New Report: LA Could Better Target Homeless Prevention Services with Predictive Analytics

Predictive modeling can be used to help address the homelessness crisis in Los Angeles County according to a recently released report by researchers at the California Policy Lab and the University of Chicago Poverty Lab at the Harris School of Public Policy. A new predictive model developed by the researchers can help identify adults in LA County who are at high risk of becoming homeless in the near future. 

Using anonymized data from seven L.A. County agencies about services they provided to LA County residents between 2012 and 2016, researchers developed models to predict which residents were most likely to become homeless in 2017. The research team then checked the accuracy of their model’s predictions against County records to see who actually became homeless in 2017. 

Results

Of the 3,000 people whom the model identified as at highest risk of experiencing homelessness in 2017, either for the first time or for a repeat spell, almost half (46%) actually became homeless, according to county records. People in this group were 27 times more likely to become homeless as compared to the average county client. Researchers also identified 3,000 LA County residents who were at the highest risk of first-time homelessness. Of this predicted group, one in three subsequently became homeless, and people in this group were 48 times more likely to become homeless for the first time as compared to the average county client.  

Policy Impact

This research informed the formation of a LA County Homelessness Prevention Working Group and related Action Plan. The Action Plan, which was submitted to the LA County Board of Supervisors on December 16th, recommends the County use the models to help intervene with adults at high risk of homelessness before they reach a crisis. Recommendations include piloting a centralized, multi-disciplinary prevention unit that will have representatives from the Department of Mental Health, Department of Health Services, Department of Social Services, and the criminal justice system. The unit will accept referrals from the risks lists generated by the models and will proactively problem solve around each individual’s needs. The Action Plan, which includes a total of 17 recommendations, is expected to receive $3m in Measure H funding in 2020, in addition to resources currently deployed by mainstream LA County departments.

“Bringing together data from multiple county agencies gave us a more nuanced understanding about what’s happening to people right before they slip into homeless and how services can be better targeted to prevent that from happening,” explained Till von Wachter, a co-author of the report, and faculty director at the California Policy Lab. 

“Last year, despite providing housing to tens of thousands of people, we saw more and more individuals and families becoming homeless,” said Phil Ansell, the director of Los Angeles County’s Homeless Initiative. “L.A. County is focused on using strategic approaches to preventing homelessness, and these groundbreaking models will make it possible to reach those who need us the most before they reach the crisis point and fall into homelessness.”

“The models suggest that sharp spikes in service use, increasingly frequent service use, and the receipt of multiple services from a single agency are all warning signs that someone living in deep poverty is at high risk for homelessness,” commented Harold Pollack, the Helen Ross Professor at the University of Chicago School of Social Service Administration, a co-author of the study and co-director of the Chicago Health Lab. “We’re now diving deeper into the models with our LA County partners to learn more and to see how these results can help focus public health and social services to this vulnerable population."

“Predictive modeling can help ensure that homelessness prevention services are getting to the right people, at the right time, before they’re in a full-blown crisis,” explains Janey Rountree, executive director of the California Policy Lab (UCLA). “We look forward to seeing its impact in connecting people to the help they need.” 

Research Findings

• Effectively serving the 1% of County clients who are at greatest risk of a new homeless spell would prevent nearly 6,900 homeless spells in one year. 

• The majority of single adults who will experience first-time homelessness or a return to homelessness are already clients of mainstream County agencies, which presents opportunities for intervention. 

• The highest risk clients are vulnerable and are interacting with multiple agencies.

• Falling into homelessness happens very fast – typically within 6 months of a precipitating event. The County and service providers must react quickly.

The researchers caution that the data used to populate the models don’t include all risk factors connected to experiencing homelessness, such as living in “doubled up” housing situations, experiencing financial shocks, or domestic violence. These risk factors could be captured in client interviews or additional screening tools once clients are connected to services. 

This research was provided at no cost to the County. It was generously supported by Arnold Ventures and by the Max Factor Family Foundation.  

Additional Background

Each year, two million single adult county residents receive housing, health, and emergency services from Los Angeles County, and about 2% of this group (76,000 people) will become homeless. This includes single adults who will be homeless as part of an ongoing episode of homelessness (42,000); after having been previously stably housed for at least six months (20,000); or for the first time (14,000). 

The anonymized data for this research project was provided by Los Angeles County and came from the Los Angeles Department of Health Services, Department of Mental Health, Probation, Sheriff’s Department, Department of Public Health (Substance Abuse Treatment & Control), Department of Public Social Services, and the Los Angeles Homelessness Services Authority.

The California Policy Lab 

The California Policy Lab creates data-driven insights for the public good. Our mission is to partner with California’s state and local governments to generate scientific evidence that solves California’s most urgent problems, including homelessness, poverty, crime, and education inequality. We facilitate close working partnerships between policymakers and researchers at the University of California to help evaluate and improve public programs through empirical research and technical assistance. For more information visit www.capolicylab.org.

The Poverty Lab at the University of Chicago

The Poverty Lab partners with civic and community leaders to identify, rigorously evaluate, and help scale programs and policies that reduce poverty and create paths to social mobility. This year, the Harris School of Public Policy became the new academic home of UChicago’s five Urban Labs: Crime, Education, Health, Poverty, and Energy & Environment. The Poverty Lab specializes in evaluating programs with the potential to improve individuals’ postsecondary success, housing security, workforce development, and household finances. Working in partnership with private, government, and nonprofit stakeholders, we bring to bear the power of data and evidence to address major social challenges. For more information, visit https://povertylab.uchicago.edu.