Having hit a pandemic high of 20.5 percent in the third quarter of last year, the effective office vacancy rate in San Francisco inched down to 19.9 percent at the end of 2021, which represents 17.0 million square feet of vacant office space in the city, including 5.5 million square feet of space which is technically leased but sitting vacant and 11.5 million square feet of un-leased space, according to data from Cushman & Wakefield.

As a point of comparison, the office vacancy rate in San Francisco was 16.7 percent at the end of 2020 and 5.7 percent, with under 5 million square feet of vacant space, at the end of 2019.

Despite the drop in the overall vacancy rate at the end of 2021, the amount of un-leased office space in San Francisco actually ticked up, both in the absolute and relatively, with 1.3 million square feet of space that was being offered for sublet in the third quarter having been leased, reoccupied or returned to the market as directly vacant space. And total leasing activity actually dropped from the third to fourth quarter of last year, with “a scarcity of large transactions,” a push back of return to office dates (yes, the surge in COVID cases is meaningful, beyond increasing hospitalizations and deaths), and under 1 million square feet of space having been leased, including sublets, for a negative net absorption.

And while the average asking rent for office space in San Francisco inched up 0.8 percent from the third quarter ($73.46 per square foot) to the fourth quarter of last year ($74.06), it’s still down nearly 12 percent from a peak of $83.83 in mid-2020 and the increase in the direct vacancy rate is expected to put pressure on landlords to reduce asking rents over the next few quarters if not year(s).

27 thoughts on “Office Vacancy Rate in San Francisco Inches Down, But…”
  1. San Francisco’s core business district will not recover to the density of workers it had prior to Covid. This is an intractable problem. The new work dynamic will impact SF more than many other cities. It already has. The SFBT recently noted that San Francisco has fallen behind other major cities in economic recovery from the pandemic. For instance, in Austin, Texas, office attendance is as high as 60% while San Francisco stands at only about 30%. Airports in Phoenix and Denver had nearly returned to pre-pandemic passenger numbers for domestic travel, while Seattle and Los Angeles had reached about 80%. But San Francisco has peaked at just 60% of pre-pandemic passengers for domestic routes.

    To make the point, AutoDesk signed one of the largest office leases in SF last year. Taking 117,000 feet at 300 Mission. Yesterday it announced it will not occupy the space as the company shifts to a remote work model. At that, AutoDesk planned to fill the space to only 30% capacity.

    The percentage of the workforce allowed to remote work will continue to grow. Office space will be more and more located in the easily accessible spots within metro areas. SF is not that for the Bay Area. Oakland and the East Bay are. Other Bay Area cities are “sheep stealing” SF companies and workers. A massive office development on El Camini in SSF is being fast tracked. It’s partially targeted at SF employers – think Stripe.

    As to rents, the coming big uptick in direct vacancy space will push rents down. Will all the empty space be filled? In 15 or 20 years? For that to happen chunks of the space need to be converted to other uses and workers/square foot will need to fall significantly – which is happening because of Covid and telework. Bottom line, the number of downtown SF office workers will be significantly smaller than pre-pandemic going forward. That has big ramifications for the downtown retail and hospitality sectors.

    1. Agree with some points, disagree with others.

      Yes, SF’s tech DNA means it’s more likely to embrace remote work. It will take longer to recover, consistent with the data you’ve cited. But don’t sleep on the fact that tech leans young, and the young want to be in dense, party-filled cities, as well as go into the office. They want to socialize with co-workers and get free food. People are also waiting longer and longer to have kids, so you may have “young” single people living in SF till 35, 40 or 45 years old. SF also has a robust housing pipeline and infill condo buildings and apartment buildings will continue…housing the young.

      I’m torn on the Oakland thing. Oakland is more convenient for the East Bay population and is an excellent BART nexus, but is less convenient for Marin, SF, Peninsula and most of South Bay. East Bay tends to be slightly older folks with young families…who probably want to work from home anyways. If you dig deeper, I don’t get the Oakland HQ use case. To me, SF is actually the “least bad” HQ option in terms of making everyone happy.

        1. Hmm…there’s no ferry service or bus service from Marin to Oakland, and it’s a backtrack to get to the Richmond bridge for much of Marin. 80 is usually very backed up.

          The only pro I can think of is that Oakland probably has easier / cheaper parking.

          1. From Novato (Marin’s largest city) to SoMA or Mission Bay, arriving at rush hour, Google Maps thinks it is a coin toss between going through SF or through Oakland. Clearly, under those conditions, a destination in Oakland will be faster for those making that trip.

        2. Compared to San Jose… and depends what you mean by Marin. Yes from Oakland you can take 580 and be in San Rafael fairly easily (though still a 30+ minute drive in good traffic)… But if you mean Sausalito, Mt. Tam, or Mill Valley, then it’s not so “close”, and certainly not as compared to anywhere in S.F., including neighborhoods such as Bernal Heights or the Outer Sunset. (e.g., Google right now showing 30 minutes to San Rafael from *downtown* Oakland (not the resi areas mostly further south or in the hills, which would add time) versus 40 minutes from Lake Merced (with a lot of surface streets). Change the destination to Tam Junction, and it’s now faster from Lake Merced than downtown Oakland).

          1. TBH I based this off meeting up at my buddy’s house in Corte Madera for a bbq. When I leave my neighborhood toward the southeast side SF and my friend leaves North Oakland he always beats me by like 15 minutes. It’s happened several times.

        3. I live in marin and would 100% rather commute to the city than Oakland. There are frequent, comfortable public transit options to SF (golden gate transit bus, ferry), and traffic can be avoided on many trips. My social circle is better accessed in SF (meeting a friend for dinner/HH), not to mention the physical safety aspect. SF has plenty of homeless and drug addicted, but I feel much less targeted by certain, full-time criminals than I would be in Oakland (say, walking from the parking garage to my office).

      1. Marin has less than barely a quarter million residents (<4% of the BA)…it's hardly going to be a deciding factor for most firms.

        1. Exactly, Marin and Sonoma strive not to be bedroom communities and fights to add housing at every opportunity. I got part of that, big sprawling single family residential doesn’t go well with wineries or Mill Valley, MT. Tam or along the coast. Where as Oakland whether it be Freeways or BART is hands down a better nexus for East Bay with still plenty of room, space for people either dense infill or single family residential all inter mixed with some great regional open and shoreline space. The one improvement is the East Bay needs a single transit agency covering Contra Costa and Alameda counties to run buses, offer transit services along with embracing some well placed BRT lines outside of BART.

  2. All the while more and more businesses are signing leases in more business-friendly states. Meta just leased up all of the space in a tower not yet complete in Austin.

    1. SF’s anti-business climate is now joined by remote work as a double threat to the City. It’s not just companies going to other states, it’s companies leaving SF for greener BA pastures. Stripe to SSF, PG&E – and a chunk of Twitter – to Oakland.

      15 million plus feet of biotech and office space is under construction or planned in the North Peninsula. At the same time the recently opened 5M tower sits empty. 88 Bluxome’s developer halted that project as it was not viable w/o a major lease commitment which was not going to happen. So, the developer drops plans for an office complex in favor for a biotech project – eliminating the tennis courts from the new proposal. The city looks as if it won’t allow the change meaning the project may be dead.

      TPTB apparently don’t see the economic threat to the city or, if they do, they have no clue. No clue how to correct things. The empty retail space tax will not lead to the filling up of SF’s plethora of vacant retail spaces.

    2. Meta signed the largest office rental deal of 2021 – in the entire US – last month in Sunnyvale (720K SQ FT)

      Meta also leased 520K SQ FT in Burlingame last month

      Apple leased 700K SQ FT in Sunnyvale in 2021

      Apple and Google are going to build more than 10 M SQ FT of office space in San Jose alone.

      While the downtown San Francisco office market will take some time to recover, tech is growing everywhere, including its birthplace.

      1. The Peninsula is on fire with massive new office and biotech developments set for this decade. More than 15 million feet proposed north of PA. As you say, south of PA through SJ millions of additional feet of space is in the works.

        Tech is growing in many US metros. Buildings not completed are being pre-leased. SF is not seeing this boom. SF’s tech boom earlier this century was driven by the SV. It was not organic to SF as can be seen with tech companies abandoning SF at the same time millions of feet of new tech space is in the works in the SV and on the Peninsula.

  3. We have been getting new variants of the virus every couple of months. The scientific community is saying that is going to continue for the foreseeable future. It is likely that some of those variants are going to be more severe than what we have seen to date.

    It’s pretty hard to know how all this unwinds in the commercial real estate market. After a couple more years of high vacancies owners start conversions to residential. My bet is it is going to take SF 5 years to get to that point – but it will…

    1. “It is likely that some of those variants are going to be more severe than what we have seen to date.”

      Why do you say that? That’s not in keeping with the general scientific consensus as to what viruses “want” to do.

      1. My main source is nature.com. . I am not a biologist. :). My understanding of what the scientist say on this is that successful virus’s are those that evolve over the long term by finding proteins with higher rates of transmission and lower fatality rates. The only way to keep the virus alive is to keep the host alive So I think we agree on that?

        But – Covid is new – and we are still in the short term – so they know some ‘unsuccessful’ versions will evolve. Those that kill the host. Those will be self limiting- since that also kill the virus which limits transmission. The unknown is the transmission rates of those varients. The scientists are looking at what we have seen and trying to head that off in the future.

        I think American society will react differently to a version that infects 1% of pop with 50% fatality rate versus a version that infects 50% of pop with a 1% fatality rate. For the first option – employers might want people to work from home more than the second. (Perceived liability)

        1. “But – Covid is new – and we are still in the short term – so they know some ‘unsuccessful’ versions will evolve. Those that kill the host. Those will be self limiting- since that also kill the virus which limits transmission. The unknown is the transmission rates of those varients.”

          I don’t know about that. Covid is “new,” yeah. It’s also at a similar stage compared to the Spanish flu as when that virus went endemic. Your example of a 1% population with 50% fatality would have seen a full on militarized response. Indeed we saw far more organized responses to virus outbreaks under better organized federal administrations, as well. But I don’t know that we haven’t already seen the “unsuccessful” outbreaks of this virus. I think back to the first wave and the scenes in Italy and New York with triage units in parking lots and whatnot. And I’ve also read that dominant forms of viruses subsume the less “successful” variants. Given that this is already worldwide where is the isolated population which a less successful virus infects with exponential fatality? That said if what you’ve read about proves to be an eventuality then hopefully that smaller outbreak can be quickly isolated.

          1. I was speaking more broadly of the potential for some catastrophic pathogen, not a recurrence of Bubonic Plague, specifically.

          2. Actually I think we – or maybe it was really you and ‘Pablito’, and I cut it…my apologies – were discussing this: “It is likely that some of those variants are going to be more severe than what we have seen to date.” to which you claimed there was a general scientific consensus twasn’t so…perhaps you can give us a clip of (at least) two scientists at a bar agreeing.

          3. Really? you’re challenging my statement that there is general scientific consensus as to what viruses “want” to do, which is widely spread without killing host organisms, thereby killing the viruses themselves? Google it. You’ll see article after article showing established evolutionary epidemiology fact.

          4. No – nor am I questioning the implication that viruses are sentient (which I’m sure you realize since you put “want” thusly) – I’m questioning that it’s impossible that a worse – however you wish to define that term – variant will develop. I seem to recall there was initially a lot of concern that this variant might be one, and while it now seems that that didn’t happen, there’s hardly a consensus that it couldn’t happen.

          5. OK I understand you but none of the language above is in the absolute, and is rather all measured with could, would, if then, eventuality, etc.

          6. “and is rather all measured with could, would, if then, eventuality, etc.”

            But then you are basically not saying anything of any significance. Anything would, could, if then…

            Obviously the virus either ends up being “unsuccessful” and essentially renders itself extinct by either by being too fatal too quickly or by generating a broad and durable enough immune response that it no longer has hosts to infect, or it becomes “successful” by being mild and transmissible enough yet not generating a long/broad immune response. Or we develop a vaccine with a durable broad spectrum of effectiveness and vaccinate enough people that we force the virus to go extinct.
            The real question is what are the probabilities of all these possibilities? And I don’t think anyone really knows at this point. Maybe Omnicon is the variant that signifies the transition to an endemic state. Or maybe there are a few ups and downs yet to come.

            I do think that it’s widely accepted that while any mutation *could* happen, recombination of existing genes is much more likely. This current Deltacron may or may not be real vs a lab error, but in general once mutations for something like increase transmissibility, antibody escape, severity,… have already occurred my understanding is that it’s much more likely that they could combine in different combinations. Some combinations could be more “successful” for the virus than others, but we don’t know how long the path to the eventual end game will be.

Leave a Reply

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