How many homes is San Francisco actually building right now?
This question is more important than ever; the city is deep in a housing crisis. Even with a few years of pandemic-related rent deflation, SF maintains one of the highest costs of living in the U.S., with housing as public enemy No. 1.
Even worse, production of homes went into a tailspin in 2022. Under threat from the state, City Hall just passed a long-term plan for tens of thousands of new homes here, but it’s just a blueprint.
Without new rules to loosen long-standing building restrictions and unstick coagulated bureaucracies, actual roofs won’t materialize over San Franciscan heads, and the so-called housing pipeline will remain as clogged as ever.
In fact, that pipeline is a key to understanding how housing works in the city. It’s a rough estimate of the number of homes currently in development. It’s a big number, and while this seems like a good thing, few issues turn the mood sour in housing circles quite like the P-word.
Indeed, when SF planners insisted last year that the big pipeline number was a good reason to expect near-term bounties of housing, state watchdogs sent a shot across the city’s bow.
The truth is, the housing pipeline does not tell us how much new housing we’re getting. Nor is it really supposed to. Just what does it tell us then? Well, that’s where things get complicated.
What’s in it?
In March, Mayor London Breed and Sup. Shamann Walton trumpeted new legislation aimed at pushing some long-stalled SF projects closer toward completion, starting with the redevelopment of the old power plant site next to Pier 70.
The law, approved unanimously by the supervisors, promises “targeted public financing” for projects that have hit a wall. “We need to get them going now,” Breed said in a statement, citing “more than 52,000 housing units that are approved but stalled” across the city.
“Unlocking the pipeline” is a key promise in Breed’s ongoing plan to goose housing development. The power plant plan is one of several new housing initiatives launched in recent weeks, including one just a few days ago aimed at cutting down on the number of hearings certain new housing projects have to have.
The idea that SF’s housing pipeline is backed up and needs intervention to give a final push to tens of thousands of new units is a fairly common complaint across the political spectrum. To understand what this means, we’ll have to examine the pipeline itself.
At first glance, the city’s own Pipeline Report seems bountiful: The dashboard tool reveals 69,898 units in the works at the end of the last quarter, of which some 22 percent (15,127) are affordable homes.
The numbers seem to suggest SF is well on its way to fulfill the state-mandated goals of some 82,000 new homes in the next eight years. But “in the pipeline” does not mean likely to be built; it just means a developer or other potential builder has filed the most basic paperwork to start the process of building a new home.
“The pipeline is the accounting of all of the development projects that are in the formal process of being reviewed for approval or have been approved and are in various phases of permitting,” says Joshua Switzky, SF Planning’s deputy director of citywide policy. “Some won’t get built for 20 years; some will get built in the next couple of years.”
On closer inspection, the pipeline report has some nuance, breaking down the number of units at various steps of the process. At the end of March, just over 19,000 homes were listed under “applications filed,” the earliest step. Nearly 15,000 were in the middling stages of permit approval, and nearly 5,000 were listed as “under construction,” which puts them closer to the finish line but still not guaranteed to make it.
Almost half of the pipeline total — more than 31,000 homes — are in a nebulous zone of megadevelopments that the city has approved, and in a few cases are underway, but that will take decades to complete. This list includes more than 9,000 homes at Hunters Point, 6,000 on Treasure Island, and 3,700 at Parkmerced. (The aforementioned power plant project is for more than 2,200 homes.)
Even within these categories, there’s gradience that’s hard to measure. “There’s going to be a healthy number of those projects in [earlier stages] that will get built before the ones in later phases,” Switzky says. That is, some homes will go from proposal to construction in a few years, and some will linger in development hell for a lot more.
But even that’s not what gives housing watchers headaches. The real problem is that some of these homes are never actually going to be built at all.
Echoes of Lewis and Clark
Housing projects get canceled in SF all the time. Sometimes the money falls through, sometimes a key approval is too hard to get, and sometimes the costs no longer make sense, which can even happen as the first shovels are about to hit paydirt, as in the case of 325 Fremont St., which broke ground in 2017 but then just sat there for years without so much as an I-beam to show for it.
A pipeline is probably not the best metaphor for SF’s development process. If water goes in one end of a pipe, unless something goes badly awry it will eventually come out the other end. Housing development is more like a wagon train.
Imagine it’s the 19th century and some 70,000 homesteaders are hitting the old Oregon Trail in hopes of making it out west. Some will go all the way; some will turn back after the first few perils; some will die of dysentery, or perhaps end up stranded in the Sierra Nevada mountains and resort to eating each other.
SF Planning’s chief of staff, Dan Sider, tells The Frisc he prefers an aqueduct analogy: A certain number of homes pour out of a big, hypothetical housing reservoir, but along the way “some of the water evaporates, some of it ends up in side canals waiting for a farmer” to divert it where it needs to go (and who knows how long that will take), “and some of it flows right through.”
In conversations with several local planning staffers, everyone expressed confidence that a bulk of the pipeline would someday be realized. But which ones, and after how long? Nobody can say for sure.
On its own, the pipeline is just a tool, inexact but helpful, to allow the city to assess possible housing futures. But when the politics of housing enter the equation, the pipeline becomes a hot-button issue.
YIMBYs and other housing hawks allege that SF politicians use the big pipeline number to tout progress that might be illusory. “I don’t think that number really means anything,” says Aaron Eckhouse, a policy manager for California YIMBY and self-proclaimed “pipeline skeptic.”
The big number sounds good, adds Eckhouse, but a relatively small number would be better — meaning projects are being completed swiftly. Numbers in the tens of thousands means, well, our pipes are clogged.
“You have a lot of these projects where these numbers have been counted for a long time, but they never move anywhere,” says Jane Natoli, SF organizing director for YIMBY Action. Some SF megaprojects, like Parkmerced and the Hunters Point Shipyard, with its controversial environmental cleanup, move at glacial pace while making up a big chunk of the pipeline.
“The number that actually means something is those buildings that exist in the physical world and a person can live in right now,” adds Eckhouse, stressing that SF should measure progress by the number of units completed, not in the works.
By that backward-looking metric, though, the city would be years behind by the time a shortage became apparent. The pipeline may be hazy and imperfect, but it does help to eyeball potential housing outcomes in advance.
“If there was a way to make the information more useful to everyone who uses the pipeline we would do it, but different people use it for different purposes,” says Sider. “Further tuning the pipeline might make it more useful for one audience but would likely have the opposite effect for other audiences.”
That’s all well and good — unless a state regulator steps in.
Element of surprise
The ambiguity of SF’s pipeline is one reason state housing officials last August scolded the city’s first attempt at a new housing blueprint, known as its Housing Element. (It was amended and eventually approved.)
In a warning letter, the California Department of Housing and Community Development essentially said putting too much faith in the pipeline was a pipe dream: “Given the element’s reliance on pipeline projects, the element must include programs with actions that commit to facilitating development and monitoring approvals,” the letter read, and demanded “a commitment to alternative actions,” such as rezoning neighborhoods to make housing easier to build if rosy pipeline projections didn’t pan out.
This plan B — called a “circuit breaker” by housing hawks, and a “dirty bomb” by people who believe the Housing Element is a giveaway to developers — made it into the final version.
Staff at HCD tell The Frisc that different California cities all have different pipeline models and different assumptions about them. In the first draft plan, SF was relying on pipeline units for as much as 50 percent of future housing goals; that’s pretty high, but the state does sometimes greenlight Housing Elements that go even further than that.
The problem with SF was that the city’s own initial analysis just wasn’t persuasive enough that the projected units would actually be delivered, and thus the state demanded more safeguards in case the pipeline proved leakier than expected. Future projections are all well and good, but not all pipelines are necessarily built equal.
Chris Elmendorf, professor of land use at UC Davis and a frequent pipeline critic, says the current plan is still “too optimistic” because it’s not based on feasible economic assumptions. But with some rezoning already called for, and a plan B on tap, Elmendorf notes that “the city ended up in a reasonable place.”
Which is lucky: As hard as it was for many SF officials to acknowledge the need for a new housing approach, at least they voted for it, albeit under the gun and at almost literally the last minute. Across the Bay Area, 85 percent of locales haven’t even gotten that far, and state watchdogs might soon be more than watching.