In early 2025, as Multnomah County, Ore., prepared to finalize its budget for the next financial year, the county’s homeless services department received some troubling data.
The department had distributed $17.6 million to “culturally specific service providers”—that is, organizations that cater to specific racial groups—in 2024, a 91 percent increase over the previous year. But outcomes for homeless minorities appeared to have worsened during that time, with more than 80 percent of the department’s programs reporting a decline or no change in the diversity of their clientele.
The trend came even as the number of “equity-focused” trainings conducted by the department jumped 42 percent over the same period, from 72 in fiscal year 2024 to 102 in fiscal year 2025.
The statistics were a red flag for the department’s budget advisory committee, which pressed the Portland-area county to provide more granular data and explain why its approach wasn’t working.
“How much better [were] outcomes for culturally specific providers vs non-culturally specific providers,” a committee member inquired in a February 2025 email. “Can you share data demonstrating that culturally specific providers produce better outcomes?”
Department officials replied that they weren’t collecting such data. What mattered, said program supervisor Breanna Flores, was the “lived experience” of the “communities being served.”
“The [Homeless Services Department] does not have expertise that supersedes that expertise in determining what would work best for a particular community, and relies on this lived expertise and experience,” Flores wrote in an email. “Specific to your question ‘How much better outcomes were for culturally specific providers vs non-culturally specific providers?’ – that is data the HSD is not currently exploring.”
Over the next several months, the department’s top brass repeatedly downplayed the importance of measurable outcomes, arguing that “qualitative data can have as much impact as quantitative.”
“We try to lean away from studying things other groups have done such as changes resulting from use of culturally specific providers,” department director Dan Field and equity manager Emily Nelson told the budget committee on March 10, 2025, according to the county’s published meeting minutes. The “goal is to create a diversity of pathways forward.”
Disturbed by the apparent indifference to evidence, a member of the committee sent the data to the Oregonian, Portland’s flagship newspaper, arguing that the equity initiatives were failing the very people they claimed to help.
“The lack of introspection and review lends credence to critiques of our equity initiatives – specifically that resources allocated for equity serve merely as patronage, or that the equity practices adopted by the county are ritualistic rather than focused on material outcomes,” the committee member wrote in an email. “If additional spending on equity staffing, consultants, and programs does not improve equitable outcomes, this represents a clear waste of resources.”
The paper never did anything with the data. And neither did Multnomah County, which continued spending millions on race-based housing programs despite the evidence of their inefficacy.
The meeting minutes offer a window into the way that bureaucratic inertia and ideological inflexibility hamstrung the county’s response to its homelessness crisis. They also show how that crisis’s impacts were hidden, in part, by the county’s determination to portray its equity initiatives in a positive light, with data discarded or gerrymandered when it did not comport with the county’s preferred narrative.
The evasions puzzled the progressive budget committee, which could not understand why, if “equity” was the goal, the department seemed so blasĂ© about the uneven impacts of its programs.
“Because I’m deeply committed to equity, I’m deeply committed to ensuring outcomes are equitable,” a member of the committee said in a written Q&A with department leadership in May 2025. “Why are we seeing declines in equity measures across multiple program offers, despite dedicating 7 [full-time equivalents] exclusively to equity work?”
Part of the reason was that, until 2024, the declines had been obscured by a simple yes/no scoring system that lumped all “BIPOC” individuals together.
Rather than track which minority groups were receiving services at rates equal to or greater than their share of the homeless population, the department only looked at the share of non-white individuals overall, masking major variations over time and between different groups.
If, for example, Hispanics were slightly overrepresented in a rapid rehousing program—but blacks and Native Americans were dramatically underrepresented—that program would still qualify as meeting its equity goals, even if the program was distributing services less evenly than it had the previous year.
“By obscuring these figures behind a YES/NO outcome, any potential fluctuations, whether positive or negative, in the rate of BIPOC assessment may be hidden from view,” the committee wrote in a 2024 report to county leadership. “This kind of opaqueness does a disservice not only to the work of our committee but to the community at large.”
The county only provided more granular data after sustained pressure from the budget committee, counting up how many “BIPOC participant groups”—black, Asian, Hispanic, American Indian, and Pacific Islander—were proportionally represented in each program. In several cases, just one of the five BIPOC groups reached demographic parity. The rest were underrepresented relative to their share of the homeless population, suggesting that the county’s equity initiatives were failing on their own terms.
Between 2024 and 2025, just four programs saw an improvement in their “equity measures.” Twelve saw a decline, while another 12 saw no change in either direction, indicating that the overall picture was getting worse, not better.
“Can you explain [why] we see a decline in three … times as many program offers … [as] we see … increase,” the budget committee asked in the May 2025 Q&A. The department replied by denying the premise, asserting that “Housing programs generally meet equity metrics year over year.”
Meanwhile, the county’s housing crisis was getting worse across the board. Portland’s homeless population rose by 26 percent between January 2024 and January 2025—a period that overlapped with the 91 percent jump in funding for “culturally specific” services—with the per-capita homeless rate now hovering around 1.3 percent.
Those statistics did not stop the county from budgeting over $200 million for race-conscious initiatives over the next year, including more than $36 million for “culturally-specific supportive housing for Native American, Black/African American, Latin/x, Somali and immigrant and refugee populations.” Nor did they spur a rethink of the race-based screening tool that the county uses to allocate housing assistance.
The tool, which awards more points for “interest in culturally specific services” than for having a disability, is now the subject of a discrimination probe by the Department of Housing and Urban Development, as well as a private lawsuit by a disabled woman.
It is not clear how, if at all, the equity data informed final budget decisions. At a March 3, 2025, meeting, the department’s planning and evaluation manager, Lori Kelly, claimed that equity was “weighted into the choices we made about what to push forward and what not to.” But the department has not cut a single program for failing to meet its equity targets, with more than $90 million earmarked for programs that scored a 0/5 on the county’s BIPOC parity measure.
Kelly also suggested that the low scores shouldn’t be taken at face value. “Some new orgs may be working with difficult to reach populations and [it] could be hard to evaluate what’s going on at this point,” she said, according to minutes from the meeting. “May be the case the program offers are not the best source for equity measures or program outcomes.”
Multnomah County did not respond to a request for comment.