Caregiver of last resort

Why does the care of an ageing parent collapse onto one sibling while the others remain willing but uninvolved?

Four adult siblings, all fond of their mother, all willing in principle to help. Within a year or two, one of them is doing nearly everything and the other three are doing very little.

The striking part is that nobody refused. Ask the uninvolved siblings and they will say, sincerely, that they would help if asked and that they had no idea it had got this heavy. Ask the involved one and they will describe a load that arrived without ever being agreed to. An outcome nobody chose and nobody defends has been produced by four people who each behaved reasonably, which means the explanation is structural rather than a story about character.

Start with the moment before anyone is the carer. A need appears — a fall, a diagnosis, a bill nobody has paid. It is urgent, it is nobody's assigned job, and there is no mechanism that allocates it.

Now ask what each sibling faces. Everyone would prefer that the thing get done. Everyone would also prefer that someone else do it, since doing it costs time, money and career. And critically, the need will be met if any one person acts — so each sibling's best move depends on whether they expect someone else to move first.

That is a recognisable structure, and it has a known and uncomfortable property: the person who acts is not the one with the most duty, the most capacity, or the most free time. It is the one with the lowest threshold — who lives nearest, worries most, sleeps worst, or simply cannot bear the thing being left. Selection is on tolerance for the situation, not on suitability for the job.

So far this explains who starts. Now follow what happens next, and this is where the outcome locks in.

Once one person has acted, the situation is no longer unallocated. There is now an incumbent, and three consequences follow immediately.

The need stops being visible. A crisis that would have prompted a sibling to act is now absorbed before they hear about it. The uninvolved are not ignoring a problem; they are not seeing one, because the carer's whole function is to make it stop being a problem.

Information concentrates. The carer learns the medications, the consultant's name, which neighbour has a key. That knowledge makes them faster than anyone else at every subsequent task — so handing a task over costs more than doing it, and the efficient move each time is for them to keep it. Efficiency ratchets the load onto the person who is already carrying it.

The comparison point moves. A sibling deciding what counts as a fair contribution compares themselves not to the need but to what is currently being covered. Since it is covered, doing a little looks adequate. The incumbent's competence has quietly redefined "enough" downward for everybody else.

Now the willingness makes sense rather than being hypocrisy. The uninvolved siblings are genuinely willing — but willingness is a disposition to respond to a request, and the structure has removed the requests. And asking is expensive for the carer: it means specifying the task, explaining the context, chasing it, and risking a conversation about fairness that could damage the relationship at the worst possible time. Very often the cheapest available move, on any given Tuesday, is to do it themselves. Every one of those cheap decisions is correct, and they sum to the outcome nobody wanted.

This is the designation problem, not the accumulation problem. The companion account of caregiving load explains why the burden ratchets on whoever is carrying it — proximity, information, and the fact that competence attracts more work. That account takes the carer as given. What is being explained here is the step before: how one particular person became the answer, without a decision, out of a group of equally willing people. The two compose — selection by lowest threshold, then ratcheting by information — and the second is what makes the first irreversible.

Willingness and availability are not the same thing, and families argue as though they were. The uninvolved sibling's "I would have helped if you'd asked" is usually true and beside the point, because the structure suppresses the asking. The carer's "I shouldn't have to ask" is also true and also beside the point, because nobody can respond to a need they cannot see. Both are describing the same missing mechanism from opposite sides.

The selection is not random. Who has the lowest threshold is patterned: by geography, by gender, by which sibling has less bargaining power, and by whose job absorbs interruptions. An outcome produced by an allocation failure still falls in a predictable direction, which is why it is a labour question as much as a family one.

What the evidence supports and what it does not. That care concentrates on one adult child is well attested, as is the gendered pattern. The mechanism above is a reconstruction fitting those observations; the volunteer-selection step has not been isolated from its rivals — pre-existing closeness, an implicit family expectation, or plain geography could each produce a similar concentration. Separating them needs data on sibling sets before the need arose, and such designs are rare.

Nothing here is advice. This describes why the arrangement forms; how any family should divide care is theirs to decide. Findings here are population patterns and say little about any individual case.

The falsification test. If the mechanism is a missing allocation step rather than differing willingness, then introducing an explicit allocation — a scheduled rota, a shared task list, a third party who assigns — should redistribute care among siblings whose stated willingness was unchanged. If siblings who professed willingness declined tasks once they were specifically and easily assignable, the account would be wrong and the difference would be in willingness after all.