Berkeley has committed roughly $15 million to a building that opens in April 2028. The budget adopted in June 2026 records what it will cost to run as three letters: TBD. Eight years earlier, a City-commissioned study put the figure at $1.5 to $2 million a year. The number did not get firmer as the building got closer. It disappeared.
This is the companion to Buy once, pay forever. That page follows a program from first appropriation to fourth to show what ownership costs when nobody counts it. This one catches the same failure eighteen months before the consequence arrives, in a document the Council adopted with the gap printed on the page.
Every figure below comes from the City’s own feasibility study, its own budget, or the Council’s own record. Where a document is characterized rather than quoted, it is identified.
The biennial budget presented for adoption on 23 June 2026 lists the African American Holistic Resource Center in its table of strategic priority projects:
Department: City Manager’s Office
Title: African American Holistic Resource Center
Description: Delivering full service community engagement and support models in the new building to be completed April 2028.
Start/End: December 2026 – Ongoing
New or Continuing: New
Estimated Budget: Staff Time; Budget TBDProposed Biennial Budget FY2027–28, strategic priority table, p. 113
Read the row across. A project the City classifies as new and ongoing, delivering services in a building with a completion date already fixed, and no number for what delivering them will cost.
The Council adopted that budget and the accompanying Five-Year Capital Improvement Program the same night. The appropriations ordinance — $921,596,116 gross — passed with zero speakers.
This is not a case of an unknowable figure. In 2018 the City commissioned a feasibility study for the facility. It was transmitted to Council as an information item in March 2019, and it contains an operating estimate:
“The projected annual budget for the AAHRC will range from $1.5 – $2 million dollars.”
AAHRC Feasibility Study, 2019
Set that against the capital numbers in the same study. It priced the building itself at between $1.16 million to rehabilitate a single-storey structure and $1.905 million for new two-storey construction — excluding land, and flagged as subject to major uncertainty.
So the study told the City, in 2019, that the facility would cost roughly as much to run every year as it would cost to build once. That is the central fact about this kind of project, it was in the City’s hands before construction money was committed, and it is the fact that went missing.
That figure is also eight years old now, and certainly too low. Inflation alone would push it up by roughly a third. But it was never an operating model — the study derived it as a percentage of construction cost, and construction cost has not held still. The project it was calculated against was a rehabilitation priced near $1.16 million. What is being built carries $13.85 million in Measure T1 authorization and a $1.6 million design contract. Applied to that project, the same rule of thumb yields a materially larger annual figure. Nobody has published the recalculation, and the 2019 range remains the only operating number in circulation.
The $1.5–2 million figure was a range in a sentence, not a model. The study did not provide a staffing plan, salary assumptions, program budgets, expected service volumes, utilization projections, revenue commitments, a cost per participant, a ceiling on the City’s subsidy, an allocation of responsibility among partners, or a multiyear lifecycle analysis.
It listed possible revenue sources — event rental, retail rental, office rental, membership fees, cooperative enterprises, grants, contracts, donations, City support. Those are possibilities. None was a commitment.
The study concluded the project was feasible on the grounds that it had community and political support, that potential funders had been identified, that possible sites existed, that similar organizations existed elsewhere, and that revenue-generating activities might contribute.
Identifying possible funders is not securing funding. Identifying possible sites is not establishing site viability. Finding analogous organizations elsewhere is not demonstrating comparable costs, governance, or outcomes. A feasibility study that concludes a project is feasible because people want it and money might be found has not tested feasibility. It has recorded enthusiasm.
The documentary trail starts in 2011, with a report prepared by the African American/Black Professionals and Community Network. Its problem statement reads:
“The Black community in Berkeley requires a place to meet in order to organize, strategize, socialize, and be at home with each other; in a welcoming environment that offers culturally congruent services for African American/Black residents.”
A Community Approach for African American/Black Culturally Congruent Services, April 2011
Its first recommendation reads: “The City of Berkeley donates a building to the African American Community.”
That is a legitimate thing for a community organization to ask for. Advocacy is supposed to name what it wants. The question this page asks is what the City did next.
What it did, seven years later, was commission a study to assess the viability of developing an African American Holistic Resource Center in South Berkeley. The building was in the terms of reference. The study asked how to deliver it, where to put it, and what it might cost. It did not ask whether a building was the intervention most likely to close the gaps the 2011 report described.
This is the crayon box at the earliest possible stage. Not four options considered and three ignored — one option carried forward from the beginning, with the analytical work arranged behind it.
Suppose the City had treated the 2011 report as a statement of the problem rather than a specification of the answer. The problem, as stated: existing services were not reaching Black residents effectively, and there was no dedicated place for the community to gather.
That admits more than one remedy, and each one has a cost the City could have compared:
| Option | What it would have cost to find out |
|---|---|
| Fix the existing services — audit which City programs underserve Black residents, and why | Staff time. No capital. |
| Fund Black-led providers directly to deliver the same services | Priceable against the $1.5–2M annual operating estimate |
| Lease or share space rather than build | Priceable against the $1.16–1.905M capital estimate |
| Co-locate in a building the City already owns | Comparable against surplus public property |
| Time-limited pilot to establish demand before committing | A fraction of either figure |
| Build the center | The option that was chosen |
None of these is obviously right. The last one may well be. The point is that the record does not show them being compared — not in 2011, not in the 2019 study, and not at any of the appropriations since.
On 10 March 2020, Council took up item 18, “Siting the African American Holistic Resource Center and Affordable Housing at 1890 Alcatraz.” The item was authored by Councilmember Bartlett, co-sponsored by Councilmember Davila and Mayor Arreguín, with Councilmember Harrison added as a co-sponsor at the meeting. It directed the City Manager to study using the site for the AAHRC and to
“report back on the costs and implementation steps to repurpose the property for the AAHRC using the AAHRC Feasibility study as a guide, including what physical improvements would need to be made, and cost for ongoing operations by a non-profit.”
Annotated Agenda, 10 March 2020, item 18 — approved as recommended
That is the operating question, asked at exactly the right moment, by the councilmember closest to the project. It is Council’s own direction, on its own consent calendar, and it was approved. Everything that follows is about what happened to the answer.
A California Public Records Act request (26-2113) sought the analysis responsive to that direction, citing the direction specifically. The City returned six documents, all Mental Health Services Act planning material. None is the requested report. Under the Public Records Act an agency must determine whether it holds responsive records and say so, and must justify in writing any decision to withhold. Berkeley produced no responsive record and claimed no exemption. Either the report does not exist or the City cannot find it; either way it was not available to inform any of the decisions that followed.
What the production does show is how the project reached a capital commitment without one.
The City scheduled the funding while saying the needs were undetermined. The 2020 MHSA plan says the Mental Health Division will propose funding for the AAHRC “once the specific needs and appropriate funding categories are determined.” The 2023 plan repeats the condition word for word — “once specific needs have been determined” — and then, in the next sentence, commits to working with the AAHRC committees “over the next month to determine proposed MHSA funding amounts.” Three years, the same condition, no determination on the record, the funding scheduled regardless.
Berkeley could not assess its own system, and said so. State law requires each mental-health jurisdiction to conduct a Capacity Assessment identifying “strengths, limitations, disparities, gaps and/or barriers” in the local system of care. The City attempted one and reported that “the response rate was very low, and the Division was unable to obtain a comprehensive assessment of the local system of care for this Three-Year Plan.”
The appendix gives the number. Three organizations responded. Asked to describe limitations affecting their ability to meet the needs of racially and ethnically diverse populations, the three answered: “No limitations”; a shortage of staff speaking languages beyond English, Spanish, Maya Mam, Portuguese, French and Russian; and a plan to hire more Spanish-speaking staff. Across the whole Capacity Assessment chapter, “African American/Black” appears only as a demographic checkbox — never as an identified gap or barrier.
Three respondents is not a valid sample, and it cannot prove no gap exists. But it is the only occasion on which Berkeley systematically asked its providers where the system fails diverse populations, and the answers did not corroborate the premise the building rests on.
The study that would have asked properly was funded and appears never to have been done. The 2023 plan allocated $60,000 to hire a consultant to conduct the Capacity Assessment through an RFP. The phrase “Capacity Assessment” then appears nowhere in the 334-page MHSA FY2025 Annual Update, nowhere in the 354-page FY2026 Annual Update, and nowhere in the 258-page Behavioral Health Services Act FY2027–2029 Three-Year Integrated Plan adopted in June 2026 — the successor document where an outstanding obligation would be carried forward. No RFP, no consultant, no deliverable, no explanation. The same plans describe in detail a different consultant study the Division did complete, awarded by competitive RFP and run from FY2021 to FY2024. Three planning cycles have passed.
The 2018 study specifies two classrooms, a multipurpose room seating 250, a dance studio, a library housing the South Berkeley Legacy Project, a children’s playroom and game room, a computer lab, a classroom kitchen, a medical screening room, and two private therapy rooms. The services it lists are “health education, health screenings, mental wellness services, educational support, cultural events, legal services, social and recreational programs, and other services as needed.”
Mental wellness is one item of eight. On the plans this is a community center that includes counseling space.
But every dollar of programming money identified for the AAHRC so far is MHSA money, and the City’s own contract reports describe what that is for: “MHSA CSS funds are primarily to be utilized for treatment services and supports for Severely Mentally Ill Adults and Seriously Emotionally Disturbed Children and Youth, and for general system development.”
A dance studio is not treatment for severe mental illness. Neither is a children’s library, a computer lab, a legal-services desk or a recreation program. There is no route from this funding source to most of the services the building is meant to provide.
What the money does buy is set out in the AAHRC’s own proposal: of the $300,000 annual budget, $87,000 goes to a Program Manager, $67,000 to an Office Specialist, $21,000 to a fiscal sponsor fee, $36,000 to office space, and $89,000 to stipends, materials and a training series. Two positions — a manager and a clerk, whose listed duties are administrative. No dance instructor, no librarian, no attorney, nobody to run the health screenings or the children’s programming.
While the City spends $15 million building 6,000 square feet, the program that will occupy it is paying rent on office space somewhere else.
The capital arrived in pieces, across four separate decisions and three different funding sources:
| Source | Amount | When |
|---|---|---|
| General Fund | $250,000 | 2020 |
| Measure T1, Phase 2 | $7,000,000 | 15 December 2020 |
| Measure T1 | $6,850,000 | 13 June 2023 |
| Federal earmark / HUD | $1,000,000 | — |
These figures were first stated by a councilmember in writing in August 2026, in response to a constituent asking what the facility would cost to operate. The City Manager’s staff report of 30 June 2026 reconciles the two largest: both the $7 million approved 15 December 2020 and the $6.85 million approved 13 June 2023 are Measure T1 funds, and its resolution cites the Measure T1 Fund (Fund 511). The table follows the staff report.
Two things about that table. The first is the arithmetic: roughly $15 million committed against a 2019 study that priced the building at under $2 million. Those are not like-for-like — the study excluded land, and scope and costs both grew over six years — but the gap is large enough to have prompted a documented re-estimate, and no such document has been located.
The second is Measure T1. Nearly $14 million of the $15 million came from an infrastructure bond, which is to say from borrowing that Berkeley is still repaying, on a program the City Auditor has separately reported is running short of what it promised.
But the most important thing about the table is what it represents. Four appropriations, from three funding sources, spread across six years. Each was a moment when the City decided to put more money into this building. Each was an opportunity to ask what running it would cost, and each was a point at which the answer would have been cheaper to act on than the last.
That is the pattern worth naming, and it is not the same as nobody knowing. Somebody knew in 2019. What the record does not show is any of the four decisions pausing to ask whether the number still held, whether the source had been identified, or what the City would do if it had not been. The project did not survive scrutiny of its all-in cost. It proceeded without receiving any.
A capital project generates its own momentum: each appropriation makes the next one easier to justify, because the money already spent argues for the money still needed. That is exactly why the operating question has to be asked early. Asked at the first $250,000, it is a planning question. Asked at $15 million, eighteen months from opening, it is a problem to be managed — and by then the honest answers have narrowed to finding the money, reducing what the building does, or asking the General Fund to absorb it.
In August 2026 a resident asked a councilmember four questions: what annual operating cost he understood the completed facility would impose, what ongoing source would cover it, whether he had reviewed a staff analysis before supporting the latest commitment, and whether Council should require a sustainable operating plan before approving a new facility at all.
He answered in writing rather than routing it to staff, which is worth noting. On the general question he agreed:
“you are suggesting that the analyses provided by the City may not fully capture such lifecycle costs — most likely because operational and maintenance costs come out of a different pot of the budget than funds allocated toward design, engineering, and construction. I believe that this concern is well-placed.”
On whether Council should require an operating plan first: “As discussed above, yes.”
On what the facility would actually cost to run, he could not say, and cited the budget line — the TBD — in answer to whether he had seen an analysis. He said he would ask the City Manager when a figure might exist.
He is right that the capital decisions predate him; he took office in June 2024 and all four appropriations were earlier. But on 23 June 2026 he moved the appropriations ordinance himself.
Fifteen years after the request and eighteen months before the doors open, the record does not establish who the final operator will be, what the annual operating budget is, what share of it the City will carry, what outside revenue is committed, what service population the facility is sized for, what it is expected to achieve, how that will be measured, or under what conditions the City would stop, redesign, or hand off the project.
Two contracts exist. Each is not-to-exceed $300,000, funded by the State through the Mental Health Services Act, and each expires 30 June 2026 — twenty-one months before the building opens. They fund services during construction. They are not an operating plan for a completed facility, and they are not a commitment by anyone to carry its costs.
The AAHRC did not begin with the councilmembers now seated. It traces to a 2011 community report, a 2019 study and capital decisions taken in 2020 and 2023. Several members took office in 2024, well after the shape of the project was set.
That is a fact about how it started, not a defense of where it stands. Every decision that keeps the project moving is being made now, on a record that has not improved. Since these members took office they have passed two Project Sankofa contracts on consent, adopted a budget carrying the facility as “Ongoing” with its cost marked TBD, and approved a $700,000 increase to the design contract on 30 June 2026 — again on consent, the staff report recording “Alternative actions considered: None.” At each of those moments the operating cost was unknown, the operator unnamed, the service model undefined, and the state-required needs assessment undone.
Inheriting a project does not transfer the obligation to examine it. The commitments now being made are this Council’s, and the building will open on their watch.
One member sits differently, and not in the direction of lesser responsibility. Councilmember Bartlett authored the March 2020 item that directed staff to report the cost of ongoing operations, and is listed in the 2018 feasibility study as a member of the AAHRC Community Leadership Committee. Nobody on this Council has a longer or closer association with the project, and nobody is better placed to have noticed that the report he asked for never arrived. He was absent from both 2025 consent checkpoints.
This is not retrospective second-guessing. The project secured its Use Permit in December 2025 and is completing Design Development, with final design anticipated in Fall 2026. Construction has not been bid and ground has not been broken.
The arithmetic favors asking now. Roughly $1.6 million in design work is the sunk cost at stake, against $15 million in capital and an operating obligation nobody has quantified. A pause before construction forfeits design fees; a pause after forfeits far more. A building the City cannot afford to operate is the most expensive outcome available, because the obligation begins the day it opens and does not stop.
Berkeley funds buildings through one process and operates them through another. Capital money is appropriated, designed, and built. Operating money arrives later, in a different budget cycle, from a different pot, decided by people who were not in the room when the commitment was made.
Nothing in that sequence requires anyone to fill in a TBD before the doors open. Not the capital appropriations, not the service contracts, not the budget adoption. The question arrives when the building does — and by then the answer changes nothing, because the building exists and somebody has to run it.
The facility opens in April 2028. There is still time to answer the question before it does.