Is Social Media a Primary Cause of the Youth Mental-Health Crisis?
Australia's world-first under-16 ban has forced a decade-old research dispute into law: whether social media is driving the rise in adolescent depression and anxiety, or only tracking a crisis whose real causes lie elsewhere.
The dispute the ban made urgent
On 10 December 2025, Australia became the first country to bar children under 16 from holding accounts on the major social platforms. By late June 2026 its government had doubled the maximum penalty for companies that let them through, after regulators watched teenagers defeat age checks with virtual private networks and drawn-on mustaches.1 The statute rests on a specific empirical claim: that the platforms are a cause of the steep rise in adolescent depression and anxiety, and that keeping children off them will help.
That claim is exactly where the science divides, and the division is not about the raw trends, which both sides accept. Roughly 95 percent of American teenagers own a smartphone, and rates of adolescent depression, anxiety, and self-harm have climbed steeply across the same decade the front-facing camera, the like button, and the algorithmic feed became universal.2 The contested step is the inference from the first fact to the second: whether the technology drove the decline, added to it, or simply arrived alongside its real causes. The inference is not academic. If the platforms are a primary cause, a ban is a proportionate response to an emergency. If they mostly mark distress rather than make it, the same ban strips away a channel some isolated teenagers lean on while the real drivers go unaddressed.
One camp, led by the social psychologist Jonathan Haidt and joined by the U.S. Surgeon General, reads the coincidence of timing, a specific mechanism, and a body of experiments as evidence of genuine causal harm, and holds that a public-health emergency justifies acting before the proof is airtight. A second camp, no less credentialed, reads the best-measured effects as vanishingly small, the largest longitudinal studies as pointing the other way, and warns that pinning the blame on screens may divert scarce attention from whatever is really hurting this generation. The disagreement runs between tenured researchers publishing in the same journals, and between two arms of officialdom, all drawing on a shared library of studies and splitting on what those studies can bear.
Jonathan Haidt and the public-health officials
The platforms are a cause, not a coincidence: the timing lines up across countries, controlled studies show harm, and an emergency should not wait for airtight proof.
To the charge that he mistakes correlation for causation, Haidt points at the experiments. His critics, he writes, "cannot say that I am drawing only on correlational studies, or that I have 'no evidence' of causation, or that I do not understand the difference between correlation and causation."3 In the running review he maintains with the researcher Zach Rausch, he counts 22 experimental studies, 16 of which found evidence of harm, or of benefit when subjects quit the platforms, and reads across them "a clear and consistent signal ... that is not consistent with the null hypothesis."3 He counts nine quasi-experiments alongside them, eight finding harm, studies that exploit a platform's staggered real-world arrival rather than a lab assignment. Critics may weigh any one of these against the correlational work, he argues, but not claim the causal designs do not exist.
His mechanism is collective, which is why he thinks studies of single users understate the damage. When an entire cohort moves its childhood onto the platforms, displacing sleep, play, and in-person friendship, the harm falls on the group; a comparison of one heavy user with one light user cannot detect an effect that acts on everyone at once. His second pillar is timing: the curves of adolescent depression, self-harm, and suicide bend upward around 2012 in the United States, Britain, and beyond, in the years when smartphones reached nearly every teenager at once. That the same upturn recurs in countries with different economies and schools is, for Haidt, the fact a purely local explanation cannot absorb.
The case for acting now is carried by the U.S. Surgeon General. Vivek Murthy called the youth mental-health crisis "an emergency, and social media has emerged as an important contributor," and asked Congress for a tobacco-style warning label on the platforms.4 His argument is clinical: in an emergency, he wrote, a doctor acts on the best available evidence rather than wait for certainty, and he cited the finding that teenagers who spend more than three hours a day on social media carry roughly double the risk of anxiety and depression symptoms. The warning label was a demand on lawmakers, not merely advice to parents.5
Jonathan Haidt, a social psychologist at New York University, stands here because The Anxious Generation set the terms of the public argument, and he has defended its causal claim study by study against its critics.3
Candice Odgers, Andrew Przybylski, and the skeptics
The measured effects are small and mostly correlational; where the data follow individuals over time, distress looks more like the cause of heavy use than its consequence.
Reviewing Haidt's book in Nature, Odgers was blunt: "the book's repeated suggestion that digital technologies are rewiring our children's brains and causing an epidemic of mental illness is not supported by science."6 Her reading of the longitudinal evidence reverses his. "When associations over time are found," she wrote, "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."6 That is the reverse-causation objection: the arrow may run from a hurting mind to the feed, and calling the feed the cause mistakes a symptom for a source.
The size of the effect is the second pillar. Applying a method built to test every reasonable model across three datasets covering more than 350,000 people, Amy Orben and Przybylski found the association between digital-technology use and well-being "negative but small, explaining at most 0.4% of the variation," and judged the effects "too small to warrant policy change."7 Two of the field's largest datasets agree: an analysis spanning 72 countries found no consistent link between well-being and the global roll-out of social media, and the Adolescent Brain Cognitive Development study, the biggest long-term look at the developing adolescent brain in the United States, has surfaced no evidence of drastic change tied to digital-technology use.6 This is not one contrarian reading, Odgers stresses: several meta-analyses and systematic reviews, she writes, "converge on the same message."6
Newer data have not overturned that. A 2026 study out of Manchester that followed 25,629 children aged 11 to 14 across three school years found no evidence that heavier social media or gaming use raised anxiety or depression the next year, for boys or girls.8 Its co-author, Neil Humphrey, drew the moral plainly: young people's online choices "may be shaped by how they're feeling, but not necessarily the other way around."9 The skeptics are not defending the platforms. Odgers argues for "considerable reforms," and warns only that age-based bans are "unlikely to be effective in practice, or worse, could backfire."6
Candice Odgers, a psychological scientist at the University of California, Irvine who has tracked adolescents' mental health and technology use for two decades, stands here because she wrote the most cited scientific rebuttal of the causal claim.6
An unsettled fact, and the natural experiment now running
What divides the two camps is not, at bottom, a value. Both want children to flourish, both accept the platforms need reform, and neither disputes that adolescents are suffering. The gap is a fact still out of reach: whether heavy social-media use causes the decline in mental health or merely travels alongside it. The disagreement is sharpest over which studies could settle that. Haidt's experiments can assign who uses what, but they are small and run for weeks; the skeptics' null results come from cohorts of tens of thousands, but those cohorts cannot randomize the exposure. Reviewing the same shelf of papers, the National Academies of Sciences "did not support the conclusion that social media causes changes in adolescent health at the population level," and held that the field's task was to run the causal tests it lacks.2
That kind of test is now running. Australia's ban is the largest natural experiment the question has had: a whole country's under-16s pushed off the platforms on a single date, with mental-health data gathered on both sides of it. If depression and anxiety fall in the cohorts raised under the ban, the causal reading gains the evidence it has lacked; if the numbers hold steady, the correlational reading survives its hardest test.1 The early signs are muddy. A study of 400 children found little change in how much they actually used the platforms and judged the ban "unlikely" to produce its intended benefits, and a large majority of under-16s were still logging in months after the ban began.10 A natural experiment answers the question only if the exposure truly changes, and randomized deactivation trials, in which volunteers are paid to leave a platform for weeks, have so far returned results mixed enough to feed both readings. Australia's first banned cohorts will not yield trustworthy before-and-after mental-health data for a year or more, which is why, for now, both readings remain standing. The verdict, when it comes, will be written in the data out of Canberra and in those trials, not in one more reading of the studies both sides already hold.1
Sources
- Semafor · Australia toughens social media ban (30 June 2026)
- National Academies of Sciences, Engineering, and Medicine · Social Media and Adolescent Health, Consensus Study Report Highlights (December 2023)
- Jonathan Haidt, After Babel · Yes, Social Media Really Is a Cause of the Epidemic of Teenage Mental Illness (2024)
- Vivek H. Murthy, The New York Times (Opinion) · Why I'm Calling for a Warning Label on Social Media Platforms (17 June 2024)
- NPR · U.S. surgeon general calls for tobacco-style warning labels for social media (17 June 2024)
- Candice L. Odgers, Nature · The great rewiring: is social media really behind an epidemic of teenage mental illness? (29 March 2024)
- Amy Orben & Andrew K. Przybylski, Nature Human Behaviour · The association between adolescent well-being and digital technology use (2019)
- Cheng, Panayiotou, Finserås, Andersen & Humphrey, Journal of Public Health · How do social media use, gaming frequency, and internalizing symptoms predict each other over time in early-to-middle adolescence? (March 2026)
- University of Manchester · Time spent on gaming and social media not to blame for teen mental health issues (2026)
- CNBC · Australia banned social media for under 16s — here's how it's going (15 January 2026)