It’s a flawed process, but it’s all they have
Counting the homeless: A simple goal, an impossible task
By now, Liz daCosta knew what to look for. She walked into a downtown Phoenix McDonald’s and scanned the early-morning crowd, searching for dirty jeans, broken shoes and stale expressions — the telltale signs of a night spent on the street.
“Are you experiencing homelessness?” she asked the first man she found, sitting by himself at a table for two. DaCosta, who oversees housing programs for the non-profit service provider Community Bridges, reached out one hand to introduce herself. In the other, she held a cellphone opened to a blank survey, ready to record the man’s answer.
“Yes,” Samuel Thompson said. He shook her hand and gulped his cold coffee, thick with 19 plastic containers’ worth of creamer.
DaCosta sat across from him and glanced at her phone, reading off the next question. “Where did you sleep last night?” she asked.
Thompson bobbed his head toward a wide glass window. “Right there,” he said. Outside, volunteers like daCosta stood on the sidewalk where he said he’d slept, asking the same set of questions of everybody they could find.
They had come to start Maricopa County’s annual Point-In-Time Homeless Count, a one-day effort to spot, survey and tally every person experiencing homelessness. The count, which happens nationwide each January, is federally mandated and often cited as the most effective way to measure homelessness across the country. It reveals trends and defines the demographics of the homeless population. The results are used to allocate funding, shape policy and craft the narrative of American poverty.
One problem: It’s usually wrong.
“There are so many problems with the PIT count,” said Maria Foscarinis, executive director of the National Law Center on Homelessness and Poverty, which serves as the legal arm of the campaign to end homelessness. “The numbers aren’t accurate. They paint a misleading picture.”
Last year, Maricopa County’s PIT count found 6,298 people experiencing homelessness.
Reactions to that figure vary. Those in the know, such as service providers and local housing officials, understand the PIT count is flawed. They say it’s best used to analyze trends, rather than to serve as an actual hard tally of people experiencing homelessness. But that nuance doesn’t always translate to politicians or the public.
“This drives the overall response to homelessness as a national problem,” Foscarinis said. “If the count tells us that the numbers are small and they’re going down, that makes it much less urgent and much less critical to respond to homelessness as a national crisis.”
It also makes it difficult to ensure that every dollar of federal funding goes where it’s most needed.
‘It’s just a snapshot’
The flaws are baked into the PIT count’s design. The process varies slightly across the country, but each count follows the same general plan: Volunteers fan across cities and countrysides, looking for people who might be homeless. If they find somebody, they conduct on-the-spot interviews, asking detailed questions about their life and their living situation: Where did you sleep the night of Jan. 22? How long have you been homeless? Have you ever been diagnosed with a serious mental illness?
In Maricopa County, most volunteers use an app on their cellphone, which marks the location of each survey. In most other parts of the state, they fill out paper forms.
There are never enough volunteers to cover every inch of American soil. Those who volunteer will rarely find every single person in their assigned area. Sometimes, they make judgments — Is that
person homeless? — that prove incorrect. Families and children, who often sleep in cars or with friends, are frequently missed. And people are under no obligation to answer the survey’s questions, or to be truthful when they do.
Arizona is particularly challenging. PIT veterans know that people sleep in the forests and just off hiking trails, but it would be impossible to search all that terrain. Cold-weather towns such as Prescott, where the homeless population often flees harsh winters and returns for mild summers, are always undercounted. Maricopa County, which holds more than half of the state’s population, is too big to cover. Instead, volunteers sample certain high-homelessness areas and a statistician extrapolates the countywide number.
“It’s just a snapshot,” said Anne Scott, human services planner for the Maricopa Association of Governments, which coordinates the county’s process. “It’s not a complete data set of people experiencing homelessness all year.”
Still, she couldn’t think of a better way to do it.
Scott and daCosta’s team of volunteers were responsible for searching most of downtown Phoenix, between Seventh Street and Seventh Avenue, Buckeye Road to Van Buren Street. Time was critical. When the sun rose, people rose with it, disappearing into offices and day centers. There, they were even harder to find.
But daCosta took her time with Samuel Thompson. He crossed his legs, sipped his coffee and answered every question. Yes, he said, he’d used drugs. Yes, a doctor once told him he was mentally ill. He’d been released from prison about a month earlier. He was white. Not Hispanic or Latino.
“I’ve been homeless my whole life,” he said.
DaCosta tapped his answers into her phone, glancing up every few seconds to let him know she was still listening. When they finished, she walked outside and came back with two pairs of tube socks and a bottle of water. She thanked him for his time, told him to check out the Human Services Campus a few blocks west and moved on to the next table.
She asked a woman in the back of the room if she’d like to participate in a survey. The woman said no. So did the man sitting at the next table, and the woman next to him. DaCosta sensed that all three were homeless — she hated to stereotype, but after a few PIT counts, she could just tell — but there was nothing she could do. She scanned the store one more time, then headed for the door.
But a woman sitting by the door caught her eye. DaCosta stopped, looked her up and down. There was an overstuffed suitcase at her feet, but her shoes were clean and her clothes seemed neat. DaCosta debated whether to ask. Was she homeless? Should she be counted?
Then the woman looked up, and saw a stranger staring her way.
“Hi,” she said, looking confused. “Hi,” daCosta replied. She decided the woman probably wasn’t homeless, and walked outside.
A few minutes later, she watched through the window as another volunteer approached the same woman and asked if she was experiencing homelessness.
She said yes.
A telltale sign, but not an answer
There was no right way to search the streets, no directions to ensure they checked in every corner. “There is no real process for us in doing our grid,” daCosta said, so she and Scott parked near First Avenue and started walking. They checked dark overpasses and narrow alleys, bus stops and the light rail station. They found a few people — an Air Force veteran, a man on his way to jail, another with kind eyes and cataracts — but mostly, they walked.
By 8 a.m., their group of four volunteers had seen 37 people.
“Wait,” daCosta said, stopping near First Avenue and Shannon Street. She pointed into a parking lot where a single car sat running. A woman was changing shirts in the front seat. “Looks like they might be sleeping in their car right here.”
Scott stayed on the sidewalk as daCosta walked to the car, her phone already open to the survey. She stopped a few feet away and looked inside. There was a boy in the back seat. Pillows were piled all around.
The woman rolled down her window. “Did you sleep in your car last night?” daCosta asked.
“No,” the woman said. She stammered that she had just left work and that they stopped to rest for a few minutes. When daCosta offered her phone number, she declined.
DaCosta waved goodbye and walked away, wondering what else she could do. She knew that PIT counts missed most families, and when they found one, parents were often reluctant to admit they were homeless. They worried the state would take their kids.
The pillows were a sign. But the woman said they were just resting.
“We have to go on what people say,” Scott said when daCosta returned. “So if she said she’s not homeless …”
DaCosta turned and gave the car one last look. “Yeah,” she said, though she sounded uncertain. She locked the phone and put it in her back pocket.
In Prescott, a test of patience
Early the next morning, Tim Laskowski took a box of paper surveys from his back seat, walked past a “No Camping” sign and started up Prescott’s Rodeo Grounds Trail, looking for people he’d never find.
“If you see somebody who looks homeless, let me know,” he told Danielle Plaxton, his search partner and colleague at the non-profit U.S. VETS. But after an hour of searching through Prescott’s northwest side, they hadn’t completed a single survey.
In rural areas such as Prescott, the PIT count is an exercise in patience. There are fewer volunteers searching for fewer people. It’s usually cold. Volunteers often end up hiking trails or wandering forests, hoping to stumble across some sign of life.
Last year, Prescott’s count turned up only 34 unsheltered people. But local service providers know the actual number is far greater — especially in the summer. The city’s largest shelter, operated by the Coalition for Compassion and Justice, sees more than that every day. That disparity, combined with Prescott’s already-low totals, makes each completed survey feel like a boost to the community’s future funding.
To make sure it counts as many people as possible, Yavapai County stretches its search to five consecutive mornings, but still asks people where they slept on the night of Jan. 22. If they can find them.
“Hey, there’s a jacket over here,” Plaxton yelled from a hidden stretch of the trail. She stretched to her toes, trying to see inside a crevice in the stones. “Hello?” she asked. No response.
Laskowski followed her voice. She pointed into the rocks, where a gray hoodie rustled in the breeze.
“That’s definitely a sign,” Laskowski said. He made a mental note to check there again later.
He remembered his first PIT count, four years ago, unsure if he was comfortable searching for people who didn’t want to be found. By now he’d grown used to the idea, but the task was no easier. He assumed they found only a third of the city’s people experiencing homelessness. But a full count was impossible. There was too much ground to cover.
“I can’t find anything up here,” he said.