Americans report having fewer close friends than at any point researchers have measured. In 1990, a third of U.S. adults said they had ten or more close friends; by 2021, only 13 percent did, according to the Survey Center on American Life’s American Perspectives Survey. The share with no close friends at all quadrupled, from 3 percent to 12 percent. Pollster Daniel Cox, who led the research, found the drop was steepest among men — the share with at least six close friends fell by half. The usual explanation is that people are too busy, too distracted, or too online. But a quieter body of psychology research points somewhere else: many adult friendships don’t fade from neglect. They erode because nobody in them ever learned how to hold a boundary.
Compassion and boundaries are not opposites
Brené Brown, a research professor at the University of Houston’s Graduate College of Social Work who spent more than a decade studying shame and vulnerability through grounded-theory interviews, found something counterintuitive in that data: the people who seemed most generous toward others were also the most specific about what they would and would not tolerate. In a 2015 Salon interview, Brown put it directly — the most compassionate people she had interviewed were also the clearest about what was and wasn’t okay with them. In her book Rising Strong, she defines boundaries in plain terms, as the lists we each keep of what’s okay and what isn’t. The absence of that list, she argues, doesn’t produce more closeness. It produces resentment, because unstated limits still get violated — the other person just finds out when you withdraw instead of when you ask.
That reframes the self-help cliché about boundaries being “selfish.” Clinical social worker Nedra Glover Tawwab, a licensed therapist and author of the bestseller Set Boundaries, Find Peace, treats boundary-setting as a core clinical skill rather than an act of coldness — the mechanism by which two people stay specific enough about needs and limits that a friendship can survive disagreement. Vague friendships, in this framework, aren’t closer. They’re just more fragile, because nothing has ever been tested.
Why you can’t be close to everyone: the math behind Dunbar’s number
There’s also a structural reason boundaries matter more as we age, and it comes from evolutionary psychologist Robin Dunbar’s 1992 paper in the Journal of Human Evolution, which found that primate species’ typical group size tracks the relative size of their neocortex. Extrapolated to humans, that produced the now-famous figure of roughly 150 stable relationships — “Dunbar’s number.” Less famous is the detail that matters more for friendship: Dunbar found this capacity isn’t one flat number but a set of nested layers, growing roughly by a factor of three each time — about 5 people in an innermost circle, 15 close friends, 50 good friends, and 150 in the outer layer of meaningful acquaintance. A 2016 study co-authored by Dunbar, using real mobile-phone call records, found layer sizes close to the theory — averaging 4.1, 11, 29.8, and 128.9 people.
The layers aren’t about popularity; they’re about a hard ceiling on cognitive and emotional bandwidth. If the inner circle of five is finite, every “yes” to a demand from someone in the outer rings is implicitly a boundary crossed against someone closer in. Declining to be endlessly available to a wide acquaintance network isn’t a character flaw. It’s what protects the five or fifteen relationships that depend on real depth.
The hours a friendship actually costs
Communication researcher Jeffrey Hall at the University of Kansas put a number on what depth requires. Publishing in the Journal of Social and Personal Relationships in 2018, Hall combined two studies — one surveying adults who had recently relocated, another tracking university freshmen — and found it takes roughly 40 to 60 hours together for an acquaintance to become a casual friend, 80 to 100 hours to become a friend, and more than 200 hours to become a close friend. Crucially, he found that time spent simply hanging out and joking around counted far more toward that threshold than time spent on shared obligations like work.
That arithmetic is unforgiving once you’re an adult with competing claims on your evenings. Every hour given to a friendship that currently crosses your limits — one that demands constant emotional labor, ignores your “no,” or treats your time as unlimited — is an hour that cannot go toward the 200 a different, healthier friendship needs to deepen. Boundaries, in this light, aren’t a wall keeping people out. They’re a budget. Hall’s data suggests that friendships without any limits on what they can ask of you aren’t more intimate; they’re just more expensive, and they quietly starve every other relationship of the hours it needs to survive.
What this looks like in practice
None of this research argues for distance as a goal. Dunbar’s layers, Hall’s hour-counts, and Brown’s interviews all point to the same practical move: naming limits early and specifically, rather than enforcing them later through silence. That might mean telling a friend directly that you can’t take a crisis call after 10 p.m., rather than letting calls go to voicemail for a month. It might mean saying you’re not the right person to weigh in on their marriage, rather than offering strained advice you don’t believe. Tawwab’s clinical framing and Brown’s interview data agree on the mechanism: the boundary has to be spoken, not just felt, or the other person has no way to meet it.
The friendship recession the Survey Center on American Life documented won’t reverse through willpower alone. But the research on what actually sustains adult friendship — finite cognitive layers, measurable hour thresholds, and boundaries that compassionate people, not cold ones, tend to hold — suggests the fix isn’t trying to be close to more people. It’s being honest, early, about what each friendship can actually ask of you, so the ones worth 200 hours get them.
