A few months ago, a young man sat across from Vicente Del Real and shared a heavy truth. He had a good job, a full calendar, and plenty of contacts in his phone. Then he said something: “Nobody really knows me.”
Del Real, who runs a Catholic nonprofit forming young Latino leaders in Chicago, wrote about that conversation in a column this June. His argument was simple and worth taking seriously: the friendship crisis is not the same thing as the loneliness epidemic. Loneliness is the symptom. The disease is that we have forgotten how to be friends.
The data backs him. In a Japanese government survey, people with no one to confide in were over 9 times as likely to report constant loneliness as people who had someone. Living alone made it less than three times as likely.
This shows a difference between having social contacts and having close friends. A person can know many people and still have nobody they can confide in about a personal problem.
Quick Answer
Loneliness tracks the quality of relationships, not the quantity. In a survey of 11,873 people, those with nobody to confide in reported constant loneliness at 23.4% — against 2.5% for those with someone. Living alone raised the rate far less. And among lonely older Americans, 52% aren’t socially isolated at all.
Having No One To Confide In Beats Living Alone, Nine To Three
Most people think living alone is a sign of social isolation. But there is a difference between loneliness and lack of friends. Some people who live alone have strong relationships. On the other hand, others can feel isolated even when they live with family or other people.
The clearest evidence comes from Japan’s statutory national survey on loneliness and isolation, fielded by the Cabinet Office on 1 December 2025 with 11,873 respondents aged 16 and over.
Table 1 — What predicts loneliness, ranked
| Circumstance | Lonely “often or always” | Multiple vs comparison |
| No one to chat casually with | 25.0% | 10.0× |
| No one to confide in | 23.4% | 9.4× |
| Rarely eats with anyone | 17.3% | 6.4× |
| Unmarried | 9.0% | 3.8× |
| Lives alone | 9.8% | 2.9× |
Source: Cabinet Office of Japan, 2025 wave, n=11,873. Multiples are derived from the published percentages. Having no confidant matters roughly three times as much as living alone.
Having no one to confide in is associated with a 9.4-fold difference in reported loneliness (23.4% against 2.5%) among those with someone. Living alone, the circumstance that attracts most of the attention, is associated with a 2.9-fold difference. The presence of a single confidant outranks the entire architecture of your living situation by a factor of three.
Two caveats belong here. The survey is cross-sectional, so it cannot establish direction. People who are lonely may lose confidants, rather than the loss of confidants making them lonely, and the likely truth is both. Also, “associated with” is doing real work in those multiples; these are gradients in a survey, not the results of an intervention.
The same pattern of loneliness shows up in the United States in a study that interviewed 6,994 older Americans and was published in Aging & Mental Health. It found that 52% of those who were lonely were not socially isolated at all. They had the contact (the events, the greetings, and the full calendar) and still felt alone.
That is what “nobody really knows me” gets at, and it is the empirical form of the argument Vicente Del Real makes in the column. The problem is not the absence of people, but the lack of a relationship in which you can be known.
The Friendship Recession, In Two Numbers
The friendship recession shows up most clearly in two figures: the share of Americans with no close friends, and the share with ten or more. Both moved sharply between 1990 and 2021. Daniel A. Cox’s report for the Survey Center on American Life tracks the change across several measures.
Table 2 — The friendship recession between 1990 and 2021
| Friendship Metric | 1990 Baseline | 2021 Survey |
| No close friends (all adults) | 3% | 12% |
| Ten or more close friends (all adults) | 33% | 13% |
| Men with six or more close friends | 55% | 27% |
| Men with no close friends | 3% | 15% |
| Women with six or more close friends | 41% | 24% |
Source: Survey Center on American Life, 2021, n=2,019. 1990 baseline is a Gallup phone survey — mode break noted in the report’s footnote 7.
In 1990, only 3% of American adults said they had no close friends. By 2021, that had risen to 12%, a shift from roughly 1 in 33 to 1 in 8 adults with no close friends.
At the other end of the scale, 33% of adults said they had ten or more close friends in 1990. But 31 years later, only 13% did. Large friendship circles fell from about 1 in 3 adults to about 1 in 8.
Simply put, the pattern is a squeeze at both ends. Deep friendships and large circles became less common at the same time.
Infographic 1: How friendship circles have collapsed
The 12% figure, therefore, isn’t just about people having fewer names in their contact lists but a lack of close confidants. And the fall from 33% to 13% is not about contact either. In 2021, Americans had more ways to reach friends and family than they did in 1990, not fewer. What declined was the closeness of the bond, not the availability of the channel.
Here is an important note. The 1990 figures came from a Gallup telephone survey of 1,226 adults. The ones from 2021 were derived from a Survey Center on American Life survey of 2,019 adults. They also used different methods of collecting responses. The 1990 data was gathered by phone, while the 2021 data was collected online. This means the figures are two comparable estimates rather than a perfectly continuous measure of change over the three decades.
Men Lost The Most, Fastest
Men have seen a particularly sharp decline in close friendships.
The share of men with six or more close friends fell from 55% in 1990 to 27% in 2021, and the share with none rose from 3% to 15%. That is close to a halving at one end and a five-fold increase at the other. No other group in the Survey Center data moved as far on both measures at once.
What the data does not tell us is why. The shift is clear; the cause is not established. One plausible driver (suggested by researchers who study how men form friendships, though not tested in this survey) is that men’s friendships tend to grow out of shared activity rather than disclosure. They can start at work, through sports, hobbies, church, school, or a shared project. Regular contact does the work. Neither man has to make a deliberate effort to keep the friendship alive, because the activity keeps bringing them back together.
But life can easily get in the way. A man changes jobs and stops seeing his former workmates. Another becomes busy raising children. Someone moves to another town. A friend gets married and has less free time. The regular meetings become less frequent, and the friendship can slowly fade.
There may be no argument or falling-out. People simply stop seeing each other as often as they once did.
So, a man may know many people but have few he can actually talk to. He may have workmates, neighbors, relatives, former classmates, and people he meets every week, yet still have nobody he feels comfortable calling when something is wrong.
What We Lost Was Friends, Not Family — And There’s Proof
When it comes to friendship, the real problem is getting people who are close. Vicente Del Real’s main argument in his essay on the friendship crisis is simple. It is not about the lack of friends, but that we have far fewer people we can turn to when life gets difficult. The General Social Survey (GSS) gives us a way to see this change in the data.
A study by sociologists Miller McPherson, Lynn Smith-Lovin, and Matthew Brashears, published in the American Sociological Review, looked at the people Americans turn to when they need to talk about important matters. The change was clear:
In 1985, Americans had an average of 2.94 people they could discuss important matters with. By 2004, that had fallen to 2.08, a drop of about one-third.
The number of both family and non-family confidants fell, but the decline was much greater among non-family members.
The share of people who talked only to family about important matters rose from 57% to roughly 80%.
The share who relied entirely on a spouse increased from about 5% to 9%.
We didn’t simply lose some people from our social circles, but became more dependent on the few relationships closest to home. Friends, neighbours, coworkers, and other people outside the family became less likely to be part of that inner circle.
The authors described a social network that was becoming smaller and more tightly connected. Instead of spreading emotional support across several kinds of relationships, more of that support was falling on spouses and close family members.
That can put a lot of pressure on one person. A spouse may be your partner, lover, and co-parent, but expecting one person to provide every kind of emotional support leaves very little room for the wider network that used to help carry that weight.
Time-use data shows the same change from another angle. Between 2003 and 2020, the amount of time people spent socially engaging with friends fell by about 20 hours a month. Time spent with other social ties fell by about 10 hours, while time with non-household family fell by just 6.5 hours.
The decline was not spread evenly across all relationships. Time with friends fell twice as much as time with other social ties and more than three times as much as time with non-household family. Friendship, in particular, has taken a significant hit.
Careful: The Statistic Everyone Quotes Didn’t Replicate
There’s a common claim that a quarter of Americans have nobody to talk to. This statistic has been quoted in numerous articles, books, and congressional testimonies. The source is the 2006 GSS paper that stated that the number of Americans with no confidant had nearly tripled to about 25% in 2004. A later study did not replicate the results.
In 2010, one of the co-authors of the original 2006 paper, Matthew Brashears, led a team of researchers on a similar test. They used the same methodology and questions to retest the earlier findings. However, what they found was significantly different from earlier results.
Infographic 2: The statistic everyone quotes that didn’t replicate
The mean number of confidants was 2.03 and was unchanged since the 2004 study
Only 4% of the respondents listed no names, not 25%
Of that 4%, 64% said it was because they simply had no important matters to discuss, not because they lacked someone to talk to.
After the second study, Brashears explained that the networks didn’t shrink to zero. Instead, most respondents classified a smaller share of their networks as suitable for “important” conversations. The question had changed meaning, not the friendships. So the authors flagged the risk themselves, stating that the original study may have overestimated the number of social isolates.
Meanwhile, other sociologists like Claude Fischer examined the GSS trends from 1970 to 2010. He concluded that Americans were no more or less likely to be friendless over the long haul. However, he conceded recent cohorts might differ. The American Friendship Project, published in PLOS ONE in 2024, notes that recent published estimates put the share of Americans with no friends between 8% and 12%, which is higher than the retest above. Its own surveys found fewer than 3%.
Table 3 — Different studies on Americans without any confidants
| Source | Findings on Americans with No Confidant |
| McPherson et al., 2006 (2004 GSS) | ~25%, nearly tripled since 1985 |
| Brashears re-test, 2010 | ~4%, of whom 64% simply had nothing to discuss |
| Fischer, GSS 1970–2010 | No long-term change in overall friendlessness |
| The American Friendship Project, 2024 | Cites published estimates of 8% to 12%; its own surveys found under 3% |
The friendship decline is well evidenced elsewhere. This particular statistic is not settled.
So, what is the conclusion? Friendship has declined in terms of time spent together, loss of casual community spaces, and frequency of gathering. However, the idea of total isolation is not supported by a later retest. While estimates of friendlessness vary, the evidence does not support treating the 25% figure as a current fact.
41 Minutes A Day, Down To 35
How much time do people spend with friends? According to the American Time Use Survey by the US Bureau of Labor Statistics, most people spent about 35 minutes socializing in 2025, down from 41 minutes in 2015. What’s more? The share of those who socialize on any given day also fell from 38% to 30%.
These figures measure socializing generally, rather than time spent specifically with friends. While a six-minute drop may not mean much in a day, it adds up to over 36 hours in a year. This is time that would have been spent venting, shared laughter, and unplanned conversations that build relationships.
Why the fall happened is not something the time-use data can answer. The survey measures what people did, not why. Plausible drivers can be longer working hours, more people holding second jobs, the scattering of friends across cities and time zones, but these are not findings from this data.
The pattern matters because friendship depends on voluntary, repeated contact to survive. It doesn’t have to be a planned outing or conversation. Think of something like meeting for coffee, talking for a few minutes, or checking on a friend. When these moments disappear, friendship fades too. When life gets busy, social time is the first to go.
Why Adult Friendship Is Structurally Harder
Making friends as an adult and keeping them is a lot harder. However, sociologists point out that it is not caused by personality but the architecture of daily life. Their argument goes back to a simple observation. For organic friendships to thrive, they need unplanned interactions, proximity, and shared settings where one doesn’t need to perform.
In school and college, those ingredients are handed to you by default. You sit next to the same people every day for a semester. All classmates suffer through exams together, walk to classes side-by-side, and hang out in dorms without an agenda. Friendship forms as a by-product of the schedule.
However, these shared moments are not available in adulthood. Your friends scatter across different neighborhoods, cities, or time zones. There are no unplanned interactions, as everything requires scheduling. Besides, you have to perform. Workplaces demand professionalism, and social media demands curation.
In the past, there were areas where people met outside work and home for casual talks and hanging out. The sociologist Ray Oldenburg called them third places. Unfortunately, these too disappeared. Neighborhoods became more car-centric, voluntary associations declined, and civic groups hollowed out. So, all the natural friendship friction points evaporated.
This is the same loss the GSS data measured from the other direction. When McPherson, Smith-Lovin, and Brashears found that non-kin confidants were disappearing faster than family ones, they identified voluntary associations and neighbourhoods as exactly where the losses occurred. The structural argument and the survey data describe the same change.
Online Community Isn’t Filling The Gap
Digital spaces are touted as the new frontier in friendships. Many people jump onto social media, Reddit, and Discord to find others with similar interests in seconds. However, the American Psychiatric Association’s Healthy Minds Poll revealed something different when it asked Americans where they feel a high sense of belonging. Here were the results:
Family: 65%
Friends: 53%
Online communities: 3%
So, do online friendships count? They do. However, it is clear that family and friends cannot be replaced by online communities. Digital spaces can help people meet others, maintain contact, and find communities built around shared interests. But, based on this survey, they were much less likely than family or friends to be identified as a source of a strong sense of belonging.
The gap between 53% and 3% is the one to carry forward. The friendship recession is not being absorbed by the internet. Nothing absorbs it in particular, and that is why it shows up in the loneliness numbers.
What The Data Says Actually Helps
The friendship crisis is structural. The losses are in voluntary associations, neighbourhoods, and the everyday contact that used to happen without planning. That means the levers are structural too, and the data is clearer on what has gone wrong than on what reverses it.
One pattern stands out. In the Japanese survey, 17.3% of people who rarely eat with anyone said they felt lonely often or always. Among people who eat with others almost every day, only 2.7% said the same — a 6.4-fold difference. That sits between the two figures from earlier: smaller than the confidant gap (9.4×), larger than living alone (2.9×).
What this does not show is that eating with people reduces loneliness. The survey is a snapshot, not a trial. The association is real, but the direction is not established. Though it does suggest that low-effort, everyday contact is one of the places the friendship recession shows up most clearly.
Del Real’s claim about vulnerability (that we perform rather than disclose, and that friendship is tested in the valley) is plausible and consistent with everything above. It is also untested. There is research on self-disclosure and intimacy, but no dataset tracking whether Americans disclose less to their friends than they used to. Naming that gap is not a weakness. It is the difference between what the evidence shows and what it merely suggests.
Questions People Ask About The Friendship Recession
How many Americans have no close friends?
Estimates range between 8% and 12% across different studies. The most cited figure is 12% in 2021, up from 3% in 1990. It comes from the Survey Center on American Life. However, the report notes a mode break between online polling and phone data collection methods.
Is the friendship recession real?
There is evidence of a significant decline in the time spent with friends. Some studies show a loss of about 20 hours a month between 2003 and 2020. While the exact number of absolute friends is contested, there’s at least one prominent sociologist finding no change from 1970 to 2010.
Why do men have fewer friends?
Data shows men reporting six or more close friends fell from 55% to 27% between 1990 and 2021, while those with zero rose to 15% from 3%. One possible explanation for this shift is the loss of activity-based socialization. When men change workplaces and shared interests, friendships tend to fade away, hence fewer close buddies.
What’s the difference between loneliness and having no friends?
The line between loneliness and having zero friends is blurred. In essence, loneliness measures that subjective gap between the relationships you currently have and the ones you crave. So, you may have friends but lack anyone you are close to. Interestingly, among lonely older adults, 52% are not socially isolated.
Do online friendships count?
When Americans were asked where they feel a strong sense of belonging, only 3% actually named online communities. This was the lowest compared to 65% for family and 53% for friends. While digital spaces offer connection, they do not substitute for local, embodied presence.
How much time do people spend with friends?
Americans spent an average of 35 minutes a day socializing in 2025, down from 41 minutes in 2015. This means they are spending three hours less socializing each month than they did. Studies also show that only 30% engage in any form of socialization on a given day.
Editorial note: If you have few or no close friends, that does not mean there is something wrong with you. Changes in work, neighbourhoods, and community life have made adult friendships harder to maintain. If loneliness is weighing on you, consider reaching out to someone you trust or seeking support.