Forty Out of Forty-Seven: What the World Happiness Report Says Social Media Is Doing to a Generation’s Mind

Forty-seven countries were asked the same question about teenage life. In forty of them, the answer got worse — and if a young person’s evening has felt different these past few years, you may already have guessed why.

You have probably stood in that doorway. The room is dark except for the blue-white wash from a screen, and the face lit by it doesn’t look up when you say goodnight. Maybe it was your own child. Maybe it was you, years ago, propped against the headboard at half past eleven, thumb moving faster than your thoughts. Either way, you have probably sensed the shift this article is about, even without a name for it: something in a young person’s evening has quietly changed shape, and nobody quite agreed to it. You know this feeling. Now it has a number: forty out of forty-seven.

This is the week that feeling got a number attached to it — one big enough that you may already have seen the headline drift past. The World Happiness Report 2026 — the same annual project, backed by Gallup and the UN Sustainable Development Solutions Network, that ranks nations by how content their citizens are — devoted an entire chapter this year to a single question: is social media making adolescents unhappier, at a scale large enough to show up in a country’s overall wellbeing? The chapter’s authors, the social psychologist Jonathan Haidt and researcher Zach Rausch of New York University’s Stern School of Business, conclude that it is. Not as a hunch. As a pattern, repeated across continents, income brackets and methods, that is now difficult to explain any other way.

What forty out of forty-seven countries are trying to tell you

Start with the plainest piece of evidence, because it is also the strangest. The OECD’s PISA survey — the same testing programme that compares how well fifteen-year-olds in different countries can read and do maths — has, since 2015, also asked those teenagers how satisfied they are with their lives. Between 2015 and 2018, that satisfaction fell in forty of the forty-seven countries measured. Not two or three outliers. Forty. Alongside it, loneliness at school rose in thirty-four of thirty-five countries over a similar span, roughly 2012 to 2018. These are not the kind of numbers a single culture, curriculum or economic wobble can explain on its own. Whatever caused a shift this size arrived in most of the world at once, and did most of its arriving between 2012 and 2015 — which happens to be exactly when the front-facing camera, the smartphone, and platforms built around endless, algorithm-fed scrolling became something almost every teenager, in almost every country, carried in a pocket. If you or a child in your life was a teenager anywhere in that window, you were living through the hinge point before anyone had a name for it.

You don’t need to take Haidt and Rausch’s word for the timing. The pattern shows up again in what economists call a “natural experiment” — moments when a new technology rolled out to some places before others, letting researchers compare otherwise similar communities. In England, researchers tracked what happened to teenage wellbeing as broadband coverage expanded street by street, exchange by exchange, through the 2000s. Where broadband arrived, adolescent wellbeing dipped — and the dip was sharpest among the heaviest users, the ones already spending more than five hours a day online. Versions of the same broadband study, run independently in Germany, Italy, Spain and the United States, found the same shape of result: a decline concentrated in young people and, within that group, especially in girls.

Does any of that sound like something you have already half-noticed — in a niece who used to text you paragraphs and now sends a single emoji, in a classroom you know that’s quieter than it used to be for reasons that have nothing to do with discipline?

The numbers behind the worry

Correlation is a cheap thing to find and an expensive thing to trust, so it is worth looking at how much weight these particular correlations can bear. The UK’s Millennium Cohort Study — one of the most carefully tracked groups of children in the world, followed from birth — surveyed 10,904 fourteen-year-olds about their social media use and their mood. Those using it for five hours or more a day were roughly twice as likely to meet the clinical threshold for depressive symptoms as those using it for under an hour. A separate 2022 meta-analysis, pooling results across dozens of studies, found the odds of depression rising by around thirteen per cent for every additional hour of daily social media use. To put that in plain terms: the difference between a teenager who dips in for twenty minutes and one who is on for four hours is not a rounding error. On these figures, it is comparable in size to some of the more established risks to a young person’s mood, including chronic sleep deprivation — which, as it happens, heavy social media use also tends to cause, since 45 per cent of American teenagers in one large survey said it was eating into their sleep. If your teenager is tired all the time and you’ve been blaming exams or a growth spurt, this is worth a second look.

None of that, on its own, proves that the scrolling causes the sadness rather than the other way round — a teenager who is already struggling might simply retreat to their phone, using it as a symptom rather than a cause. If you’ve quietly wondered whether that’s what happened in your own house, you’re asking exactly the right question: it is the single most common objection to this entire field of research, and a fair one. It is also why the next kind of evidence matters more than any survey.

What happens when the phone comes off

If social media were only a mirror, reflecting a sadness that was going to arrive anyway, then taking it away should change nothing. That is testable, and it has been tested — repeatedly, and under proper experimental conditions where people are randomly assigned to cut back rather than choosing to. A 2025 meta-analysis pooling thirty-two randomised controlled trials, covering 5,544 participants, found that reducing social media use produced measurable improvements in depression, anxiety and general wellbeing, with the effect on anxiety the largest of the three. In one of the trials feeding into that pool, 220 distressed young people aged seventeen to twenty-five were asked to cap their use at one hour a day for three weeks; by the end, their depression, anxiety, fear of missing out, and sleep had all improved. Even Meta’s own internal research, in a project given the mundane name “Project Mercury,” found that people who deactivated Facebook for a single week reported feeling less depressed, less anxious and less lonely than before.

It was not a hunch, in other words. It was there when you looked for it, more than once, in data the platforms themselves generated.

If you are the parent who took the phone away for a week and watched a fight erupt that seemed to be about everything except the phone, none of this will feel simple. You were not imagining the fallout, and you were not wrong to worry about it. Removing something a young person has built their social life around is genuinely disruptive in the short term, even when it helps in the medium term — which is exactly why the family agreements later in this piece work better as a conversation than an ambush.

Before you take this to the family WhatsApp group

If you are reading this with a letter from school still on the kitchen counter, or waiting on an assessment for a teenager whose anxiety seemed to arrive alongside a phone, this is the part worth reading most carefully. Here I want to slow down, because hope without honesty helps no one, and this evidence has real limits that deserve saying plainly rather than glossing over. Not every serious researcher in this field agrees with the strength of Haidt and Rausch’s conclusions. Yale’s Christine Cha and colleagues, reviewing the same broad literature, have pushed back on the idea that “more use simply means more harm,” pointing out that what social media use actually looks like — an hour spent messaging close friends versus an hour spent doomscrolling strangers’ outrage — varies enormously from person to person, and that the effects researchers do find tend to be stronger in young people who were already struggling with other things, such as behavioural or emotional difficulties. The U.S. Surgeon General’s own advisory on the subject, while raising real alarm, also acknowledged that social media carries possible benefits for some young people, particularly those who find community online that they cannot find anywhere else — including, often, LGBTQ+ teenagers and those managing conditions like ADHD or autism, for whom online spaces can be where they first feel understood.

So before you treat any of this as a verdict on your own child, or on yourself: most of the studies behind these headlines are observational, meaning they show a strong association rather than an airtight cause; the randomised trials, while genuinely reassuring, tend to be smaller and shorter than we would ideally want; and averages describe a population, not a person — your teenager, or the version of you that used to stay up scrolling, might sit well outside them in either direction. What the evidence supports is not “social media is uniquely evil” but something narrower and, if anything, more useful: that heavy, unstructured, night-displacing use carries a real and now well-documented cost, and that most young people are getting far more of it than either they or the adults around them intended. None of this is a reason to feel you have failed, or that your child is broken. It is a reason to make a few specific changes — which is a very different thing.

What you can actually do this week

None of the research above is much use to you as a set of statistics to worry about. It is only useful as a set of decisions you can actually make — and the encouraging finding buried in all of it is that fairly modest changes produced real, measured improvements in the trials above, not just theoretical ones. So what would you try first, tonight?

• Move the charger out of the bedroom. The single change with the most consistent evidence behind it is keeping phones out of bedrooms overnight — it protects sleep directly, and sleep is doing much of the heavy lifting in the mood research above.
• Delay, rather than ban, the first smartphone. Waiting until secondary school, and starting with a basic phone or a smartphone with social apps switched off, mirrors what several of the “natural experiment” studies suggest reduces exposure during the years the brain is most sensitive to it.
• Write a short family media agreement together, not as a punishment. One page: which apps, which hours, what happens at mealtimes. Young people who help write the rule are far more likely to keep it than those who are simply handed it.
• Use the platforms’ own limits as a starting point, not a solution. Screen Time and Digital Wellbeing tools are blunt instruments, but setting a one-hour daily cap — the same figure used in the youth trial above — is a realistic, evidence-backed place to begin.
• Model the boundary yourself before you enforce it. If your own phone is the last thing you look at at night and the first thing you look at in the morning, say so out loud, and change it alongside them. Credibility here is earned, not assumed.
• Ask about the account, not just the hours. “What’s your feed actually showing you lately?” tends to open a more honest conversation than “how much time did you spend on that.”

Worth a closer look	Probably still healthy
Mood visibly dips after time online; irritability when the phone is removed	Uses it to stay close to specific friends, then puts it down
Sleep consistently pushed later by scrolling	Occasional late night, not a nightly pattern
Withdrawal from activities they used to enjoy in person	Online interests running alongside offline ones, not replacing them
Distress when comparing themselves to images or posts online	Follows accounts that generally leave them feeling neutral or good

You will notice none of this requires confiscating a phone in a rage on a Tuesday evening. It was not there, in any of the trials that actually worked — what was there was a boundary, held calmly, and repeated.

The point of all this

You cannot re-run the last decade and see what a childhood without a smartphone in every pocket would have looked like. That experiment already happened, on you or on your child, without anyone’s informed consent. What you can do is treat the evidence now in front of you as usable information, not another reason to feel guilty at half past eleven at night. Forty countries’ worth of teenagers reported feeling worse over the same few years that a device changed what their evenings were made of. That is not proof that the story is finished. It is closer to a map — imperfect, argued over, still being redrawn — of where the ground gave way. With a few deliberate changes this week, you can start putting some of it back under a young person’s feet.

---

Source: Jonathan Haidt & Zach Rausch, “Social Media and Youth Mental Health,” Chapter 3 of the World Happiness Report 2026, Wellbeing Research Centre, University of Oxford, in partnership with Gallup and the UN Sustainable Development Solutions Network (2026); additional data from the UK Millennium Cohort Study; a 2022 meta-analysis on social media use and depression risk; a 2025 meta-analysis of 32 randomised controlled trials on social media reduction; and commentary from Yale School of Medicine (2026).