At a kitchen table in almost any city, a version of the same scene is playing out tonight. A parent is speaking; a teenager is nodding; a phone lies face-up beside the plate, lighting briefly every few minutes with a message from someone not in the room. Neither party is being rude. The parent is worried, and the teenager is tired of being the subject of a worry. Between them sits a question that has accumulated enormous emotional weight over the past decade: is the glowing rectangle the reason young people are less happy than they used to be?
For years, that question was argued largely with anecdote and alarm. In March 2026, the World Happiness Report—the annual publication of the Wellbeing Research Centre at the University of Oxford, Gallup, and the UN Sustainable Development Solutions Network—devoted its entire thematic section to answering it. Nine chapters, drawing on data from dozens of countries and on researchers who disagree with one another, set out to ask whether the hypothesis survives rigorous analysis. The answer, as is often the case when a serious question meets serious evidence, is more interesting than either the panic or the dismissal.
This article walks through what the report found: where youth wellbeing has fallen and where it has not, what the evidence says about heavy use, why platform design and a sense of belonging seem to matter at least as much as screen hours, and where the science remains still uncertain. It also draws on older traditions—Indian, Buddhist, African—that have long thought about attention, craving, and community, because the questions the report raises are not new. They are ancient questions in a new medium.
The report opens from an observation that has been troubling researchers for some time. Young people in North America and Western Europe are considerably less happy than they were 15 years ago, and over the same period social media use has risen dramatically. Many people have drawn the obvious connection. The report’s first task was to test whether the pattern is as clean as the connection assumes.
The geography is the first complication. Using Gallup World Poll data for 136 countries, the report’s international-evidence chapter, led by John Helliwell, Lara Aknin, and colleagues, found that in 85 of those countries, people under 25 are happier now than they were between 2006 and 2010. Youth wellbeing fell, both in absolute terms and relative to adults, only in two groups of countries: Western Europe, and the four English-speaking nations of the United States, Canada, Australia, and New Zealand. In that second group, life evaluations among the under-25s fell by more than 0.8 of a point on the 0–10 scale—a very large movement by the standards of wellbeing research, where national averages typically shift by hundredths of a point from year to year. In Latin America, by contrast, young people’s life evaluations rose.
The pattern is hard to reconcile with a simple story in which smartphones and social platforms uniformly drain the happiness of those who use them. By 2022, cellphone internet access and social media use were nearly universal among young people worldwide, and usage patterns were very similar across regions. If screens alone were the cause, the decline should have been global. It was not. One further detail from the report is worth holding on to: in New Zealand, Australia, and North America, young women now report lower life evaluations than young men, and this is the only region in the world where that is true.
The wider rankings supply context. Finland was named the world’s happiest country for a record ninth consecutive year, with an average score of 7.764 out of 10; Iceland, Denmark, and Costa Rica followed, with Costa Rica’s fourth place the highest ever achieved by a Latin American country. No English-speaking country appears in the top 10 for the second year running: New Zealand is 11th, Ireland 13th, Australia 15th, the United States 23rd, Canada 25th, and the United Kingdom 29th. The countries where youth wellbeing has fallen furthest are, in other words, mostly the wealthy ones where social media use is also heavy—a correlation the report treats as a clue, not a conclusion.
None of this means the concern is misplaced. The report’s strongest evidence points to real harm in specific conditions, and it is worth stating clearly.
The chapter on population-scale harm, by Jonathan Haidt and Zach Rausch, assesses seven lines of evidence: surveys of young people, surveys of parents, teachers, and clinicians, corporate documents, cross-sectional studies, longitudinal studies, reduction experiments, and natural experiments. The authors conclude that the evidence of direct harms, such as cyberbullying and sextortion, is overwhelming, and that the evidence of indirect harms, such as depression and anxiety, is compelling. They argue that the adoption of social media in the early 2010s contributed substantially to rising mental illness among adolescents in many Western nations. It is the most forceful statement in the report, and it comes with an explicit policy implication: that platforms should be treated like other consumer products and held to standards of safety.
The international data support a more cautious version of the same claim. In a survey of 15-year-olds in 47 countries, conducted through the OECD’s PISA assessment, life satisfaction was highest among those who used social media lightly and lower among those who used it heavily. Social media, gaming, and browsing for fun were all linked to lower life satisfaction at higher levels of use, while communication, news, learning, and content creation were linked to higher satisfaction at moderate levels. A chapter on gender differences, drawing on more than 270,000 students aged 15 and 16, found that among girls life satisfaction was highest for light users—under an hour a day—and declined as use increased. Among boys, the same pattern appeared mainly in Western Europe and English-speaking countries. As Oxford’s summary of the report puts it, young people who use social media for under an hour a day report the highest wellbeing, higher even than those who do not use it at all.
A further chapter used an unusual method. Looking at 30 European countries between 2016 and 2024, researchers used local internet speed as an instrument to separate cause from correlation, and found a significant negative effect of internet use on wellbeing—one that standard statistical approaches had missed. The effect, the chapter reports, was strongly negative for Generation Z, moderately negative for Millennials, close to zero for Generation X, and slightly positive for Baby Boomers. It is important to note that this chapter concerns internet use in general, not social media alone, but the generational gradient is striking, and the declines in trust and social activity were largest among Gen Z and Millennial women.
Another chapter turns to behavioural economics and finds something quietly unsettling: a pattern of what it calls product traps. Several studies suggest that many people use social media largely because others do. Users would pay little to keep using a platform but demand far more to give it up, and young people said they would pay to have platforms such as Instagram and TikTok removed from their communities altogether—a “negative non-user externality,” in the report’s phrase. In a study of Facebook, users were more anxious and depressed while using it, even after a happier month without it. The chapter’s conclusion is that many users would be better off if these platforms did not exist. That is not the claim that a screen harms a child; it is the claim that a social arrangement can trap people into an activity they would collectively choose to abandon.
A responsible reading must also take seriously what the report itself says about the limits of the evidence, and what earlier research has found.
The report’s executive summary is explicit: social media is not the sole explanation for the decline in youth wellbeing, and heavy use “provides an important part of the explanation,” particularly in some countries. Much of the evidence it reviews is correlational. The authors of the international-evidence chapter note that associations may overstate causal effects because of feedback loops: a young person who is already unhappy may turn to social media more, and heavier use may deepen the unhappiness in turn. The usage measures in the PISA data are an index, not an exact count of hours. And the amounts of variation explained are small. The chapter calculates that higher US usage accounts for only about eight per cent of the larger decline in American girls’ life satisfaction, and about five per cent for boys. In the same data, the correlation between hours of social media use and life satisfaction is -0.084 for girls and -0.019 for boys—weak by any standard.
Those figures echo a landmark 2019 study by Amy Orben and Andrew Przybylski, published in Nature Human Behaviour. Analysing data on more than 300,000 adolescents and parents in the United Kingdom and United States, they found that technology use explained at most 0.4 per cent of the variation in adolescent wellbeing—an association they likened, with some wit, to the negative association of regularly eating potatoes. Factors such as being bullied, they noted, were associated with several times more harm, and getting enough sleep and eating breakfast had stronger associations with wellbeing than technology use did. Their approach, specification curve analysis, tested a large sample of the very many reasonable ways one might analyse the same data, and produced positive, negative, and null associations alike.
The two bodies of work are not as contradictory as they first appear. Average effects across a large population can be small while effects on a vulnerable subgroup—girls in the middle of adolescence, say, or those with problematic patterns of use—are large. The 2026 report’s own findings show that the association between problematic social media use and wellbeing was present in all 43 countries examined, was stronger for adolescents from lower-income families, and intensified between 2018 and 2022. The sensible conclusion is not that one camp is right and the other wrong, but that “social media” is too coarse a category to carry a single causal verdict. The report’s chapter on evidence synthesis adds a sobering note: three high-profile expert reports on this very topic cited similar kinds of research but shared fewer than one per cent of their sources, and differed in their accuracy of citation, acknowledgement of limitations, and strength of conclusions. Reasonable people reading the same literature can arrive at very different places, and the public deserves to know it.
Perhaps the most practically useful finding in the report is also the least dramatic. When the authors set the effect of screen time beside the effect of something else, the something else won by a wide margin.
In the PISA data, the correlation between a sense of school belonging and life satisfaction is +0.339 for girls and +0.271 for boys, compared with -0.084 and -0.019 for social media hours. Put another way, for girls in the United Kingdom and Ireland, moving from low to high belonging at school raises life satisfaction four times as much as moving from high to low social media use; across the full 47-country sample, the ratio is six to one. Oxford’s summary of the report states the general point simply: social connections and a sense of belonging are associated with much larger changes in how people feel about their lives than social media use is.
This reframing carries a quietly important implication. If a young person’s difficulty is loneliness, exclusion, or the absence of a place where they are known, reducing screen time may help a little but will not address the underlying lack. Conversely, a young person who is deeply embedded in friendships, family, and community may use a phone heavily without apparent harm. The finding aligns with decades of research on relationships and happiness—from the Harvard Study of Adult Development, which has found that the quality of close relationships predicts later-life health and wellbeing more reliably than wealth or status, to Julianne Holt-Lunstad’s meta-analysis linking social connection to survival. Social media harms young people, where it does, partly by displacing and degrading the very thing that protects them.
The report’s closing remark from Jan-Emmanuel De Neve, director of the Wellbeing Research Centre, makes the same point from the design side: “we should look as much as possible to put the ‘social’ back into social media.”
Treating social media as a single thing hides what may be the most useful distinction of all. The report suggests that platform design matters.
In Latin America, analysts examined data from Mexico and from 17 countries in the Latinobarómetro survey. They distinguished between platforms built primarily for social connection, such as WhatsApp and Facebook, and platforms built around algorithmically curated content, such as Instagram, TikTok, and X. The connection platforms were linked to higher life satisfaction and more positive emotion; the algorithmic platforms were linked to lower happiness and more negative affect and mental health problems at high levels of use. In 11 of the 17 countries, the connection platforms scored higher than the algorithmic ones. And Latin America combines heavy social media use with relatively high youth wellbeing, which the authors read as evidence that design and social context matter.
The Middle East and North Africa offer a similar lesson. Between 20 and 40 per cent of users in the region reported more than five hours of daily use in 2023 and 2024, and heavy use was linked to lower wellbeing, higher stress and depressive symptoms, and more unfavourable comparisons with parents. Harm was greater among those who used multiple platforms, relied on social media for news, or followed influencers. Yet youth wellbeing across the region has not declined despite the heavy use. Whatever is eroding young people’s happiness in the English-speaking world, it is not simply the number of hours spent on a screen. It may be something about what is on the screen, what the screen replaces, and what surrounds the child when the screen is off.
One more finding deserves mention: the harm from internet use, one chapter suggests, depends on one’s peers. Internet use appears beneficial when few of one’s peers use social media heavily, and increasingly harmful as such use becomes common. That is the signature of a collective-action problem, and it helps explain why individual abstinence is so difficult and why the report’s editors take policy seriously.
The policy response has already begun. In December 2025, Australia raised the minimum age for 10 major platforms from 13 to 16, and Denmark, France, and Spain are planning similar rules. The report’s editors state that the evidence should inform such policies. But they add a caution that deserves attention: because the quality of evidence synthesis varies so widely, there is a risk that policy built on poorly synthesised research will be ineffective or even harmful. A rule that removes a young person from a platform without replacing what the platform provided—friendship, information, a sense of identity—may fail to help.
The more promising directions, as the report’s findings suggest, concern design as much as access: reducing the pull of algorithmic feeds, building features that foster real connection, and strengthening the communities, schools, and families into which young people belong. These are matters for governments and companies, and also for ordinary households, where the most useful interventions are modest.
The report’s findings read, in places, like a modern gloss on very old teachings.
The Bhagavad Gita, in its second chapter, offers one of the most precise descriptions of compulsive attention in any literature. One who dwells on objects of the senses, it says, develops attachment to them; from attachment springs desire; from desire, anger; from anger, delusion; from delusion, the loss of memory and the ruin of discernment. The sequence is a psychological map of the scroll—the lingering, the craving for the next item, the irritation when interrupted, the muddle that follows hours of fragmented attention. The remedy the Gita proposes in its sixth chapter is neither asceticism nor abandon but moderation: yoga, it says, is not for one who eats too much or too little, sleeps too much or too little, but for one who is balanced in food, recreation, effort, and rest. This is, in effect, the report’s finding that low and moderate use is associated with the highest wellbeing.
Patanjali’s Yoga Sutras name a related practice, pratyahara, the withdrawal of the senses from their objects, as a discipline in its own right. It is not suppression but the cultivated ability to choose where attention goes. And the Buddhist teaching of the three poisons—craving, aversion, and delusion—describes what product designers call engagement: the engineered cultivation of craving. These traditions did not think the answer was to be rid of the world. They thought it was to regain sovereignty over one’s attention.
They also thought, with equal emphasis, that the self is formed in community. The Southern African principle of Ubuntu—”I am because we are”—and the Sanskrit Vasudhaiva Kutumbakam, the idea that the world is one family, both place belonging at the centre of a good life, and the Indian practice of satsang, the company of the good, treats the circle of one’s companions as a primary determinant of character. The report’s discovery that belonging outweighs screen time by a factor of four to six would not have surprised the people who wrote those texts.
What follows from all this for those living with the question every day? Several modest, evidence-aligned practices emerge.
For young people themselves, the evidence suggests paying attention to how rather than only how much. Communication, learning, and creating appear to be associated with better outcomes at moderate levels than passive scrolling, gaming, and browsing for fun. Light use appears to be associated with the highest wellbeing, and heavy use, especially more than five hours a day, with the lowest. It is worth noticing, over a week, which uses leave one feeling more connected and which leave one feeling smaller. The Greater Good Science Center’s research on gratitude and connection offers a complementary habit: a daily note of one person one is thankful for tends to shift attention from comparison towards relationship.
For parents, the data counsel against making screens the sole battleground. A conversation that asks about belonging—who at school makes you feel at home, where do you feel known—addresses the factor the report found to matter most. Shared meals without phones, sleep protected from devices, and a household where adults visibly manage their own attention probably do more than any filter. The evidence that being bullied and lacking sleep are associated with far larger harms than technology use itself, from the earlier Oxford work, reinforces the point.
For schools and teachers, the finding that belonging outweighs screen use is an invitation to invest in the unglamorous work of making every pupil feel part of something: clubs, mentoring, cooperative projects, and a culture in which exclusion is taken seriously. Emotional wellbeing education that teaches young people to understand their own attention, emotions, and relationships can be part of that effort, and it is a better-evidenced investment than a blanket ban.
If belonging matters more than screen hours, then the places where young people learn together matter a great deal. That is part of the reason the Rekhi Foundation for Happiness works through universities and institutions. Its Science of Happiness course, offered with more than 90 institutions across six countries, blends scientific insights with practical exercises, giving students a shared vocabulary and a community in which to practise attention, gratitude, and connection—the very capacities that the World Happiness Report suggests are worth protecting.
According to the World Happiness Report 2026, it is part of the story but not the whole of it. Youth wellbeing has fallen sharply only in Western Europe and in the United States, Canada, Australia, and New Zealand, where under-25 life evaluations dropped by more than 0.8 of a point on a 10-point scale; in 85 of 136 countries, young people are happier than they were around 2006 to 2010. The report finds that heavy use is linked to lower wellbeing, especially for girls, and that platforms built around algorithmic feeds show a negative link at high use. But the effects depend on platform design, how it is used, and social context, and the evidence is largely correlational. The report's editors conclude that social media is not the sole cause, but heavy use "provides an important part of the explanation." For anyone interested in happiness research, the lesson is to ask which uses, for whom, and in what circumstances.
The report offers a pattern, not a precise threshold. In the PISA data covering 15-year-olds across 47 countries, life satisfaction was highest among light users and declined as use increased; among girls, those using social media for under an hour a day reported the highest satisfaction. Oxford's summary notes that young people who use social media for under an hour a day report higher wellbeing than those who do not use it at all. In the Middle East and North Africa, use of more than five hours a day was linked to lower wellbeing, higher stress, and more depressive symptoms. Adolescent use averages about 2.5 hours a day according to one estimate. These are associations, not prescriptions: the type of use matters, and communication, learning, and creating tend to look better than passive scrolling. A sensible approach for families is to focus on the quality of use, protect sleep, and watch how a young person feels afterwards.
Yes, according to several chapters in the report. In an analysis of more than 270,000 students aged 15 and 16, girls' life satisfaction was highest among light users and fell as use rose, while among boys the same pattern appeared mainly in Western Europe and English-speaking countries. The decline with heavy use was steeper for girls than boys in all regions. In the United States, Canada, Australia, and New Zealand, young women now report lower life evaluations than young men, the only region where that is true. A chapter on internet use in 30 European countries found that declines in trust and social activity were largest among Gen Z and Millennial women. The report cautions that these are associations, and that school belonging mattered more than screen hours for girls in several samples. For parents and educators, emotional wellbeing education and a strong sense of belonging are therefore as important as limits on devices.
The report points to belonging and social connection. In the PISA data, school belonging correlated with life satisfaction at +0.339 for girls and +0.271 for boys, while social media hours correlated at -0.084 and -0.019. For girls in the United Kingdom and Ireland, moving from low to high belonging raises life satisfaction four times as much as moving from high to low social media use; across 47 countries the ratio is six to one. This fits decades of research, from the Harvard Study of Adult Development to Julianne Holt-Lunstad's work on social connection, showing that relationships are among the strongest predictors of wellbeing. For families and schools, the practical message is to invest in friendship, mentoring, shared activities, and a culture of inclusion. The Rekhi Foundation for Happiness builds this into its Science of Happiness course, which pairs research with practice in a community setting.
The report does not issue a verdict, but it frames the choice carefully. In December 2025, Australia raised the minimum age for 10 platforms from 13 to 16, and Denmark, France, and Spain are planning similar rules. The editors say the evidence should inform such policies, while a chapter on evidence synthesis warns that three high-profile expert reports shared under one per cent of their sources and differed in the accuracy and strength of their conclusions, so poorly synthesised evidence could lead to ineffective or harmful policy. Other chapters suggest that platform design matters, with connection-focused platforms linked to better outcomes than algorithmic feeds at high use, and that peer effects make individual abstinence hard. The likeliest productive path combines age limits where justified with safer design, support for belonging, and careful evaluation of results. For those studying human flourishing, the debate is a reminder that technology shapes wellbeing through the communities it either builds or erodes.
The phone was never the whole story. A young person who is known, welcomed, and needed can hold a screen lightly; the work of our time is to build the rooms in which that is possible.Discover the science behind a flourishing life at rekhifoundation.com.
Rekhi Foundation, founded in 2016, promotes Happiness Science via university centers, collaborating globally across six countries.
Table of Contents
6/79, S/F, Gurudwara Road, Karol Bagh, New Delhi – 110005, INDIA
2051, Last Chance Court, Gold River, CA 95670, USA
