the validity and reliability of a study's findings. A few well-designed, large-sample studies finding no effect are more evidentially powerful than many small, methodologically weak studies finding harm. In their view, Haidt's approach systematically overstates effect sizes by ignoring these crucial details.
Orben and Przybylski have also raised concerns about relying too heavily on self-reporting for measuring usage. Study participants consistently over- and under-estimate their social media usage. When researchers instead look at device-level log data they find substantially weaker associations than those relying on self-reports.
Christopher Ferguson situates The Anxious Generation within a well-documented historical pattern of moral panics about new media — a pattern that has, in every previous instance, been eventually deflated by evidence. He was one of the researchers Haidt consulted before publication.
"My advice to him was to not publish it. Look at kids today — they're less violent, they smoke less, they use drugs less. All that happened during the social media smartphone age… We're probably in the midst of another moral panic."
In the 1950s, Fredric Wertham's bestseller Seduction of the Innocent claimed comic books caused juvenile delinquency. Congressional hearings followed. Then it was rock music in the 1960s. Then D&D in the 1980s. Then violent video games in the 1990s. In each case, a small number of persuasive voices drove public alarm. In each case, large numbers of experts disagreed. Tellingly, none of those previous examples was ultimately found to cause large-scale harm and the proposed regulations were abandoned as soon as the moral panic died down.
"Perhaps the best historical parallel to Haidt's 'anxious generation' is Seduction of the Innocent… People took all this seriously: a US congressional inquiry was even launched into the comic book industry. Wertham went on to appear before a Senate subcommittee, where he compared the comic book industry to Hitler."
Ferguson also raises a crucial point that focusing on social media displaces attention and action from more powerful predictors of youth distress. In his 2025 peer-reviewed analysis, Ferguson finds youth suicide patterns correlate with multiple environmental factors including income inequality, family structure, and adult suicide trends. He warns that attributing these to any single cause risks ecological fallacy. Ferguson highlights that Haidt's trend graphs do not control for these confounds.
One of Haidt's most rhetorically powerful arguments is the international scope of the crisis. If social media is causing the problem, he argues, we should see the same patterns wherever smartphones spread. But the most comprehensive global analysis ever conducted tells a different story.
In a large-scale study spanning 72 countries, Vuorre and Przybylski (2023) found no consistent global association between internet or social media adoption and changes in well-being. This is the most geographically comprehensive and systematic analysis of the question conducted to date.
If social media were the cause of rising distress among youth, this should appear in every country where smartphones + social media spread rapidly. The 72-country data shows no such consistent pattern.
"It's a collision between a knowledge-discovery process, which is necessarily messy, and a narrative-generation process, which is unencumbered by critique."
Critics have noted that Haidt draws heavily on American data, the country with the most marked trends, while citing international comparisons selectively. The specific patterns most prominent in his analysis may reflect domestic American factors: the opioid crisis, racial inequality, gun violence and school safety culture, and healthcare access, rather than a universal effect of the phone-based childhood. Large cross-national analyses (notably Vuorre & Przybylski's 72-country Facebook adoption study) find no consistent global pattern of youth well-being decline tied to the timing of platform rollout. In this and other studies international comparisons show several high-technology countries outside the US with no comparable youth suicide trend during the social media era.
Critics observe that Anxious Generation embraces cross-national similarity as proof of a universal cause where and when the data supports that claim, but does not adequately account for the many countries where the pattern does not appear. The evidentiary standard is asymmetric.
Patti Valkenburg is Distinguished University Professor and founding director of the Center for Research on Children, Adolescents, and the Media (CCAM) at the University of Amsterdam. She is one of the world's most cited researchers on media and child development. Her work using intensive longitudinal and within-person research designs produces some of the most methodologically sophisticated evidence available on how social media actually affects different young people.
Her core finding challenges the very logic of Haidt's universal prescriptions: social media effects are highly variable across individuals. Most young people show little to no relationship between social media use and well-being. A minority show negative effects. Some show positive effects.
"What has often been ignored in [media effects] debates is that the effect sizes are just what they are: statistics observed at the aggregate level. Such statistics are typically derived from heterogeneous samples of adolescents who may differ greatly in their susceptibilities to the effects of environmental influences in general [53] and media influences in particular."
"A person-specific approach [can] explain why social media browsing leads to increases in well-being among some adolescents and decreases in well-being among others. It shows, for example, that adolescents more often experienced a negative effect of browsing on well-being when they felt envy during browsing, and that they more often experienced a positive effect when they enjoyed their social media experience.."
The studies Haidt draws on are predominantly between-person and cross-sectional: comparing heavy social media users to light users at one point in time. Valkenburg uses within-person designs — tracking how the same individual's social media use relates to their own well-being over time. These designs are far better at isolating actual effects, and they consistently find smaller, more variable, and more context-dependent associations than Haidt's literature suggests.
In a series of person-specific studies, Valkenburg et al. found that the "passive social media use harms well-being" hypothesis was rejected for 80% of adolescents in within-person designs. For most young people, social media use simply does not affect their well-being in the ways Haidt claims.
If effects are highly heterogeneous, uniform age-based restrictions remove a resource that benefits many young people — including those who rely on online communities for support unavailable offline — in order to protect a vulnerable minority who may, in any case, face their greatest risks from pre-existing conditions rather than the platforms themselves.
Sonia Livingstone OBE FBA is Professor of Social Psychology at LSE, founding director of Digital Futures for Children, recent director of the Global Kids Online project (with UNICEF), and one of the world's foremost authorities on children's digital lives. She has advised the UN Committee on the Rights of the Child, the European Commission, UNICEF, and the OECD, and is now appointed to the UN's inaugural Independent International Scientific Panel on Artificial Intelligence. She has published 20 books on media, children, and digital rights.
"I think the argument for a ban is an admission of failure that we cannot regulate companies, so we can only restrict children."
"It's the 15 years in which we don't let our children go outside and meet their friends. It's the 15 years in which we stopped funding parks and youth clubs for them to meet in. So a ban now is to say to children: 'We can't make the regulation work. We can't update it fast enough. We haven't built you anything else to do, but that's just tough. We've terrified your parents into feeling that there's nothing they can do, and we're going to take you away from the service where you hoped you would feel some sociability and entertainment.'"
Livingstone co-authored a 2026 study examining school smartphone bans across multiple European countries, finding they had little impact on pupils' mental wellbeing or quality of life. Her 2025 paper in the British Journal of Developmental Psychology critiques the "ages and stages" model underlying current regulatory frameworks as theoretically dated. Her stance is that children's digital needs vary enormously within age groups, making rigid age thresholds both over-inclusive and under-inclusive.
Grounding her argument in the UN Convention on the Rights of the Child, Livingstone argues that age bans collapse protection rights and participation rights entirely on the side of protection — failing to account for genuine benefits of digital participation, especially for LGBTQ+ youth, young people with disabilities, and those in isolated communities whose primary support may be online.
The scholarly critique of The Anxious Generation does not come from a single source but a from an array of distinctive and intersecting arguments advanced by researchers with deep domain expertise across adolescent psychology, media studies, developmental science, public health, and children's rights. These world-renowned, peer reviewed, and widely respected experts raise multiple and substantive issues with how the evidence was selected, analyzed, and represented. These issues become especially problematic when they are being used to drive major legislative change affecting millions of children.
The critiques converge on several key points. The empirical foundation for a social media age ban is shaky at best. Effects are small, correlational, context-dependent, and highly variable across and even among individuals. The international data do not consistently support a universal narrative of social-media-driven decline in youth mental health, which is also inconsistently and imprecisely defined and measured. Meanwhile, blanket age bans are poorly calibrated to the heterogeneity of adolescent experience and risk, infringe on children's rights, and may especially harm the most vulnerable young people most.
The framing of social media as the primary cause of a youth mental health crisis risks displacing attention and resources from structural factors including but not limited to inequality, family disruption, educational pressure, climate anxiety, and deteriorating mental health infrastructure, for which the evidence of causal contribution is considerably stronger. The proposed social media ban would essentially enroll Canadians in an unproven, population-wide social experiment without required participant consent or opt out.
None of this establishes that social media is teleologically safe or beneficial for all children. It does however, establish that the question is substantially more complex than the current discourse allows. A more defensible policy framework would centre platform design accountability, age-appropriate design codes, genuine enforcement of existing protections, targeted supports for vulnerable youth, and the active involvement of children and young people themselves in shaping the digital environments they inhabit.