Berkeley has no rule requiring a siting decision to name the site, or consultation to precede the money. It has an understanding. This is an account of what happened to it.
On November 18, 2025, the Berkeley City Council approved a $7,920,000 contract. Inside it was a three-year master lease of a hotel, worth $4,002,750, for a post-arrest diversion and reentry program.
The staff report describes the property as “a 30-room, master-leased hotel.”
It does not say which hotel.
Under “ALTERNATIVE ACTIONS CONSIDERED,” the same report says: “None.” The stated reason is grant compliance.
The item was on the consent calendar. It passed with dozens of others in a single batch vote, with no discussion. It was not pulled.
Seven weeks earlier, on September 30, the Council accepted the $8,000,000 state grant that funds the program. Also on consent. Also unpulled. That report did not name the site either.
The Council committed $15.92 million and a three-year presence in a specific neighborhood, and at no point in the authorizing record did the neighborhood appear.
There is no step in Berkeley’s process where a report that omits the location is returned as incomplete. No requirement that consultation precede the appropriation that makes a siting effectively final. No rule that an externally funded item with a deadline receives more scrutiny rather than less. No default pulling an item with district-specific consequences off consent.
Each of those is discretionary. Berkeley’s City Council passes most business in a single batch vote called the consent calendar. Dozens of items are approved together, in one motion, with no individual discussion. It exists so routine business — minutes, small contracts, ceremonial resolutions — doesn’t consume a meeting, and it is a reasonable tool. Any councilmember can “pull” an item off consent to have it debated separately.
But pulling is a choice, and to make it on behalf of a neighborhood a member must already know which neighborhood is involved. When the report omits the site, that knowledge has to come from outside the record.
So the only safeguard against a neighborhood learning about a facility after the money is committed is a member choosing, with nothing requiring it, to volunteer the information and interrupt a batch vote.
That is not a safeguard. It is a norm — and Berkeley has one. The understanding that you tell the district first is well enough established that when it was disregarded, the councilmember involved acknowledged it in writing without anyone having to cite a rule.
What follows is what happened to that understanding. Both money votes described below were on the consent calendar. Neither was pulled. Every quotation is verbatim from Councilmember Tregub’s own emails and newsletters or from the City’s agenda record.
On the morning of November 19, 2024, District 4 residents received their monthly newsletter. Partway down, under “District 4 Updates”:
“At our Council meeting tonight, November 19th, we will vote on approving a contract with Dorothy Day House to operate interim housing at the Howard Johnson motel…
Our office plans to hold an in-person community meeting to discuss this project in more detail… Please save the date: Wednesday, December 11th, 5:30–7:00 PM.”
Councilmember Igor Tregub, District 4 newsletter, November 19 2024, 10:42 AM
Notice of the vote and notice of the consultation arrived in the same email, hours before the Council met. The community meeting was scheduled for 22 days after the funding decision. It could explain the project and address implementation concerns, but it could not influence whether Council approved the contract.
A constituent wrote to ask why. The reply, the next day:
“I acknowledge… that the timing of this matter was less than ideal. We had a narrow window of opportunity to accept the Governor’s grant once awarded to us.”
Tregub to constituent, November 20 2024
The grant window was the explanation given. It matters because urgency appears again in the second episode.
Eleven months later, a second project: the Berkeley CareBridge Program, a post-arrest diversion and reentry program funded by an $8 million state Proposition 47 grant, to run January 2026 through December 2028. The site selected was the Signature Inn at 1761 University Avenue — roughly half a mile from the Howard Johnson.
Here is the sequence, combining Tregub’s own account of when he learned the location with the City’s funding record.
| Date | What happened | Source |
|---|---|---|
| ~June 2025 | Prop 47 grant application submitted. Preparing it involved City staff, the City Manager’s office, Options Recovery Services, Dorothy Day House, and BPD. | correspondence |
| late Aug 2025 | “I first became aware of the specific location in late August of 2025.” | Tregub, Oct 30 |
| Sep 30 2025 | $8,000,000 grant accepted by Council — on the consent calendar. Not pulled for discussion. | agenda record |
| Oct 21 2025 | Neighbors told about the project and the site. A community meeting is promised “in the coming weeks.” | Tregub newsletter |
| Nov 18 2025 | $7,920,000 contract with Options Recovery approved — on the consent calendar. Not pulled. | agenda record |
The neighborhood was informed 21 days after Council accepted the $8 million grant, and four weeks before Council approved the $7.92 million operating contract funded from it. By the time residents were told, the City had already accepted the funding structure around which the project would proceed.
Asked about the timing, Councilmember Tregub sent two emails on October 30, 2025, roughly six hours apart.
“My office became aware of the proposed location very recently. We shared this information with the community shortly after it became available to us.”
“I first became aware of the specific location in late August of 2025.”
Late August to October 21 is eight weeks. Inside those eight weeks, on September 30, the Council accepted the $8 million on the consent calendar — and the councilmember for the district in which the money would be spent did not pull the item, and had not yet told the district it existed.
Asked why this site, the answer was that there wasn’t much choice:
“While the alternative to this project would likely be a long-term vacant motel, I understand and share your concern about concentrating non-congregate shelter services predominantly within the downtown and West Berkeley districts.”
Tregub to constituent, October 26 2025
The premise there is worth examining. The Signature Inn is privately owned, on a commercial corridor, and by the City’s own account the owners approached the City to offer it. A motel whose owners approached the City about a transaction was not self-evidently destined to remain vacant. Other private or public uses were at least possible, yet no comparison of alternative sites appears in the record examined here. No comparison of alternative sites appears in the record.
Pressed further, the councilmember was direct about the constraint he now accepts:
“I am not inclined to support additional projects for prior low-level offenders or other potentially higher-risk populations in our district absent a commitment of other parts of the region or other Berkeley districts to shoulder the load.”
Tregub to constituent, October 30 2025, 10:58 AM
That is a substantive commitment, made in private correspondence after two projects had already been sited.
Taken alone, either could be bad luck — a grant window that closed faster than a calendar could be cleared. Taken together they describe a method:
| Howard Johnson, 2024 | CareBridge, 2025 | |
|---|---|---|
| Outside grant with a deadline | Governor’s grant | Prop 47, $8M |
| District told | Morning of the vote | 21 days after the money vote |
| Community meeting | 22 days after | “In the coming weeks” |
| Money votes on consent | — | Both funding actions: acceptance of the $8M grant and approval of the $7.92M contract funded from it |
| Reason given | “Narrow window… to accept the Governor’s grant” | “Became aware… very recently” |
The second episode happened after a constituent had objected to the first in writing, and after the councilmember had conceded the first was “less than ideal.” Whatever explains the first episode, the consultation problem had already been identified in writing before the second occurred.
A single occurrence is consistent with bad luck: a grant window that closed faster than a calendar could be cleared.
The second occurrence removes that reading. The norm had been named, breached, acknowledged in writing, and objected to by a constituent. Every input an informal constraint needs in order to bind was present. The same sequence ran anyway, in the same district, on the same justification.
That is the finding, and it is not a finding about a person. It establishes what the norm is worth. A rule that is breached produces a consequence. A norm that is breached, conceded, and breached again produces a second community meeting after a second vote. The mechanism did not fail because it was ignored by someone unusually inclined to ignore it. It failed because nothing in the process gave the acknowledgment any operative consequence — no requirement to change the sequence next time, no record of the concession anywhere the next agenda would encounter it.
The concession was made in an email. Emails do not calendar items.
Two things are established by that phrase, and the second matters more than the first.
The first is that the norm exists. You cannot fall short of an ideal that isn’t there. Nobody had to cite a rule, because there was no rule to cite — the standard was understood well enough to be conceded on request.
The second is the shape of the concession. “Less than ideal” is not an admission that a line was crossed. It places the conduct on a scale running from ideal downward, where the top is aspiration and everything below remains within bounds. It concedes that something better was available while denying that anything was breached. Paired with a reason — the grant window — it does not commit the office to doing differently. It explains why this was acceptable on this occasion.
That is a justification schema, not a correction. And a justification remains available the next time the same conditions arise.
It also describes a disposition, not merely a defense. “Less than ideal” is what a member concludes when the project continues. The vocabulary and the act travel together: a councilmember who characterizes the sequence as suboptimal has already decided it proceeds, because suboptimal things proceed all the time. The alternative characterization is available on the same facts and points somewhere else entirely — this is a breach of my constituents’ trust; halt it until they have been consulted. That is a sentence a member can say to staff, on the record, with no rule requiring it and nothing preventing it.
Neither episode contains that sentence. What the record contains is the milder one, twice.
This is the general case, and it is worth stating plainly because the last decade has taught American civic life how much of government runs on understandings rather than requirements, and how little happens when one is disregarded. That lesson tends to get filed as somebody else’s problem — a failure of the other party, in some other institution, at some other level of government. It is not. A norm binds only those already inclined to follow it, and that is as true in a progressive city council chamber as anywhere else. Berkeley is not exempt because its intentions are good.
This page names a councilmember, which these explainers ordinarily avoid. The reason is that the documentary record is the correspondence: the dates come from his own emails and newsletters, quoted verbatim, alongside the City’s agenda record. An account that described it anonymously would be less checkable, not more principled.
But the finding is about the norm, not the man. Nothing in either sequence required breaking a rule, because no rule applies. A successor in that seat could run the identical sequence next term — accept a grant on consent for an unnamed site, schedule the meeting after the vote, cite the window — without violating a single procedural requirement. The conduct was available to any of the nine, in any district, on any externally funded siting decision. What made it available was not one member’s disposition. It was a gap in the rules and an expectation the body had stopped enforcing on itself.
The Berkeley Budget Framework names this exact failure mode. On grant money:
“A grant changes the financing of an undertaking; it does not determine the undertaking’s category or priority… Dollars attracted is not a success metric if the award creates obligations that outlive the money.”
And on deadlines specifically:
“Urgency can justify a faster process. It does not justify pretending that quality gates do not apply… Missing a funding window may be less costly than accepting a durable obligation the City cannot execute well.”
A three-year program in a fixed location is a durable obligation. The grant that pays for it expires in December 2028; the neighborhood does not. Whether the projects are good is a separate question from whether the district was given any opportunity to weigh in before the money was committed — and on the second question the record is unambiguous.
There is a further problem, and it is separate from the timing. Neither staff report named the site. Council authorized a three-year master lease of “a 30-room, master-leased hotel” without the record identifying which hotel — the single fact that determines which neighborhood absorbs the program. The November 18 report also records “ALTERNATIVE ACTIONS CONSIDERED: None,” justified by grant compliance.
That is why this page treats the episodes as more than a courtesy failure. A member who wanted to pull the item on behalf of a specific neighborhood would have had to know, from outside the record, which neighborhood it was. And when asked afterward why this site, the answer given was that the alternative “would likely be a long-term vacant motel” — a comparison no document in the record performs.
A community meeting held after Council has accepted the funding cannot give residents the same influence as consultation before commitment. It can still shape operations, mitigation and neighborhood communication, and sometimes implementation — all of which have value. What it cannot do is place residents on equal footing in the original choice of whether and where to proceed.
The councilmember has since said he is “not inclined to support additional projects… absent a commitment of other parts of the region or other Berkeley districts to shoulder the load.” That commitment was made in an email to one constituent. Whether it is repeated on the record, and whether it governs the next vote, is checkable.
The case for writing these down is not that officials cannot be trusted. It is that the informal version was tried here, under favorable conditions — named, conceded, objected to — and did not hold. Codification is what you reach for when a norm has been shown not to bind.
Name the site in the authorizing record. An appropriation with a specific location should not be votable while the location is absent. Absent it, the report is incomplete rather than brief, and incomplete items are returned, not approved.
Consultation precedes commitment. For a siting decision with durable neighborhood effect, the affected district gets documented notice before the appropriation that makes it final — not after, and not in the same email.
District-consequential items come off consent by default. Invert the burden. Rather than requiring a member to know enough to pull an item, require that items with identified district-specific siting consequences be calendared for discussion unless Council affirmatively places them on consent.
Deadline pressure is disclosed and priced, not used as an answer. Where an external deadline compresses the process, the report should say what the compression cost — which steps were skipped, what would otherwise have been examined, and what happens if the window is missed. Missing a funding window is sometimes cheaper than accepting a durable obligation the City cannot site well.
“None” is not an answer to alternatives considered. If grant compliance genuinely forecloses comparison, the report should say what the grant required and why that eliminated the alternatives, rather than recording the absence of analysis as though it were its result.
A member needs a way to stop the clock, and the standing to use it. Every rule above can be complied with formally and still produce this sequence, because rules govern documents and someone has to notice in the moment. What worked against that here is that a member who objects is asking the City to forgo money it has been offered, on behalf of residents who do not yet know the project exists, against a deadline someone else set. The institutional answer is to make the request ordinary rather than heroic: a district councilmember’s demand for consultation before a siting appropriation should pause the item by right, once, for a defined period, without requiring a majority or an accusation. A pause a member can invoke alone is a pause a member will actually invoke.