Who the internet hurts most — and why your peers matter

Two people on the same train are scrolling their feeds: a 20-year-old student and a man of about sixty. They spend roughly the same amount of time online. Does it affect them in the same way?

A study published as part of the World Happiness Report 2026 suggests the answer is no. And the reason is not only age. The data point to something else as well: how much of their peers' lives happens online.

What the researchers studied

The World Happiness Report is an annual report on wellbeing around the world. It is produced by the Wellbeing Research Centre at the University of Oxford, Gallup, the UN Sustainable Development Solutions Network and an independent editorial board. The 2026 edition came out on March 19 and focuses on social media, trust and social connection.

One chapter was written by Zeynep Ozkok, Jonathan Rosborough and Brandon Malloy of St. Francis Xavier University in Canada. They used data from the European Social Survey for 2016–2024: nearly 200,000 respondents in 30 countries, including the United Kingdom and Ireland.

Wellbeing was measured with two questions on a 0–10 scale: how happy people are, and how satisfied they are with life overall. Internet use was measured by how many minutes a day people say they spend online.

Why a simple correlation is not enough

The main trap in this kind of research is reverse causality. Perhaps the internet makes people feel worse. Or perhaps people who already feel low simply find it easier to spend the evening online. A plain comparison cannot tell these apart.

The authors used a technique common in economics: an instrumental variable. Their instrument was regional internet speed, based on median download speeds measured by M-Lab. The logic is simple. Faster internet encourages people to spend more time online. But the infrastructure itself, the authors argue, does not depend on how happy local residents are.

The result: across everyone, more time online is linked to somewhat lower wellbeing. Notably, a simpler analysis without this technique found almost no effect.

Different generations, different effects

The most interesting part comes when the data are split by generation. According to the authors, the internet is most harmful for Gen Z and moderately harmful for Millennials. For Gen X the effect is close to zero, and for Baby Boomers it is even slightly positive.

Gen Z's time online rose sharply between 2016–19 and 2020–24. The authors estimate that this rise alone implies a drop of roughly 0.3–0.5 points on the 10-point wellbeing scale. That is not a catastrophe, but it is not trivial for a measure that usually moves slowly.

The researchers also tracked what happened to people's social lives. The biggest declines were in trust in other people, trust in institutions, how often people meet others, and the sense of being as socially active as one's peers. These fell most sharply among young women in Gen Z and the Millennial generation, especially in Western Europe. These are changes over time, not proof that the internet caused them.

Why your peers matter

Back to the train. The authors divided respondents into "peer groups" by country, age group, gender and survey year, about 2,700 groups in all. For each, they used Eurostat data to find the share of people using social media.

The effect of the internet turned out to depend on that share. In groups where fewer than half of people used social media, extra time online was actually linked to higher wellbeing. The wider the reach, the worse the effect became. In groups where more than 90% of peers used social media, the effect was strongly negative and statistically significant.

So what matters is not only how much time a person spends online. It is also how their social world lives. For the 20-year-old student, nearly all her friends are in the feed. For the man next to her, far fewer are. The internet is a different environment for each of them, even if the apps are the same.

The authors do not offer a final explanation. But their data point to several possible mechanisms: online contact crowding out in-person meetings, eroding social capital such as trust, and an environment where everyone is constantly on display, which invites comparison.

What the rest of the report says

The chapter fits with the rest of the report, though the overall picture is more complex.

The sharpest warning concerns English-speaking countries. In the United States, Canada, Australia and New Zealand, the happiness of people under 25 has fallen by an average of 0.86 points over the past two decades. Meanwhile, in 85 of 136 countries, young people are happier now than they were about twenty years ago. In the overall 2026 ranking, the US is 23rd, Canada 25th and the UK 29th; no English-speaking country made the top 10.

How people use social media also matters. In 47 countries, 15-year-old students who spend more than seven hours a day on social media have much lower wellbeing than those who spend less than an hour. In data from Latin America, platforms built around algorithmic feeds were associated with lower happiness, while those focused on connecting with people you know were associated with higher happiness.

Jan-Emmanuel De Neve, director of the Oxford centre, put it this way: "Heavy usage is associated with much lower well-being, but those deliberately off social media also appear to be missing out on some positive effects."

Why researchers still disagree

The effect of social media on young people is one of the most contested questions in social science today.

In The Anxious Generation (2024), psychologist Jonathan Haidt argues that the shift to a "phone-based childhood" is the main cause of declining adolescent mental health. He also frames it as a collective action problem: one family or one teenager can do little if everyone around them is online.

In a review for Nature in March 2024, psychologist Candice Odgers countered that the available evidence does not support such a strong causal claim. In her view, the associations are mostly small, and many other factors shape young people's mental health.

The World Happiness Report chapter does not settle this dispute. But it adds data that try to answer the question of cause, not just correlation. And its finding about peers echoes the collective action argument in an unexpected way.

What we still don't know

• What people actually do online. The authors themselves acknowledge that the survey does not measure which platforms people use, what time online replaces, or the quality of their relationships.
• How reliable the instrument is. The method assumes that internet speed affects wellbeing only through time online. One could object that faster internet tends to come with bigger cities and different ways of life. This is our caveat, not the authors' conclusion.
• Self-reports. People rate both their own happiness and their minutes online.
• Geography. The data cover Europe, including the UK and Ireland. They should be applied to North America or Australia with care.
• Averages. A generation is millions of different people. For some, the internet is a source of loneliness; for others, it is their only link to a community.

What follows from this

The authors conclude that efforts to improve wellbeing cannot focus solely on individual screen time. They must address the wider environment: declining trust, weakening community bonds and the highly comparative nature of online spaces, especially for young people.

That conclusion lands in the middle of a debate about regulation. Since December 10, 2025, Australia has required social media platforms to deactivate accounts of users under 16; penalties fall on the platforms, not on children or parents. On 16 September 2026, the European Commission proposed banning social media for children under 13; the proposal still needs the approval of the European Parliament and member states. The chapter's authors do not assess specific laws. But their data are a reminder that the internet's effect is not only a private matter. It is also a property of the environment we all share.

For readers, perhaps one observation follows. It helps to look not only at how much time your phone takes, but also at whether there is still room for meeting people in person, and whether "everyone is online" has become the only way of being together.

Questions for discussion

1. Have you noticed the internet affecting you differently from people of other ages around you? What is the difference?
2. If the harm depends on how a whole peer group behaves, who should address it: individuals, families, schools or governments?
3. Which online spaces, in your view, bring people closer, and which tend to pull them apart?