Digging into Harrell's "They Aren't From Here" Homelessness Talking Point

The mayor says most homeless people in Seattle became homeless somewhere else, so the rest of the county needs to do its fair share. But his talking point is based on skimpy evidence.

By Erica C. Barnett

On the campaign trail and in his slickly produced Seattle Channel budget video, Mayor Bruce Harrell has been touting a new statistic: “Most of the most of the people that are homeless in Seattle did not become unhoused in Seattle,” he said at a recent debate. He’s made a similar claim at other campaign forums and events.

“About 70% of people experiencing homelessness on Seattle streets became homeless outside of our city, and even with Seattle providing over six providing over 60% of the region’s shelter beds and 85% of the region’s tiny homes, we cannot do this alone,” Harrell claimed in his budget video, adding that it’s time for the rest of the region to step up and take care of their homeless residents. He also called on the King County Regional Homelessness Authority to identify land in areas with “fewer land constraints than we have here” where new shelters could open away from the city.

Leaving aside the fact that unhoused people congregate in cities, rather than far-flung suburbs, for obvious reasons—not only are there services here, there’s also community, opportunity, and access to transit—Harrell’s “they’re not from here” stat is extremely misleading, because it’s based entirely on responses to about 240 surveys with people who showed up at KCRHA’s survey tents over several days in January and February 2024.

A bit of background: Since 2022, the KCRHA has based its Point In Time Count of the region’s homeless population not on a physical count (which generally results in undercounting) but surveys with people who travel to fixed sites to provide demographic data and answer questions. This technique, known as respondent-driven sampling, was used to extrapolate some of the data in the full 2024 report, including an estimate (16,868) of the total number of sheltered and unsheltered homeless people across the county.

However, Harrell’s “about 70 percent” number (actually 69 percent) isn’t based on that sampling method; it’s based on raw data that represents less than a third of about 800 surveys across all parts of King County. In Seattle, the KCRHA interviewed around 240 people.

This is important, because statistics based on small, self-selected samples of people aren’t generally considered reliable guides to demographic and population-level trends; that’s why KCRHA used statistical sampling for all its top-level data, rather than just extrapolating directly from a few hundred surveys.

It’s also to important to know that Harrell’s new talking point completely ignores the 43 percent of people who said they were last stably housed outside King County, which is one reason the “last housed in Seattle” and “last housed in King County” sample sizes are so small. If the city and county based their “fair share” of homelessness spending on where people were housed most recently, it would make sense for both Seattle and King County to refuse to fund services for nearly half their homeless residents, a heartless policy neither would be likely to adopt.

During a “deep dive” on the PIT numbers on Friday, KCRHA data staffers confirmed that researchers intended for the raw survey information to be taken “at face value,” not used to “make any further inferences” about the homeless population as a whole. The staffers also noted that KCRHA didn’t ask how long it had been since people had stable housing; given that the number of chronically homeless people increased in the 2024 count, people could have been talking about housing they had years in the past, which makes the question of which city or region they “belong” to murkier.

Like previous point-in-time surveys, the 2024 PIT count suggests not just that people are mobile—relocating around the region for various reasons, including local laws designed to keep them moving—but that “last stably housed” statistics aren’t a a good way of deciding whether people are worth spending local dollars on—even ignoring the obvious anti-humanitarian implications of making policy this way.