The Women Holding the Shelter Together Aren’t on Any Staff List

She filed other people’s housing applications on her own computer skills. Another woman offered to teach craftwork in the shelter. Others walked each other to medical appointments, searched for flats together, and kept talking daily long after one of them had been rehoused across the city. None of that appears in a case file as a service delivered. A new study from the University of Calgary argues it should count as something precise: social capital, built by eleven women aged 57 to 65 who were themselves waiting for housing, care and benefits that kept not arriving.

The study, published in Frontiers in Global Women’s Health, followed the women through two in-depth interviews each and three focus groups between February and May 2025. Every participant lived with multimorbidity — two or more chronic conditions at once — spanning diabetes, heart failure, cancer, chronic pain and post-traumatic stress. Ten of the eleven reported past intimate partner violence. That health profile matters for how you read what they did for each other. This wasn’t spare capacity. It was labour performed by people who were already ill.

Hurry up and wait, then make lemonade

The researchers name two storylines. The first, “Hurry up and Wait,” is bureaucratic rhythm as harm: shelters described as unsafe yet nearly impossible to leave, a waitlist place lost if a case manager’s voice mailbox happened to be full, mandatory workshops repeated regardless of need. One composite figure, Marilyn, had a stroke at 54, lost work and driving, and faced rents above $1,200 against income support of less than $1,000 a month. Seniors housing sat behind an age-65 threshold she hadn’t reached. Most participants, aged 57 to 64, fell straight through that same gap.

The second storyline, “Making lemonade,” is the mutual care itself. The women formed bonding ties with peers and bridging ties with providers who treated them with dignity, repaired family relationships where they could, and advocated at building meetings in transitional housing. The authors borrow a phrase from urban geography: people as infrastructure. Informal networks patched over fragmented, under-resourced formal systems.

Hold on, though. The researchers are explicit that mutual aid cannot substitute for housing and healthcare, and they note other scholars’ warning that homelessness itself erodes the reciprocity such networks need. Celebrating resilience while leaving rents and waitlists untouched would be a misreading the paper itself blocks.

The same pattern, measured in Korea and Sichuan

Two other studies published this month point the same way from very different settings. In South Korea, researcher Sok An analysed 1,557 older adults living alone from the Korean Welfare Panel Study. Poor housing was linked to more depressive symptoms, but the interaction did the real work: as social capital rose, the depressive effect of housing deprivation weakened, and for people with high social capital depression stayed essentially flat whatever their housing quality. Rural residence made the same bad housing worse. An is careful — the design is cross-sectional, so causal inference isn’t possible from this data — but the shape is hard to ignore.

In Sichuan, a survey of 540 farming households across three flood-prone counties found social risk — thin networks, weak standing, missing safety nets — was the risk most consistently tied to lower resilience. Environmental risk, strikingly, showed no statistically significant association with any resilience outcome in that sample. The authors warn against reading that as proof floods don’t matter; measurement and timing may explain the null. Still, across Calgary shelters, Korean single-person households and Sichuan villages, the variable that keeps surfacing is connection, not concrete.

Analysis: Policy loves countable units — beds, vouchers, retrofits. The Calgary study suggests the counting itself is part of the problem. If a woman keeps three peers housed by sharing forms, rides and warnings about a dangerous floor, a system that records only her own file will call her a cost and miss her output entirely. Recognising that labour doesn’t mean paying her in praise instead of housing. It means designing caseloads, continuity and meeting structures so the network she built isn’t destroyed every time a worker rotates or a shelter placement scatters a floor of women across a city. The Korean and Sichuan results give that design question empirical weight beyond one small, deliberately intimate sample — while also marking its limit: eleven women, with Black, Indigenous and LGBTQ participants underrepresented, can’t stand in for everyone. What they can do is show, in close detail, work that aggregate statistics have been stepping over.

The women’s own fear, named in the study, was institutionalisation. Not abstraction. A locked ward, a managed life. Against that fear they built something unglamorous and daily: a phone call, a shared application, a joke at the shelter desk that made a staff member smile for the first time. Systems will keep failing to see it until they are built to look.

Sources

1. Science Magazine summary (Scienmag) — “Older Women Facing Homelessness Turn Mutual Care Into Unrecognized Social Capital” (16 October 2026), reporting Nixon et al., Frontiers in Global Women’s Health: https://scienmag.com/older-women-facing-homelessness-turn-mutual-care-into-unrecognized-social-capital/
2. Scienmag — “Poor Housing Hits Hardest When Community Ties Are Weak, Study of Older Adults Finds” (9 October 2026), reporting An, BMC Psychology: https://scienmag.com/poor-housing-hits-hardest-when-community-ties-are-weak-study-of-older-adults-finds/
3. Scienmag — “Social Ties, Not Floods, Decide Whether Rural Households Bounce Back, Sichuan Survey Finds” (19 October 2026), reporting Liu et al., Frontiers in Sustainable Food Systems: https://scienmag.com/social-ties-not-floods-decide-whether-rural-households-bounce-back-sichuan-survey-finds/