The advice is remarkably uniform and it arrives from everyone. Get out more. Join something. Put yourself out there. It is offered kindly, it is easy to remember, and for the most common version of loneliness it is close to useless, because it adds more of the one thing that was not running short.
Loneliness has been defined in the research since the 1970s as a gap, meaning the distance between the connection a person has and the connection they want. A gap can be closed from either side. New acquaintances close it only on the side most people are already fine on.
The separation between loneliness and solitude, and how the layers of that gap read against one another, is a real question and a different one. This piece is about what the gap actually responds to.
1 in 6
people worldwide report feeling lonely, on WHO's 2025 estimate
871,000
deaths a year the same report links to loneliness
20 min
a day Americans spent in person with friends in 2020, down from about 60 in 2003
46,000
people across 237 countries in the largest survey of loneliness yet run
Why meeting more people does not close the gap
Contact really has fallen, and the figures are not small. The United States Surgeon General's 2023 advisory put time spent in person with friends at about twenty minutes a day in 2020, down from roughly an hour two decades earlier. So there is a genuine quantity problem in the population, and for some people it is their problem.
It is not most people's problem. The largest survey of loneliness yet conducted, the BBC Loneliness Experiment analysed by Manuela Barreto and colleagues with more than forty-six thousand respondents across 237 countries, found loneliness highest among the youngest adults and in the most individualistic societies. Those are not the places or the ages with the least contact. Maike Luhmann and Louise Hawkley, working with a representative German sample of over sixteen thousand people, found loneliness falling across midlife and rising sharply again only after seventy-five, which does not track the amount of company available at those ages in any straightforward way.
What the advice is actually doing becomes clearer once it is set beside the thing it supplies.
| The prescription | What it supplies | When that is the shortage |
|---|---|---|
| Meet more people, join something | Quantity of contact | Contact is genuinely near zero: a new city, work done alone, a year indoors |
| Work on confidence in company | The ability to use contact once it exists | Fear is what is stopping it, not the diary |
| Say one costly true thing to somebody already in your life | Being known | There is contact and nothing in it is risky to say |
| A fixed weekly thing rather than a planned evening | Frequency, at a price low enough to survive a bad week | Arrangements keep dissolving before they happen |
Most people who type a question like this one are in the third row and are being handed the first.
What being known is made of
The useful research here is not about friendship at all. It is about how intimacy gets produced, turn by turn, inside a conversation.
In 1998 Jean-Philippe Laurenceau, Lisa Feldman Barrett and Paula Pietromonaco had people record their interactions as they happened and found two things worth holding on to. Disclosure of emotion predicted intimacy considerably better than disclosure of facts and information, so the content that counts is not news. And the effect ran through perceived partner responsiveness, which means intimacy was not produced by handing something over. It was produced by what came back.
Nancy Collins and Lynn Miller's meta-analysis of ninety-four studies found three separate effects sitting inside the same phenomenon: people who disclose are liked more, people disclose more to those they already like, and people come to like those they have disclosed to. The third one is routinely left out of summaries and it is the load-bearing one. Disclosure does not merely report closeness that already exists. It manufactures some, in the person doing the disclosing.
There is observational evidence too, and it is worth reporting with its limits attached. Matthias Mehl and colleagues put recorders on seventy-nine undergraduates for four days, collected nearly twenty-four thousand snippets of their waking lives, and found wellbeing correlated with the amount of substantive conversation at around .28 and with time spent alone at around minus .33. A pooled replication in 2018 across four samples, four hundred and eighty-six people in total, confirmed the substantive conversation association and failed to replicate the small talk one. Read honestly, that pair of studies does not say small talk is corrosive. It says the amount of talk that goes somewhere tracks how people are doing, and the amount that goes nowhere is roughly neutral. Both studies are correlational and the first was students.
None of this depends on the relationship being a friendship. The entry on platonic relationships has the broader picture of what these ties are for and what they cost.
The sentence that does the work is one you already have somebody to say to
Here is the move, and it sounds too small to be the answer, which is most of why it does not get made. Take one person who already knows you, and the next time the question comes, replace fine with something true.
Not a confession. Fine is not a lie, it is a formatting convention, and the thing wrong with it is structural rather than moral: nothing can be returned to it. A sentence with an actual subject in it gives the other person something to respond to, and the response is where the intimacy gets made. That is the whole finding from the Laurenceau work, turned round and pointed at a Tuesday.
The step that is missing is almost never a person. It is a sentence that could have been answered.
Two things make this harder than it looks. The first is that closeness is built by exchange, and exchange is sequential, so somebody has to go first with no guarantee. The second is that people are systematically wrong about what happens next. Michael Kardas, Amit Kumar and Nicholas Epley ran twelve experiments on this and found that participants expected deeper conversations to be more awkward and less enjoyable than they turned out to be, and consistently underestimated how much the other person would care. The caveat matters: most of that work was done with strangers, so it is evidence about the expectation rather than a promise about a specific person on a specific Tuesday.
Repetition does more than a good evening
The other thing that grows depth has nothing to do with what gets said. It is frequency, at a low enough threshold that nobody has to decide.
Richard Moreland and Scott Beach ran a quietly remarkable study in a large lecture course. Four women of similar appearance attended either zero, five, ten or fifteen sessions, speaking to nobody and doing nothing. At the end of term students rated the ones who had turned up most as more attractive and more similar to themselves. Nothing happened between them. Attendance alone moved the dial.
That is the mechanism behind the fixed class, the same five-a-side, the standing Wednesday. A planned evening has to be decided on every single time, and the deciding is where it dies, because there is always a week when nobody has it in them. A recurring thing happens unless somebody cancels, which is a completely different default.
The last row is the one that matters most. Quality time, as a model, requires the hard thing to be the point of the occasion, which is more or less unbearable and therefore does not happen. Recurring contact lets the hard thing arrive in the middle of something else, which is the only way most people ever say it.
Why the advice survives being wrong
It is worth asking why the weaker prescription is the one everybody reaches for, because the answer is not stupidity.
Meeting more people is executable. It has a venue, a date and a cost. It is countable afterwards, in events attended and conversations had, and when somebody is lonely what they want most urgently is to be doing something about it. Say one true thing to a person you already know has none of those properties. It cannot be scheduled, it cannot be audited, and there is nothing to report at the end of the week. Advice gets selected for legibility at least as much as for effect.
And sometimes it is simply correct, which is the part that gets lost when this argument is made too hard. If an honest count of the conversations in the last fortnight that were not purely practical comes out at or near zero, then quantity is the shortage, and it is first, because there is no existing relationship available to deepen. People who have moved somewhere with nobody in it, who work alone, who have come out the far side of a decade inside one relationship: for them the standard advice is not a misdiagnosis. How people actually get met is a real question. It is a different one from this.
The loneliness that arrives with a change of address
There is a whole category of this that is situational, and the people in it are the most likely to read it as a fact about themselves.
Moving. A relationship ending. A death. A new job, retirement, a first baby. What these have in common is that they remove, quickly, a structure that had been supplying both the contact and the being known, and they remove it without announcing that this is what has happened. The Luhmann and Hawkley analysis found exactly this shape in the data: for young adults loneliness tracked income, full-time work and relationship status, and the rise in the oldest group was attributable mostly to the loss of a spouse and to health. Circumstance, in other words, not character.
The reason it gets misread is that it so often follows something chosen. The promotion, the city, the baby. Loneliness that arrives after a decision gets filed as evidence about the person who made the decision, and the filing happens automatically.
New parents get the sharpest version of it. Surrounded, touched constantly, spoken to all day, and not asked a single question about themselves for a month. Retirement produces a close relative of the same thing, because the job had been quietly supplying both halves and only the contact half tends to get replaced. Bereavement produces the version nobody argues with, which is also the one where people most often decline to mention it, on the grounds that the grief is the legitimate complaint and the loneliness is self-indulgent.
What situational loneliness has going for it is that it generally does lift as the structure gets rebuilt. It takes longer than anybody budgets for. The thing to distrust in the meantime is not the feeling but the explanation that comes attached to it.
What a reading does and does not settle
Three bars over one month is a sample, not a verdict. A quiet four weeks in which nothing needed anybody produces the same numbers as four weeks in which somebody was genuinely held, and the real test of who is there comes after something hard happens rather than from a questionnaire. There is also nothing to be diagnosed with at the end of it, because loneliness is a description of a situation rather than a condition.
What a reading is good for is narrower and more useful than a score. It separates the quantity question from the being-known question, and that separation is the one almost nobody makes unaided, which is why so many people spend a year working hard on the wrong half.
Sources
- World Health Organization (2025). From loneliness to social connection: charting a path to healthier societies. Report of the WHO Commission on Social Connection.
- Office of the U.S. Surgeon General (2023). Our Epidemic of Loneliness and Isolation. U.S. Department of Health and Human Services.
- Barreto, M., Victor, C., Hammond, C., Eccles, A., Richins, M. T., & Qualter, P. (2021). Loneliness around the world: Age, gender, and cultural differences in loneliness. Personality and Individual Differences, 169, 110066.
- Luhmann, M., & Hawkley, L. C. (2016). Age differences in loneliness from late adolescence to oldest old age. Developmental Psychology, 52(6), 943–959.
- Laurenceau, J.-P., Barrett, L. F., & Pietromonaco, P. R. (1998). Intimacy as an interpersonal process. Journal of Personality and Social Psychology, 74(5), 1238–1251.
- Collins, N. L., & Miller, L. C. (1994). Self-disclosure and liking: A meta-analytic review. Psychological Bulletin, 116(3), 457–475.
- Mehl, M. R., Vazire, S., Holleran, S. E., & Clark, C. S. (2010). Eavesdropping on happiness. Psychological Science, 21(4), 539–541.
- Milek, A., Butler, E. A., Tackman, A. M., Kaplan, D. M., Raison, C. L., Sbarra, D. A., Vazire, S., & Mehl, M. R. (2018). "Eavesdropping on Happiness" revisited: a pooled, multisample replication. Psychological Science, 29(9), 1451–1462.
- Kardas, M., Kumar, A., & Epley, N. (2022). Overly shallow? Miscalibrated expectations create a barrier to deeper conversation. Journal of Personality and Social Psychology, 122(3), 367–398.
- Moreland, R. L., & Beach, S. R. (1992). Exposure effects in the classroom: The development of affinity among students. Journal of Experimental Social Psychology, 28(3), 255–276.