Think Out Loud

California homeless study may provide insight for Oregon

By Sage Van Wing (OPB)
July 24, 2023 4:19 p.m. Updated: July 31, 2023 9:28 p.m.

Broadcast: Monday, July 24

A new study from the University of California, San Francisco is the largest study of homelessness in the U.S. in several decades. The California Statewide Study of People Experiencing Homelessness includes 3,200 administered questionnaires and 365 in-depth interviews with adults experiencing homelessness in urban, rural and suburban areas. We talk to Margot Kushel, principal investigator on the study, about what researchers learned and how the data might be relevant outside of California.


Note: The following transcript was created by a computer and edited by a volunteer.

Dave Miller: This is Think Out Loud on OPB. I’m Dave Miller. We start today with the largest study of homelessness in the US in several decades. It came from researchers at the University of California, San Francisco. They collected 3,200 questionnaires and more than 350 in-depth interviews with adults experiencing homelessness around California in urban, rural and suburban areas. Doctor Margot Kushel is the principal investigator of the study. She is a director of the UCSF Fending Off Homelessness and Housing Initiative.

She joins us now to talk about what researchers learned and how it might be relevant outside of California. Doctor Kushel, welcome.

Margot Kushel: Thank you for having me.

Miller: What was the purpose of this study?

Kushel:  We really wanted to get a handle on who was experiencing homelessness in California. How did they come to be homeless? What were their experiences, and what was  keeping them there? We really wanted to design this, to help policymakers design policies and to try to get above the rhetoric a little bit and do really careful science to really understand the problem.

Miller: How is this different from the once every two years, ‘point in time count.’

Kushel: So, ‘point in time’ counts are designed as a way to count the number of people experiencing homelessness. Each jurisdiction, which can be a county, or part of a county, or several counties, uses slightly different methods and some do surveys, some don’t. But this was really meant to give us a sense, not of how many people in California are experiencing homelessness, but to be able to represent the experience of the state as a whole rather than a specific jurisdiction.

Miller: Can you give us a sense for how this was carried out?

Kushel: Yeah. Faced with a state as huge as California, we knew that we couldn’t go to every place. And so we divided the state up into eight sections that were based on some other ways people have divided the state. And then we learned everything we could about each of the counties in California, and we basically threw it all into a big computer model and had it choose for us one county from each of those eight sections that we would be able to represent the state in each of those counties.

We spent months ahead of time learning every place where adults experiencing homelessness might be –  so homeless shelters, encampments, free and low cost food lines and the like, and then we learned a lot about them and developed a sampling scheme. So we basically randomly chose for amongst those and had a plan, like we would talk to every third person or every fifth person in those areas. We sent teams that included our own researchers who traveled around as well as people that we hired from the area. We had them accompanied by people with lived experience who really knew the areas, knew the ins and outs, could help make other people comfortable.

We had IPADS and portable WIFI and went in and asked a survey which lasted about 45 minutes. Did it in English and Spanish, used interpreters for others. And then based on people’s answers to questions, we chose about one in every eight people for what’s called an in-depth interview where we just turn on a tape recorder and ask a series of open ended questions to really get at the full depth of their experience.

Miller: How did the researchers along with the people with lived experience get people to actually say ‘yes’ to tell their stories? I mean, even just the questionnaire alone, that’s 40 minutes. That’s not nothing, let alone the in-depth interview with the tape recorder. What was their way in?

Kushel: I mean, we were really blessed, we did this study and had a group of people with lived experience advising us from day one, helping us to craft the questionnaire and the in-depth interviews, but importantly, helping us to strategize on what would make people comfortable, honest, willing to accept. And then we had local folks with lived experience on the ground, sort of going ahead of us and saying, ‘Hey, I’m coming in with a team of researchers, they’re good people, I trust them. They’re, they’re not here to judge you.’

We did it anonymously in all but one county where we voluntarily, at the end of the interview, asked for people’s names so we could link data, but mostly we did it anonymously. We spent a lot of time helping people understand that we were not ever gonna name them. We’re not ever gonna know about them, but that it was important for us to see the experience of people who had been homeless. And we offered a small gift card, usually in most places, it was like a gift card that was turned in that could be used at just about any vendor rather than a specific store, because we heard from our lived-experience-four that many people felt uncomfortable in certain stores that they might be followed around. So it was like a card that they could use in just about any place they wanted. And people said yes.

We kept really careful track of who said ‘yes,’ who said ‘no.’ And used that because we basically weighted the answers just like the census does for what we call non-response. But we were looking to see if any particular types of people might have said ‘no.’  But really, we had an overwhelmingly positive response because we used all of these methods.

Miller: And that double checking was a way to try to prevent some version of a self-selection effect where the results would be skewed by virtue of who agreed to take part.

Kushel: Absolutely. You know, when someone doesn’t agree to take part, you have only limited information about them, but we had agreed upon several observable characteristics. We also were really careful that if someone couldn’t consent because it was a research study, so people had to give informed consent. And if someone, let’s say, couldn’t give informed consent because they were intoxicated or because they had dementia or something, we were able to keep track of that and then take that into account in the statistics at the end.

Miller: So let’s turn to the meat of this. And I’m curious before we get to any of the nitty gritty specifics, you’ve been working with people experiencing homelessness, first as a physician, and now as both a physician and the head of a much larger research institution for decades now. Were you nevertheless surprised by anything that came out of this study?

Kushel: There were a few things that would probably surprise most people that didn’t surprise me as much, but even I was surprised by a few things. I think I was surprised by how optimistic people were that relatively small amounts of money would have prevented their homelessness. I think I was surprised by a few other aspects of peoples’ experiences. I was surprised, for instance, by how many of the women were pregnant during this episode – that really jumped out at me, where 26% of all people who were women or ciswomen or assigned female at birth had been pregnant during this episode of homelessness. I think I was maybe a little surprised by how disconnected folks who were unsheltered were, how little help they had been receiving to get out of this predicament.

Miller: Let’s talk about geography for a second. There is an idea that liberal cities or liberal states or places that have less punitive approaches to homelessness are magnets - magnets to people who don’t have a place of their own to live. What does the data show in terms of where people were before they became homeless?

Kushel: So this would be one of those things that might surprise most people, didn’t necessarily surprise me, but I think it’s really important to know. We did not find any evidence of people moving towards ‘magnet regions,’ and certainly not moving towards California. So 90% of everybody lost their stable housing in California. California had been where they were really living, not living homeless, but living in a house, before they became homeless. 75% were in the same county that they lost their last housing in. Even amongst the 10% who had come from outside California, that was generally because they had deep ties to California. And so we drew this from the surveys and from the in-depth interviews, people really were Californian and we did not see any evidence of a ‘pull’ towards areas that might sort of have more resources.

We saw some movements, like people who had been disrupted by wildfires, were a good example of people who had often fled one region because they just realized there was no housing coming anywhere near there. But overall, we did not see a lot of movement and certainly not movement in the way that the public views it might be happening.

Miller: How do you explain that disconnect between the reality and a very common public understanding?


Kushel: I think it’s important to note that all over the country people believe this, that if you go to other regions of the country, everyone thinks that people experiencing homelessness are from elsewhere. I think on one hand, it’s a natural way of disbelief that we could be doing this to our own neighbors, our own friends. I think it’s a way to subtly let ourselves, the housed community, off the hook for what has happened. And I think it ultimately becomes dangerous because it becomes a way to dehumanize people to treat people as ‘less than’ or ‘other.’

But what we know and what all our research, but many other forms of research have shown, is that people do stick close to home and that what people said is they were homeless but their home was California or their home was a county that they were in, that, first of all, when people lose their housing, the last thing that they’re really able to do is garner the resources to pick up and leave everything that they knew behind. It’s actually just hard to do. I think the reason though people have this idea is that cities and particularly often the same cities that have a lot of benefits, tend to be cities that have a tax base. Cities have a tax base, often, because they have a lot of wealthy people who can pay more taxes on, you know, property taxes.

And so what we know is that homelessness varies between regions very, very much by the cost of housing in that area, and particularly the disconnect between the cost of housing and the incomes of low income people. In our study, the median monthly household income in the months before homelessness was $960 a month. That’s about 15% of the area median income within California. People were desperately poor and the wealthier the area was, often the higher the housing costs, and the bigger the disconnect between what people could afford and what they could find.

Miller: This gets to the bigger issues of the reasons for people becoming homeless and the connections between homelessness and housing and housing affordability.

But you’ve said in the past that for every 100 people who are at risk of homelessness, only one will become homeless; does this massive study give you a better understanding of what makes the difference between that one and the 99?

Kushel: Yeah, really what we see when we look at people experiencing homelessness, unfortunately, is the tip of a much larger iceberg. And this makes things like preventing homelessness harder, because it’s hard to predict who’s gonna be there. We found for most people that it was a slow slide or, things fell apart slowly and then all at once.

For most people, almost 20% came from institutions. So these are folks who were discharged from jails or prisons, hospitals, drug treatment programs, without any place to go. Of everyone else? They came from housing; 60% of those were what we call non-leaseholders. They were already doubled up with friends or family. They were staying on the good graces of someone else who they could stay with. And this turns out to be a very, very high risk situation, because they have no legal rights. They often are really dependent on friends or family and they are often living in overcrowded tense situations. We found that these fell apart very quickly, within a day’s notice. 40% of people came straight from a situation for which they had legal rights – were tenants, mostly, a few were homeowners. What we found that was really interesting is even in that group, they had had a slow slide down to homelessness. Meaning that they told us, they had lived in a two bedroom apartment and then somebody lost their job, and then they couldn’t afford it. So they all piled into one room in a house and then something went wrong there.

So I think in terms of finding that secret sauce, we learned there were certain key events - the biggest one was loss of household income. So what you need to do to prevent [homelessness] is to look at the people who are really at risk. Those are people who are spending more than 50% of their household income on rent. Those are people who are very poor, less than 30%, maybe less than 15%, of the median income of the area. Ever having been homeless before is a huge risk factor for being homeless again. But we also need to look for people who are going through these life events that we know could be triggers- someone in the household losing income. [For] someone who has these characteristics, we should be jumping in to prevent that episode from destabilizing the whole household…

Miller: How would that ‘jumping in’ work? Because we’re talking about huge systemic policy questions, different pots of money. And there’s even the question of how the people who might have that money, if it were set aside as a policy, how they would even know who to give it to, when. What’s the mechanism that you imagine?

Kushel: So there’s some really good research and practice coming out of New York City, where there’s a right to shelter. So what that means is they actually know as soon as people become homeless because they present to shelter, where they have set up focused-on areas where people are most at risk, which are often low income areas, formerly red lined areas, areas that are dense in people who had lots of structural depression, People of Color, et cetera. They have these neighborhood hotspots and they do things like post on the subway and buses, ‘Hey, are you worried about losing your housing? If you are, come visit us.’

When people present there, they actually ask them a series of questions that are actually scored because they know the things that  have in the past predicted who’s going to become homeless. Basically, they give resources really only to the people who score high enough. So some of the people come and they’re unfortunately turned away, but they’re turned away because they really want to be sure that they concentrate their homeless prevention activities in the people who are at the highest risk.

Miller: And in the end, are we really talking about cash, about emergency rental assistance, say?

Kushell: That seems to be the most important thing. And in our study, you know, 70% of people said $300-$500 a month of rental assistance would have meaningfully prevented their homelessness. 82% said $5,000-$10,000, one time, would have, and a much higher percentage said something that sounds like what we call a Housing Choice or Section 8 Voucher would have prevented it. So these organizations in New York do a lot of that – catching people up on rental assistance. But they do a lot of things, like they will provide mediation with the landlord or property owner. They will buy new furniture because maybe your sister is willing to take you in, but there’s no place for you to sleep, and so maybe they’ll buy a bunk bed. They really will do a lot of different things. They’ll make sure that you’re signed up for benefits. They’ll help you get a job, they will make sure your legal rights are exercised, but cash is certainly a very important part of it.

Miller: Just to stick with this question of preventing homelessness for a second. What is the argument that you would make, the coldhearted financial argument you would make, if we just set aside human misery for a second, preventing human misery as a good in and of itself. If that’s not enough for people who are looking at the use of public funds, what’s the justification for emergency rental assistance in terms of saving money down the line?

Kushel: Clearly, the most important thing is the personal cost of people experiencing it, but it is extremely expensive for societies to do. Things like emergency shelter costs between $100-$300 a night and they don’t end homelessness. Public works, policing costs, excess use of jails, excess use of hospitals, we pay for homelessness, just the impact on businesses. You see this in downtown San Francisco - there’s a lot of concern about visible homelessness as decreasing tourism. That’s what people are afraid of, decreasing peoples’ sense of public safety.

There’s so many ways that society pays for homelessness that it would make a lot more sense to put those funds into both preventing it happening and really ending it, by solving our affordable housing crisis. California right now has 24 units of housing available and affordable for every 100 extremely-low income households. Oregon has 23. Really dealing with the roots of this problem is going to cost money. But we have to be honest, we’re spending that money, we’re just spending it in not very effective ways.

Miller: I want to turn to behavioral health. I’m thinking about untreated mental illness or serious substance use disorder. How often were these issues a result of homelessness as opposed to the other way around? And is that even a question that you can answer?

Kushel: Yeah, we try to get at it in different ways, asking about, for people, let’s say who have had a psychiatric hospitalization, did the first hospitalization happen before or after their first episode of homelessness? We talked to people about their substance use and how it changed during this episode. And I would say it is both a contributor to the cause and it is an effect and it’s hard to sort those out. There’s no question that if you ask the question, ‘Why does one region have more homelessness than another region? Why do California and Oregon have so much higher per capita rates of homelessness than Ohio and West Virginia?’ That’s 100% because of housing costs, the higher the housing costs, the more homelessness.

If you ask a different question, ‘Why is this person in Oregon? Why is this person in California experiencing homelessness?’ There’s no question that having a substance use or a mental health disability increases one’s personal risk for homelessness. But there’s also no question that homelessness complicates and worsens those. We saw incredible amounts of trauma - 10% of everyone in our study was sexually assaulted during this episode of homelessness, more than a third were physically assaulted. People were exhausted, hungry, scared, very few people were getting any support for mental health problems or substance use problems even though many wanted it and had actually tried to access it.

I think the other important point is…first of all homelessness worsens mental health problems, worsens substance abuse, and complicates or lessens treatment or the ability to address these problems. The flip side has also been true and we’ve been, and it’s been shown empirically is that if you offer people housing, you don’t make it contingent on their sobriety, but you offer them robust and low barrier treatment, two things happen: people get housed and stay housed no matter how severe these disabilities are, but they then take you up on your offer for mental health health and substance use treatment. Really at the heart of this problem is we don’t have the housing. It’s not to say that folks with behavioral health disabilities aren’t way overrepresented. It’s a complicated circle, but to interrupt that we have to get them housing and then we need to provide those services and make them really easy for people to get.

Miller: You mentioned the discrepancies currently between states like Oregon and California, in terms of homelessness and other places, I think you mentioned Ohio and West Virginia, and pointed to housing affordability as the clearest empirical reason for that difference. But often in American society, what happens in California eventually happens in other states. Do you think that’s going to be the case for homelessness as well?

Kushel: Yeah, I’m afraid that what we’re seeing in California is definitely a model in a not a good way, up and down the west coast and throughout other parts of the country as well. Any place that has these high housing costs that are places where people want to live - that’s why we love Oregon and California. There are places that are wonderful places to be in many ways, many people come in. But California has done a terrible job keeping up with their housing supply. Housing shortages, housing scarcity leads to increased housing costs.

As a society, we’ve really backed away from a commitment to subsidize housing or to see housing as both a responsibility of government to help make sure everyone has it. Only one in four people across this country who qualify for rental subsidies, receive them. There has been very limited spending from the federal government, states, et cetera on subsidizing housing for low cost individuals. So I see our results as really similar to what we’d see in many parts of the country and in some ways, it is a warning, ‘Don’t become like us,’ address your housing crisis before it becomes as bad. We need to get the federal government bought in to help fund affordable housing, and we need to let go of some of these myths that we’re attached to that aren’t helping us solve the problem.

Miller: Margot Kushel, thanks very much.

Kushel: Thank you so much for having me.

Miller: Margot Kushel is a physician and the director of the UCSF Fending Off Homelessness and Housing Initiative, also the principal investigator behind this major new study that collected the experiences of thousands of people experiencing homelessness all across the state of California.

Contact “Think Out Loud®”

If you’d like to comment on any of the topics in this show, or suggest a topic of your own, please get in touch with us on Facebook, send an email to, or you can leave a voicemail for us at 503-293-1983. The call-in phone number during the noon hour is 888-665-5865.