While it makes for good headlines and viral sound bites, the notion that there are “more than 40,000 vacant homes in San Francisco” is not only misleading but represents a fundamental misunderstanding, if not misrepresentation, of the underlying data and market at hand.

Yes, 40,458 units of housing in San Francisco were classified as “vacant” in the U.S. Census Bureau’s American Community Survey (ACS) for 2019. But the devil is in the details, methodology and footnotes.

First of all, an occupied unit is actually classified as “vacant” for the purposes of the ACS if the person living in the unit is planning to move within two months of being surveyed. On top of that, a unit that has been already been rented to a new tenant, but the new tenant hasn’t yet completed their move, is also considered “vacant.” And along with all the units that are actively being offered for rent or for sale, over a quarter of the “vacant” units in San Francisco are simply markers of normal turnover.

Second homes, including pieds-à-terre maintained by people who work in the city but live elsewhere, are also classified as “vacant,” even if they’re frequently occupied during the week or on weekends. And 21 percent of those “40,000 vacant units in San Francisco” were seasonal, recreational, or occasional use (i.e., second/third/fourth) homes.

Twenty (20) percent of the “40,000 vacant units in San Francisco” are units that hadn’t yet been reoccupied since trading ownership, but that included units that were in the process of being renovated, either to flip, rent or be occupied by their new owners.

The remaining 32 percent of the “vacant” units in San Francisco included those that were occupied but used for corporate housing or short-term rentals; units that were in the process of undergoing needed repairs; units that were being prepared for a sale or change in ownership; and a host of “other” reasons as well.

All told, 10.0 percent of San Francisco’s housing units were classified as “vacant” per the ACS in 2019. And while that’s “up from 8.5 percent in 2015,” having dropped from 10.8 percent in 2010, it was right in line with a median vacancy rate of 9.7 percent for nine of San Francisco’s peer counties, per the City’s recent Residential Vacancy Report that was produced for Supervisor Preston.

Keep in mind that the average ACS vacancy rate in San Francisco was 9.95 percent from 2005 to 2010 as well, as broken down by a plugged-in reader nearly a decade ago.

41 thoughts on “There Are Not 40,000 Vacant Homes in San Francisco”
  1. Dean Preston is the very worst socialist on the market. Anyone who takes this fool seriously is a fool themselves.

    1. Dean Preston – one of the worst. I wouldn’t say socialist. That gives socialists a bad name. Opportunist. Doesn’t let truth get in the way of what he wants. Don’t get started on this one. He and Shamann. Surprised he hasn’t hopped on this one either. Likely just waiting to see which way the wind is blowing.

  2. This is great information, and it’s the first time I’ve seen anybody really dive into this data. For more than twenty years I’ve been hearing about all of the thousands of rental units held off the market for various reasons, and I never completely believed it — while there are undoubtedly some of them, the economic value to be gained by utilizing those assets in some way made it hard to believe that the scope of the issue was as large as described.

    Now we see that, indeed, our vacancy rate is completely unremarkable. But I don’t imagine this will cause certain factions to cease remarking on it anyway.

  3. Well done.

    Side note, how many of these technically vacant units are occupied by renters who own a property, or properties, and even other tenant statuses?

    I think many of us know people who fall into these categories. I’m not judging but the Dean Prestons of the world could begin looking at the “gal owns a house in Sonoma where the couple spends most of their time,” not to mention the various familial dodges herein demographics before thinking about some bogus vacancy tax.

    This is as bogus as Mandelman and his crap “I’m not gonna allow any more monster houses in my district,” thing. Never mind that he never made it easier to build multiple units instead.

    Bunch of fakes. Only in it for the sound bites.

  4. How does this percentage compare to Oakland? They collected $7m last year in vacancy taxes.

    Also surprising to think that businesses (ie corporate housing) would respond to the American Community Survey, I suspect the “other reasons” group is a lot more slushy than legit.

    I don’t have a dog in this fight as I haven’t been anointed to reside in the 7×7 (but I do subsidize your laughable bus service with bridge tolls) so I can’t scoff at Dean Prester’s latest junket. I do suspect the readers here are more connected to slushy “Other Vacant” than the hoi polloi

  5. It’s still worth considering that 8,000 homes are only used occasionally, and some portion of 12,000 are unoccupied investment shells. We need far more than 20,000 additional units, but 20,000 is still far more than zero.

    1. Keep in mind that the number of “second home” units, the use of which isn’t actually quantified in the ACS (in terms of occasionally versus frequently) has only increased by 750 since 2010, despite the fact that an additional 45,157 units of housing have since been built, and the number of “other vacant” units has actually dropped by 5,542 or 29.9 percent.

      1. Well, to be fair, no where does it say who produced/published it. Someone mentions “Dean Preston”. Where did this come from? The US Census Bureau?

  6. Great post highlighting SF’s normal vacancy rate. So really it about 2% of the 400k housing stock.

    How about multi-unit bldgs (2-3units) where it’s occupied by the whole extended family, where no rental agreement is in place or money changes hands – because they are family. I don’t think those units should count as vacant either, or end up on some future city registry and be taxed. They’re are tons of immigrant families in Sf (with all kinds of legal status), and not everyone can file for the home ownership exemption, I suppose this would apply the US citizens too. (Adding a family member to the deed to avoid a vacancy tax, would trigger another type of tax when the unit isn’t really vacant).

    If anything it’s Sups like Dean w their policies that take housing off the market in the first place. I’d like to see him add an ADU and subject himself to the rules he’s forced on others.

    I know one family that had 2 1-bedrooms, (aside from their marriage was just better with their own space), with kids they were hardly taking up room that someone else might be in.

    1. This is an excellent point that I think is going unnoticed. According to the 2018 Housing Needs and Trends report, most “vacant” housing is older housing (built pre-1939), not new construction. It’s anecdotal, sure, but we all know Chinese-American families that own multi-unit buildings or whose entire family occupies a 3+ unit property. Are these units considered technically vacant? Also, how many “vacant” units are immediately inhabitable given the number in century old buildings? If the vacancy fines disproportionately target Chinese-Americans and AAPI, it could be problematic; a scenario needs to be examined prior towards implementation.

      Also, there was a report online that the Oakland Vacancy Tax revenue dropped from 7mm to 4mm the second year, but I can’t confirm this. But I believe other studies have shown income from the tax drops fairly precipitously once some units come back on the market (again, read it online, so take what I write with a grain of salt w/o the data to back it up). Still, it stands to reason that it’s essentially a one shot legislation and not an endless let a lone consistent stream of revenue. The Chron reports that the tax may only affect up to 5k units and I would imagine a large number of those will be exempt for the reasons mentioned above. It will be interesting to see how it plays out. But as with anything in SF, setting up a Rental Registry and having a fully staffed city agency poring over I guess PUC records etc. to sleuth out vacancies is going to be costly and time consuming not to mention the inevitable push back once the tax notifications start going out.

      1. Keep in mind that the “up to 5k units” that could be taxed, as reported by the Chron and parroted elsewhere, was based on the flawed vacancy report and calculus.

        1. Any idea what the actual number could be? It’s hard to trust all the vague numbers floating around online from competing sources. It’s frustrating, because without actual facts not to mention a comparatively accurate revenue projection, people will be voting on egregious misinformation.

          Maybe you’ve already covered this elsewhere….

          1. Will the Planning and Building departments have to pay the vacant unit tax on a property if the years long vacancy is because it is awaiting permits that are in the city backlog?

  7. 45,000+ new units in SF in just the last decade. Everything apparently is measured here in 40’s of thousands. Sheesh. That’s revealing as well because it belies the fundamental fraud of the “housing crisis” bs.

    The only crisis is when Scott Weiner misses his build build build quota and his financial backers put the squeeze on. No wonder this city has gotten so crowded.

    1. finally some common sense around here. i hate reading these pro-developer blogs, they’re so full of it. you’re telling me you think Preston is full of it because we have more vacation homes than homeless people? give me a break

  8. great… then why don’t we do an actual survey of landlords to see the real vacancy rate. Are your stats true? Let’s find out. Full disclosure – declaration – by landlords under penalty of perjury – Bet we’d see some different numbers then.

    1. Well your wish is coming true – beginning in 2023 all landlords are required to report everything about their units to the newly established “Rental Registry”. Failure to comply results in no business registration or the ability to implement the meager rent increase allowed for that year. Happy now?

    2. Apartment registration will commence in SF in 2023, data will be collected by the Rent Board. This will finally help clear the air on the actual rate of vacancy and what the real rate of tenancy is in the city.

  9. The details regarding the supposed 40K vacant units are revealing. BTW, did Supervisor Preston vote to kill the 400 plus unit development on Mission near the 5M tower? Is he part of the BOS majority that tries to slow/stymie new housing development in SF? I suspect so and this makes his proposal little more than grandstanding – trying to appear as if he is serious about addressing the housing shortage in SF.

    The “occasional” use category is common in my area. I’m aware of 7 homes minutes from my place that fall into this category. A longtime neighbor retired as an SF employee a year ago. Very nice pension. To avoid state taxes, he retired to the Nevada side of Tahoe. He keeps his SF home as the capital gains/transfer tax would be huge. It’s cheaper to leave it empty for 20 years than to sell. He occupies the home 5/6 weeks/year.

    SF has a medium-term mandate to add 80K housing units. Preston’s effort does nothing to address that. Unlike the SF PTB, their SSF counterparts are fast tracking new housing. A 200-unit mid-rise complex is nearing completion at El Camino and Chestnut. Just the other day I noticed steel girders going up directly across the street which I assume will be another 200-unit complex. These two mid-rise structures are just blocks from the old one story detached SFH neighborhood that makes up most of Chestnut and Grand in SSF. SF would never allow this kind of medium density housing development near its old SFH neighborhoods. A tale of two cities.

    1. > It’s cheaper to leave it empty than to sell

      Are you stating that the amount of long-term capital gains tax paid on a $1M+ transaction is greater than the property tax rate which is locked in by prop 13? That ignores the massive windfall from the sale of the house. My heart breaks for this Nevada boomer with an excellent pension and a vacant house in SF, this tax would surely put him out on the street.

      Also if there’s steel girders in SSF, it’s for biotech office space, not housing. Everything going up in SSF is 6 story stick frames.

      1. Anyone who has driven past Chesnut at El Camino in the past year plus will have seen the steel girders going up. The building now has its facade completed. It is a 172-unit apartment complex called Summerhill. Directly across Chesnut steel girders are just now going up. I’m guessing that is likely another apartment complex. It’s a perfect location for new housing as less than 10 minutes south on El Camino (across from See’s) a massive office complex is bring fast tracked by SSF. Not to mention Tanforan which is being developed as an office complex and could add several million feet of new office space to the rapidly developing El Camino corridor.

        1. The building (at Chestnut and Arroyo) with the “steel girders”, that both you and “Plugger” are referring to, is the new “Community Civic Campus” project for South San Francisco.

          It’ll contain a library, community theater, parks & recreation facility, etc.

          Its going up right next to the new police headquarters that have been recently completed.

  10. SPUR looked at the number of vacant units in the city around 1996, and then again in 2014, each time for slightly different purposes than the current study. But the results are very similar all three times…around 10% and nothing remarkably different from other cities.

  11. Altho the point is made well enough, I suppose, it seems a little not apples to apples to compare San Francisco County to Los Angeles (or Cook)County the former being , essentially, (only) the core of the Metro area while the latter is(are) almost the entirety of it; assuming, of course that the rates tail off as one moves into the fringes.

  12. Spreading misinformation and lies are meant to serve only one purpose. Kickback for the kick back kings. $ talks democrats or republicans, progressive or communist. It’s all about the $. The dumber the voting class the better it is.

  13. It would still be far easier and cheaper to free up 8K+ units than to build them from scratch given material costs.

    It is still good policy to force speculative units back onto market regardless if 8K or 40K or somewhere in-between which is most likely fact.

  14. Let’s tax second, third or fourth homes. If what an unsigned socket site post claims is true, then the real estate boosters should have no problem supporting the vacancy tax that will have no effect whatsoever.

  15. dean preston blocks all new housing, no matter how big or small, no matter the location and no matter the % of affordable units. he is the worst

  16. I’m surprised they didn’t included situations like “single resident, currently at work”, “owners went to the store” and “occupied but not, like… ‘living'” into their vacant metrics.

    1. We should have an end of year accounting within a week or two, but at the end of the third quarter there were over 18,000 units for which building permits have already been issued, approved or requested but have yet to break ground, along with another 33,000 units which have been approved by Planning but for which a permit application has yet to be filed.

  17. If you live on the East side of town, the original report to the BOS is correct. Crap, there are still units built when Willie Brown was mayor that are empty. People demand more housing (affordable), and we get luxury condos. We need to do what Vancouver did. +5% property tax on foreign owned residential property, +12% property tax on vacant housing units. That will do (2) things.. bring in about 900mil annually in taxes, lower rental rates, and correct overpriced real estate. Then maybe we can fill some potholes. How about this – Housing for people who LIVE in San Francisco. Then maybe we can get some local business’ back.

    1. “If you live on the East side of town, the original report to the BOS is correct.”

      Unfortunately, that’s incorrect. But the newly built units on the East side of town, which are closer to offices, Caltrain and SFO, are more likely to be pieds-a-terre than homes on the west side of town. There’s also a higher percentage of non-rent controlled rentals on the East side of town, the normal turnover of which are misclassified as “vacancies” in the ACS-based analysis, as well.

Leave a Reply

Your email address will not be published. Required fields are marked *