What’s the most you would pay for nothing at all?
For the city of San Francisco, that price has gone as high as seven figures over four years. In fact, for the city to meet its new housing goals, paying millions of dollars a year for effectively nothing might have to become standard practice.
Part of SF’s state-mandated plan to build 82,000 new homes in the next eight years calls for land banking — essentially, the city buying empty lots to keep them off the market while lining up future affordable housing at the sites.
Land banking proponents say that without it, the biggest chunk of the new homes — 46,000 below the market rate — will never happen. But as of now, the city has secured almost no properties for future development, and the city’s housing agency doesn’t seem terribly enthusiastic about pursuing more.
There is no list of potential sites to acquire, at least not that city housing officials want to make public. But SF has secured one critical site: the former McDonald’s lot in the Upper Haight, at the corner of Haight and Stanyan streets. Pressed on why SF isn’t buying more, the Mayor’s Office of Housing and Community Development (MOHCD) points out that even passively owning land costs money as we wait for development to start.
The city has paid more than $1 million since 2018 just to hold the old McDonald’s lot, and while it never pays to be cavalier about public money, these costs are marginal compared with the value of the planned development (current estimate: $166 million). What’s holding San Francisco back from buying into its own future?
A critical tool
“Is San Francisco land banking?” Sup. Dean Preston demanded to know at a January hearing about the city’s new housing plan, known as the Housing Element, which calls for the city to identify sites and funding for land banking.
The answer he got from MOHCD director Eric Shaw: The plan for acquisitions is still in development.
“Does MOHCD have a list of sites?” Preston asked. It’s still being compiled, Shaw replied, adding that the department is “resource constrained” and “focused on production, not buy and hold.”
“Buy and hold” — in other words, land banking — is exactly what city planners singled out in the Housing Element as a critical tool for creating 46,000-plus affordable units over the next eight years.
The term land banking originated in the 1960s, as local governments either bought or seized vacant land and derelict homes with plans to turn them over to private developers for future construction.
The concept really caught on after the mortgage crisis and Great Recession left vast swaths of abandoned properties blighting neighborhoods. Eighty percent of U.S. land banks have sprung up since 2008, according to the Bay Area urban planning think tank SPUR, which held a 2020 forum on the subject.
Waiting for market-based relief, as a SPUR presentation noted, isn’t an option: “Whether because of lack of demand or lack of supply, a housing market will not meet human needs without public sector interventions.”
SF’s idea of land banking is different from most places. “If you look at the states that have land-bank legislation, they rarely occur in stable or hot markets,” Frank Alexander, Emory University law professor and the country’s foremost land banking evangelist, tells The Frisc. “The focus for the past 30 years has been properties where the owners walked away.”
In a hot market, private investors snatch up and rehab those lots themselves, hence the phenomenon of the million-dollar San Francisco teardown. This makes an SF land bank a more complicated endeavor. The city wouldn’t seize properties, so it might have to compete with other bidders, adding even more expense.
The city will be getting pressure from both sides: mow or don’t mow, but a neglected lot often costs more to clean up than a well-maintained one.
frank alexander, emory university
Affordable housing boosters have pushed for SF to bank available land for years, and with the new Housing Element, SF Planning agrees.
So why isn’t SF doing it? “There’s a huge carrying cost when we buy and hold,” Shaw said at last month’s hearing. Even empty lots cost the city money, and the longer development takes, the more it costs.
Or, as Alexander says, “doing nothing is not a costless enterprise.”
You deserve a break
If you’ve ever been to the Haight, you probably remember the 3,500 square-foot McDonald’s that opened in 1975, which thanks to its strategic location across from Golden Gate Park was always a magnet for activity.
But the fast-food depot consistently served up some less-than-happy meals, with frequent calls to the cops to break up fights, drug dealing, and other alleged crimes on the premises. In 2015, the city attorney sent a complaint to the McDonald’s corporation saying the police received 1,100 calls about this location in just three years.
In 2017, the city bought the lot at 730 Stanyan for $15.5 million, demolishing the golden arches and pledging to build 100 percent affordable housing there. Other than a temporary “safe sleeping” homeless tent site for a year of the pandemic, the lot has sat empty.
When The Frisc asked MOHCD why we’re not banking more land, spokesperson Anne Stanley cited 730 Stanyan as a cautionary tale: “Since 2018, MOHCD has spent $1.24 million on holding costs for 730 Stanyan,” Stanley said via email. And that might be a bargain: Stanley estimates similar sized lots could cost SF up to $500,000 annually, depending on variables.
That’s the bill just to keep the site from deteriorating: putting up fencing, removing trash, covering up graffiti, pulling weeds, exterminating vermin, and generally keeping things presentable. The biggest cost is security; it’s $300,000 per year for round-the-clock surveillance.
“The level of security necessitated to ensure the site is well maintained and to mitigate community impact is based on previous experience with a site,” according to Stanley. In other words, that space was a hot spot for casual criminality for so long that some people would keep using it for such dealings, even if they can no longer get fries with that.
FEMA covered some costs of the tent site. Otherwise, while SF waits for developers to break ground on the 160-unit project in 2024 (as of now), taxpayers are on the hook.
These holding costs will have to be factored into future affordable housing budgets, which most housing-aware folks in SF acknowledge need to rise. If this was just about one parking lot, it might not be such a big deal.
To mow or not to mow
When asked how cities with extensive land-bank holdings keep costs down, Emory’s Alexander notes that while a “city could do nothing” to maintain a site, there’s usually neighborhood pushback. “The city will be getting pressure from both sides: mow or don’t mow,” he says, but a neglected lot often costs more to clean up in the long run than a well-maintained one.
There’s also liability. What if the 730 Stanyan lot became host to homeless encampments, and say, a fire started and caused damage to neighbors (in this case, Amoeba Records on the Haight side and residences on Waller Street)?
Alternatively, SF could seek out developers who are willing to reimburse the city for the costs of holding the land for them. But Whitney Jones, deputy director of operations at Chinatown Community Development Center, one of 730 Stanyan’s two developers, told The Frisc via email that chipping in for these interim costs is “more appropriate” in for-profit developments.
However, CCDC’s Jones agrees that land banking is smart: “Given that the Housing Element creates the expectation that 46,000 below-market-rate units will be created during the next eight years, the city will need opportunity sites.”
A discussion about paying these costs — in the case of 730 Stanyan, $1.24 million and counting — would be less consequential somewhere other than San Francisco, because of two very San Francisco things. First, everything here is more expensive, from pulling weeds on an old McDonald’s parking lot to hiring security guards. And second, everything in SF takes longer to build. If we’re lucky, the desperately needed 100 percent affordable housing at 730 Stanyan, which the local district supervisor himself (yep, Dean Preston) adamantly supports, will have taken nearly a decade to build from the time the city bought the lot.
If development didn’t take so long, holding costs wouldn’t matter as much, and the city could afford to start buying up more land when opportunity knocks.
Part of the problem is that “we’re reliant on state funds” to jump-start these projects, says Preston aide Kyle Smeallie, and they may sit fallow for years while the cash trickles in.
Preston is still feuding with the mayor’s office over Proposition I funds, a higher property transfer tax that Preston authored and voters approved in 2020. Smeallie says those funds should be used to break ground on sites just like 730 Stanyan. Preston also argues that the mayor isn’t fostering more temporary uses for the lot.
There’s always the chance that the “$1.24 million empty lot” could turn into the next $1.7 million toilet. But if the city blueprint holds up, SF has to bite the bullet on some of these costs sooner or later. “Carrying costs are a real concern, but if we’d done a buying spree 20 years ago, we’d be saving a huge amount in the long run,” says Smeallie.
As city planners have laid out, San Francisco has to create a record-breaking number of homes, and to do so, it needs a land bank. It also needs an acquisition plan, and yes, a plan to keep costs down. Right now, MOHCD says we need more money for these goals, but City Hall hasn’t yet settled on how to tap into the funds we already have.
In development terms, eight years is a blink of an eye. We need to break new ground on this one lot — and on a lot of other sites.