During a frigid January night, teams of volunteers and outreach staff spread across Allegheny County, searching streets, underpasses and hidden corners to capture the best possible snapshot of homelessness. It’s part of a federally mandated effort to survey homeless people nationwide on a single night. 

While these annual “Point In Time” counts have shown that homelessness has been rising in Allegheny County since 2021, this year’s results showed a higher than expected jump. And it led the county, which conducted the January survey, to take an unusual step: to conduct a recount in March. 

As the nation’s largest and most regular housing survey, the “Point in Time” count offers an annual snapshot of people in shelters, temporary housing and outdoors, although it has been criticized for underestimating the true scale of the crisis, as it only captures data from a single night.

“We found ourselves in a situation where I thought it would be really, really difficult to explain [the results], so we wanted to collect another data point,” said Erin Dalton, executive director of Allegheny County Department of Human Services.

Advertise with us

The city and county have made major moves to address the ongoing homelessness crisis in the last year — transitioning 300 people into housing through the 500-in-500 initiative, adding more than 100 shelter beds during the winter and dismantling large encampments near Downtown.

Still, the January survey found a significant number of people sleeping outside Downtown, as well as 120 people who never engaged with any county street outreach prior. In an email to the county’s Homeless Advisory Board last month, Ms. Dalton described the results as “unexplainable.”

A key change made in the second count in March was ensuring that every person identified as unhoused had a completed survey, Ms. Dalton told the Post-Gazette. In January, 80 people were counted without one, making it difficult to verify their housing status or collect any additional information.

In the first count, Ms. Dalton said there wasn't a consistent strategy for identifying places where unhoused people might be staying and where outreach teams were surveying. That made it difficult to verify results and identify cases where people may have been overlooked or mistakenly identified multiple times.

And at times, outreach staff also marked down people they knew to be living in encampments, but didn’t actually lay eyes on during the actual survey. The March count required outreach teams to document exactly where people were found, adding a new feature to the survey where they could drop a pin on a map.

Advertise with us

While Ms. Dalton acknowledged that these issues may have happened in past surveys, she said previous results aligned more closely with DHS’s expectations.

“It didn't catch my attention because the results made sense,” she said. “They were consistent with the data that we were collecting and publishing weekly. They made logical sense, so we didn't do a deep dive on our methods until this year.”

Local shelters like Light of Life say their numbers have always indicated a greater homelessness crisis than what the PIT surveys ever showed. 

The 2024 PIT count identified just over 1,000 people experiencing homelessness in Allegheny County. Yet, Light of Life served over 3,000 people alone over the last year. The nonprofit’s executive director Jerrel Gilliam said county officials have discouraged him from speaking out about these discrepancies.

“They were saying it's not good to be implying that the numbers are higher, that it will cause fear,” Mr. Gilliam said. “It was like it was a PR issue. But we want to know the truth. We want to know what the actual numbers are, and then we can begin working it.” 

Advertise with us

Advocates nationwide have long argued that the PIT count has fundamental flaws. The federal survey excludes individuals who are "doubled up" with friends or family due to economic hardship, as well as those in the jail, hospital and university systems. Research indicates that the actual number of people experiencing homelessness in the U.S. is between 2.5 to 10 times greater than what PIT surveys show. 

But Linda Gibbs, a former NYC commissioner of homeless services who has launched PIT counts worldwide, said it’s not designed to be a precise count. That’s why HUD gives so much flexibility to local municipalities on how they conduct the surveys each year.  

There are fewer homeless camps around the city on Thursday, April 3, 2025. But where did they go? This is one of two tents at the end of the Hot Metal Bridge, along South Water Street on the South Side. (Lucy Schaly / Pittsburgh Post-Gazette)

“By definition, it's an estimate,” Ms. Gibbs said. “When you're looking at an entire city, you're doing your best to deploy methodology that giving you a really strong sense of what the number of unsheltered individuals is. But it's still an estimate.” 

Allegheny County DHS recently invited Ms. Gibbs to meet the team in Pittsburgh and discuss best practices for PIT counts. While she hasn’t examined Pittsburgh’s methodology in depth yet, she described her meeting with DHS as “very positive.” 

She said local leaders are trying to determine whether their methodology led to misleading results or if there were genuine changes in the number of people experiencing homelessness. 

Advertise with us

“A lot of places do these counts with a kind of hesitancy and bury the numbers,” she said. “This is the opposite, which is quite refreshing.”

Still, some questioned the need for a recount, including those who were involved in the January effort.

“I feel like outreach was rather upset that they were having to do this again,” said Ben Talik, program manager with the Pittsburgh Office of Community Health and Safety’s Reaching Out On The Streets (ROOTS) program. He has assisted the county in conducting PIT counts since 2010.

While Mr. Talik said the March count led to improved coordination among outreach staff, he questions how much actually changed.

“It was like, let’s interact with each other so we know we're not counting the same person three or four times,” he said. “I think that was beneficial. But I also don't think we really changed the methods really much at all from the first time around. It did kind of feel like a waste.”

Advertise with us

He wonders if expanded outreach efforts had something to do with the results. 

In the last year, he said ROOTS has significantly expanded its team, allowing them to assign staff to specific neighborhoods and reach more unhoused communities. By the time the count took place, the team was familiar with more encampments, making it easier to identify people during the survey than in previous years.

“I don’t think this increase in numbers necessarily correlates to homelessness itself increasing,” Mr. Talik said. “I just think we got a little bit more of an accurate count.”

With two separate counts completed, DHS plans to invite a small group of Homeless Advisory Board members, staff and research experts to review the findings. The department said it’s not hiding any information, with plans to publish findings from both surveys.

“None of this will be hidden,” Ms. Dalton said. “We’ll do our best to explain it all.”