Why do so many people end up with a noticeably smaller circle of friends the older they get? The easy answer is that making friends simply gets harder with age.
People retire, move, lose the built-in settings, like a job or a school, that used to put them in a room with new people on a regular basis. It’s a tidy explanation, and according to researchers who have actually tracked what happens to people’s social networks over time, it’s also largely the wrong one.
The better-supported explanation is less flattering to that assumption: people aren’t losing a skill so much as changing what they want from the people around them. That distinction sounds small, but it reframes a shrinking social circle from a symptom of decline into the gradual result of a decision, made mostly without anyone announcing it out loud.
What actually shrinks, and what doesn’t
A study led by psychologist Tammy English alongside Laura Carstensen at Stanford followed adults across the full span of adulthood and tracked how their social networks changed over time. The pattern that emerged wasn’t a general decline.
Reductions were observed almost entirely among what the researchers call peripheral partners, the more casual, looser ties in a person’s life. The number of close partners, the people someone would actually call in a real crisis, stayed relatively stable across the years.
That distinction matters more than it might first appear. If getting older genuinely made it harder to form or keep any relationship, close ties would be expected to thin out too, since maintaining a close friendship takes at least as much effort as maintaining a casual one, usually more. Instead, the people who mattered most stuck around, while acquaintances, work contacts, and looser social ties were the ones that fell away. That pattern reads as active editing, the kind a person does on purpose, rather than a general fading of the ability to connect.
Picture the ordinary version of this happening to someone over twenty years. In their thirties, a person’s contact list might include a dozen work colleagues, a handful of parents from school pickup, a couple of old college roommates, and two or three people they’d call at 2 a.m. if something went wrong.
By their sixties, the colleagues have mostly scattered along with the jobs, the school-gate acquaintances ended when the reason for seeing them did, and even a few of the old roommates have drifted into holiday-card status. The two or three people they’d call at 2 a.m., though, are often still there, sometimes joined by one or two new names added along the way. Nothing about that pattern looks like a person who forgot how to make friends.
Why smaller can mean happier
The theory behind that editing has a name: socioemotional selectivity theory, developed by Carstensen over decades of research. The core idea is that how much time a person believes they have left, not their age itself, shapes what they want out of relationships. “When time horizons are relatively short, people focus on emotionally meaningful goals,” Carstensen has explained.
When the future feels wide open, expanding a social network for information, opportunity, or status makes sense. When it feels shorter, the payoff shifts toward relationships that already deliver comfort and connection right now, and less toward relationships that might pay off eventually.
What makes the theory more than a tidy story is that the shift shows up even when age itself is held constant. Carstensen’s broader research program has found that younger adults placed in situations that shrink their own sense of time left, such as facing a serious illness or living through a period of widespread uncertainty, start making the same kinds of relationship choices that older adults typically make: fewer new acquaintances, more time protected for the people who already matter most. Age is one common way a person’s time horizon gets shorter, but the research points to the shorter horizon’s effect on priorities as the actual mechanism, with the narrowing itself mostly along for the ride.
That shift shows up in the numbers, not just the theory. English and Carstensen’s research found that the emotional tone of a person’s social network predicted their daily emotional experience, and that older adults’ surviving networks skewed toward people who brought them more positive feeling and less negative feeling than the networks of younger adults. Fewer relationships, in other words, wasn’t the problem. A network trimmed down to the people who actually make someone feel good was doing real emotional work.
What this means for the rest of us
None of this means every older adult with a small circle is thriving, or that everyone who prunes a friendship is following some wise internal clock. Loneliness is still real, and it isn’t the same thing as selectivity. The distinction the research draws is specific. An older person quietly letting a few acquaintances drift while holding tightly to two or three real friendships is closer to curating a life than losing the knack for connection. With enough practice, that person has simply gotten better at recognizing which relationships are worth the limited time they have left, and choosing accordingly.
That’s a strange thing to envy, since it usually gets framed as a loss. Getting quicker at recognizing which relationships pull their weight, and which ones you’re maintaining out of habit rather than actual want, might be one of the more useful things that comes with age instead of something that happens to you despite it.
It also suggests the smaller circle is less something to brace for later in life than a skill that can be practiced earlier. Nobody has to wait until their seventies to start asking which relationships in their life are running on genuine want and which are running on inertia. Carstensen’s research didn’t set out to hand people a life hack, but it points to one anyway: the people who seem to handle aging best tend to be the ones who figured out, sooner rather than later, which few connections were worth protecting, and let the rest go without much guilt about it.
