A new report estimates about 2% of the Portland metro area’s population, or 38,000 people, were homeless at some point in 2017.

The report provides a simple, if rough, estimate of the scale of homelessness regionwide – in Clackamas, Multnomah and Washington counties — over a single year.

And it’s seven times higher than the approximately 5,200 homeless people in the three counties in the official Point-In-Time tally used by the state and federal government.

The report estimates it would cost at least $2.6 billion to fully fund housing and social services for 10 years to reach 38,000 homeless people.  

 Marisa Zapata, director of Portland State University’s Homelessness Research and Action Collaborative, is the lead author of the report.

Zapata believes her team’s calculations show the scale of homelessness in Portland is magnitudes greater than what most people perceive – the population of homeless people sleeping in tents or in parked cars.

“The great majority of people experiencing homelessness in the region aren’t visible,” she said.

“What we think we’re seeing that’s shaping and driving our problem definition, and it’s not even close to understanding the scope of the problem,” she said.

Zapata said the idea for the report came from Robert Stoll and Mitch Hornecker, attorneys who are influential in Portland’s business and philanthropy circles.

It was funded using part of a $910,000 donation to PSU’s new collaborative with Columbia Sportswear CEO and billionaire Tim Boyle.  He is also the most significant donor to the effort to launch a new homeless shelter at the edge of the Pearl District.

Coming Up With A Better Estimate

Estimating the population of homeless people is notoriously difficult.

Most advocates consider the counting method the federal government requires counties to use – a point in time count in January – flawed in ways that consistently undercounts the homeless population.

 That count, known as the PIT, doesn’t account for the flow of people in and out of homelessness over the course of a year, and it tallies people at a time of year — the dead of winter — when they are most likely to be seeking shelter and are out of sight.  

The new report uses the PIT data from Multnomah, Clackamas and Washington counties as a starting point – but attempts to correct the undercounting problem by expanding on it.

Zapata took the number of unsheltered people tallied by volunteers during the point in time count and multiplied it by 1.9,  to produce an annual estimate of the number of people living outside or in their cars.

Why 1.9? That’s a factor that Multnomah County’s Joint Office of Homeless services uses to extrapolate from the one day point in time count to an annual number, based on federal research into veterans homelessness.

Then, Zapata added in the number of individuals the three counties reported were in shelter or transitional housing in 2017.

Finally, she combined the county data with an additional source – data compiled by the Department of Education on the number of homeless students in 2017.

The DOE tracks a population that isn’t captured in any other official count of homelessness: students who have no stable address and are couch surfing, living doubled up with friends or relatives. Zapata used that data to derive an estimate of the number of children and parents living doubled up.

Zapata acknowledges that her 38,000 number is an estimate.

The methodology may double-count some homeless people while missing others — for example,  homeless adults without children who are staying with friends.

Mark Jolin, the Director of the Joint Office for Homeless Services, views it as a solid estimate.

 “There’s going to be some duplication,” he said. “I don’t know there’s a better way to get at a number that gives you a good ballpark sense of the scale of the problem.”