We Wrote That Third Places Were Coming Back. The Census Data Says They Were Disappearing.

On 30 June 2025 the World Health Organization Commission on Social Connection published its flagship report, naming social connection a global public health priority. It found that about one in six people worldwide, some 16 per cent, experienced loneliness across the period from 2014 to 2023, and that more than 871,000 deaths a year are associated with it, which the Commission framed as roughly a hundred an hour. Loneliness was most commonly reported among adolescents, at almost 21 per cent of thirteen to seventeen year olds.

The word associated is doing necessary work in that sentence, and it should not be upgraded to caused. Loneliness travels with poverty, illness, disability and age, all of which independently affect mortality, and disentangling them is genuinely hard. What the figure establishes is scale, not mechanism.

What do people say about their own lives?

That the difficulty is in meeting people at all.

Survey work published this year found that around 66 per cent of Europeans think it is harder to meet new people now than in the past, and that more than half are dissatisfied with their social relationships. Seventy two per cent said they would like more real world experiences, and 63 per cent expressed a wish to belong to a community organised around a shared interest.

The last of those is the most practically useful finding in the whole literature. People are not asking for more socialising in the abstract. They are asking for a reason to be in the same room repeatedly, which is what a shared interest provides and what an invitation does not.

What actually works, then?

Repetition and low stakes, rather than events.

The defining property of a third place is not that it is sociable. It is that you can go without arranging anything, that the same people are there most times, and that leaving early costs nothing. That combination is what allows an acquaintance to become something more without anyone deciding it should.

Almost every designed solution to loneliness fails on one of those three. An event requires arranging. An app supplies novelty rather than repetition, which is why the category has grown into a market without solving the problem, as we looked at in companion apps and the loneliness market.

The workplace used to supply repetition by default, which is one of the least discussed consequences of the change in where people work, covered in what the evidence shows about hybrid arrangements. Removing the commute removed a great deal of unplanned contact along with it, and nothing was designed to replace it.

The correction we owe

Our revival piece was not wrong, but it was partial.

Third places are having a cultural moment and are being rebuilt in the places best able to pay for them. The infrastructure version, the kind you do not have to buy anything to use, has been shrinking, and shrinking fastest where it was already thinnest.

Reporting only the first is how a genuine trend becomes a misleading one. The interesting question for the next few years is not whether third places are back. It is whether the ones coming back are open to the people who lost theirs.