Assessments that determine who gets help with homelessness sometimes prioritize white people over people of color, according to a study based in part on data from King and Pierce counties.
The results spurred one Seattle nonprofit to call for communities around the country to rethink how they triage people seeking help with housing.
"Race is an underlying factor in determining who the homeless system refers to a housing program — and who doesn’t get those referrals," said D’Artagnan Caliman, executive director of Building Changes, which funded the study. "That’s simply unfair."
Governments use "coordinated entry systems" to determine who gets limited housing resources. The systems often use "vulnerability index" surveys to determine who is most in need of help so those people can get resources first.
The scores are based on the problems people self-report. People who report long or repeated periods of homelessness, mental illness, substance use, and other conditions get higher "vulnerability scores" and are prioritized for getting help.
But those triage tools can worsen existing racial disparities in the homeless population, according to the study by C4 Innovations, a Massachusetts-based research and training company.
Among single adults, white people had higher mean scores than black and indigenous people, according to an analysis by C4 Innovations. The difference was "small" but "statistically significant," the company said.
Race was not a "significant predictor" of how homeless families are prioritized, C4 Innovations said.
The study was based on data from "coordinated entry systems" in King and Pierce counties; Multnomah County, Oregon; and the Blue Ridge area of Virginia.
Building Changes sought to fund the study after Pierce County officials, in an internal analysis, found "racial disproportionality" in who was referred to services for homelessness, said Daniel Zavala, director of policy and strategic communications for Building Changes.
The nonprofit wanted to find out if there were similar patterns elsewhere in the country, he said.
People of color, specifically those who are black, are much more likely than white people to be homeless in the U.S.
A 2018 study by C4 Innovations pointed to a theory called "network impoverishment." It's the idea that communities and families of color have fewer safety nets, such as friends or relatives with resources, so a financial crisis is more likely to spiral into homelessness.
But racial disparities in homeless services point to another possible factor.
The C4 Innovations study said differences in culture or experiences could make people more or less likely to volunteer facts about themselves, and that could affect vulnerability scores.
The study pointed to a question about drug diversion, the practice of giving a legally prescribed drug to someone else for illicit use.
"One might understand why someone experiencing homelessness, who is seeking services and likely already experiencing discrimination in some form in their lives, might be unwilling to report this," the report said.
"When we use tools, we often think that they're objective," Zavala said. "And when you see a study like this come out and actually show that there are differences in terms of how folks are scoring based on race, that demonstrates that that's just not right."
Zavala said the study pointed for the need for "targeted and tailored practices or questions" that "specifically address racial inequities" in who receives help with homelessness.
King and Pierce counties have already taken steps to refine their systems for assessing people seeking homeless services, according to Building Changes.
