Arrow-right Camera
The Spokesman-Review Newspaper
Spokane, Washington  Est. May 19, 1883

Why King County homeless count continues, even with accuracy in doubt

Andrea Suarez dismantles a tent as garbage lies piled at a homeless encampment on March 13, 2022, in Seattle.  (John Moore/Tribune News Service)
By Greg Kim Seattle Times

SEATTLE – With $63 million of federal funding on the line, months of planning and thousands of paid and volunteer hours have gone into conducting King County’s 2024 Point-in-Time count, a biennial snapshot of the homeless population.

But despite the time and money involved, some in the homelessness system say the effort serves no useful purpose.

The Point-In-Time count, a biennial snapshot of the homeless population, has long been criticized as an underestimate.

The King County Regional Homelessness Authority set out to fix that in 2022, switching from physically counting tents to a statistical sampling method that could enumerate people outside even if they were hidden.

Reception was mixed at the time, but the authority doubled down on that method this year, making a few adjustments to appease critics.

Some experts say that switching methods of counting, or even slightly tweaking them, can make year-over-year changes difficult to compare.

The federal agency that requires the count says the main purpose of the Point-in-Time count isn’t to get the most accurate numbers, but to see increases or decreases over time.

Meanwhile, the Regional Homelessness Authority downplays the Point-in-Time count altogether, saying other, larger numbers more accurately reflect the scale of the issue.

Counts can show success

Vanessa Krueger, spokesperson at the U.S. Department of Housing and Urban Development, wrote in an email that the amount of funding is not directly correlated to the number of people identified in the Point-in-Time count, but that conducting one is a condition for receiving federal homelessness dollars.

“It is important because it is the only national data source that captures people in both sheltered and unsheltered settings in a consistent way that enables year-to-year and community-to-community comparisons,” Krueger wrote.

Stephen Metraux, a homelessness researcher at the University of Delaware, said that decreases captured in the Point-in-Time count are useful to correlate with programs and interventions that are working. He highlighted how a national drop in veteran homelessness has created a blueprint on how homelessness can be addressed through housing subsidies and wraparound services.

Even consistent increases in homeless numbers “prick the conscience” of the public and make it more difficult to ignore that the problem is growing, he said.

A data point among others

2020 was the last time the count was done in King County the way most other places do it. In a single day in January, volunteers fanned out across the region, knocking on RV and tent doors, canvassing food banks and other places homeless people gather.

Anitra Freeman, a board member with the homelessness organization SHARE/WHEEL, volunteered for the Point-in-Time count for years using this method, knowing it was an undercount.

“My husband called it the one night count of homeless people who don’t know how to hide,” Freeman said.

Since the Regional Homelessness Authority was created, it has preferred to cite results showing that 53,000 people experienced homelessness in 2022, which is calculated by the Washington State Department of Commerce using records of different systems homeless people have interacted with.

“Because we are where we are in King County and have administrative data to understand those things, I think that it is less important for us to produce the one-night count,” said Owen Kajfasz, data lead at the Regional Homelessness Authority.

The agency has also moved to only counting the unsheltered population once every two years, the minimum requirements from HUD.

Mayor Bruce Harrell’s office said both homeless censuses are likely inaccurate and that changing the methodology frequently limits how they can be used.

“Given these factors, the City considers the (Point-in-Time) count to be one data point among others that can be used in assessing the effectiveness of our regional response,” a spokesperson for Harrell’s office wrote in a statement. “It does not, however, directly determine funding amounts from year to year.”

King County relies on both numbers, and says that both point to a growing homelessness crisis due to a lack of affordable housing options.

Tweaks made for accuracy

In 2022, the authority tried to skip the Point-in-Time count until it was chastised by the federal housing department. In a rush to get one done on time, the agency switched from a literal count of visible tents to a statistical survey that involves a lot of math and projections.

Kajfasz, who’s leading the Regional Homelessness Authority’s Point-in-Time count, said the old approach only worked well in urban areas, where homelessness is most visible, whereas the new method can also count people in suburban and rural areas and greenbelts. He added that its sampling model is a more rigorous method of counting that the World Health Organization uses and HUD has approved.

But there were critics. Some local leaders felt that homeless people in their city weren’t being counted. Chief Seattle Club argued that the count in 2022 didn’t adequately reach Native people. And some advocates said people living in their cars were undercounted.

In 2024, the Regional Homelessness Authority made adjustments in who and how they reached out to the people they surveyed, and where they did it, consulting with the officials and groups that complained in 2022.

Kajfasz said that the changes the authority made between the 2022 and 2024 counts should not affect their comparability because the first survey group can largely be filtered out of the data, removing the bias of who is initially surveyed.

Metraux, at the University of Delaware, said the new method will likely capture more people, but making changes opens the count up to criticism if people don’t like the numbers produced.

“Little tweaks have the danger of having big changes in the results,” Metraux said.

Officials aren’t losing faith in the Point-in-Time count yet. Redmond Mayor Angela Birney, a Regional Homelessness Authority governing committee member, wrote in an email that she will believe the results of the 2024 Point-in-Time count and that the city’s outreach staff helped make sure homeless people showed up to the survey.

“Refinement to this year’s (Point-in-Time) are based on lessons learned and are intended to result in more accurate sampling,” Birney wrote.

Money for information

Over the course of an hour on a Thursday morning, about a half-dozen homeless people entered the Georgetown Food Bank, one of 17 locations around the county where the 2024 King County Point-in-Time count is being conducted. The location was partially chosen because it’s close to areas where many people live in their vehicles.

Raymundo Ortiz, who lives in a tent on Fourth Avenue in Sodo, brought a flyer an outreach worker had given him that contained a coupon code for a $20 gift card for participating.

He sat down at a table with a volunteer who asked him where he stays and where every other homeless person he knows stays. The ratio of the people he knows staying in shelters versus on the street is key to calculating this count.

The Regional Homelessness Authority already knows how many people stay in shelters. If it can calculate the ratio of people staying outside versus in shelters by talking to hundreds of people like Ortiz, it can estimate how many people live outside without counting them all.

The agency has completed more than 1,350 surveys with people, double what it did last year, which should yield a more accurate ratio between sheltered and unsheltered homelessness. The authority expects to have a final count early this spring.

Ortiz also answered questions about his demographics, his health and details about being homeless. He received his gift card and three flyers to hand out to his friends. He said he would use the money on groceries.

Freeman, who volunteered for this year’s count at the Georgetown Food Bank, is skeptical the new methodology is more accurate. But she said it might not matter.

“The numbers don’t count unless they’re going to do something with it,” Freeman said. “They keep getting these numbers. They keep saying there’s thousands of people out there that don’t have shelters, and then they don’t create thousands of units of shelter.”

Regional Homelessness Authority spokesperson Anne Martens agreed, writing in an email, “Our system does not currently have the resources to meet the need.”