According to city data that tracks homeless individuals who seek healthcare and other services, the number of people experiencing homelessness in San Francisco jumped 30 percent between 2018 and 2019 — and the total number tracked was more than twice the number counted in the January homeless census.

The census, known as the "point-in-time" count, is mandated by the federal Department of Housing and Urban Development (HUD) and such counts happen in cities across the country around the same time, every other year. As we learned when the final numbers were released in July, based on this point-in-time method, the city saw a 17 percent uptick in the number of homeless, up from 6,858 in January 2017 to to 8,011 in January 2019.

The rationale for the methodology of counting individuals on one given night, using volunteers who purportedly canvas every corner of the city and make a visual count, is to create an apples-to-apples comparison and identify trends between each count.

But critics have long held that the point-in-time method leads to chronic under-counting — and it didn't help that this year's count turned out not to be an apples-to-apples comparison at all. Days after that 8,011 number came out, "homeless czar" Jeff Kositsky admitted that there was a change in the counting protocol this year compared to 2017 in that his department stopped counting homeless people in rehab facilities, hospitals, and jail — instead relegating those numbers to an appendix. The switch was made because HUD doesn't define those in hospitals or jails as homeless, though SF had been counting them along with other state and regional entities. If you counted those individuals, the number of homeless was 1,800 people higher, or 9,784.

Now, as the New York Times reports, the number is nearly double that — and more than double the point-in-time count — if you use the city's database on people seeking services. That number for the city fiscal year that ended in June was 17,595 people. And while some people might seek services more than once in a year, they are only counted once if that is the case.

Some will argue that this number isn't any better or more "accurate" than the point-in-time count because it includes transient people who may have stopped through SF in the course of the fiscal year, used city services, and moved on somewhere else. I would argue it probably is closer to a "real" number given the fact that a) it doesn't include the many people who did not need services or who are "service resistant," as social workers call them; and b) it does include more individuals who, for whatever reason, were not visible to volunteers on the night of the point-in-time count — like they were staying in an SRO for a week, crashing with a friend, or were otherwise "housed" outside the shelter system that one chilly night in January.

One clue that points to the new number's relevance: It marks a 30-percent uptick in the number of homeless seeking services over the previous fiscal year. That is the same percent rise shown between the 2017 and 2019 point-in-time counts if doing the apple-to-apples comparison with those jailed, in rehab, and hospitalized (6,858 vs. 9,784).

Rachael Kagan, spokesperson for the city’s Department of Public Health, tells the Times that one reason for the rise in the department's count was that they undertook a "blitz" in the last fiscal year in which staffers proactively sought out homeless people in shelters and hospitals, entering 1,272 new people into the city's databases. Kagan still says that the 17,595 figure is the "most complete picture that we have" and is likely an under-count in itself.

Previously: SF Homeless Number Nearly 1,800 People Higher If Same Criteria Used As 2017