Last Thursday night, January 26, from the passenger seat of a friend's Prius, I examined a paper map of Russian Hill and Fisherman's Wharf, giving sometimes confusing and wrong directions to her while also tallying, with pencil and paper on a clipboard, anyone who appeared to be homeless in the 15-square block radius we drove, as spotted and called out by two more volunteers in the back seat who had eyes on the street.

That's one lesser known aspect to San Francisco's homeless census, the Homeless Point-In-Time count conducted every two years on the same January night in every US city that receives federal homeless assistance funding: A significant portion of it is conducted by car.

I don't mean to draw attention to this as criticism, necessarily, and in fact it's a relief, as the section of the city to which we were assigned is one of its steepest, which is saying something. We even drove the twisty section of Lombard Street — not exactly known as a haven of tent encampments. Walking a section even a quarter the size of the one were assigned would have required oxygen tanks. I also suspect it wouldn't be terribly worthwhile. In the somewhat affluent corner of San Francisco that we scoured, we tallied fewer than 10 people we deemed to be homeless.

The Point-In-Time Count, conducted last week by a record high number of volunteers, is actually a patchwork of different counts. From St. Ignatius College Preparatory school in the Outer Sunset, which served as the staging area, car units fanned out across the city from 8 p.m. to midnight to roam large swathes of hilly and residential neighborhoods. In the Tenderloin and Mission Districts, meanwhile, groups on foot tallied the homeless people they saw in smaller, more precise areas. A designated unit conducts the census in BART stations, where many seek shelter for the evening, and elsewhere, in areas like Golden Gate Park and Ocean Beach, part of the Golden Gate National Recreation Area, park officials conduct the count themselves. As one training official explained to us, nobody from our group would be walking around on the rocks at the beach that night.

Still, for three small park spaces included in our section — Russian Hill Park and Open Space, George Sterling Park, and the lawn area of San Francisco Maritime National Historic Park — we found parking and continued the count on foot. It's obviously easier to be precise this way, but precision is far from guaranteed. That's in part because we were instructed not to speak with any homeless people during our training session. Volunteers are told this would take too much time and interfere with the count.

Nonetheless, volunteers are asked to note the observed ages — under 18, 19-24, or over 25 — and genders — male, female, transgender — of the homeless individuals they observe. "Unknown" is an option here, and we used it for several people like one wrapped in a sleeping bag. The point: Without engaging with people, much of this is a guessing game.

Similarly, volunteers are told to mark down tents or RVs that seem to serve as shelter for homeless people even if it's not possible to look inside. We were told not to shine flashlights on tents or knock on doors, but instructed that if we could peer into shelters and count people without causing a disturbance, we should do that. Volunteers are also asked to note if homeless people are members of a family.

For us it was easy, which I assume was lucky. We saw two RVs parked with people in them. In one, there was a couple in the front, and written on their vehicle were requests for donations for their disabled son. We guessed they were living out of the RV and marked them as a homeless family of three. In another RV, we found a couple playing cards. We assumed they were on a road trip.

These are assumptions, and as such, could very well be wrong. In recapping her participation in the count last week, Chronicle columnist Caille Millner writes that "the unspoken difficulty with raising a small battalion... of volunteers to try to count all of the homeless people on one night in San Francisco: The city’s government is relying on the judgments of those volunteers." That troubled Millner, and it worried me too. "Most of us aren’t experts. We’re just regular people, bringing our own prejudices about class and race to a crucial task that can determine everything from San Francisco’s level of federal support for homelessness to city leaders’ decisions on solutions," she concluded.

To Millner's concern, I would add that engaging with people to determine if they're homeless isn't always effective either. In one of our park segments, in the dark, two men sat on a bench. Breaking the rules, I asked them if they were spending the night there, told them what I was doing, and emphasized that I wouldn't keep them from doing so or tell anyone where they were. The two men said they weren't sleeping there, and I believed them — they just had a small backpack, probably not enough for the night, I assumed. But would they have told me if they were homeless?

In 2015, the Point-In-Time Count came to 6,686 people, a 3.8 percent increase from two years prior and a number that's been relatively static for a decade. The San Francisco Coalition on Homelessness considers 13,000 a more realistic count, since not all those who are homeless are visible — some aren't tallied, some are marginally or temporarily housed, etc. Another figure sometimes invoked is the 9,975 homeless people who availed themselves of services provided by the city's Department of Public Health between 2014 and 2015.

Jeff Kositsky, who heads the city's Department of Homelessness and Supportive Housing,
was among the group surveying the Tenderloin on Thursday night, the Chronicle reports accompanied by his 14-year-old daughter. In fact, he was one of a number testing out a new Google form to perform the tally. This year, the Chronicle writes, could be the last that paper and pencil are used for the count. Orange County has already made such a switch. But electronic systems won't ensure accuracy, they'll just speed up tabulation, sending along results faster. Those are due later this spring — in 2015, after the last January count, they were available in May.

Critical as so many are of the count, Kositsky tells the Chronicle it's still of value. "It's still important because it provides us an apple-to-apple comparison to other counties and locally over the years," he emphasizes.

Human Services Agency Director Trent Rhorer, an official in charge of services for homeless people who precedes Kositsky, was encouraged by another count: The number of volunteers. A record 768 people showed up for the one-night count. “I think this much turnout, this many people shows the increased interest in the whole subject of homelessness and, really, the compassion of the people in this city,” Rhorer told the paper.

A final note: This year, as it has in so many spheres of American life, the Trump Administration has caused concern regarding potential cuts to the federal funds correlated to the Point-In-Time count. Trump is “going to be a disaster” with regard to homelessness, Paul Boden of the Western Regional Advocacy Project in SF told the Chronicle. “A guy who’s a brain surgeon ... will now be in charge of housing,” Boden said, referencing Ben Carson. “That pretty much tells you everything you need to know. It’s all going to get worse.”

Former SF Mayor Art Agnos agreed with Boden. Carson, Agnos told the Chronicle "is ignorant and uninformed about homeless and housing issues ... and, quite frankly, in the HUD bureaucracy there is no strong inclination toward dealing with homelessness, either. If anything, services will diminish. At best they will stay the same."

Related: First Homeless Navigation Center Abandons Permanent Housing Placement Promise