Shrinking social circles

Why does an older person let a wide circle of friends fall away while a young one keeps collecting new ones?

A twenty-five-year-old will go to a party where she knows one person and come home with three new numbers. A seventy-five-year-old, offered the same evening, will very often decline and spend it with her sister. Count their networks and the older one is smaller — fewer names, fewer casual ties, fewer new entrants.

The obvious story is subtraction. People die, retire, move away, get ill; the circle shrinks because the world removes people from it and old age is not a good time for making replacements. That is true as far as it goes. But it makes a prediction we can check: if narrowing were purely loss, older people should be less satisfied with their social lives, and should take any replacement on offer. Neither is reliably what the data show — older adults report relationships they find at least as satisfying, often more so. So what if some of the shrinking is not something happening to them, but something they are doing?

Ask what a social contact is for. Not romantically — functionally. Some of what we get from other people is information: how this industry works, who to talk to, what the world outside my street is like, what kind of person I am turning out to be. Some of it is emotional: being known, being at ease, feeling something good in the next hour.

Now, when is information worth paying for? Only if you are going to use it. Information is an investment in a future — it buys a better position later. So a person with a long runway ahead should be willing to spend an evening on a stranger who might teach her something, be dull company tonight, and pay off in five years. Not a bad trade for someone with fifty years to run.

What happens when the runway shortens? The emotional return is immediate and still available. The informational return has nowhere left to be spent. So the same person, without changing a single preference about what feels good, should weight the two differently — pruning exactly the ties that are informationally promising but emotionally thin, while keeping and deepening the ties that are emotionally rich. This is Laura Carstensen's socioemotional selectivity theory, and the narrowing it predicts is selective rather than uniform: the periphery goes and the core stays.

Which raises the sharpest question available: if the driver is really perceived time, and not age itself, then the effect should be detachable from age. Can you produce it in a young person?

You can, and this is the finding that makes the theory more than a plausible story. Ask young adults to imagine they are about to move far away in a few weeks, and their preferences among potential social partners shift toward the familiar and emotionally close — toward the pattern typical of older adults. Run the manipulation the other way, asking older adults to imagine an extension of twenty healthy years, and the age difference is reduced or erased. Carstensen's group also studied Hong Kong around the 1997 handover, when an ending sat on everyone's calendar, and found preferences shifting as that date approached and relaxing again afterwards. Age was not doing the work. The perceived horizon was.

A related pattern comes from the same root: as horizons shorten, attention and memory tilt somewhat toward positive material over negative — the positivity effect. If the goal is how the remaining time feels rather than how well it prepares you, that is what you would expect.

The theory is easy to over-read in two directions, so let us fence it in.

It does not say the losses are illusory. Structural attrition is real and substantial — bereavement, retirement removing a whole daily cast at once, reduced mobility, illness, relocation. Both mechanisms operate, and in any individual case it is genuinely hard to say how much of a smaller circle is choice and how much is circumstance. The claim is not that selection explains the shrinkage, but that selection is present, and that a pure-attrition account fails to explain the shape of what remains or the maintained satisfaction.

Nor does it say older people are withdrawing. Fewer ties, differently composed, is not less connection — if anything the theory predicts more contact concentrated in fewer places. Loneliness in later life is a real and serious problem, but it is not the same phenomenon as deliberate narrowing, and the two get conflated constantly.

And the perceived-horizon mechanism, while unusually well supported for a theory in this area, is not the only account on offer. Some narrowing may reflect accumulated knowledge of which relationships work, some the plain fact that maintaining ties costs energy. The manipulation studies are what make the horizon claim hard to dismiss — they show the effect is producible, not merely correlated with age.

• How the positivity effect changes what older adults notice and remember, and where it fails.
• Why loneliness in later life often tracks the loss of a single core tie rather than overall network size.

Every term the collection defines is gathered in the glossary.