Ep 65: Reducing Homelessness with Unconditional Cash Transfers with Jiaying Zhao (Pathways Home pt. 5)

What happens when you provide unhoused people with a large sum of money? Jiaying Zhao shares the results of a study in Vancouver, BC, which include reduced shelter use, more spending on food and rent, and no increase in spending on “temptation goods” like drugs and alcohol.

Show notes:

• Dwyer, R., Palepu, A., Williams, C., Daly-Grafstein, D., & Zhao, J. (2023). Unconditional cash transfers reduce homelessness. Proceedings of the National Academy of Sciences, 120(36), e2222103120.
• Mullainathan, S., & Shafir, E. (2013). Scarcity: Why Having Too Little Means So Much. Macmillan.
• Brisson, D., Calhoun, K. H., Coddington, L., Flaxman, Z. J., Johnsen, M., & Locke, S. (2023). Denver Basic Income Project Interim Report. Center for Housing and Homelessness Research, University of Denver Graduate School of Social Work.

Transcript

Shane Phillips  0:04  
Hello, this is the UCLA Housing Voice podcast and I'm your host, Shane Phillips. This is the fifth installment in Pathways Home, our series on research into the causes of and solutions to homelessness. Our guest this time is JZ Zhao to talk about a study she ran in Vancouver, BC, which provided a one-time unconditional cash transfer of $7,500 CAD to people experiencing homelessness and compared their outcomes to a control group receiving typical homelessness interventions. One of the ideas behind a lump sum payment is that it might help people get out of survival mode and free them to think about their longer term future, or to make big purchases like a car or a security deposit on a rental that they need to get back on their feet. JZ and her colleagues tested that theory and a few others. They found that the lump sum cash transfer dramatically reduced the time people spend in shelters and increase their days in stable housing compared to usual treatment. And recipients did not, on average, increase their spending on "temptation goods" like drugs, cigarettes, and alcohol. They did spend a lot more on durable goods, rent, food, and transit. The researchers also found that the savings from reduced shelter use was enough to fully offset the $7,500 payment, meaning a well-targeted program has the potential to save the government money. Over the next hour we will get into those results, findings from a survey testing how public opinion is affected by different messages about this intervention, and what comes next for this area of homelessness research. 

Before we get to the interview, a quick rundown on the remainder of the Pathways Home series. In our next episode, we'll be talking with Tim Aubrey about a major study testing the effectiveness of the Housing First model for assisting chronically homeless individuals. And I can now share that our seventh episode will feature a conversation with Monica Diaz and Shawn Liu, the Executive Director and Director of Communications, respectivel,y for the Veteran Homeless Programs office at the U.S. Department of Veterans Affairs. We'll be talking to them about how the U.S. managed to cut veteran homelessness by more than half since 2009, even as overall homelessness has not really budged nationwide. And finally, we are adding yet another episode to this series. In the last episode — the really last episode — we will be summarizing the main takeaways from this series of conversations and some of the concepts, insights, and messages that really stuck with us. If there's anything you want to make sure that we cover, or any burning questions you want answered, send them my way at shanephillips@ucla.edu. The Housing Voice podcast is a production of the UCLA Lewis Center for Regional Policy Studies, with production support from Claudia Bustamante, Gavin Carlson, and Jason Sutedja. Again, you can send comments and questions to shanephillips@ucla.edu. With that, here's our conversation with JZ Zhao.

JZ Zhao is associate professor in the Department of Psychology and the Institute for Resources, Environment and Sustainability at the University of British Columbia. And she's here to discuss her research on the impact of one time unconditional cash transfers of $7,500 CAD on the lives of homeless people in Vancouver, BC, JZ. Thanks for joining us, and welcome to the Housing Voice podcast.

Jiaying Zhao  3:36  
Thanks for having me. 

Shane Phillips  3:37  
And my co host today is Mike Manville. Hey, Mike. 

Mike Manville  3:41  
Hey, great to be with both of you. 

Shane Phillips  3:42  
So as always, we start our show by asking our guests to give us a tour of somewhere they know and maybe even love often the place they live now. Or maybe a hometown urbanist, or housing related stops are welcome, but not necessary. JZ, where would you like to take us and our listeners? 

Jiaying Zhao  3:59  
Sure, I can talk about Vancouver. I don't know this is British Columbia, Vancouver, not Washington. 

Shane Phillips  4:07  
Your Vancouver is more interesting than ours.

Jiaying Zhao  4:11  
So I've been here 10 years. I love the city. We do have a housing crisis, and that's something I don't enjoy. But the city is beautiful. It's vibrant. I love the food culture here. I really enjoy where I live. I invite people to visit if you can. In the summertime, it's the best season of the year. Don't come in the wintertime. It's just depressing and raining and cold. So between May and August, I think that's the best time.

Shane Phillips  4:39  
So the article we're discussing today was published this year in August in the Proceedings of the National Academy of Sciences and it's titled, "Unconditional Cash Transfers Reduce Homelessness." We do love a declarative title on the show. Your co authors on this we're Ryan Dwyer, Anita Palepu, Claire Williams, and Daniel Daly-Grafstein. The study was a cluster randomized controlled trial in which 50 unhoused adults in Vancouver were provided a one time unconditional cash transfer of $7,500. Canadian. And they were compared to a group of 65 unhoused people who did not receive a cash transfer but still received other support services that we'll talk about. The results of this study were that over a one year period, cash recipients spent fewer days many fewer days homeless, increased savings and spending with no increase in spending on what are known as temptation goods, things like alcohol and drugs, and likely actually saved the government money because of reduced shelter use. Something I personally love about this study is that it was paired with public opinion research, and it revealed two important findings. First, perhaps unsurprisingly, the general public tends to be skeptical of enhanced people's ability to manage money effectively. And they significantly overestimate how much money will be spent on temptation goods. Second, support for cash transfer policies increases when people are presented with evidence showing that most spending goes toward things like rent, food, and clothing, and that cash transfers can save the government money that would otherwise be spent on shelters, as Dr. Zhao and her colleagues acknowledge, and we will no doubt talk about. There are also limitations to what cash transfers can accomplish, or at least what this relatively small study can tell us about the impacts of cash transfers. But it is very promising research. And so we're excited to talk about it in some detail here. JC to start off, it would be great to hear about your motivation for this research project and your team's motivation. As with a lot of homelessness research, most of your team has a background in psychology and medicine. So I think you approach the subject from a somewhat different perspective than urban planners or housing researchers like us. And with respect to the intervention itself. What did you imagine that a large one time unconditional cash transfer might achieve that smaller transfers, or in kind assistance, like rent vouchers that we commonly use might not achieve? What did prior research have to say about that?

Jiaying Zhao  7:17  
Yeah, so we actually came into the study with very low expectations. This is actually reflected in our pre registrations, we actually didn't think that the gas transfer will help people get into housing necessarily, because of the housing crisis in Vancouver, we have a very low vacancy rate, our rent is really high. So we thought, you know, $7,500 is a big chunk of money, but it's still not large enough in the context of Vancouver. So that's why we actually, when we pre registered our study, we only predicted that maybe the cash will make people feel better. So this is the increases in subjective well being, as in a lot of past studies on cash transfers have shown, we also predicted that it may increase cognitive function, as my past research has shown, we actually didn't even talk about housing in the pre registration. So maybe there are reasons why we opted for one time versus regular, like monthly payments, as some basic income studies have done. And the reason is, we think the one time cash transfer might put people in the long term planning mode, where they may think more strategically about how to spend it, as opposed to monthly payments will be a lot smaller, will just help them firefight every month, maybe getting food, buying necessities, right just enough to get by essentially, exactly instead of allowing them to actually move into housing and paid posit rent, furniture, etc. 

Shane Phillips  8:50  
And even think about, you know, finding employment, returning to education, those kinds of things. 

Jiaying Zhao  8:56  
Yeah, I mean, we did measure all of these things, but which didn't think that this 7500 would be large enough to enable that. That was our low expectations going in.

Shane Phillips  9:07  
So one important caveat about this study that we should know early on is that you looked at a relatively high functioning subset of the unhoused population, only 31% of the people initially screened ended up being eligible, and then only about half of those participated in the study. So to the extent we can generalize these findings to a broader unhoused population, it should be to people with similar backgrounds and characteristics. Could you say a bit about your screening process and criteria, and the reasons for those fairly selective criteria? What were you looking for?

Jiaying Zhao  9:41  
Yeah, we chose the sample. For many reasons. The biggest reason is to mitigate any risks to our participants from the Casper, and this is actually coming from UBC ethics board. We actually spent six months getting approval for the study, because no such study has been done. In the past, with this kind of transfers with individuals and homelessness, I think the bottom line is to ensure participants safety and ensure there's no harm of outcasts transfer. This is why we opted for the higher functioning subset. And we screened people based on several criteria. One is substance use. So we screen people without severe levels of substance use. So they can still use with light to moderate levels, just not severe. By severe I mean, we have a threshold in the questionnaire in this clinical questionnaire that we administer. So they just cannot exceed that threshold. 

Shane Phillips  10:37  
And just to be explicit about this, this is about concerns about overdosing essentially?

Jiaying Zhao  10:45  
Yeah, potentially. Even though this concern is not warranted, because you don't need $7,500 to overdose, you need 10 bucks to overdose. Yeah. So this is kind of just to make the I think, Ethics Committee feel at ease. And again, we want this study to succeed, as this is, this is the first one of their, we offered for, you know, no severe levels of substance use. Then we also had other criteria, like alcohol use, as well as mental health symptoms. So these are all measured by validated tools that have been used in this population in the past. And again, we have certain thresholds for selection or screening. So as you said, collectively, 31% of these people's past, you know, this is a low rate, it's only a, you know, less than a third of the broader and house population. But it's, it doesn't necessarily suggest that this approach will only work for these people. And the reason I'm saying that is we're actively running an expansion project. So we're basically scaling this approach to include more people. So this case, 400 people in the expansion, and we're more than halfway through this expansion. And our screening rate, this time is actually 80%. So I think it's hard to say that, you know, take the screening rate as a representation. And the reason is, it really depends on where you go to find these people. So in this initial project, we went into shelters, we only visited shelters to recruit people. And the shelter population is actually only a subset of, you know, the water and house population, and they tend to be lower functioning in general, because they're usually, you know, living on the street, and then they move in the shelter and then went back to the street. So there's no, they, they tend to experience episodic homelessness. Whereas in the expansion project that we're running right now, we're actually recruiting people from the so called Hidden homeless population. So those are folks that are couchsurfing that are living in their cars. You know, they are not in shelters, but they're still homeless based on the definition of homelessness in Canada. So they tend to be higher functioning. and that's why the screening rate is a lot higher now. So it's not going to generalize based on this small study that we ran first, I think it really depends on the homeless population, broadly speaking, and which subset you're looking at?

Shane Phillips  13:16  
Yeah, yeah. No, I think that's important context to know that it's sort of a high functioning or strict thresholds within the the sheltered homeless population, but that that population itself often has more barriers to getting back into stable housing than some other at least, almost populations, like those living in their vehicles, for example. So in terms of these interventions, you initially had two treatment groups and two control groups. Can you just tell us about what those groups looked like? This was not clear to me in the article, but just before this recording, we were talking a bit about the controls and how even they were receiving some support. And so in a way, I think this kind of amplifies the impact of the cash transfer because the folks who received this cash transfer did better than folks who also were receiving kind of above average services or kind of direct supports as part of participation in this program. So can you just talk a bit about the difference between the treatment and the control groups? 

Jiaying Zhao  14:21  
Sure. So initially, we had four conditions to cash to control. So there's the cash transfer group that received the cash transfer, but they also received workshop and coaching. Another group received the cash transfer, but only received the workshop and no coaching. So these are the two cash groups, the two control groups in one group they received workshop and coaching and the other group did not receive workshop or coaching. So there's other four groups that we initially started with. Now everyone in the studies of this includes control on cash participants received additional supports. Everyone received a free checking account from Um, the van city, this is a local credit union in Vancouver, the motivation is we can eat transfer the cash transfer to the cash recipients, instead of giving them just cash in an envelope. They also received free ID replacement services. So if they didn't have an ID to start with, we would offer to replace the ID for free. So we cover that cost as well. That's actually the cost was covered by advanced city as a as a as our income partner. On this project. Everyone received the resource booklet, this so this is basically a kind of a small booklet that we put together that outlined the local services and resources for people in our study. So depending on where they live, they can find the health, social or employment or education services that they are eligible for. Finally, everybody also received a used smartphone, so that we can contact them to set up follow up surveys and interviews. 

Shane Phillips  15:58  
And things like the checking account, I can see how you would want to, you know, provide that to everyone, even to the folks who are not receiving cash transfers, just so you can be sure that any effects you observe are not just due to providing people with a checking account that they may not have been able to access otherwise, or, or those ID services or that kind of thing. Were there other resources available to the control group just in terms of things that could normally access, like if the local housing agency is providing housing vouchers, and maybe there's a waitlist or not everyone's eligible, but like, they're still eligible for those things, and probably using many of those services or resources if they can,

Jiaying Zhao  16:39  
Yeah, so everyone is still eligible for the surfaces they can access. I mean, this is this is an important condition to run the study in the first place. Because getting the $7,500 are immediately disqualify you for social and income assistance in British Columbia, because you suddenly have way too much cash. So to avoid this benefits, Cliff, I think this is what a lot of cash transfers face in the US. If you give them too much cash, then they disqualify for a lot of the services. So that's kind of a perverse incentive, they actually don't want to receive the cash transfer. So to avoid that we negotiated with the British Columbia provincial government, so that we actually had a big exemption where all participants in our study can keep the supports that included in the cash transfer while still being eligible for income and social services. So I think this is a an important condition, so that we're actually not inadvertently hurting the participants by giving them cash. 

Shane Phillips  17:43  
Yeah, it sounds like you've had a lot of negotiating to do amongst different with the epics and the provincial authority, and the bank and everyone else. So I think we're ready to get into the results, we already shared a summary at the top. But let's get into the details a little bit, starting with the overall results for the whole year, focusing on where the impact was fairly large or statistically significant, or both. You also had results at specific time periods within that year. But let's just start with the whole year. And we can come back to those specific time periods and a little bit. 

Jiaying Zhao  18:19  
Sure. So those significant results are actually from our exploratory analyses. So none of this is pre registered, or pre registered analyses. We're only focusing on I think, one month outcomes on subjective well being and Coggan function. That was our very conservative expectation of what will happen.

Shane Phillips  18:38  
Could you explain for the listeners what it means to pre register? 

Jiaying Zhao  18:42  
Oh, sure. I mean, pre-registration is an open science practice where you predict what's going to happen in your study before you start running the study. So this is where we, we say, oh, you know, we think that these outcomes will become significant, but those outcomes may not be, it's where we outline the variables, the measures we collect in the study, and where we think, you know, the significant results will lie. So that's purely our guests. And the, again, our initial guesses were very wrong. I think we just based our guesses on past studies, with cash transfers, but in very different contexts, mostly in developing countries with low income villagers. So I mean, extremely different contexts from this current study, which were very happy went wrong. So basically, the preregister results are a completely no, none of the professional results were significant.

Shane Phillips  19:43  
So this was the subjective well being like just how people feel about their life that kind of stuff.

Jiaying Zhao  19:47  
This objective well being cognitive function at one month, yeah, I one month after the cash transfer. We also you know, compared to specific groups, so compare the two cash groups with each other, and the two control groups with each other to See if there's a difference due to workshops or coaching. That didn't pan out. So workshop and coaching was largely ineffective. So we combined the two cash group into one broad cash condition, and the two non cash group into one control condition. And this is how we did the exploratory analyses, just to compare the cash participants with the non cash participants.

Shane Phillips  20:23  
And so what were the kind of the headline findings there.

Jiaying Zhao  20:26  
 Right, so over one year, we found that cash recipients reduced homelessness by 99 days per person, on average, there was actually no marginal increase in days in stable housing, even though that wasn't significant. There was a significant increase in spending. Obviously, the cash was spent more specifically on on rent, food, transit and durable goods, like furniture or used car. And over one year, cash recipients relied less on social services, if we just look at shelter use, in particular, they actually avoided shelters, I think, as I said, by nine days per person per year. And that led to cost savings for the government. So we calculated the cost savings was actually at $777, on average, per person per year. So subtracting the cash transfer, which is $7,500, each person actually generated a net savings of $777 per year. So this, this turned out to be a net positive on a cost effective approach compared to the status quo. 

Mike Manville  21:38  
Would that change a little bit? If you then factored in the cost of providing the smartphone and all that?

Jiaying Zhao  21:44  
Well, we've provided that to all participants. Right? So there's no difference between cash and control group.

Mike Manville  21:50  
Right. But it's the expenditure on the treated group would rise? Right? 

Jiaying Zhao  21:56  
No, because the baseline, the control group is still have the same amount of cost. So the only difference between the two groups is the cash transfer. Right. So so this is the cost savings was always in context with in comparison with the cash group. So the 8277 figure was the difference between the cash group and the control group in their access to social services over one year.

Shane Phillips  22:19  
In a way it sort of depends on whether those non cash benefits that were provided to everyone had any impact themselves on on shelter. I mean, they certainly had a cost. But I see what you're saying, though, right.

Jiaying Zhao  22:31  
Maybe what Mike was trying to get to, I think was the Okay, what what if you include the overall costs, so not just the cash transfer, but you know, the other supports that you give people, as well as even the research costs and the overhead, right, the operation costs. So when you include all of that it's no longer cost of, you know, it's no longer net positive. But understandably, that's, that's going to happen, because this project was very labor intensive. I'm not sure if this will reflect the actual costs, if this were to be implemented as a government program.

Mike Manville  23:05  
Yeah. And none of this invalidates the validity of, you know, using a cash transfers, which is, you know, I guess the question boils down to, if this was to scale, and it sounds like you are trying to scale it, that does sound very promising. To what extent is, for instance, getting someone to smartphones separable from administering a cash transfer, because if it if it scales, and you also have to add the cell phone, and there's always that capital cost, and I, you know, you can imagine a government saying, Well, what matters is not just how much we give the recipient and cash, but when we do a sort of net cost benefit. It's just our total expenditure.

Jiaying Zhao  23:42  
Yeah, good point. That's a good point. I mean, I wouldn't say that the smartphone is necessary for the cash transfer. I mean, the smartphone was just so we could reach them to do follow up surveys. So the government may not need to do that. Right. Right.

Mike Manville  23:56  
One answer might be this is just something that was very important in early evaluation, right, so that we could know what it did. Once we become comfortable that this is an intervention that works, the smartphone can fade away, in which case your cost benefit starts to look much better.

Jiaying Zhao  24:13  
Right? I think that's a really important point. I mean, even the checking accounts, I think most low income folks have a checking account. And this is also how they are able to file for taxes and get their tax returns. But our participant sample, actually, most of them didn't have a checking account or didn't even have a bank account. So So that's that could be another additional costs for the government to consider is, how do we even give them the cash transfer in the first place? 

Mike Manville  24:43  
Yeah, and I think, you know, it's in raising this, it's not to throw any cold water on it, because a lot of these interventions, I think, we might in the context of this study, consider them sort of, you know, it's not just the cash transfer, but they could well be on their own One sort of valuable one time expenses that we give to this population, right, like above and beyond whether you you get a sort of one time cash grant, it might be very useful that people suffering from homelessness have access to a checking account and access to a smartphone. The fact that you mentioned about relatively few of your participants having a checking account does raise this question for me. You mentioned earlier about the previous literature about cash transfers really being heavily focused in sort of very low income countries where, you know, presumably that cash transfer that they study there, when it's done in Sub Saharan Africa, or places like that just represents, you know, sometimes just many multiples of what those residents would earn otherwise, like in a year, right, yeah, just a, just a life transforming amount of money. That seems very small to someone in an OECD country, but is a huge change. And so I'm curious, you know, did you guys have a sense of what the magnitude of this transfer meant, relative to how much the typical respondent was earning beforehand? 

Jiaying Zhao  26:11  
Yeah, we would put that the $7,500 was about 60% of their annual income.

Mike Manville  26:16  
Okay. Yeah. So still big for sure. But not the same at to your point earlier, as, you know, an intervention in a rural village in Kenya, where you're sort of doubling someone's annual income. That's right.

Shane Phillips  26:28  
Yep. Yeah. And where cost of housing especially are much lower than Vancouver, BC. 

Mike Manville  26:34  
Yeah, and where the entire infrastructure of needs is very different. Yeah. Yeah.

Shane Phillips  26:38  
So if I had to guess which results got the most attention here from the media and the public, assuming, you know, this is has gotten some attention. I think I read about this article, this study initially in kind of a mainstream media outlet. And so I imagine it's it's made the rounds a bit. But I would think that the thing that got the most attention, or one of the things is that people did not increase their spending on temptation, goods, like drugs or alcohol. Tell me if you know that intuition is wrong or right, but assuming I'm on track, how do you explain that to people who might be skeptical about that finding, in previous episodes in this very series, actually, that we're doing? We've already touched on how substance abuse and mental illness have this bidirectional causation with homelessness, where these things are not just risk factors for becoming homeless or staying homeless, but also maybe triggered by or exacerbated by homelessness, it seems like even if some cash recipients may increase their spending on these kinds of temptation goods, others may be in a position now where they're able to relieve some of the stresses and traumas that they've been enduring living in shelters or on the street and maybe less reliant on you know, things like self medication.

Jiaying Zhao  27:53  
Yeah, I mean, so we didn't see any increased spending on temptation goods, like alcohol and drugs. And that's consistent with, you know, the vast majority of cash transfer studies in the low income, lower middle income countries. And I think we actually saw this is one of the other effects that we saw was reduced substance use severity over one year. So that was marginal decline. But it was not significant. What a man was over 12 months, they actually cash recipients showed less substance use compared to the control participants. And this, I think that speaks to the point that you raised, which is people need to use or even use more when they're unhoused. Right. So the homelessness may exacerbate your substance use. And what we heard from our participants is that they actually need needed to use drugs to stay warm, that's a very common reason to self medicate, because they couldn't access other mental health services. So the only way to cope living on the street was to use substance. And I think, as in many US countries, Vancouver has an raging opioid crisis. And at this point, it's mostly fentanyl, which is incredibly cheap. I think it's like $1, a pill or something at this point. So it's way cheaper than alcohol. Alcohol in Canada, was actually really expensive. And I think that seems to be the substance that unhoused folks resort to on the street, and moving them into stable housing actually alleviates the dependence on that. So that's what we found in this project. And it's very, I was really encouraged to see that result. 

Mike Manville  29:35  
Everything JZ just said made a lot of sense to me. And it's consistent not just with this bigger literature on cash transfers, but also a bigger literature on poverty and experimental evidence from what poverty does to people's decision making, you know, the sort of conservative trope about poverty is that you know, a big contributor to it is is poor self control, you know, and this is why use See lower income people being more likely to smoke or gamble or drink to excess or things like that. But I think a lot of the literature has has actually demonstrated that it really, a lot of it does run in the other direction, which is that when you don't have many resources, it's just harder to make the decisions. And the main book I think about when I think about this, as it's called scarcity by Sendhil, Mullainathan and elder Shaffir, it's an economist and a psychologists that summarize a lot of this, and just demonstrate in a wide variety of ways that, you know, when you have a situation where virtually every situation you're in qualifies as an emergency, you know, if you have like seven hard decisions every day, you can hit the first three out of the park. And then the fourth, and the fifth, you know, you have sort of a cognitive depletion that happens. And that one of the points they make is just that people who are relatively affluent, it's the sheer number of things we aren't responsible for, that makes our lives easier. Like we don't we just day to day, we just don't have that many hard decisions to make. And I think that to circle back to one of the elements of Jaycees, and her team's research design that I really like, I think that the decision to make that one big transfer could be quite meaningful in the context of this literature, because what that allows you to do is just if you're a person in these dire straits, that big chunk of cash, if necessary, could just let you solve some problems. You know, rather than here's, you know, I have a, I have a need that, you know, to Jason's point, maybe it's just putting a deposit down on a place, maybe it's getting my car fixed, I have a $500 need. But if I'm getting doled out $250, a month for 12 months, or something like that, I'm never actually going to get to the point where a huge float is taken off my mind. And I do think this is something that's under appreciated among sort of lay people who think about this correlation between poverty and what we might call temptation goods, or bad decision making, which is that sure if you make bad decisions, like the chance of a bad outcome rises, but also if you find yourself in a bad situation, the chance of making a bad decision goes up, but you know, a quote unquote bad decision.

Jiaying Zhao  32:11  
I completely agree with what Mike just said, although and Central, actually my PhD advisors. So I did, I did the scarcity work was was a big chunk of my PhD work. That was 10 years ago. Well, they wrote a great book, why I, you know, put the cognitive function in there? Well, we saw with the cognitive results was actually there was a transient improvement over three months, right, right recipients, it was not sustainable over long term. And there are many reasons for that. One is that data if you look closely at the month by month data that are really noisy, and the reason being, we ran these surveys and interviews with participants on a tablet, and those images flashed quickly. So those cognitive tests are not designed for people in homelessness, or poverty. They're designed for mostly undergraduate students in psychology, who are familiar with the computers. And so these are all computerized tests. And they don't work very well for our participants, because our participants couldn't see on the screen, or they were not familiar with this interface. So we actually couldn't, we didn't even have a lot of data from the count of measurements, unfortunately, because of that. And I just think that tool was not great. The implementation of that tool was not great for our participants. So I mean, that also speaks that we need to develop more appropriate measurements for this population.

Mike Manville  33:34  
I mean, that's that's an I hadn't thought of that. But that is a, that would be a legitimate limitation for sure. 

Jiaying Zhao  33:39  
Yeah, we just we just didn't have enough data for these measures. But yeah, I think that we have a lot of stories or qualitative data from our participants that we didn't include in this paper. So the common theme in their response was, this transfer was life saving, even though it was was small, in the context of Vancouver, they thought that was their one chance to get out of homelessness, this is their one chance to actually do something and get their life back on track. So a lot of them said, this actually gives me peace of mind. I got into housing, and now I can properly make decisions. So this is exactly verbatim what our participants told us after getting the cash transfer.

Shane Phillips  34:22  
Yeah. So on that subject, actually, you know, something that stood out to me is how the positive effects for people receiving cash transfers mostly diminished to non significance, statistically speaking by the 12th month, I think that could be misinterpreted as the transfers having no effect. But obviously, that's not the case, because we just talked about the treatment group spending an average of 99 fewer days homeless and 55 more days in stable housing. So what does it mean that the outcomes for the treatment and control group were broadly similar by the end of the first year even if You know, the the cumulative outcomes were quite different. How do we think about those two different ways of measuring this? And what do you think explains this convergence at the specific time points as as the study went on?

Jiaying Zhao  35:14  
Yeah. So what happened was the cash group got into housing much faster than the control group. So if you're in a homeless shelter in Vancouver, you were put on a waitlist for transitional housing for six months. So you have to wait six months to get into housing, that's typically the the average the wait, but cash recipients moving into housing, even within one month, actually have more than half of them move into stable housing within one month. So that's where the savings, the fewer days homeless came in. And then the cash flow participants actually remained house, this is not that they went back to housing after they spend the money back to homelessness back to the sorry, homelessness. Yeah. So most people spend the money after six months. I mean, $7,500 wasn't, again, wouldn't sustain you, after a couple months in Vancouver, given the high living costs, and they actually were able to stay housed over 12 months. And the reason was, they actually worked more hours, they got higher paying jobs, etc. So that's how they were able to sustain their income or livelihood after the initial transfer. And then after six months, you know, the control group caught up. So they they move it into housing, and no longer in the shelter. This is why later, especially after six months, the two groups, the difference between the two groups disappeared. So when you average that over one year, the overall effect was much weaker than the initial couple months of the cash transfer. 

Mike Manville  36:42  
That makes sense. To just do a quick follow up, I don't mean to get us off track. But when folks did move into stable housing, you guys have a sense that I think our listeners would love to know, where respawn has tended to end up is this did they manage to lease up on their own and really inexpensive places? Were they able to just kind of put some money toward crashing with some friends? Or what in general a fatty patterns Did you notice?

Jiaying Zhao  37:08  
So for the cash recipients, most of the time they spend in stable housing was in apartment rentals. That's 74% of the recipients. So that meant they moved in to actually market housing, and they tend to share the apartment with roommates. Sure. So that's the type of home housing they moved into the 17% moved into single room occupancy units. So these are affordable social housing units that are much smaller, and rent was cheaper. The quality is less good than apartment rentals. But it's still stable housing. So that's where the cash recipients went.

Shane Phillips  37:48  
Yeah, yeah. The the fact that the effects of the cash transfer had dissipated by month 12 Sort of makes me think that instead of thinking about this policy, necessarily as a way to reduce homelessness, maybe we should think of it as a way to relieve suffering as people get back on their feet and, you know, potentially save money on government expenditures in the process. I don't think that's the only way to see this. But, you know, even if folks end up in pretty much the same place by, you know, the end of the year, the ones who received the cash transfers spent more than three fewer months homeless during that time, on average, and the cost to the government was essentially the same. So that strikes me as a win, even if it isn't a permanent solution by itself. And, you know, it's worth mentioning that the effects at month 12, were still mostly in the positive direction, the kinds of outcomes we would want to see for that cash transfer treatment group. They just were not as strong as they had been in earlier months and had lost their statistical significance. So it seems entirely possible that a larger future study could reveal stronger or more statistically significant and maybe potentially more durable effects as well.

Jiaying Zhao  39:01  
Yeah, absolutely.

Shane Phillips  39:02  
So I wanted to come back to the cost here a bit, as you shared, the annual savings on shelter costs was about $8,200 per person in the treatment group compared to cost of the cash transfer of $7,500. That's clearly really great news. And so you know, I just want to complicate that a bit in actually both directions. So in the maybe we're under counting the benefits category, it seems like there are probably other savings besides shelter costs, such as reduced visits to emergency rooms or fewer interactions with law enforcement, which, you know, these are things that tend to be major costs associated with homelessness, certainly here in the US. And so I'm curious to hear how you approach that you do have something about service costs, but I feel like it's not as emphasized in the study, or at least in the write up, and then in the maybe we're over counting category. I'm wondering You know, there's a lot of different ways to measure shelter costs. And I had a few things noted down here, but one I'll just focus on is, if you have a shelter with capacity for 50 people, and now you have less demand for shelters, and you're only using 40 beds, like, does the cost really fall by 20%? Or is a lot of that cost essentially fixed, so you're not necessarily saving as much money there. And then also, the study found that the cash transfer recipients spent 99, fewer days homeless, but not necessarily 99, fewer days in shelter. That was my reading of it anyway. And so maybe those costs wouldn't have been incurred? I understand these are all difficult things to measure, or to answer with any certainty. But I thought they would be worth a little bit of discussion just to really, if anything, underscore their complexity of these systems, and how many nuances and caveats and complications come up when we're trying to do these kinds of analyses? 

Jiaying Zhao  40:57  
Yeah, That's a very good point, we did measure costs or services beyond just shelter use. We looked at their hospital visits, police encounters, court appearances, nights in jail, etc, etc. So, ambulance visit 911 calls, you know, all of that. And when we included all the services, they accessed over 12 months, the numbers were a little lower than that, just to shelter us, as you can see from our figure, but it was still, there's still positive, it's still significantly higher than the cache other than control group. So the cache group still use less overall than the control group. The the number was actually a little smaller than 8277. The reason being, we had couple of mothers in the cache group that gave birth after the cash transfer that resulted in more hospital stays for these women, whereas the control group actually didn't. This is just purely by chance.

Shane Phillips  41:56  
Yeah, yeah. And especially when you have a small sample, 

Jiaying Zhao  41:59  
Like, exactly. And then each night in the hospital in Vancouver was, you know, $1,000, when you count up just like three nights or four nights, that sort of just reduced the savings from shelters a little bit. Yeah, yeah. 

Shane Phillips  42:14  
People here in the US are crying and hearing that it was only $1,000 A night, and it's all free. 

Jiaying Zhao  42:19  
So they all didn't have to pay out of pocket. 

Yeah, I mean, I think, you know, it's that sounds very much as you indicated, like the idiosyncrasy of a small sample. You know, I do think that the shelter costs are probably in some ways, the most intuitive and predictable costs that we think about when when someone is unhoused. But you know, at the tails, it is these medical emergencies, like, you can just as easily imagine someone who didn't get the cash having an accident or a medical emergency, and that they're in the hospital for four days, and the cost and the control group spikes. And I just think those are, they're a little bit harder to predict a priori, but they're absolutely they're huge costs of having a large, unhoused population. And I think the other thing that's just very hard to measure is the well being mental well being of people who worry about homeless people. And then sometimes that's, you know, neighbors who just, they worry about their neighborhood, because there's a lot of homeless people there. But I think even more to the point, it's the the family members of people who know that they have family who are on the street or on the verge of being on the street, and it takes a toll on them to know day in and day out, like, Well, I'm not really exactly sure where my brother is, I don't know if he's okay, and things like that. And in the grand scheme of things, a relatively small expenditure of money that puts that person under a stable roof is, of course, first and foremost of huge benefit to that person, but it does have ricochet effects that reverberate around a lot of other people that are very hard to sort of put into dollar terms. But I think if he ever talked to anybody who has a relative or a close friend who struggles with with staying housed, I mean, there, there are real benefits.

This is what we call the spillover effects. Mike just mentioned, those are very hard to quantify, especially if it's just psychological measures like worry or concern, we did look at some spillover effects with spending. So again, as I mentioned, there are mothers with dependents in our in our sample. And then for those participants, they actually spent a lot more money on their children, and not not on themselves. So I think that's that's the kind of the benefit of the cash transfer, the spillover to their family members.

Mike Manville  44:41  
Right. And I do think, you know, it's absolutely the case that the psychological benefits are very hard to quantify. And I would hesitate to try and put them in a paper with a hard number as well. I think, though, maybe to go back to Shane's point about the easing of suffering. You know, any of our listeners could just imagine if you had a a close family member who, you know, struggled with an illness for a year, like he would probably spend that whole year worrying about them. And if there was some treatment that let them get better three months earlier, right, the amount of you know, fine, you have a hard time putting $1 amount on it. But I think all of us would understand like, wow, like that would be a huge benefit to everyone around this person. And I think it's similar in many respects to saying, Well, you could you got this person into a more stable housing situation three months earlier, those are, are going to be fairly substantial benefits for the folks who care about that person.

Shane Phillips  45:35  
So in the results, you found positive effects for some outcomes and no effect for others. But in one area, you found a negative effect or impact for the cash recipients relative to the control group. And that was in social connection. Can you talk about that?

Jiaying Zhao  45:52  
Sure. That was an unexpected negative result. And that only showed up let me just be very clear here, the social connection showed up at nine months, where the cash recipients felt less connected socially than the control group. I think, again, there are several reasons for this. One is by nine months, they already moved into a new community, new environment, most of them are not no longer talking to the old friends and people they know in the shelter. So I think they felt lonelier at that point, whereas the control group probably didn't have to move that big of a change or move that far, or their life didn't change substantially. So that could be one reason. Another reason is, before we gave them the cash transfer, we actually asked the participants to carefully think about how to keep this new safe, or how to keep themselves safe, from potential assaults, thefts, targeting, from others. So a lot of them actually chose not to tell anybody but one or two of their confidants, right, so about the cache news. So you can imagine this is not a, even though this is great news for themselves, but this is not something that they want to broadcast. So they have to keep that news to themselves for a long time. And I think that also that may have contributed to this lower social connection, feeling.

Shane Phillips  47:12  
Yeah, it's super interesting. And looking at the table of information about the demographics of both the treatment group and the control group, these were folks who, on average, had been homeless, in their current instance of homelessness, an average about 20 to 30 weeks, and had spent a total time homeless, you know, of almost two years and had been homeless multiple times. And so I think there's, this is a group who maybe had a lot of connections, in many cases to other unhoused people. And so that might have become their community in many ways. Whereas you might expect something a little bit different for folks who are more newly homeless and receiving, they wouldn't necessarily need this kind of lump sum payment. But if they were to receive some kind of intervention, the intervention actually might allow them to retain their social connection with other, you know, house, friends and family. 

Mike Manville  47:13  
Just to add to that, it's it that is not a sort of threat to validity that would have occurred to me until JC mentioned it, but it is it makes sense. I mean, it's both the security concern, but also you can just imagine, to Shane's point, like, you have a bunch of friends, and you don't want to trigger envy or resentment in them, it probably in some situations for her having come by this windfall. And and that's a very human reaction. And so obviously, it's a net benefit to have the resources as opposed to not have them. But for any of us, regardless of our socio economic status. Anything that sort of threatens to upend our relationships with other people is a source of turmoil for us. 

Shane Phillips  48:51  
Yeah, I would love to see a study on that just that subject more broadly with homelessness, you know, because I'm sure the same thing applies just to getting you know, a housing voucher or placed in supportive housing, that kind of thing for folks. And then probably, again, like more of a an issue for people who have been homeless longer. Moving on to you had a few other kind of smaller studies, about this research that are not just about policy, but really politics in a way, by way of public opinion. So we all know that politics and public opinion play a big role in determining which policies get implemented. And public opinion about homelessness in particular is rife with misconceptions and misunderstandings. And our first two episodes in the series with Greg Coburn and Margo Kachelle are really about addressing a lot of those misconceptions because they can be a real barrier to implementing solutions, or at least effective solutions. So in this article, you have two other studies aimed at addressing the public's views on unhoused people and their expectations about how an enhanced person would spend money that's granted to them unconditionally, and then testing how people's support for cash transfers is influenced by different messages about your research findings? Could you summarize those two studies and what you found with them? Sure.

Jiaying Zhao  50:10  
The reason that we looked at public perception of cash transfers is to figure out ways to increase public support for potential policies going forward. This is because I've been working with policymakers very closely enacting policy change at a government level to enable programs like this to benefit more people in homelessness, we're in poverty more generally. So what we found from these two studies was one, that public actually don't trust people in homelessness in their ability to manage money. So we describe our study to people, this is people in the US, from the US public, we describe our project where we gave, you know, $7,500, one time to people in homelessness, but they have no, you know, severe levels of substance use alcohol use and psychiatric symptoms. So just like in our study, and we asked them to predict how much money do you think they were spent on alcohol, drugs and cigarettes every month, and we describe the recipient as somebody who's homeless, or who's not homeless, or somebody who's yourself, homeless or not. So we had different kind of descriptions of that recipient. And what we found was people think that if you just if this is someone who's homeless, then they will spend a lot more money, actually, I think, almost double the amount of money than if somebody's who's now homeless or is themselves. So that difference is sort of captured over reflects the the amount of mistrust that the public have toward people in homelessness in how they manage money. So that's what I suspected all along. Whenever I present this study to people in public or policymakers the first responses the No, why would you do that they will just squander it on drugs, that our study shows the opposite, right? There was no increase in temptation, good spending. And yet people think that recipients will spend more money on alcohol and drugs. So we, we again, we run another study, where this is, again, recording people from the US public, where we said, we ran this cash transfer study with people in homelessness. This is what the study showed. So we said there was no increase in spending on temptation goods. And actually, the cash transfer led to net savings for society because of the reduced use on social services. And cash recipients actually spend more money on rent, food, clothing, etc. So we found that framing the cash transfer, with the benefits of a cash transfer, in a utilitarian way, like the cost saving way, significantly increased public support for cash transfer policy. So the utilitarian aspect of the cash transfer, helped people support the policy more. And another framing was the counter stereotype frame, which is no increased spending on temptation goods, but rather increased spending on rent, food, and transit and all that, that also increased public support for cash transfer policies. So that's kind of a our motivation to include that perception piece, in addition to the cash transfer itself. This is actually a debate we had with the reviewers of this paper, who said, why don't you just focus on the paper on the cash reference study and remove the opinion, public perception studies. So the reviewers thought it was not necessary to include the public perception studies. And I had to kind of fight back and say, we could easily remove the two studies, that's not a problem. But that would diminish the impact of the paper, especially when when it comes to you know, communicating with public policymakers. How to Move policy forward.

Shane Phillips  53:40  
Right like, it's it's one thing to know that the policy itself is effective in these different ways. It's another to know whether that actually persuades the public to support the policy in any way.

Jiaying Zhao  53:52  
That's right. I mean, that's what kind of what public policymakers asked us specifically, like, Okay, we love your results, but how, you know, how do I use it to convince my colleagues who are against this policy? So that's, that's the helpful piece for them. And I think that that actually has changed a lot of people's minds to know, oh, well, this is what you found. And this is what people thought, and this is how you can sort of mitigate that gap. So yeah, I thought, you know why I thought this combination of the three studies is probably going to be the most impactful. And actually, I don't know if you know that in Canada, we at the federal level, we are actively studying, we're discussing this basic income bill. It's actually you know, it's a cash transfer to people in poverty. And I think knowing the results of the study actually helps policymakers a lot.

Mike Manville  54:44  
Yeah, I mean, I think the inclusion of those two studies also just functions as an important backdrop for the experiment itself. It sort of gets that kind of the motivation for it and the obstacles that proposals like that Let's face you know, certainly in the United States, we have this long history of our welfare state being, you know, pretty paternalistic, and sort of systematically under estimating the extent to which people find themselves in dire circumstances out of bad luck. And I think that's it's a it's an under appreciation of that fact that leaves the typical Amazon Mechanical Turk, you know, survey respondent to say, well, of course, they're gonna use more drugs and use more alcohol. I mean, you I guess we don't know for sure. They might be just saying to themselves, Oh, my goodness, they have to self medicate, because they're homeless. But I strongly suspect that the typical respondent is saying, Well, you know, that they wouldn't be homeless to begin with if they weren't irresponsible. So you can't just give them money. And of course, right, it would be wrong to say that that's categorically and never correct. But I do think that a story that runs through American and perhaps Canadian social services as well is just a, an underestimation of the the extent to which people can find themselves in very bad circumstances through a few bad breaks. And it's not because they don't know how to manage money well, compared to the typical person or don't know how to discipline themselves compared to the typical person, they really just don't, as you mentioned, in the first paragraph of your article, they just don't have any money, and having a little bit more money would just would be a huge boost.

Jiaying Zhao  56:22  
Yeah, that's all speaks to the scarcity point is that it's really just a lack of money and not a lack of character or abilities.

Mike Manville  56:29  
Yeah, I mean, look, you know, it's just, I think this is an important point in general, I mean, if I just look around my university, there's plenty of people walking around who I'm pretty sure don't manage their money very well. And I know for a fact, they don't manage their time very well. And they're not the most disciplined people on Earth. And I fall into some of these categories. The difference is just that, like, they all have a pretty good paycheck. And so these little errors in judgment are just sort of intrinsically forgiven. And that's a, that's not true when you're one step above the street or when you're on the street. And so that, you know, one of the great arguments for a cash transfer is like you get to that point where there's just a few fewer crucial, you know, monumentally important decisions and everyone's day. That's right.

Shane Phillips  57:13  
I do want to clarify that there were some differences in how effective those two messages were the, you know, unhoused people do not increase their spending on temptation goods versus the it saves government money to actually provide these cash transfers. So that it saves the government money message was actually I think, get increased support about twice as much as the the other messages that roughly correct JC. No,

Jiaying Zhao  57:38  
so the there was a marginal difference between the two messages, not significant. And if not twice as much, it's like a public support. So policy support is measured on a scale from one to five, the utilitarian message had the strongest support of 3.95 out of this five point scale, whereas the common stereotype message had a slightly lower support. So 3.78, so it's a little lower, but not not too bad.

Shane Phillips  58:07  
Yeah. And the baseline was around 3.5 or so. So it was actually baseline 3.5. Yeah, fairly high. It's surprisingly high baseline support. But both

Jiaying Zhao  58:17  
three is neutral. Right? So three is like one is strongly opposed. Five is strongly support. Three is neutral. So most people are like, just a little bit, a little bit more supportive than neutral. Okay,

Shane Phillips  58:28  
got it. So last question here, you know, what is next for the study of unconditional cash transfers as a tool for reducing homelessness, or at least reducing suffering during homelessness? You mentioned a little bit about where you're going with this. But is there anything else you want to share and also anything you wish you had done differently in this first project?

Jiaying Zhao  58:49  
There's so many things I wish everything differently. So we're actively expanding this project to include more people with a larger transfer. So that as you said, we're actually increasing the cash amount. But now we've literally reduced the conditions to to just cash versus control, that that project is underway, and I can't disclose too much, because they don't want to influence the integrity of the study at this point. But more broadly, I do think, I wish more projects like this can be done. I mean, the the other project that is actually actively ongoing is the Denver Basic Income demonstration project. Actually, we talked to them before they started and we told them our lessons learned from this project. So they're actually running a very similar project in Denver, with homeless people in unhoused. They actually provide them with a larger amount of cash transfers. I think it's $10,000 us over one year,

Shane Phillips  59:46  
In a somewhat more affordable city as well.

Jiaying Zhao  59:49  
Less less expensive, right. So and they're actually they just published their six months interim report, which I find really encouraging. Actually, they're finding bigger effects than we did. So that's great. And they actually saw the biggest surprise I saw in their report was the increase in full time employment. Whereas we didn't see that in our sample. We did see increases in work hours and pay, but we didn't see a, you know, a significant increase in full time employment, whereas they saw that they actually saw a 50% increase in full time employment, which was really surprising to me. I'm talking to them in a couple of weeks. So I would love to find out what they did that was so effective. 

Shane Phillips  1:00:32  
All right, JZ. Zhao, thank you so much for joining us today, and good luck on the next steps of this project.

Jiaying Zhao  1:00:38  
Thank you so much.

Mike Manville  1:00:39  
Fascinating paper. Thanks for coming on the show.

Shane Phillips  1:00:45  
You can read more about JZ's work on our website, lewis.ucla.edu. Show notes and a transcript of the interview are there too. The Lewis Center is on the socials, I'm on Twitter @shanedphillips, and Mike is there @MichaelManvill6. Thanks for listening, we'll see you in two weeks with the next installment of Pathways Home.