It’s just after 5AM on Thursday, January 22. It’s dark outside but streetlights light the way. I’m at the headquarters of a nonprofit called the East Oakland Collective, at the corner of Macarthur Boulevard and 78th Avenue, with roughly 100 people. They’re eating breakfast burritos, gathering supplies, going over forms and policies, and preparing to head out. Starting at 6AM, they’ll hit the streets to count homeless people.
It’s all part of a nationwide initiative called the Point-In-Time (PIT) count, a census of homelessness to guide how federal funds are dispersed among regional Continuums of Care (CoCs)—which are coordinating entities between local governments and organizations that provide housing and services to homeless individuals and families. Oakland’s CoC, which includes the rest of Alameda County, is called EveryOne Home. Overseen by the U.S. Department of Housing and Urban Development (HUD), the count is mandated for COCs to receive federal homelessness funding. It must be conducted on a single morning—between 5AM and 12PM—every other year, sometime in the last 10 days of January.
Here in East Oakland, the local count has brought out a relatively diverse crowd, at least in terms of age. There’s a lot of volunteers in their 20s or 30s, but also plenty of older folks, including some who appear to be in their 60s and 70s. About half the folks appear to be White, a disproportionately high number compared to Oakland’s 30 percent White population, according to the most recent census. But there’s also folks who appear to be Black, Latinx, and Asian.
People tell me they’re tired, but they also seem excited.
“I’m tired, but it seems like a good thing to do,” said Pamela, a White woman who appeared to be in her 70s. “I know they need an accurate count, so I hope I can help.”
When asked if she thought the count could be accurate, Pamela was skeptical.
“There’s so many nooks and crannies where homeless people live,” she said, “I think it’s difficult to get an accurate count.”
To conduct the count, staff and volunteers disperse into smaller assigned areas to record what they find, including individuals who appear homeless, tents, structures, RVs, or vehicles that appear to be used as shelter. They also gather more information from homeless people who are willing to take a survey, such as their race, gender, disability status, and how long they’ve been homeless.
But people involved with organizing the count, and those who have studied it, concede it isn’t all that accurate. A 2017 report from the National Homelessness Law Center, said the PIT count “systematically undercounts unsheltered adults and youth.” In 2024, former HUD Secretary Marcia Fudge acknowledged that it’s “just a sampling,” and “not an exact science.”
From the large showing at East Oakland Collective at such an early hour—one of several hubs where volunteers gathered that morning—it’s pretty clear many folks in Oakland think the count is important and are willing to work to support it. Oakland spokesperson Jean Walsh told me that, overall, more that 560 people signed up to help across the city.
According to city spokespeople in neighboring cities, 2026 PIT volunteer numbers were drastically lower. In Fremont, Hayward, and Alameda, 110, 62, and 30 people participated in the January 22 count, respectively.
Even when you account for city population and size, Oakland folks showed up for the PIT count far more than most surrounding cities. In Oakland, over 10 people per square mile of the city helped out. Alameda had just under three people per square mile, while Hayward and Fremont had about one and a half people covering each square mile. Berkeley had the second most volunteers, about 85 according to city spokesperson Seung Lee, or about eight people per square mile.
When more people participate in the count, it’s logical that they’ll find more people. Not having enough volunteers is one factor the National Homelessness Law Center’s report cited in determining the PIT count is an undercount.
Counting through the Pitfalls
At 6AM, I hit the road with a group of four people to start counting in their assigned area, roughly one square mile near the Coliseum BART Station. David Modersbach, a worker with Alameda County Healthcare for the Homeless, was the de facto leader. He’d done the count twice before, and with over 30 years of experience conducting outreach in unhoused communities, Modersbach knew a number of people we encountered on the streets.
It was pretty clear from the get-go that this was not to be an exact science. Sunrise was over an hour away, and in the sleepy pre-dawn we saw a man who appeared to be Black, walking on the shoulder of a street and carting multiple bikes. Given that there was no safe place to pull over, we did not stop to talk with him.
Modersbach said that the man appeared homeless—adding that if he were a White guy walking with a briefcase he would have felt differently—so he thought he should be counted. But according to PIT count policies, tallying him wasn’t required and came down to a judgement call. It’s easy to imagine how another group might have chosen not to count him, or simply missed him in the dark just after 6AM.
Beyond the simple tally of the man’s unhoused status, the group was unable to include any additional information. The PIT count theoretically measures race, gender, disability status, and a host of other factors, but it mandates people be surveyed in order to determine their identities. So the group couldn’t record that he was Black and male—even though he appeared as such—or other data markers such as how long he’s lived unsheltered, what type of shelter he uses on the streets, or how long he’s lived in Alameda County.
The group was given $10 Safeway gift cards to incentivize participation in the count, and, although a few declined, most folks agreed to take the survey when asked. Unfortunately, the group saw several other people who appeared to be homeless in areas they couldn’t reach, and had no chance to speak with them.
Overall, the majority of what the group recorded were the people they couldn’t even see. Waking up people is forbidden, so gentle knocks on RV and vehicle doors, or calls of “outreach” outside of tents and structures had to suffice. When they received no response, the group just recorded tents, structures, and vehicles instead of actual people.
John Janosko, who was homeless in Oakland for over a decade before getting into housing about two years ago, said he only remembered being counted in one PIT count.
“I had heard about it a couple of times,” Janosko said. “But most of the time I was out and about when they came by. It wasn’t like we rushed back to an encampment to get counted.”
Janosko said he thinks the count needs to happen several times a year, at different times of day, in order to be more effective. He also thinks homeless people should be hired to help with the count, as they have access and knowledge others don’t.
As the group drove through their assigned area, I asked Fiona, one of the PIT count volunteers, why they decided to participate. They told me their own experience with homelessness in Los Angeles made them want to make homeless folks’ experiences more visible.
“I’ve lived here for more than a decade and I wanted to take note of folks who are going through this,” Fiona said. “When I was unhoused I felt very invisible.”
Fiona also said that they were inspired to join out of a desire to help bring more funding for homeless services into Oakland, which many local governments advertise in order to court potential volunteers. There’s not, however, a direct formula for how much each jurisdiction gets awarded per homeless person.
According to Emma Welty, who works for the Homeless Action Center in Berkeley as an Outreach Administrative Assistant, HUD allocates funding by theatrically assessing two sets of data: The CoC’s bi-annual PIT count alongside its “housing inventory count,” a record of shelter spaces and permanent supportive housing units, as well as its performance in filling those units. But Welty also said it’s not always so simple.
“Of course there’s always political factors and economic interests that are leveraging for support in a given jurisdiction,” Welty said.
HUD’s funding streams often shift between election cycles. Since late last year, the Trump administration has attempted to drastically reduce the amount of federal funds areas can allocate to permanent supportive housing. Attorney generals across 20 states have been challenging this mandate in court, but it’s still unclear how local governments will—or won’t—be able to use federal homelessness funding going forward.
An Opportunity for Connection
As the sun began to rise, the group I joined had increasingly friendly interactions with unhoused residents. About 10 homeless people agreed to take the provided survey. Volunteers said it was a bit rigid and bureaucratic, but homeless folks still answered the questions with a frank openness. On two different occasions, people answered “Where did you sleep last night?” with a stark truth: “I didn’t sleep.” When asked about employment, one person said he did sex work, but there was no place to mark his answer on the survey. Luckily, all of those surveyed spoke English, as no language interpretation was provided.
At the very least, the January 22 PIT count served as an excuse for Oakland residents to connect with each other, many of whom may have never connected otherwise. Fiona smiled as she surveyed Derrick Soo, a longtime resident of the 77th Ave. Rangers community, an encampment in the group’s assigned area. I’ve known Soo for several years through my work as a journalist, and he’s always been talkative and happy to share stories. He’s a leader within the camp and has even run for mayor.
Soo told me he’s been leveraging connections he has with a local Safeway to create a grassroots foodbank at the encampment. It’s been serving not just residents of the camp, but also housed people throughout the area.
Hidden and Tucked Away
One demographic group—children—was entirely absent from the volunteers’ count that morning, even though the Oakland Unified School District counted over 2,350 students as having experienced homelessness during the 2024–25 school year, roughly 95 of whom were living on the street, and 195 of whom were living in temporary shelters.
There’s likely a few reasons for this. One is that school districts measure homelessness in a totally different manner than the PIT count. Families who “double-up”—by living with friends or family without stable housing of their own—or live in hotels, are counted as homeless by school districts in California. That’s not the case with the PIT count.
Then, there’s weather. The federal government’s decision to conduct the count in January relies on the assumption that unhoused people are more likely to use homeless shelters to get out of the cold, and therefore will be easier to count. But shelters are often full, not to mention dangerous or unappealing to some, and those who have the ability to temporarily double-up with a friend or relative might choose relative privacy—which is especially important for children—over a temporary shelter bed. The National Homelessness Law Center’s 2017 report calls out both of these decisions as factors in why PIT count data is inaccurate.
It’s also possible that children living unhoused are just hard to find. They could be in school when the count is taken, or are purposely hidden from view for safety.
Janosko noted that he lived among several children in encampments, but they were often kept out of view by their parents, especially when people associated with the government would come by.
“The kids are always hidden and tucked away,” Janosko said. “If you see two kids at an encampment there’s probably 10.”
Less Visible and Easily Missed
As we drove through the coverage area, Modersbach decided to do his own personal tally, one that’s totally separate from the PIT count. He counted large concrete blocks that had been dumped on along the curbs of public roadways, eliminating space where vehicles and RVs could have parked. He said each block represented a displaced homeless person, and he counted 376 of them total.
Anecdotally, I’ve noticed strategically placed objects like blocks, logs, and planters on the streets of Oakland for at least eight years, seemingly to prevent homeless people from living there. It’s not always clear who dumps these objects, but the city never seems to move them.
Concrete blocks weren’t the only evidence of displacement I saw that morning. In the 77th Ave. Rangers community, I saw bright pink notices informing over a dozen residents of the encampment’s eviction in early February.
According to data analysis recently published in Street Spirit, encampment closures in Oakland have surged since April 2025. Since then, the city’s scheduled interventions started covering larger areas—including adjacent blocks where people were not encamped—and were posted for longer periods of time.
Around 10AM, after making its way through the assigned survey area, the group decided to head back to the East Oakland Collective and turn in their findings. But just before we left, I saw a structure with a tarp hanging over it, tucked away in the distance, up on a hill next to the freeway. I pointed it out to the group and they recorded it—an example of a dwelling, far off the beaten path, that easily could have been missed.
It made me wonder how many folks the group didn’t count, simply because we didn’t see them.
Once we made it back, East Oakland Collective’s Executive Director Candice Elder talked to me about the PIT count, emphasizing how difficult it was to conduct, and how she suspected it was inaccurate.
“The PIT count is flawed, but it’s the only count the government really goes off of,” Elder said. “The time is an issue. In the morning, people may or may not be there. Also, over the last few years, it’s been harder to find people because the city has increasingly removed encampments, which displaces people.”
Elder said that East Oakland Collective regularly distributes food to homeless people in the area, but she emphasized how difficult it’s been to reach people as the city continues to evict larger encampments, which causes individuals to spread out in smaller numbers to less-visible places. She also said large objects dumped in the road, such as concrete blocks, limit where people can go.
“If the numbers end up low for this count,” Elder said. “I don’t think that means people went into housing or into shelter. I just think it means we couldn’t find them.”
Zack Haber is a journalist and writer who lives in Oakland.