Vacant Nuance in the Vacant Housing Debate
We should want more/less vacancies depending on the kind of vacancies we're talking about but nobody likes to specify
Much of the California housing discourse has been embroiled about vacant housing for the latter half of the 2010s. The vacancy debate appears numerous times on Twitter, presumably whenever the beams cross and Housing Twitter and (non-housing) Left Twitter intersect. One of the easiest ways to get over a thousand likes is to simply post there are 18 million vacant homes in the United States, but only 500,000 homeless people. It’s also a very easy way to start a housing war in your comments section with people pointing out those vacancies are largely in the middle of the country or in depopulating areas.
It seems to be a very common refrain on the Left and I understand why. It invokes similar levels of anti-Capitalist sentiment I felt working in corporate retail. Dumping beverages down the drain or tossing perfectly good food in the trash that was a day over expiration, while the homeless were starving just a few feet away. It really wakes you up to the inequity of distribution and that’s how leftists view everything.
Proponents of building more housing don’t really have a snappy response. Some people will retort by saying the statistic is wrong but it’s not, it is completely correct. It’s not hard to pull up vacancy counts from the Census website and homeless statistics from any jurisdiction. Fact is, there are more vacant homes than counted homeless people in virtually every city, every county, every state and yes, in every nation.
Some YIMBYs will state we actually need more vacancies, and I agree with this but the problem is that it lacks nuance on what kind of vacancies we want more of. Some leftists hear this and think: “Why should a home even be vacant at all?” We tend not to disaggregate what types of vacancies there are, or which ones they have problems with in the housing discourse.
So we’re going to do that here.
The first problem with the vacancy statistic is that invokes imagery of empty houses and apartments waiting for use, years on end, or an old house that needs fixing. But not all vacancies are created equal. Vacancies in the Census break down into five categories:
On the Market
Leased, but Not Occupied
Vacation or Seasonal Vacancies
“On The Market” means properties for rent or sale currently being looked at on like Craigslist or Zillow. “Leased, but not Occupied” means there’s someone on the lease but nobody in the unit. The most common example would be the interval of time between the securing of one’s lease and the actual arrival at one’s home. “Vacation and Seasonal Vacancies” are consistently vacant at a certain time. So if there’s housing for students, we’d expect it to be vacant when school’s out. Or a summer home or resort home that a wealthy person or tourist lives in only during the summer. “Migrant housing” is housing for migrants but there are none in or near major cities. Lastly, “Other” means virtually anything else that doesn’t qualify. Dilapidated housing, housing under renovation, housing under construction, and housing just sitting empty for whatever reason. My neighbor has an Other vacancy because her family moved into a house with a garage ADU and never rented the ADU as the previous owner did.
So let’s use San Francisco as an example. 2020 Census data with this level of specificity is not yet available, but we can use the Census’s American Community Survey (5-year) for a clearer, pre-pandemic picture.
Market: 8,040 (combined rental + owner MOE: +/- 1,294)
Leased: 7,414 (combined rental + owner MOE: +/- 1,240)
Seasonal/Vacation: 8,316 (MOE: +/- 992)
Other: 11,688 (MOE: +/- 901)
Well that seems pretty cut and dry: there’s 8,035 homeless in San Francisco, more than enough Other vacancies alone including maybe some fat in remainder categories. But a more accurate homeless count conducted using a variety of sources than just a nighttime survey suggests 17,595 homeless in SF (2019)1. That’s still a bad statistic though, 2 vacancies for 1 homeless person overall, but we can see how the standard Point-In-Time counts underestimated those in need. Additionally, Other vacancies are quite a few thousand less than the amount of homeless so decrepit or abandoned housing isn’t sufficient by itself.
Another problem is that many units reported vacant are certainly occupied by now, and some occupied units are certainly vacant by now. Remember, the vacancy count is just a snapshot at the time of the survey. Americans move on average 11 times in their life2. As you read this, there’s probably a departing occupant closing the door permanently on a now vacant unit. By the time you finished this sentence, there’s probably a new occupant somewhere else walking into their now no-longer-vacant home for the very first time.
The Census provides the duration of vacancies in their American Housing Survey3, although only at the metro level, so in San Francisco’s case this would include the metropolitan area. I have no reason to think the duration wouldn’t be fairly consistent for SF unless it were a college town or some such thing.
60% of SF metro area vacancies were vacant for 0 to just under 6 months. These are the characteristics of just plain old transitional vacancies from one occupant to another, with the higher end duration usually housing under repair inbetween occupants. (Vancouver’s vacancy tax triggers at 6 months for this reason.)
40% are literally everything else. Vacant 6, 7, 8 months, 1 year, 2 years, forever, or never occupied (meaning just finished construction), or just unidentified. Vacant for any reason, including occupant turnover repairs stretching out longer than a few months.
7% of all vacancies are vacant for 6 months to a year
12% of all vacancies are vacant for 2 years onwards
If we were to apply this 40% to the SF vacancy count, you’re at 14,000 long-term, abnormal or unidentified vacant homes. If we were to subtract out the 7%, then it’s about 11,700. That’s a couple thousand homes short of the homeless count. Assuming an average household size of 2.3, you’d need just over 7,000 homes for the homeless however. But that’s a very generous—perhaps flawed—household size assumption because 70% of the homeless are men4 and it’s not a given that they would choose to live with other homeless rather than family members or by themselves. So the vacant housing needed is somewhere in-between 7,000 to 17,000 depending on people’s households.
It’s not the eye popping ratios people often deploy, you’re just barely at enough homes or barely short if the usable vacant count is 11,700. That number of possibly usable vacant homes includes a lot of ineligible properties undergoing long term rehabilitation or under construction so that it’s likely an overestimate as well.
With the super long term vacancies comprising 12% of all vacancies, assuming a household size of 2.3 and that all the homeless would live with each other, you could house 2/3rds of the homeless, assuming those units could be identified and aren’t just some old basement units in someone’s house. Short answer: we don’t know but it’s not these super gluts constantly portrayed. To be clear though, that’s not to say it’s not worth focusing on these vacancies. I’ll recommend how at the end.
Another problem is that vacancies are almost always stated in total numbers rather than the percentage, so we don’t know what it’s relative to. The home owning vacancy rate is at its lowest in recently recorded post-war history5 and the last time the national vacancy rate for rentals hovered around 6.2% was 35 years ago6.
Now what’s the ideal vacancy rate? Who knows, economists aren’t even certain but they’ll generally argue it should be 7-8% for rentals, and 2% for owners. In California, our rental vacancy rate is 4% 7, and our owner is 0.7%8. In other words, you’re going to pay a lot in rent here, and you can forget about buying a home here. The recently released 2020 Census count says that California’s overall vacancy rate is the lowest vacancy rate of all 50 states.
If you think the vacancy rate should be 0%, the only country I’m aware that’s “accomplished” that, supposedly, is the cities of the Soviet Union around mid-20th century9. But in the U.S.S.R., you moved on permission of the government, people got rejected a lot, household crowding was common as were housing black markets. I also found that vacancy statistics from the Soviet states were not very reliable. This wasn’t a point of pride for Soviet leadership, but a call to build more housing than any nation on earth in the span in 20 years.
Another problem with having very few vacant homes on the market is that you give monopoly power to property owners over entire neighborhoods. Landlords now force tenants to compete for their vacant unit, when under housing abundance, tenants would force landlords to compete for their occupancy. This leads to humiliating nonsense, like being forced to put in dumb things about yourself that doesn’t reflect on your ability to pay in your apartment application. Those with criminal records, however small, and those with low incomes, or non-white skin, can be weeded out much easier despite fair housing laws. This is why some twitter yimbys say that more vacancies equals more tenant and buyer control. But what they mean is available vacancies—not derelict properties or randomly vacant homes.
You know who also loves low vacancy rates? Private equity speculators. Of them all, Blackstone is the most famous, and their subsidiary, Invitation Homes, is the largest owner of single-family houses in the United States. In their SEC filings10, they explain clearly to the government and investors their profit model:
We have selected markets that we believe will experience strong population, household formation and employment growth and exhibit constrained levels of new home construction. As a result, we believe our markets have and will continue to outperform the broader U.S. housing and rental market in rent growth and home price appreciation. As measured by the September 2016 Case Shiller Index, home price appreciation in our markets was 6.2% for the twelve months ended September 30, 2016, compared to growth in the broader U.S. market over the same period of 5.5%. We believe home price appreciation is a leading indicator of future rental growth. Within our markets, we have focused on highly desirable in-fill locations with multiple demand drivers, such as proximity to major employment centers, attractive schools and transportation corridors.
And they even explained how they lose money.
We could also be adversely affected by overbuilding or high vacancy rates of homes in our markets, which could result in an excess supply of homes and reduce occupancy and rental rates. Continuing development of apartment buildings and condominium units in many of our markets will increase the supply of housing and exacerbate competition for residents.
This is all public, I’m not making it up, check out the footnote at the bottom. All the other real estate speculators and corporate landlords speak the very same way. You don’t have to believe in supply and demand but real estate speculators do. If I’m a pension fund like Black Rock, I’m buying housing in areas with population growth and low vacancies per Blackstone’s strategy, because that’s guaranteed dividends as I’ve monopolized the neighborhood.
It’s like the movie Goldfinger. Instead of being a gold hoarder nuking Fort Knox, it’s speculators buying houses and relying on NIMBYism and construction costs to nuke housing for more people.
Some will suggest that the solution to tamper these speculator’s exploitation is rent control, and I agree it’s part of the solution. But in the absence of enough vacant housing at best we’d become Stockholm where the waitlist for rent-controlled apartments stretches 9 to 20 years long.11 Although I have heard those wait times have declined thanks to new construction. You can take down that “Immigrants are Welcome Here” sign on your front lawn if the only way to get housing in your city is to be born into it or wait 1/4th your life. Look at how housing shortages in California hurt Afghan refugee settlement right now.
The point is: mobility matters. Some leftists who are very vocal on the vacancy problem seem to not even realize the importance of mobility. People move for a whole host of reasons and they search for vacant housing. The lower your vacancy rate, the lower your mobility is. This was recently confirmed in a study showing the decline of people being able to move within their own regions, as was common, with housing constrained Los Angeles showing double the mobility decline of the national average12.
The problem with the “vacant divided by homeless” equations as an excuse to not build more housing is that it narrows down demand for housing to those without a house, excluding millions in precarious housing situations. While the homeless should be prioritized for shelter, hundreds of thousands of people every second are looking for housing who are not technically homeless, but are often on the brink. Think about it:
List of People Searching For Housing At Any Moment:
People who get a job in another city
People fleeing spousal or parental abuse (high cause of homelessness for women13)
People who are immigrants
People whose houses must be repaired
Kids moving out of their parents’ house when they come of age
People who move away to pursue an education
People who find a partner and need a bigger bedroom or two bedrooms
Couples expecting children and need a bedroom for their kid
Parents whose kids grow up, move away and need to downsize
Seniors suffering from arthritis and need an elevator
Seniors whose partner has passed away and needs to downsize
People who need to live close by their relatives
People who live in multi-generational households with their relatives
People whose houses have been destroyed in a disaster
People who are fleeing marginalization or violence from their community
People who got allergies
People who don’t like their housemates or don’t want housemates
People who want to have pets
People who get evicted
People who just want to move. For any reason.
You can take your pick about which ones are deserving and which ones aren’t but you can’t stop any of them from moving except maybe where they’ll end up. When your vacancy rate is low and housing isn’t available, then people:
Live in overcrowded conditions
Live in structurally unsafe, physically abusive, or mentally unsustainable households
Have to leave their communities for better housing elsewhere (i.e. gentrification)
You can put all the homeless in vacant units and you can make the vacancy rate ~0%, but there will be more homeless the day after. Homelessness is not static but dynamic. There’s constantly new people, all of us, at risk of becoming homeless. Fires in California destroying communities. People sleeping under laundromats in San Francisco. 12 people crammed into 3 bedroom homes in Silicon Valley. 10% of students at Cal State and UC Berkeley experiencing homelessness at any given time. Those have to be factored into the denominator too when you do the division of vacant homes.
You know who understands the importance of mobility the most? The homeless. There was a proposal at Berkeley City Council one night to evict a community of homeless folks living in RVs. A Black man spoke to the city council, saying that he used to live in a house outside the city, but came to Berkeley and could only afford an RV. His intent in coming here was to enroll his child in the city’s world class K-12 schools, thus affording his child a better chance at life—even if that meant becoming homeless.
This is also why many neighborhood community groups like those in San Francisco’s Mission District always emphasize that new affordable housing contain multiple bedrooms for families. Because they know that a primary cause of Latinos leaving SF for bigger houses in Oakland is that new families need space, and a 1 or 2 bedroom apartment, even with stable rents, doesn’t work.
I don’t believe that people who say we’ve built enough housing actually want these things. I just think they haven’t considered this. Any city that’s truly equitable ought to have four types of mobility:
Those who want to stay in their city can stay
Those who want to move in can move in
Those who need different housing and living accommodations can get them
All within a timely manner
California is really none of these unless you’re rich. Los Angeles leads the nation in overcrowded households (which fueled coronavirus spread14), as does the state of California15, which also ranks 49th out of 50th in homes per person. All these things are very closely related to coming in dead last in 2020 vacancy rates.
But fellow YIMBYs should also understand that there’s likely two problems here that do not contradict each other. The vacancy rate is historically low in the United States but there are still properties that shouldn’t be vacant. If we can speak in favor of a new development providing more housing, why not speak in favor of making a derelict apartment provide more housing too?
Reducing those longterm vacancies is supply-based housing policy. Policies such as ending the mortgage interest deduction on second homes would remove a lot of vacant vacation houses. Proposals like vacancy taxes can push landlords to rent out their properties faster to acquire income as quickly as possible. Project Roomkey has proven highly effective at housing the homeless, and cities should invest in acquiring existing properties like motels for homeless housing.
I’m in favor of pulling out eminent domain and acquiring vacant housing complexes that have sat empty for years. Scenes like these are more common than they should be, and urbanists should get real that people won’t ignore them just because some economist’s statistic says they’re few in numbers. Thanks to capped property taxes in California, property owners can sit on dead houses without a need for income, land bank for decades, and are immune to property tax increases the increased land values bring. All while people sleep under freeways.
In a housing shortage, we should be exhausting all policies to ensure that as much housing can be made available as possible. However, we should make no mistake that vacancy taxes like Vancouver’s, while successful at pulling vacancies down by a huge magnitude16, have not shown really any impact on housing prices17. The underlying problem is the total amount of housing, but it’s an incremental step that helps someone get housing, and to me that’s what matters.
There are plenty leftists who understand the complexities of vacancies though and this article is not about them. Even Alliance of Californians for Community Empowerment acknowledges in their report on speculative and wasteful vacancies that overall they’re rather complex, and places like the Bay Area in aggregate have “extremely low vacancy rates.”18 These are nuanced takes that are completely non-contradictory while still criticizing speculative and negligent vacancies.
There are many cities around the world that have made huge strides in affordability, household stability, reduced household crowding, and increased mobility. None of them did it by acquiring or reducing vacant houses as a primary affordability strategy. Vacant housing is nuanced but it’s not hard.
Thanks for reading my substack. Subscribe for some exclusives coming soon!
Morton, H. (1984). Housing in the Soviet Union. Proceedings of the Academy of Political Science,35(3), 69-80. doi:10.2307/1174118