Fund the Roles, Not Just the Room

Two proposals land on your desk. Both ask for money to run a weekly group for isolated adults. Both describe a room, a time, refreshments, and an expected headcount. One of them also describes who greets a newcomer at the door, who keeps a simple attendance log, and what happens to the group if the person currently running it leaves. The other does not mention any of that.

Fund the second kind, or send the first kind back with questions. That is the practical content of this note.

Why role structure is the thing to check

A room with chairs in it is not a programme. It is the precondition for one. What determines whether a weekly gathering turns into sustained connection, rather than a slow-leaking attendance sheet, is largely who is responsible for what inside it.

A 2022 qualitative meta-synthesis of social prescribing found that participants describe the benefit of group activity as extending well beyond having someone to talk to — they describe restored purpose and meaningful participation, and the synthesis notes that structured, purposeful activity appears to work better than contact alone. Unstructured contact — people sitting in a room, hoping conversation happens — is the weaker version of the same intervention. A 2021 systematic review of social prescribing similarly ties reported gains in self-esteem and confidence to programmes with defined activity and roles, not to informal drop-in contact.

The strongest available causal evidence points the same way, even though it comes from a different intervention entirely. The HEAL-HOA trial, a randomised controlled trial published in the Lancet Healthy Longevity in 2024, tested prosocial engagement and volunteering — giving lonely older adults a defined role helping others — against a control condition. It is one of the only randomised trials in this literature, which is otherwise dominated by uncontrolled programme evaluations. The design itself is a signal: the researchers built the trial around giving people something to do for someone else, not simply somewhere to be.

None of this proves that assigning roles inside an ordinary social group causes better outcomes than leaving it unstructured — that specific comparison has not been run as an RCT. But the pattern across social prescribing research and the one strong trial available all point the same direction, and it costs a funder nothing to require applicants to think about structure rather than assume it will emerge.

What to require in the application

Ask every group-based proposal to answer four questions before you fund it.

1. Who welcomes a first-time attendee, by name, and is that a rotating job or one person’s job? Robert Putnam’s account of declining associational life describes exactly this kind of informal welcoming role disappearing along with the clubs and lodges that used to assign it. If the proposal has no answer, the welcome is happening by accident, if at all.
2. Who tracks who stopped coming, and what do they do about it? A follow-up call after two missed sessions is cheap and rarely specified in advance.
3. Is there a task, beyond attending, that a member can be asked to take on? Setting up chairs, bringing something, leading a short segment. A 2025 study of college students found that the quality and regularity of connection predicted wellbeing more than sheer volume of social contact — a task gives a reason to show up regularly and a reason to matter, which passive attendance does not.
4. What happens to the group if the person currently running it leaves? This is the question applicants answer worst, and it is the one that predicts collapse.

The founder trap

Name the failure mode: the founder trap. One capable, generous person builds a group, becomes its host, greeter, treasurer, and emotional center, and the group runs well for as long as they are in the room. Then they move, burn out, or get a different job, and the group does not survive them because no role was ever separated from that one person.

Funders see this constantly and rarely ask about it in advance, because a thriving group with an energetic leader looks like a success story in year one. It is also the single most common reason a funded group disappears in year two. Ask for a named deputy or a rotation schedule as a condition of a second-year grant, not as an afterthought.

Evidence status

Claim	Evidence status
Structured, purposeful group activity produces better reported outcomes than unstructured contact	Supported — consistent finding across qualitative synthesis and systematic review of social prescribing
Giving people a defined prosocial role (e.g., volunteering) reduces loneliness	Supported by RCT evidence, though from one trial in a specific population (older adults in Hong Kong)
Rotating or shared roles specifically prevent group collapse when a leader departs	Plausible, consistent with the social infrastructure literature, not directly tested
Network diversity and physical proximity predict lower loneliness	Supported — large national survey evidence

What this means in practice: treat “who does what” as a fundable line item, not a detail. A proposal that names a rotating host, a newcomer-welcome role, and a succession plan is describing infrastructure. A proposal that describes only a room and a schedule is describing a hope.

What this does not solve

None of this addresses reach. Structured roles make a group more durable for the people already inside it; they do nothing for the person who never walks through the door in the first place. Eric Klinenberg’s argument about social infrastructure and the AARP Foundation’s 2018 survey both point to the same prior barrier — physical isolation and thin networks keep people from finding the group at all, long before anyone gets to argue about how it is run once they arrive. Fund outreach and fund structure as two separate line items, because they solve two different problems.