In June of this year The Business Times covered a story about a Chicago tech worker moving to San Francisco who lost out on an apartment due to a competitor who was willing to pay a full year’s rent upfront in cash. The market for renters has gone bonkers again following a massive wave of capital and high-earning workers flooding the city and colliding with a frozen housing supply.
After a few quiet post-pandemic years, the rental market is back in hyperdrive. Long lines at unit viewings and bidding wars over rent are actively happening across the city. This renewed demand has reignited familiar arguments over the reality of housing inventory in the city. Do we have a true shortage of housing or would there be plenty of apartments to match demand if we simply bring existing vacant units to the market?
In San Francisco numbers are more than just data. They are weapons to tell entirely conflicting stories about why housing is so expensive. A notable statistic comes from the Census Bureau’s 2021 survey, which stated that San Francisco had over 61,000 empty apartments. It is still being quoted in 2026 to make the claim that there is no housing shortage in San Francisco at all. But market conditions have since changed and the reality behind this figure is much more complicated. Search for these empty ghost apartments and you will find projects in the mid-remodel stage, protracted family estate feuds, and tech bros holding a condo for occasional use.
The history of the Empty Homes Tax
The progressive/tenant-action camp argues that housing supply exists and is being hoarded. They reject the housing shortage narrative that calls for accelerated building. They argue that building market-rate housing increases gentrification and fuels corporate profits. This group includes people like former District 5 Supervisor Dean Preston who was the main proponent of the Empty Homes Tax designed to tax owners of 3 or more units left empty for more than 182 days per year. The Empty Homes Tax was approved by voters in 2022.
Legal Update: Although Prop M (Empty Homes Tax) was slated to collect its first payments in early 2025, a Superior Court judge struck down the tax as unconstitutional in late 2024 (ruling it violated the Takings Clause and the state's Ellis Act). The city appealed, and the tax remains completely frozen in the Court of Appeal as of 2026.
The pre-pandemic inventory figure determined for the Empty Homes Tax is from the 2019 ACS Survey by the U.S. Census Bureau, which placed the figure at 40,548 vacancies, with a margin of error of ± 4,400 using federal sampling methods. The 2021 ACS counted 61,473 vacancies, an increase driven almost entirely by units sitting listed for rent as renters left the city during the pandemic. By 2024 the count fell back to roughly 48,000.
The 2021 vacancy number has continued to spread and even increase, despite the fact that the rental market has changed dramatically since then. In 2026 Abe Woodliff (who runs Bay Area Memes) said on a podcast that San Francisco's housing crisis "is fake," citing 63,000 empty units. Then on another show, he says there were "damn near 70,000" and said that "people are just hoarding that shit." While you may wonder why anyone should care what meme accounts say, there is a familiar name behind the messaging. While San Francisco is done with Dean Preston as a supervisor, he apparently isn’t done with us. He has since built a media network called PUML (Punch Up Media Lab) to work with these influencers to change the narrative on housing in San Francisco. The SF Gazetteer has quoted Woodliff: “People with money spend a significant amount of money on propaganda. We’re basically using their tactics of sensationalism and propaganda to get out a message that actually affects people positively.”
The problem with the ACS figure
The ACS figure has been used to make the claim that San Francisco doesn’t have a housing shortage. It was the primary catalyst used to justify putting the Empty Homes Tax on the ballot. It is a tally of every single housing type in San Francisco and that includes single-family homes. But the inclusion of single-family homes is highly misleading when used in the context of the vacancy tax (Proposition M).
Once you strip out those single-family homes and duplexes from the initial figure you are left with roughly 30,000 units. That figure is further reduced by filtering out those affected by the 182-day rule and legal exemptions which include a long list of exclusions where the tax is legally waived. In 2022, the city’s controller’s office calculated that only about 4,000 units would ever actually trigger the Empty Homes tax criteria in a given year.
When proponents of the tax say there are over 60,000 units we can unlock they are being disingenuous because the law as written specifically exempts duplexes and single family homes from the tax. They are also still using the 2021 number from a time period when the city emptied out, even though a new ACS figure was published in 2024, and the rental market has only gotten tighter as of 2026.
The ACS classifies a unit as vacant if it is listed for rent and waiting for a tenant, if it has been rented but the tenant has not yet moved in, if it has been sold but the buyer has not yet arrived, if it is a second home used occasionally, or if it is a new building whose exterior is finished but whose interior is not. As noted by The Frisc, it is a snapshot of units that are unoccupied, not a count of units that are available.
The “apartment warehousing” argument
A common argument of the progressives is that corporate developers and housing investors are intentionally leaving a significant portion of their inventory open to artificially choke supply and manipulate the market. Corporate landlords once had algorithmic revenue management software that could instruct property managers to accept a higher vacancy rate rather than lower rental rates for an entire building, thereby resetting the market rate for everyone. But San Francisco banned that software in October 2024, and California banned it statewide in January 2026. The rental market has only gotten tighter since.
The city's own data never supported the broader claim anyway. The Controller's Office found that among units vacant longer than six months, the only ones the tax could reach, just 11% were listed for rent, compared to 53% of short-term vacancies. The apartments sitting empty the longest are overwhelmingly the ones not on the market at all: second homes, estates in probate, units sold but not yet moved into, buildings mid-renovation. The Controller found that the opportunity cost of lost rent far exceeds the tax.
The 2026 reality check
For 2026, we can look at the Yardi Matrix Multifamily Market Report for institutional-grade research on vacancy and occupancy rates in major U.S. cities. The February 2026 report shows a property occupancy rate of 95.8% equating to a tight 4.2% vacancy rate. In July 2026, rental platform Apartment List reported just 2.2% of multifamily units on the platform were vacant, making it the tightest market of any U.S. city tracked. Both figures tell us that the units a renter can actually move into are scarce: a vacancy rate under 5% represents an incredibly tight market where demand outstrips supply. Clearly the problem is not improving for those who wish to rent an apartment in the city.
So the reality of the housing inventory issue is that there are far fewer units available to prospective renters to meet the current demand. Once you strip away probate court deadlocks, second homes, units mid-renovation, and active tenant turnarounds, the city’s true vacancy rate is razor thin. And the market continues to stay hot with return-to-office mandates and an influx of workers due to the local AI boom. San Francisco’s housing crunch is driven by a true supply deficit, leaving prospective tenants to fight over a thin pool of available options. It’s important to look deeper than the numbers stated by policy makers to be able to make informed opinions on housing policy, and to vote accordingly when these issues come to the ballot.