King County’s homeless count revised up to 16,868 due to improved survey method

(The Center Square) – King County’s 2024 homeless population has been revised upward by 483 people after new data was derived from a unique survey method used by the King County Regional Homelessness Authority.

Unlike other agencies, KCRHA conducts its U.S. Department of Housing and Urban Development’s required Point-in-Time, or PIT, unsheltered count in even- numbered years. Sheltered counts are done every year. When the agency initially reported its count in May 2024, 16,385 people were tallied.

However, the count was recently revised to a total of 16,868 homeless people in King County. This is a result of a subset of the unsheltered homeless population being added to the count as data continued to be processed.

In a Zoom call with The Center Square, KCRHA Chief Research and Data Officer Tim Thomas and Joel Bernstein, data system coordinator at KCRHA, explained the change in PIT count statistics.

KCRHA utilizes a more advanced method for its PIT count than other continuum of care systems in Washington called “respondent-driven sampling,” in which the agency collaborates with the University of Washington to hand out flyers for unhoused people to distribute to their peers who are also living unsheltered. This method is sometimes called “snowball” sampling.

This peer-to-peer approach allows KCRHA to reach unsheltered people who might otherwise be missed – especially those living in remote encampments or who avoid formal shelters. KCRHA said the method provides more accurate data than the traditional method, which involves volunteers going out in the middle of the night in January, counting people, and then coming to an estimate on demographics based on that, which is why the federally-required PIT count is widely considered to be an undercount.

To encourage participation, KCRHA offered $20 Visa gift cards to unhoused people who completed the survey. Participants could also earn up to $15 more by referring peers, creating a chain-style outreach model intended to tap into homeless groups.

This method saw as many as 14 people get surveyed based off of one person who initially took the survey, according to Thomas.

The 2024 count was the second time the agency used the respondent-driven sampling method. In both counts, PIT data increased from initial results.

Bernstein explained that the nature of unsheltered families makes it difficult to find and survey them because they are much less likely to come to a hub to be counted. To improve data collecting, KCRHA set up a phone line and partnered with the nonprofit organization Mary’s Place to have people call and finish their surveys.

“The nature of that data requires a little extra analysis and when [the delayed data] got added in, that’s the majority of that increase that you see,” Bernstein said in the Zoom call.

Thomas pointed to a study by the National Law Center on Homelessness and Poverty, which found that the PIT count is between 2.5 to 10.2 times lower than the actual number of homeless people due to people entering and exiting out of unhoused shelters throughout a year.

This chain sampling method better records how many people are homeless and not in the shelter-system, Thomas said, noting it cuts the inaccuracy factor down from 10 to

“If somebody lives deep in the woods, but they have a friend that could refer them to the survey, then that’s a much more effective way to get them to come out than trying to go and find them,” Bernstein said.

The respondent-driven sampling method is still not perfect due to its opt-in process, meaning people living in their vehicles and homeless families are under-represented in the count, according to Bernstein.

Thomas added that KCRHA faces challenges counting unhoused people in rural parts of King County.

For 2026, Thomas said KCRHA is considering expanding its reach by setting up mobile data hubs in areas with high concentrations of unhoused residents, aiming to improve accuracy by ensuring more people are counted.