The Washington PostDemocracy Dies in Darkness

A survey was meant to help the homeless. Some say it hurts Black people.

William Pearson recently took a survey that assesses vulnerability among homeless people in D.C. and elsewhere. (Bill O'Leary/The Washington Post)
Placeholder while article actions load

William Pearson came to D.C. about a year ago by way of North Carolina. After a stint living unhoused in the South, he thought the District would be friendlier to homeless people.

Then earlier this year, Pearson, 39, had to deal with some red tape: To qualify for a housing voucher, he had to complete a 27-question survey that was “a lot about homelessness and demographics,” he said. He has filled it out twice — once around the end of winter and again in April at the shelter where he’s staying — but isn’t sure when he’ll get his results and doesn’t expect to hear back about a housing voucher for up to a year.

“I don’t think the test itself was too bad,” he said. “It’s the process of them getting the results back and having the structure to utilize these results. … The system is overwhelmed.”

By seeking a housing voucher, Pearson, who is Black, stumbled into what some advocates say is an obstacle to getting unhoused people housed — a survey that was designed as a tool to gauge vulnerability but has been disowned by its co-creator after studies showed it disadvantaged minorities.

Despite these concerns, the survey — known as the “Vulnerability Index — Service Prioritization Decision Assistance Tool,” or VI-SPDAT (pronounced VEE-EYE-SPID-DAT) for short — has been used in at least 39 states as well as the District. And though the U.S. Department of Housing and Urban Development (HUD) once encouraged communities to use the tool, the agency cannot provide information on how many use it or how they use it.

Number of homeless residents in D.C. lowest in 17 years, mayor says

Jurisdictions are using the VI-SPDAT to help them determine which homeless people are most vulnerable and should get scarce affordable housing. In the District, it’s used to inform this decision but is not the deciding factor, officials said.

According to advocates, this test meant to measure vulnerability may be leaving vulnerable people behind.

D.C. Department of Human Services Director Laura Zeilinger said she was aware of concerns about the VI-SPDAT. However, she said, the District doesn’t deploy it as a “math formula.”

“It is knowledge of people and their needs,” she said.

Iain De Jong — a longtime advocate for the unhoused and president of a Canadian company specializing in “homeless system transformations,” according to its website — said he helped develop the survey because advocates a decade ago were looking for a “triage tool” to figure out which homeless people needed housing the most. HUD regulations issued in 2012 pressed for “an initial, comprehensive assessment of the needs of individuals and families” as they first sought housing assistance.

De Jong’s company, OrgCode Consulting, worked with the New York-based housing advocacy group Community Solutions to combine two tools the organizations developed independently to create the VI-SPDAT, which launched in 2013.

Designed to be completed in seven minutes, a 2015 version asks about clients’ health, daily functioning, and history of abuse and trauma. The survey tallies points for problems the client faces, such as hospitalizations and prison stays. The more points are recorded, the more-dramatic housing intervention is recommended.

But as the VI-SPDAT was deployed, it drew criticism. A 2019 study, for example, found that people of color received lower scores than White people and that the survey did not “equitably capture vulnerabilities.” Another study last year found the tool was racially biased, with White women scoring consistently higher than Black women because Black women are less likely to seek health care and more reluctant to report risky behavior.

Despite these criticisms, HUD on its website points to the VI-SPDAT as an example of a community assessment tool — with the caveat that it should be used as “a starting point.” While declining to comment on VI-SPDAT or any other assessment tool, HUD spokesman Michael F. Burns said the agency “encourages communities to use the coordinated assessment process to ensure that homelessness assistance is provided in the most equitable manner possible.”

However, the organizations that developed the VI-SPDAT have left it behind.

In an OrgCode blog post last year, De Jong said his organization would “phase out” the survey, transitioning to an “approach that also addresses racial and gender inequities — which the VI-SPDAT was never intended to do.”

De Jong said many communities didn’t use the VI-SPDAT as intended: As care providers determined who should get scarce housing, the survey designed as a “triage tool” became the only tool in the tool kit.

“Some nights I stay awake and think it’s catastrophic,” De Jong said. “As much as I tried to put the genie back in the bottle, I couldn’t right the ship.”

Community Solutions, the survey’s co-creator, has also stopped promoting it.

“We believe deeply that we all — including Community Solutions — have to be accountable for addressing disparity,” said Beth Sandor, one of the organization’s principals.

Nonetheless, the survey — the first tool created to respond to the government’s demand for a coordinated assessment of people seeking assistance — remains embedded in North America, said Tim Aubry, a psychology professor at the University of Ottawa who studies homelessness. Alternatives to the VI-SPDAT should have been developed and deployed, he said.

“In the world of consulting, if someone can respond to the need … people eat it up,” he said. “I don’t think the tool should have been used without proper research. I don’t understand why HUD is behind it.”

Julieanne Turner, a longtime social worker in D.C. working for the Downtown Cluster of Congregations, said she had administered the VI-SPDAT dozens of times. She noted that surveying people who were traumatized or mentally ill was problematic.

She offered an example: The VI-SPDAT asks about “mental health or brain issues” that might prevent a client from living independently. Clients might deny having these problems even though they recently complained about people following them or trying to hack into their bank accounts — classic symptoms of paranoia and schizophrenia. Thus, their care providers must prompt them to answer accurately.

However, if care providers do not know clients well enough to surface their problems in the VI-SPDAT, the clients’ vulnerabilities go unmeasured — resulting in a lower score that recommends less-serious intervention. The assistance they get depends upon their relationship with their care providers, Turner said.

“It’s not fair,” she said. “Not everyone is going to have an advocate.”

Not all care providers share Turner’s view.

Before the VI-SPDAT, housing vouchers were awarded haphazardly, said Adam Rocap, deputy director of the homeless outreach nonprofit Miriam’s Kitchen and a member of the District’s Interagency Council on Homelessness. Rocap helped bring the survey to D.C. Before, he said, a dying person who had been homeless for years might not get a voucher while a newly homeless healthy person might.

“It came down to who did you know and how good was your case manager,” Rocap said. The goal, he said, was to build a system where homeless people don’t have “to work very hard to be recognized.”

Used with other metrics such as length of homelessness, the VI-SPDAT facilitates this system, Rocap said. The survey is criticized, but the problem isn’t the survey, he said — it’s the scarcity of affordable housing.

“Whatever tool you pick … people are going to end up hating it,” he said.

Some jurisdictions have turned to other tools.

Quiana Fisher, strategy director for the Ending Community Homelessness Coalition, or ECHO, Austin’s main provider of homeless services, said that when HUD encouraged the use of the VI-SPDAT, communities fell in line.

“We have kind of built our systems around this tool that now we find out had monumental unintended consequences,” she said.

In October, the city launched the Austin Prioritization Index, a new survey developed with input from unhoused people that asks whether clients are natives of the city and whether they have been displaced from gentrified Austin Zip codes.

These new queries can surface vulnerabilities a generic tool developed elsewhere might not, Fisher said. She also said agencies will try to make sure Black and Brown service providers administer the index, which the community can update at any time if disparities arise.

“That research should be done with people with lived experience of homelessness,” Fisher said. “They are the experts in our system.”