Unconditional cash for people without homes
What Happened When Researchers Handed Unhoused People Cash, and Why Vancouver and Denver Disagree
Most help for people without homes arrives as supervision: a shelter bed with rules, a caseworker, a voucher that can only be spent one way. The assumption is that the scarce thing is good judgment. Two randomized trials that gave unhoused people money with no conditions tested that assumption directly, and they point to a sharper idea. For many people, the scarce thing is money itself, and the second scarce thing is housing that money can actually rent. This page uses those trials to study how basic income ideas apply to one of the hardest problems in living, and what it would take to move people from scarcity toward abundance rather than only managing them inside it.
The Vancouver lump sum
In 2016 a team from the University of British Columbia and the charity Foundations for Social Change began planning what they described as the first randomized trial of cash for people experiencing homelessness.[1] Each person in the cash group received one payment of 7,500 Canadian dollars, equal to a year of provincial income assistance at the time. The researchers chose a lump sum on purpose, so that people could pay a deposit, buy furniture, or cover first rent, which small monthly amounts would not allow.[2]
Two design choices matter for anyone thinking about delivery. First, the team negotiated an agreement with the British Columbia government so that the money would not reduce anyone's existing benefits, which removed the benefits cliff that often punishes people for receiving help. Second, the local credit union Vancity provided free accounts to receive the funds.[1] In practice, the trial was a small model of semi-private basic income: a nonprofit raised and sent the money, a credit union held it, and the state agreed to stay out of the way.
What the money did
Over one year, cash recipients spent 99 fewer days homeless and 55 more days in stable housing than the control group. They kept about 1,160 dollars more in savings and raised their monthly spending by about 429 dollars, mostly on rent, food, transit, and durable goods such as furniture or a car. Spending on alcohol, cigarettes, and drugs did not rise.[2] About half of recipients moved into housing within a month, in a city with a vacancy rate near one percent.[1]
Because recipients used fewer shelter nights, the authors estimated that the shelter system saved more than the transfer cost, for a net saving of about 777 dollars per recipient over the year.[2] The charity reports a similar figure of about 600 dollars per person after counting the transfer.[3]
The study also tested something the money could not fix: public opinion. In separate experiments, members of the public expected unhoused people to manage money poorly. Messages that described what recipients actually did with the cash, or that stressed the cost savings, raised support for a cash policy.[2]
What the money did not do
The same team reported that the cash did not produce substantial gains in food security, employment, education, or well-being. They suggest 7,500 dollars was simply too small against Vancouver living costs, since recipients remained below the poverty line. A workshop and coaching offered to some participants had no measurable effect; most people stopped attending after the first month. The authors concluded that people needed practical help, such as identity documents and job applications, rather than motivational support.[1]
The screening question
The most important limit is who was eligible. Of 732 people screened in 22 shelters, only 229 passed, about 31 percent. The criteria excluded people with severe levels of substance use, alcohol use, or mental health symptoms on standard screening scales, and included only people homeless for less than two years. The team then could not reach 114 of the eligible people again, often because they had no stable phone or address, leaving 115 participants.[2][1]
So the Vancouver result is strong evidence about people recently pushed out of housing, mostly by lack of money. It says less about people who have been on the street for years.
Denver, a larger test with a surprise
The Denver Basic Income Project randomized about 807 unhoused people into three groups. Group A received 1,000 dollars a month for a year. Group B received 6,500 dollars at enrollment and then 500 dollars a month, the same 12,000 dollars total. Group C, the comparison group, received 50 dollars a month plus a phone with service, a debit card, and contact with the program.[4]
At ten months, between 43 and 48 percent of each group reported living in their own house or apartment, and stable housing roughly doubled in all three groups. Some differences appeared: 19 percent of Group A spent at least one night unsheltered in the prior week, compared with 29 and 30 percent in Groups B and C. Among people who started out unsheltered, 45 percent of Group A reached their own housing, compared with roughly a quarter in the other two groups.[4]
The Coalition for Evidence-Based Policy criticized how the project was reported. Its review notes minimal differences between the 12,000 dollar groups and the comparison group on the main housing measure, while coverage emphasized improvement from baseline, which a comparison group exists to rule out. The review also notes that about half of the sample was lost to follow-up at ten months.[5]
Reading the two trials together
Taken together, the trials undercut two opposite stories. One says people without homes will waste cash, and Vancouver found no rise in spending on temptation goods. The other says cash alone ends homelessness, and Denver found that a phone, a debit card, and a small stipend went along with nearly the same housing gains on its headline measure as twenty times more money.
A useful working thesis is that cash helps most when three conditions hold at once: the person's main barrier is a money gap rather than something money cannot buy, there is housing available to rent, and the payment does not trigger the loss of other benefits. Vancouver arranged the third condition by agreement and partly selected for the first. Neither trial could create the second. Where homes are scarce, extra money in many hands can bid for the same few units, which is why supply sits underneath every cash program. The Kenya evidence on local prices makes the same point from another direction: cash raised prices little where sellers could expand supply.
Design ideas worth studying
- Treat the benefits cliff as part of the program, not a footnote. Without a protection like the British Columbia agreement, a transfer can cost a person more than it gives.
- Count connection costs. Half of eligible Vancouver participants could not be reached again, and Denver gave even its comparison group a phone.
- Pair cash with supply, for example through community land trusts or other routes that add units people can rent.
- Use member-owned institutions, as discussed in credit union basic income.
- Report against the comparison group first. Improvement from baseline is the wrong headline in a randomized trial.
Limits of this evidence
Both trials are modest in size and short in duration. Vancouver had 115 participants, screened out many people with the most complex situations, and lost about 30 percent of contact during the year.[1] Denver lost about half of its sample at ten months.[5] Neither study shows what happens after the money ends. This page is educational and is not a recommendation for any specific program or a guide to applying for one.
Why this matters for abundance
Scarcity thinking treats poor people as risks to be managed, so help arrives in narrow, supervised forms. The Vancouver trial suggests that for many people the respectful option, plain money, also cost the public less. The Denver trial is a reminder that money moves people only when there is somewhere to move. An abundance approach to homelessness therefore needs both halves: trust people with resources, and build enough homes that those resources can buy one.
Discussion questions, essay ideas, and learning related AI prompt ideas
- Why might a lump sum help someone leave a shelter more than the same amount paid in small monthly installments?
- Describe how a benefits cliff works and why removing it was a precondition for the Vancouver trial.
- Compare the Vancouver screening criteria with the people most visible in public debates about homelessness, and discuss what the results can and cannot tell us about them.
- Design a cash program for people without homes that a credit union and a local nonprofit could run together, and identify the hardest part to get right.
- Discuss why public distrust, rather than recipient behavior, may be the main barrier to cash programs, and what evidence could change that distrust.
- Analyze how housing supply limits what any cash program can achieve in a city with a one percent vacancy rate.
See also
- Ending homelessness
- Basic income experiments
- Semi-private basic income
- Village cash transfers and prices
- Community land trusts
- Savings circles
- Registered reports
References
- ↑ 1.0 1.1 1.2 1.3 1.4 1.5 Jiaying Zhao, Anita Palepu, and Daniel Daly-Grafstein, "We gave $7,500 to people experiencing homelessness, here is what happened next," The Conversation, September 17, 2023. https://theconversation.com/we-gave-7-500-to-people-experiencing-homelessness-heres-what-happened-next-212960
- ↑ 2.0 2.1 2.2 2.3 2.4 Ryan Dwyer and others, "Unconditional cash transfers reduce homelessness," Proceedings of the National Academy of Sciences 120(36), 2023. https://www.pnas.org/doi/10.1073/pnas.2222103120
- ↑ Foundations for Social Change, "Our Impact." https://forsocialchange.org/impact
- ↑ 4.0 4.1 Center for Housing and Homelessness Research, University of Denver, "Denver Basic Income Project: Year One Report," 2024. https://hpri.usc.edu/wp-content/uploads/2024/07/CHHR_DBIP_Report.pdf
- ↑ 5.0 5.1 Coalition for Evidence-Based Policy, "Denver Basic Income Project," study review. https://www.evidencebasedpolicy.org/study-reviews-1/denver-basic-income-project