Do Homelessness Prevention Programs Work?
Santa Clara County, California is one of the most affluent counties in the country. Encompassing the heart of Silicon Valley, it is home to many high-profile tech companies including Google, Apple and Facebook. But this wealth comes with a cost; Santa Clara supports a growing number of people without a home. A 2019 study found nearly 10,000 people lacking housing, a 31 percent increase from 2017. With an extremely tight rental market and some of the highest average rent prices in the nation, Santa Clara County had the fourth largest homeless population in the country in 2019, behind only New York, Los Angeles and Seattle.
The problem of homelessness is well-known, but the solutions are far from obvious. Social workers who interact with families in precarious housing situations have identified one promising approach known as targeted homelessness prevention. Adhering to the maxim that “an ounce of prevention is worth a pound of cure,” this intervention emphasizes a data-driven approach to identifying families who might most benefit from immediate but short-term financial assistance. Program workers then provide families with money to cover rent, other household expenses such as utilities and some other costs that contribute to housing stability (e.g., vehicle repairs necessary to maintain employment). The total amount of assistance is modest, equivalent to roughly one or two months of rent for the average recipient.
While this is popular among practitioners, there had been little rigorous research on the practice to date. The most compelling evidence comes from a 2016 study in Chicago conducted by the Wilson Sheehan Lab for Economic Opportunities at Notre Dame (LEO for short). LEO is a research lab that partners with social service providers across the country to evaluate interventions intended to help alleviate poverty and improve the lives of those living in disadvantaged communities. LEO researchers tracked individuals who were experiencing an immediate housing crisis and reached out to a homelessness prevention call center. The center would connect the caller with organizations able to provide emergency financial assistance. If funding was available at the time an individual called the center, the caller was 88 percent less likely to be homeless after three months and 76 percent less likely after six months. While this study was encouraging, it had several limitations. It was non-experimental and retrospective, and conducted in a context (Chicago) with relatively low housing costs compared with places like New York and California.
Seeking to more rigorously evaluate this approach, LEO researchers teamed up with a nonprofit organization in Santa Clara County, Destination:Home, that had been considering a targeted cash assistance approach after 40 percent of those surveyed in in the county reported that receiving rental assistance would have prevented them from becoming homeless. Because the program did not have sufficient funding to serve all eligible households, Destination:Home and LEO were able to implement an experimental evaluation of the program in which a randomly selected subset of eligible applicants were offered the cash assistance.
The study, published in the May 2023 issue of The Review of Economics and Statistics, evaluated individuals and families at imminent risk of being evicted or becoming homeless between July 2019 and December 2020. These families eligible had moderate risk scores according to an index used to measure vulnerability and were ineligible for other prevention programs as they could not demonstrate the ability to pay rent in the future. A randomly selected subset of households were offered emergency financial assistance as well as non-financial services such as credit counseling and landlord dispute resolution. The control group only received the nonfinancial services. During the period of study, the average treatment household received about $2,000 to pay rent, utilities and other housing-related expenses. LEO used data on individuals from the county’s Homelessness Management Information System (HMIS) and data on addresses from Infutor Data Solutions to compare shelter entry and address changes across the families assigned to receive assistance with this same information for families who were not.
So, what did they learn? The authors of the study, David Phillips, a research professor at LEO, and James Sullivan, a professor of economics and co-founder of LEO, found that people offered the emergency financial assistance were 81 percent less likely to become homeless within six months of enrollment and 73 percent less likely within 12 months. This means that providing financial assistance to those at imminent risk of homelessness decreases the chance of a recipient needing to enter a shelter and decreases homelessness over time rather than just delaying it.
In our exchange below, Phillips and Sullivan discuss what they learned from their collaboration with Destination:Home, and how it fits into the broader efforts to tackle homelessness in the U.S.
With an increasing demand for identifying what works in tackling homelessness, what are some of the other interventions we know are effective and where do we still have gaps?
If we look at homelessness interventions on a spectrum, we’d have prevention on one end and what’s known and permanent supportive housing on the other. Permanent supportive housing provides long-term rental subsidies to people who are already homeless. Permanent supportive housing has considerable evidence behind it. One example is the HUD-VASH program, which is a federal effort to reduce homelessness among veterans through the distribution of housing vouchers. A LEO study found that the expansion of the program that happened from 2008 to 2017 cut homelessness among veterans in half. The lesson we can take from this is that these more intensive solutions are effective for chronically homeless people. But when we talk about this issue, we have to keep in mind that the experience of homelessness varies among the homeless population, so the solutions we offer need to reflect that. We now have evidence that targeted intervention for at-risk individuals works and we have evidence that permanent supportive housing for the chronically homeless works. So, where we really need to build evidence is the space between those two points. Things like rapid rehousing subsidies and shelter diversion.
Why is it so important to focus on prevention in addition to other efforts?
The first and most important reason is the toll that homelessness takes on those who experience it. As you can imagine, losing your home is a traumatic event in itself. And this trauma can be lasting. Once someone becomes homeless, they are immediately facing additional problems like finding housing, basic necessities, and healthcare. They are more likely to lose a job and have more frequent hospital visits. If we can prevent that suffering, we should.
Temporary financial assistance seems like an expensive but temporary band-aid. Is it really worth it in the long-term?
Financial assistance prevents homelessness rather than just delaying it. Even 12 months later, people offered prevention services are less likely to be homeless. And prevention is also less expensive. Based on individuals not facing the adverse effects of homelessness mentioned above, we conservatively estimate in our study that communities get $2.47 back in benefits per net dollar spent on emergency financial assistance. This frees up those dollars to go towards other supportive services, like those that help the chronically homeless.
If this intervention is effective, why are homeless numbers still rising and what can be done?
Short answer: Get this intervention in the hands of more people who can benefit from it. The challenge comes in offering a preventative measure to families that currently aren’t interacting with the system until it is too late. How do we identify the individuals or families at risk and proactively reach out to them to increase awareness of services available to help in their specific situations? LEO is currently studying implementing a new risk assessment tool in Texas that could lead to a more equitable tool to use than the VI-SPDAT (the current tool used across much of the country). We know some factors increase one’s chance of becoming homeless, such as prior exposure to homelessness or being released from incarceration. If a public entity with access to that information and those populations were to take over an emergency financial assistance intervention, could they systematically and effectively offer the intervention to populations with a naturally higher risk of homelessness?
How are the lessons from this study being used currently?
Since emergency financial assistance has now been tested in two different cities during different frames of time, it is safe to say that it can be effective. Other communities can learn from these examples. There is an organization in King County, WA that has worked to learn from Destination:Home. They are now putting to use lessons learned from Destination:Home as they offer emergency financial assistance in their own community.
If an RCT gives such clear evidence about effectiveness, why isn’t it always done?
It’s hard. It can be a difficult process to design and takes time and money. To effectively run an RCT, you have to have researchers in it for the long-haul and, more importantly, a community or organization who is dedicated to the RCT from top-to-bottom. They must be interested in the impact of their intervention, and willing to accept the time it will take to properly test their intervention and different possible research findings. This is where we were so fortunate to partner with Destination:Home. They were truly all in from the start to the end and we continue to work with them closely sharing the impact of their intervention with as many people as possible.
Does LEO focus exclusively on homelessness? How do you find partners and what makes an intervention a good fit for a LEO study?
LEO’s mission is to identify and share effective poverty solutions and policies. Poverty is complex, so our research focuses on a wide range of issues. Our main focus areas are homelessness, education, self-sufficiency, health, and criminal justice. We also have some emerging focus areas, like foster care. We’ve been fortunate to partner with more than 80 organizations across the country and have connected with them through a referral of some sort, we reach out to them or they find us. Either way, they go through our partnership process which kicks off with a two-day workshop on campus where they learn all about impact evaluations, get to know their research team, and start the design process. We work closely with them to plan, launch and implement the project. Once a project is launched, we regularly check-in to problem solve and share information. After results, the partnership continues with our policy and impact team because our ultimate goal is to not just build evidence, but see that it’s used.
Brian A. Jacob is the Walter H. Annenberg Professor of Education Policy and professor of economics at the Ford School at the University of Michigan, and is co-director of the Youth Policy Lab. He is an editor of The Review of Economics and Statistics.
David Phillips is a research professor at Notre Dame’s Wilson Sheehan Lab for Economic Opportunities (LEO).
James Sullivan is the Gilbert F. Schaefer College Professor of Economics at Notre Dame and a co-founder of LEO.
Phillips and Sullivan’s 2023 paper “Do Homelessness Prevention Programs Prevent Homelessness? Evidence from a Randomized Controlled Trial” was published in the May 2023 issue of the Review of Economics and Statistics.