How big is California’s homelessness crisis? Inside the massive, statewide effort to find out
by By Marisa Kendall, CalMatters
cover photo: Deidra Perry, program financial manager for Alameda County Healthcare for the Homeless, takes part in Alameda County’s 2024 point-in-time count in Berkeley, Jan. 25, 2024. The PIT count, which included a voluntary survey, gathers data on the county’s homeless population. Photo by Loren Elliott for CalMatters
Thousands of volunteers fanned out across California this week, peering down alleyways, into parked cars and along creek beds in a mass effort to count the state’s homeless population.
The federally mandated census, done every two years and dubbed the point-in-time count, serves as the main framework Californians use to understand their state’s homelessness crisis. The data it produces influence everything from allocations of state funding, to local policy decisions, to the way politicians talk about homelessness in campaign speeches.
The U.S. Department of Housing and Urban Development, which requires the counts every other year, compiles the data from across the country into an annual report submitted to Congress. Last year, the department tallied 181,399 unhoused Californians — 28% of the nation’s total homeless population. That’s up nearly 40% from five years ago.
But the counts, despite being mandated by federal law and serving as the basis for any number of decisions on California housing and homelessness policy, rely on unpaid volunteers and are far from an exact science. Different counties in California tally their numbers differently. Some attempt to talk to each person they count, while others use algorithms to estimate how many people live inside each tent and RV. Experts agree the real number is higher than the count.
“It is not a finite count,” said Tamera Kohler, CEO of the Regional Task Force on Homelessness in San Diego County. “It is at a minimum, you know you have this many people. …You’re going to miss some folks. You’re just not going to be able to get in every little area in that period of time.”
Several counties tested out new methods this year in an attempt to improve accuracy. In the last count, Alameda County volunteers were instructed to tally tents and RVs without bothering the people inside. This year, volunteers instead tried to talk to everyone inside those tents and RVs, and get them to answer a 15-minute survey.
Surveying the homeless
The sun hadn’t yet risen Thursday morning when Andrea Zeppa, homeless services regional coordinator at Alameda County Healthcare for the Homeless, led her team down a row of RVs parked along the side of a street in Berkeley. Each volunteer was equipped with a map speckled with red dots that signified known encampments. Bundled up against the early morning chill, they picked around piles of trash as a fluffy cat with a bell on its collar darted back and forth between the vehicles.
“Hello, hello,” Zeppa called softly. “I’m here for the homeless count. Anybody awake?” A generator powering one of the RVs droned in the background.
Across the street, 61-year-old volunteer Deidra Perry met a young woman in a long peasant skirt. She was living in a silver SUV adorned with a “Zombie outbreak response team” sticker in the rear window. As Perry asked her a series of increasingly personal questions about everything from her mental health to her HIV status, the woman revealed a glimpse into her life, and how she ended up on the street. She survived domestic violence in 2012, she said, and her post-traumatic stress disorder makes it impossible for her to sleep in a crowded homeless shelter. She sleeps on the streets when she isn’t working as a live-in nanny or house-sitting. At one point she’d had a tent, but it was stolen before she even had a chance to set it up.
As the sky began to lighten and birds started chirping, Perry typed the woman’s answers into an app on her phone. At the end of the interview, Perry handed her a $10 Safeway gift card. The woman thanked her, and Perry moved on.
“They’re tough questions to ask,” Perry said afterward. It was her first time participating in a point-in-time count. “It feels very intrusive.”
Most of the vehicles they approached remained dark and silent. Starting the count before dawn helps ensure people aren’t moving around town and at risk of being counted twice, but it means many people are still sleeping.
Even so, by mid-morning Zeppa said the surveys were going “really well.” The people who were awake were receptive to sharing their stories — which Zeppa hopes leads to more accurate information about the region’s homeless population, and helps agencies like hers plan better programming and spend their money more effectively.
After the last census, there were concerns that Black and indigenous unhoused people were undercounted in Alameda County, said Sharon Cornu, executive director of St. Mary’s Center, which provides housing and support services for homeless seniors in Oakland.
“We are hopeful this year that the count will provide us better data that helps us design and run a stronger system to bring people home,” Cornu said.
St. Mary’s Center uses the point-in-time count data in its applications for funding, Cornu said. The data also helps them learn about the people falling through the cracks in the area’s net of social services.
“It really busts a lot of the myths,” Cornu said. For example, the 2022 count showed that 65% of people surveyed had lived in Alameda County for 10 or more years, compared to just 8% who had lived there a year or less — contrary to speculation that unhoused people are traveling there from out of the area to take advantage of social services.
But lingering impacts from the COVID pandemic have complicated efforts to paint a statewide picture of the homelessness crisis. Counts that were supposed to happen in 2021 were delayed until the following year over fears that they could put unhoused people and volunteers at risk of catching the virus.
The counts went ahead in 2022. But then things got messy. Some counties, such as Santa Clara, opted to count again in 2023 — reasoning counts in the past had happened in odd-numbered years. Others, including the rest of the Bay Area and Sacramento, chose to count this year instead.
As a result, parts of the state are out of sync. Even the federal government doesn’t know which counties are counting when.
“It is not easy to predict which (continuums of care) will conduct a count every year,” Andra Higgs and Andrew Ten said in an emailed statement on behalf of the housing department. “COVID-19 complicated the regularity of counts. Most of Southern California has historically counted every single year. However, the Inland Empire upwards to Northern California follow less-clear patterns.”
In Los Angeles County, about 6,000 volunteers spent three nights counting.
The “coolest feature” of this year’s count was new geo-fencing technology on volunteers’ smartphones that alerted them if they strayed outside of their assigned area, said Ahmad Chapman, communications director for the Los Angeles Homeless Services Authority. It also alerted count organizers if streets were skipped, so they could go back and count them later.
Volunteers in Los Angeles County don’t try to survey every unhoused person they meet. Instead, they tally the number of people, tents and vehicles they see. That data then goes to researchers at USC, who use it to estimate how many people are living in each tent and vehicle.
The result isn’t a precise count, said Ben Henwood, project lead for the homeless count at USC. But the estimate can pinpoint trends, such as places where the population is going up or down.
And the count has other benefits, Henwood said.
“This is a mass civic engagement project that gets people out of their house and involved in thinking about homelessness,” he said, “which serves its own purposes.