Nika Bartoo-Smith
Underscore Native News + ICT
KING COUNTY, Washington — Data collection methods on Native communities continue to be flawed, especially by the federal government. The latest example is King County’s 2026 Point-In-Time Count (PIT Count) of unsheltered individuals.
The King County Regional Homelessness Authority report showed that 4.2 percent of unsheltered individuals self-identified as American Indian, Alaska Native or Indigenous. The count happens every two years.
However, leadership at both Chief Seattle Club, a Native-led housing organization, and the Seattle Indian Health Board, a Native-led community health center, agree that that number is likely closer to the 2020 count which was 15 percent.
While these two Native serving organizations that are on the frontlines of addressing homelessness believe that the count method itself is flawed, others argue that it’s not about the method, but the collection process itself.
This latest count emphasizes the larger issue of data collection and the tangible harms it can cause in Native communities, according to Abigail Echo-Hawk, a citizen of the Pawnee Nation of Oklahoma.
Echo-Hawk is the vice president of the Seattle Indian Health Board and director of the tribal epidemiology center there, the Urban Indian Health Institute.
“The relatives that I serve are going to suffer as a direct result, as funds will be misallocated based on bad data because of the lack of representation of American Indians, Alaska Natives in this report, and Native people will suffer, and Native people will die,” she said.
A flawed method?
Since 2022, the PIT Count in King County has been conducted using a data collection method called respondent-driven sampling.
The PIT Count, mandated by the federal department of Housing and Urban Development, is conducted in a single night and is supposed to provide a snapshot of homelessness across the country.
Essentially, this method involves setting up hub sites around the city then finding people experiencing homelessness around each site. Those people are asked to take a survey, then given coupons and asked to recruit other people experiencing homelessness, according to Zack Almquist, a professor of sociology at the University of Washington who designed King County’s respondent-driving sampling method.
Following the surveys, those responses are used to estimate a larger population figure based on the collected sample.
During the count this year, the King County Regional Homelessness Authority had more than 150 community volunteers surveying at nearly 30 locations across the county.
“[The method is] specifically built for getting hard to reach populations and underserved populations,” said Almquist.
But some people are concerned with the respondent-driven sampling method as a whole and argue it does not work for the urban Indigenous population in King County.
“The methodology is flawed. Period,” Echo-Hawk said. “The flaw starts out with the beginning of the recruitment process.”
For Chief Seattle Club, there is a three-pronged issue with the respondent-driven sampling methodology, according to Margaret Faliano, Chippewa Cree, communications director for the organization. First, the sampling is an approach that relies on individual incentives.
“We’ve seen this in all messaging, like for instance in COVID, to get Indian Country vaccinated,” Faliano said. “Messages that focused on yourself did not work. Messages that focused on the collective, on the elders, on the youth, on the next seven generations, that worked.”
Second, she said that the data collection sites themselves were largely not spaces where Native people experiencing homelessness are accessing services, as there was no coordination with organizations like Chief Seattle Club and the Seattle Indian Health Board.
Third, as with other federally compiled data, there is a multiracial category that does not break down further. This means that if someone identifies as Native and any other race, they are put into the multiracial category, not the Native category.
According to the King County Regional Homelessness Authority 2026 count, 9 percent of individuals identified as multiracial.
“The multiracial category was a throwback to when they considered the United States to be a melting pot. The intention with that category was to eliminate culture, and when we look at that multiracial category, it is from what I consider, and have told every epidemiologist I know, a trash data category,” Echo-Hawk said. “There is no ability to truly analyze what is happening within that multiracial category, and that’s intentional.”
The federal reporting standards are the issue, not the method itself, according to Almquist.
The federal government’s definition for American Indian and Alaska Native is the first challenge.
“When they’re collecting data, the American Indian, Alaska Native definition is inclusive of all people Indigenous to North, Central or South America,” Robert Maxim, Mashpee Wampanoag citizen and a fellow at Brookings Institution in Washington, D.C., told ICT. “And so that becomes a challenge because that’s a much broader definition than who the federal government offers service to and has a government-to-government relationship with, which is just enrolled citizens of federally recognized tribes.”
Almquist did also acknowledge a need to build trust and work closely with the Indigenous community in preparation for the next count in 2028.
“I think a lot of the pushback is maybe being mischaracterized on the method and [it’s] more on the federal reporting standards that are required. The federal definition for some of the subgroups, especially the way that it handles multiracial categories, excessively hides the size of many of these populations, and one of the things that really needs to happen is a local report that disambiguates that, so we can see how big the populations really are,” Almquist said.
What can be done
While there is disagreement on if the respondent-driven sampling method itself is suitable for the PIT Count, the fact remains that the Native community experiencing homelessness in King County is being drastically undercounted which can have major consequences for funding.
“The math might work if the appropriate trust existed in order to get the sample size or the unit engagement from the respondents themselves,” Echo-Hawk continued. “The math doesn’t matter if the input of the data is bad.”
Almquist, Echo-Hawk and Derrick Belgarde, chief executive officer of Chief Seattle Club, all agree that there needs to be more engagement with the Indigenous community in Seattle prior to the next PIT Count.
While Almquist proposed unpacking the numbers on a local level, particularly when it comes to the multiracial category, Echo-Hawk and Belgarde believe the respondent-driven sampling method should be done away with all together, as the methods should be Native-informed from the beginning.
“They’re always just focused on creating these one size fits all tools of whatever they’re doing in this case, data collection,” Belgarde, Siletz and Chippewa Cree, said. “The issue is that they developed a tool, put it out there that we said ain’t work, and now instead of scrapping it and figuring out a way that’s more equitable for all communities, they’re just focused on using it and trying to get our input on how they can make it better. But they can’t. It’s fundamentally flawed.”
For now, both Belgarde and Echo-Hawk agree that the King County Regional Homelessness Authority should issue an acknowledgement that this data is flawed and this should not be considered or used when it comes to any funding or other analysis.
UNN + ICT reached out to King County Regional Homelessness Authority for comment, and did not hear back from them prior to publication.
This story is co-published by Underscore Native News and ICT, a news partnership that covers Indigenous communities in the Pacific Northwest.
