Ever since City Planner Andrew Thomas began working on a draft of the 2023‑31 Housing Element for the City of Alameda, he has made clear that his goal is to prepare a document that the state Department of Housing and Community Development will not hesitate to certify as being in compliance with state housing law.
Well, it appears Mr. Thomas now is only a couple of steps away from crossing the finish line.
The City submitted the “final” draft of the Housing Element to HCD in May, and it has delivered two sets of revisions since then. On August 25, HCD sent Mr. Thomas a letter stating that, “HCD is pleased to find the [most recently revised] draft housing element in substantial compliance with State Housing Element Law.” The letter goes on to suggest that, once the Alameda City Council adopts the draft and it is sent to HCD, the department will certify it as “comply[ing] with the State Housing Element Law.”
The next step, Mr. Thomas told us, will be to present the HCD‑validated draft to the Planning Board at its September 12 meeting.
Thus far the chronology.
Much to the delight of the anti-Measure A crowd, the draft blessed by HCD commits the City to allowing multi‑family housing in all six residential zoning districts, removing density limitations for every remodeled building in all of them, and increasing the maximum permissible density for newly constructed buildings in four of them.
At the same time, however, the draft projects fewer new housing units – and fewer new housing units affordable by lower‑income households – resulting from residential rezoning than the draft originally submitted to HCD did.
Therein lies a story.
(The draft also creates a new zoning category sanctioning multi‑family residential development with unlimited density on parcels located within a “transit corridor,” but for the present we’ll focus on the rezoning of the existing residential districts.)
We’ll begin by dialing back the clock.
After the Association of Bay Area Governments handed the City an allocation of 5,353 new housing units as its share of the Regional Housing Needs Assessment for the period from 2023 through 2031, the major challenge for the City’s planners was to identify “opportunity sites” that would enable the City to meet the mandated target.
The City is divided into six residential zoning districts. The two largest are R‑1, which contains 7,763 parcels, and R‑4, which contains 3,393 parcels. They are followed by R‑2 (1,168 parcels); R‑5 (840 parcels); R‑3 (415 parcels); and R‑6 (81 parcels).
From the beginning, rezoning these residential districts was one of the actions Mr. Thomas advocated to make additional housing units available. The first draft of the Housing Element, published in October 2021, proposed that the City remove the existing prohibition on multi‑family housing and eliminate the existing density limitation throughout the city. Under the draft, two units would be permitted on every existing residentially zoned lot, and at least five units on lots of 5,000 square feet or more in the R‑2 through R‑6 districts. These actions, the draft stated, could be expected to produce 63 new units per year or about 500 units over the eight‑year period.
The problem, of course, was that the multi‑family prohibition and the density limitation were set forth in Article XXVI of the City Charter (aka Measure A), which had just been overwhelmingly affirmed by Alameda voters. But the obligation to satisfy the RHNA target was imposed by state law. And as long as the rezoning was necessary to enable the City to comply with that obligation, state law preempted the Charter.
For a while, residential rezoning did indeed appear necessary. Under both the October and November drafts of the Housing Element, the City still would fall about 300 units short of the 5,353‑unit target absent additional units in the existing residential districts. But the planners kept at it. By March 2022, they had come up with a site inventory totaling 5,891 new units without any residential rezoning. After further tinkering, that number rose to 6,143 units in the April draft.
Yet the draft continued to provide for rezoning the residential districts to create a “buffer” between the site inventory and the RHNA target. Having gone through several permutations, the rezoning scheme now called for removing the multi‑family prohibition in every district and increasing the permissible residential density in the R‑3 through R‑6 districts from 21 units per acre to 30 units per acre in R‑3, 40 units per acre in R‑4, 50 units per acre in R‑5, and 60 units per acre in R‑6. The R‑1 and R‑2 districts would remain zoned at 21 units per acre, but the former is subject to SB 9, which allows as many as four units on a single lot.
(Why wasn’t the density increased for R‑2? We wondered, so we asked Mr. Thomas. “We are trying to reinforce an urban form that in many ways already exists because of development between 1900 and 1972, which also reinforces diversity of character in different neighborhoods,” he said. “R‑2 at 21 just seems to fit.”)
The proposed residential rezoning, the draft stated, would lead to 34 new units per year (270 units over eight years). Moreover, according to the site inventory, 120 of the 270 new units would be affordable by very‑low‑ or low‑income households.
Like the original rezoning proposal, the April version on its face violated Article XXVI, and it was no longer possible to claim that state law preempted the Charter on the grounds that rezoning was necessary to enable the City to meet its RHNA target (it wasn’t.) But now a new rationale for preemption emerged: “affirmatively further[ing] fair housing.”
Under state law, a housing element must contain a “program” designed to “affirmatively further fair housing.” The latter phrase “means taking meaningful actions, in addition to combating discrimination, that overcome patterns of segregation and foster inclusive communities free from barriers that restrict access to opportunity based on protected characteristics.” Not exactly clear, but what state law is?
Mr. Thomas, of course, was well aware of this provision, and many of his staff reports referred to the fair‑housing law. Not until March, however, was the duty to “affirmatively further fair housing” set forth as an explicit rationale for residential rezoning (and thus for finding a state‑imposed obligation that preempted Article XXVI).
As the March 14 staff report puts it, the “goal” of residential rezoning was not only to “help accommodate the RHNA” target, but also to “make the necessary adjustments to comply with State fair housing law.” Mr. Thomas drove home the point at the Planning Board meeting in response to a question from Board member Hanson Hom:
So [the] staff recommendation is, we should not be sending a housing element to HCD that says oh, we’re going to continue to prohibit multi‑family housing in our residential zoning districts. I guarantee you the very first comment we will get back from them, among seven pages of comments, will be, “You’re prohibiting multi‑family housing in residential zoning districts? No, that’s not going to work, Alameda.”
A few weeks later, Mr. Thomas doubled down in response to a softball thrown by Councilman John Knox White. If the Housing Element didn’t rezone the residential districts to remove the multi‑family prohibition and the density limitations, HCD would “send it right back to us,” he said, and continued:
What [the fair‑housing law, as interpreted by HCD is] saying is you cannot prohibit multifamily housing and residential densities that support affordable housing in all of your residential districts. That is unfair to people who need affordable housing. . . . What [the law] means is [that erecting] multi‑family housing barriers [and] reducing or not allowing residential densities that support affordable housing . . . you’re saying people who need that are not allowed in. And that’s one barrier to furthering fair housing.
Neither the staff report nor Mr. Thomas explained in so many words exactly how residential rezoning would “affirmatively further fair housing.” Presumably, the argument was that, as a result of the residential zoning changes, there would be more affordable housing more widely dispersed across the city. But this was just a prediction, not an inevitable consequence. To the anti-Measure A crowd, “everyone knows” that the prediction would come true. But to more analytic observers, it risked confusing aspirations with reality.
One more more round of changes was made before Mr. Thomas sent the draft Housing Element to HCD. (He maintained that he could do so without a prior vote by Council endorsing the document.) But it was a crucial one: Under the April draft, the new density limits for the R‑3 through R‑6 districts applied regardless of whether additional housing units were created by new construction or by remodeling. The May draft, however, added a provision “exempt[ing] adaptive reuse of an existing building from residential density limits . . . in all residential districts.” (Emphasis added.)
What this meant was that, if a developer or homeowner wanted to redevelop a parcel in, say, the R‑4 district with a new building, she’d be limited to 40 units per acre (which translates to four units on a 5,000 square‑foot lot) – but if she wanted to remodel an existing building, she could add as many new units as the structure would bear.
Once Mr. Thomas submitted the “final” draft to HCD, the department had 90 days to review it and provide comments. But the City planners didn’t sit on their hands – nor did interested parties like the Alameda Citizens Task Force. In August, Mr. Thomas submitted revisions in response to a list of questions from HCD. The submission didn’t change the residential rezoning program, but it added a “Residential Infill Analysis” with more detailed, and more enlightening, information.
But it was the response by ACT that got Mr. Thomas’s (and HCD’s) attention. In a letter to HCD, ACT’s Paul Foreman argued that state law requires that, if a city included non‑vacant sites in its inventory, it must “describe the realistic development potential of each site” and list them by assessor’s parcel number. But the draft Housing Element did not do this for the parcels in the “infill residential district sites” category. Rather, it lumped all of them together. This, Mr. Foreman argued, did not comply with state law.
Perhaps after consulting counsel, Mr. Thomas concluded that Mr. Foreman’s legal conclusion was correct! (By admitting that an opponent might have a point, the City Planner forfeited his honorary membership in the Alameda “progressive” society, which undoubtedly would have simply told Mr. Foreman to sod off.)
And Mr. Thomas promptly took corrective action. Rather than undertake the onerous, if not impossible, task of site‑by‑site description, he revised the inventory. The total units in the “infill residential district sites” category now includes only new units attributable to additions to existing buildings. The statute cited by Mr. Foreman, Mr. Thomas maintained, does not require a more extensive description in such a case.
The effect of this revision was to reduce the number of units resulting from residential rezoning from 34 to 20 per year. And that was the number in the draft Housing Element that HCD validated. (Mr. Thomas told us the draft, as revised and approved, will be posted on the City website this week).
So the picture has gotten a little clearer. But a few questions still need to be asked to flesh it out.
First, what kind of new housing units is the residential rezoning likely to produce?
In theory, the zoning changes would allow a developer to buy an improved lot, knock down the existing structure, and then build – depending on the zoning district – a three‑, four‑, five‑, or six‑unit apartment building in its place. But this would be an extraordinarily expensive enterprise, and Mr. Thomas expects it would happen rarely, if it all.
More likely, a homeowner – or a developer who buys the property – would put one or more new units on the lot without demolishing the existing structure. As the “Residential Infill Analysis” points out, this could happen in one of three ways: a new stand‑alone structure could be built in the backyard; a second story or basement could be added to an existing home, or a single‑family home could be converted into a multi‑unit building.
Of these alternatives, Mr. Thomas told us he thought the last was the likeliest. The City’s experience with Accessory Dwelling Units supports this conclusion, he said. Since the ADU ordinance was enacted in 2017, the “vast majority” of new ADUs consisted of additions to existing homes. The same factors that led to this outcome, Mr. Thomas argued, would cause the new units permitted by the zoning amendments to be added in the same way.
Second, where in the city will the new units go?
In theory, a homeowner or developer could put one or more new units on a property located on any residential lot in the city. But Mr. Thomas told us that his “gut feeling” was that the bulk of the new units resulting from the zoning changes would go into the R‑2 and R‑3 districts. Those districts had a lot of land, a lot of parcels, and a lot of large backyards, he said. By contrast, there weren’t as many lots in the R‑4 through R‑6 districts, and many of them already had been developed with multi‑family buildings in the pre‑Measure A era.
And now for the harder questions.
First, realistically, how many new units are likely to be built over the next eight years?
As we discussed earlier, the number of new units projected to result from residential rezoning has proved to be something of a moving target. At the outset, the site inventory in the draft Housing Element showed 500 new units over eight years; now, it’s 20 units per year or 160 units all told.
We suspect that even Mr. Thomas will admit that we’re talking about best guesses here. For the sake of argument, we computed the maximum number of new units that could be added if every parcel in the R‑3 through R‑6 districts was developed to the fullest extent permitted by the zoning amendments. The total – hold your breath – came to 6,128 units. But that ain’t gonna happen. What’s a more realistic number? It depends on the clarity of your crystal ball. And we’re in no position to opine whether Mr. Thomas, or, say, Mr. Foreman, has the clearer instrument.
Two reasons, however, can be cited in support of a low number. One is economic: a homeowner or developer won’t add units to a property without a level of confidence that the addition(s) will, to use the time‑worn phrase, “pencil out.” Yet, in the current real‑estate market, can anyone be confident that building a bevy of new units would pay off? The other reason – cited to us by Mr. Thomas himself – bolsters the first: since Council implemented SB 9 in March 2022 by passing an ordinance allowing four units on a single‑family lot, there have been no applications filed for new SB 9 units. Why should one expect an influx of applications for other newly permitted units?
Finally, how affordable are the new units likely to be?
Mr. Thomas insisted to us that the “vast majority” of the new “in‑fill” units projected for the City’s residential districts will be affordable by lower‑income households. He directed us to an “ADU affordability survey” done by ABAG, which found that 30 percent of ADUs were affordable by very‑low‑income households, 30 percent by low‑income households, and 30 percent by moderate‑income households. The new units that would be added as a result of the residential zoning changes resembled ADUs, he said, so the same mix could be expected.
Yet, for Mr. Thomas to be right about affordability, the new units would need to rent, according to a table in the draft Housing Element, for no more than $1,028 per month for an extremely low‑income, four‑person household; $1,703 for a very‑low‑income four‑person household, and $2,740 for a low‑income four‑person household.
But consider the state of the current market. According to another table in the draft Housing Element, the average monthly rent in Alameda in 2021 was $2,175 for a studio; $2,775 for a one‑bedroom apartment; $3,228 for a two‑bedroom apartment, and $3,878 for a three‑bedroom apartment. The new units permitted by the zoning amendments thus would need to be rented at significantly below‑market rates in order to be affordable by lower‑income households.
In any event, as we’ve noted, a homeowner or developer must charge rent sufficient to cover the cost of adding the unit. As construction costs continue to rise, the odds of keeping a new unit within the affordable range diminish accordingly.
(There is one other point to note about the relationship between cost and affordability. The cost of adding new units not only may dictate rents that are unaffordable to new tenants, it may also force the owner to raise rents for existing tenants. If so, an apartment that once was affordable by a lower‑income household will no longer be so.)
Of the four questions we’ve raised, the last two matter the most to the issue of whether, and to what extent, the residential rezoning program “affirmatively furthers fair housing.”
Suppose that the program in fact produces only 20 new units. Is that number large enough to substantiate the claim that the program actually promotes additional housing? Or suppose that, contrary to Mr. Thomas’s belief, the new units in fact are affordable by only a handful of lower‑income households. Is that benefit significant enough to support the contention that the program actually “fosters inclusive communities”?
If the answer to these questions is no, the City may find itself under legal attack from both ends of the political spectrum.
On the one hand, the “progressives” could challenge the residential rezoning program on the grounds that it produces so little new affordable housing that it can’t be deemed to satisfy the City’s obligations under the fair housing laws. On the other hand, their opponents could attack the program on the grounds that rezoning produces so little new affordable housing that it can’t be considered necessary to comply with the housing element law and therefore does not preempt Article XXVI.
We suppose that, from a public‑policy perspective, it is possible to defend the residential rezoning program even if it produces only 20 new affordable units per year. Even one new affordable unit, it can be argued, represents a positive step toward addressing the housing crisis. Moreover, a program that induces so few new units, all of them in existing buildings, is unlikely to change the “character” of the built environment. Indeed, the ordinary single‑family homeowner might not even be aware that it existed.
But if Mr. Thomas’s analysis is wrong, the program could lead to the parade of horribles – Victorians being torn down and replaced with multi‑family apartment buildings with units affordable only by the wealthy – conjured up by its opponents. And remember this: once the zoning is changed, it’s going to be very difficult to go back.
The first set of issues can only be adjudicated in a court of law. The second set will be resolved in the court of public opinion. Either way, it’s fair to say that, alas, for Mr. Thomas (and Alamedans generally), the saga may be just beginning.
October 2021 Housing Element draft: 2021-10-25 Ex. 1 to staff report to P.B. – October Draft Housing Element
November 2021 Housing Element draft: 2021-11-16 Ex. 1 to staff report – (November) Draft Housing Element
April 2022 Housing Element draft: 2023-2031 Housing Element_Public Review Draft_April 2022
May 2022 Housing Element draft: 2023-31 Housing Element (May 22 draft)
August 2022 Housing Element draft: 2023-31 Housing Element (submitted 5.27.22; revised_8.12.22) (redlined)
August 25, 2022 letter from HCD to Andrew Thomas: 2022-08-25 HCD letter