Viewpoint: Homeless response must include congregant shelter
Jill Bonny
There’s been a lot of talk lately about emergency shelter, especially congregate shelter, not being a best practice. I want to take a moment to set the record straight and to re-center this conversation on the inherent dignity of all people, regardless of their housing status, behavioral health, or substance use.
Congregate shelter typically refers to shared living environments, like dormitory-style shelters, where multiple people sleep in the same space. Non-congregate shelter , by contrast, provides individual rooms or units, such as motel rooms, tiny homes, or converted apartment spaces, offering more privacy and autonomy for residents.
It’s true that national research and best practice literature, including from the U.S. Interagency Council on Homelessness and the National Alliance to End Homelessness, shows that non-congregate, housing-focused models tend to produce the best long-term housing outcomes. However, that research reflects only part of the picture. Those outcomes often come from programs that serve individuals who are already relatively stable, who can meet high behavioral or programmatic expectations, or who are selected because they’re most likely to succeed.
What’s often missing from this conversation is recognition that many non-congregate models, particularly high-barrier ones, are not designed for, or accessible to, individuals experiencing acute mental illness, active substance use, or high levels of trauma. Research from the University of California, San Francisco, and others has shown that these models frequently lack the onsite supports, clinical staffing, and design considerations necessary to engage people with the most complex needs. In fact, a 2024 study in BMC Health Services Research found that while non-congregate settings offer privacy, they can inadvertently isolate individuals with serious mental illness and fail to support their recovery and housing transitions.
This is why low-barrier, congregate emergency shelter remains an essential part of a balanced homelessness response system, especially in communities like ours, where non-congregate capacity is extremely limited and existing options are too high-barrier for many who are unsheltered today.
When the Johnson Street shelter closes as planned, it will leave a glaring hole in our local shelter system, removing more than 170 beds and dramatically reducing the number of low-barrier options available. We will see more people outside, not because they’re unwilling to engage, but because we haven’t created accessible options that meet them where they are.
This closure will also place significant added pressure on the Poverello Center’s main shelter, which is already operating near capacity. As demand increases, we will be called to stretch farther and do more, with the same limited space and staffing. We are committed to continuing to provide safe, low-barrier shelter and support to as many people as we can, but we will need the community’s support now more than ever to do this work safely, compassionately, and effectively.
Yes, we should absolutely be working toward more trauma-informed, non-congregate, housing-focused programs. But those take time and money to scale. And in the meantime, people need a safe place to be tonight. Congregate shelter can fill that gap when it’s done thoughtfully, with low barriers, strong engagement, and a housing-first approach.
We are not arguing against non-congregate models. We need those. But we also need to acknowledge that they’re often not accessible to individuals with the highest levels of vulnerability. These are often the folks we see sleeping outside, the people most at risk of harm or death. And they’re the people who will continue to be left out if we only build high-barrier models that work well on paper but exclude those most in need.
One of my colleagues recently described a man who had cycled through both congregate and non-congregate shelters, yet still couldn’t find a place that could truly meet his needs. He was, in her words, “a person whom society failed.” And sadly, he is not alone. Until we build systems that are truly accessible to everyone, including those with the most complex challenges, we will keep failing people like him.
Our broader systems of care, especially mental health treatment, substance use services, and accessible long-term support, are deeply under-resourced. Many people experiencing homelessness are doing everything they can to survive and move forward, but they’re navigating complex challenges without the help they need. If it takes someone time to stabilize, it isn’t because they’re not trying, it’s because the resources that should be there to catch them too often simply aren’t.
Until we build systems that are truly accessible to everyone, including those with the most complex challenges, we will keep failing people like him.
If we are serious about reducing unsheltered homelessness, protecting public health, and moving people toward permanent housing, then we must invest in both low-barrier congregate and non-congregate shelters. The research supports a systems approach. Our policies should, too.
Jill Bonny, LMSW, is the executive director of the Poverello Center