Funding by Mechanism, Not by Symptom

A commissioner with a loneliness line item usually gets pitched the same three things: a befriending scheme, a men’s shed, and a walking group. All three get funded under the same rationale — “tackles loneliness” — and all three get evaluated the same way, with a pre/post loneliness scale. This is how a lot of loneliness money gets spent, and it is close to the wrong way to spend it.

The 2015 meta-analysis by Julianne Holt-Lunstad and colleagues gives a reason to be more precise. It treats social isolation, loneliness, and living alone as three separate risk factors, with different odds ratios for early mortality, and notes the effects held up even after controlling for health status. These are not three names for one problem. They are three different deficits, and an intervention built for one does not automatically fix another.

Three deficits, not one

Isolation is a structural fact: how many people you have contact with, and how often. Loneliness is a subjective gap: the difference between the connection someone has and the connection they want. Living alone is a household configuration that correlates with both but is neither. The AARP Foundation’s 2018 survey of adults 45 and older found that a person’s loneliness is best predicted by the size and diversity of their social network and by physical isolation — but it also found that people who talk to their neighbours are lonely at roughly half the rate of those who never do, which is a contact effect, not a network-size effect. Different levers move different things.

The consequence for commissioning: a programme that increases contact frequency (a lunch club, a phone check-in service) is treating isolation. A programme that increases the sense of being understood or valued (peer support, structured group activity with shared purpose) is treating loneliness. A programme that gets someone into shared physical space regularly (a library programme, a community centre, what Eric Klinenberg calls social infrastructure) can do either, depending on design, but doing either well requires knowing which one you’re aiming at.

The failure mode here is worth naming: the check-in trap. A funder buys a weekly phone call service for isolated older adults, measures loneliness before and after, and finds little movement. The service wasn’t wrong — it increased contact, which is what it was built to do — but it was evaluated against a loneliness scale, which measures something more subjective and slower-moving. The programme gets marked as ineffective and defunded, when the actual failure was matching the wrong outcome to the wrong mechanism at the funding stage, not at delivery.

What the social prescribing literature actually supports

Social prescribing gets funded as a general loneliness remedy, and the evidence doesn’t fully back that framing. A 2022 qualitative meta-synthesis found that participants describe benefit extending well beyond social contact — toward restored purpose and meaningful participation — and that structured, purposeful group activity outperforms unstructured contact. A 2021 systematic review reported gains in self-esteem and confidence but flagged limited trial evidence and heavy heterogeneity across programmes. Read together, these point to a specific claim, not a general one: structured activity with a purpose attached does more than contact for its own sake, for people whose deficit is the subjective gap rather than raw frequency of contact.

That is a narrower claim than most social prescribing pitches make, and it is the one worth funding against.

An evidence-status table for common interventions

Intervention type	Deficit it plausibly addresses	Evidence status
Regular check-in calls / visits	Isolation (contact frequency)	Plausible mechanism, thin outcome evidence
Purpose-built group activity (skills, volunteering)	Loneliness (subjective gap)	Reasonably supported for self-esteem/confidence; heterogeneous trials
Befriending / peer matching	Loneliness, secondarily isolation	Mixed; depends heavily on match quality and duration
Shared physical space (library, community centre)	Isolation, structural contact	Strong observational case (Klinenberg), few controlled outcome studies
Household composition changes (co-housing, home-sharing)	Living alone	Rarely funded as a loneliness intervention at all; mechanism is different from the other two

What this means in practice: before writing a request for proposals, decide which of the three deficits — isolation, loneliness, or living alone — the money is meant to move. Then require applicants to state their mechanism and pick an outcome measure that actually tracks it. A contact-frequency intervention should report a contact-frequency measure, not just a loneliness scale.

Why targeting matters more for scarce money

The National Academies’ 2020 consensus report calls for the health care system to routinely assess isolation and loneliness — precisely because they are different things needing different responses, and a single screening question won’t tell a clinician which lever to pull. The 2023 Surgeon General’s advisory frames the whole area as needing a multi-pillar national strategy rather than one intervention type scaled everywhere, which is an implicit admission that no single programme design fits all three deficits.

Funders with limited money face a choice most guidance skips: fund fewer things well-matched to a defined deficit, or fund more things loosely labeled “anti-loneliness.” The second option spends the same money against a vaguer target and gets vaguer evidence back. The American Enterprise Institute’s 2021 survey found the share of men reporting no close friends rose from 3% in 1990 to 12% by 2021 — a friendship-network collapse, structurally closer to isolation than to subjective loneliness. A programme designed to address that needs to build network size and diversity, not just make people feel briefly less alone in a room once a week.

What this does not solve

None of this addresses reach. Sorting interventions by mechanism improves what happens to people who show up; it does nothing for the isolated person who never enters a commissioning conversation at all, because they have no link worker, no GP referral pathway, and no contact with any service that might notice. Precision targeting is a tool for spending well among people already inside the system. It is not a substitute for outreach, and funders who mistake better-matched interventions for wider reach will fund excellent programmes that serve the same population that was always going to be served.