Hot Topics10 Ballot MeasuresU · V · Z
Seven local measures are on the November 3 ballot and four of them are taxes. I am voting no on three: the $300 million bond, the half-cent sales tax, and the parcel tax for a public bank. I am voting yes on three, and I am taking no position on the arts parcel tax. I do measurement and verification for a living — thirty years of answering one question for building owners, which is whether the money delivered what was promised. Every one of these three fails that question in a different way, and the failures are in the city’s own documents rather than anybody’s campaign literature.
Last updated 21 September 2026 · Every figure below links to a public document in Look it up.
No ballot measure appears on either September agenda. One item on the September 29 agenda speaks directly to the first ask on this page.
On September 29, Mayor Ishii — with Councilmembers Taplin, Tregub and O’Keefe — asks the City Manager to develop an updated cross-departmental Program Plan and Strategic Asset Management Plan under Realize Vision 2050, and to report progress to Council every six months. The estimated cost is $200,000 in staff and consultant time. Ask 01 on this page asks for that plan, with the ranked project list, before the bond. On September 29 the work is being commissioned. Ballots are due November 3.
Council agenda, 29 Sep 2026, Item 26, Council consentThe referral gives the number as approximately $1.5 billion. The 2020 Framework counted $882 million. The 2022 Program Plan counted $1.802 billion. The Task Force report of December 2025 said about $2 billion. Four figures in six years, each counted for a different purpose. The bond asks for $300 million against a need the city has stated four different ways.
Item 26 referral, p.1 · Vision 2050 Framework, 2020 · Program Plan, 26 Jul 2022Item 26 asks staff to stand up a transparency dashboard tracking infrastructure project schedules and budgets and carrying future capital planning, modeled on the capital improvement program hub that UC Berkeley students built in the Transportation Planning Studio course. Publishing the score is the condition this page attaches to a yes on U. It now exists as a sentence in a referral that has not yet been voted.
Council agenda, 29 Sep 2026, Item 26, Council consentSeptember 15, Item 4 approves $25,822,884 in bid solicitations and requests for proposals in one motion. September 29, Item 6 approves $17,585,000 the same way. That is $43.4 million of purchasing authority across two consent votes, from various funds, with the itemized list in an attachment to the staff report. Neither figure is on a ballot. Both are worth seeing beside a $300 million ask.
Council agenda, 15 Sep 2026, Item 4 · Council agenda, 29 Sep 2026, Item 6The September 15 agenda was posted on September 3 and is final. The September 29 agenda was posted on September 17 and is final; the item numbers here follow it. Sources: the September 15 Council agenda, the September 29 Council agenda and the Agenda & Rules packet of September 14, 2026, which carried the September 29 draft.
The council item that started Measure U, on December 2, 2025, lists 31 potential projects totaling $307,317,742 — 19 in Parks, Recreation & Waterfront, 8 in Public Works, 4 in Fire. It does not prioritize them. It commits to none. It says Council will “provide input and direction on project priorities” and that staff will “develop a community survey to aid decision making.” Ten months later the survey has happened, the bond is on the ballot, and the prioritized list still does not exist. You are being asked for $300 million against a menu.
Council Special Item 02, 2 December 2025The same item notes that general obligation bonds “May be used only for capital improvements, not for ongoing operational costs.” It then says nothing about what the operating costs will be, or who pays them. This is the failure I care most about, because Berkeley has already shown it knows better: the 2010 pools measure was a $22.5 million bond plus an operations tax, both prices on the same ballot, decided together. Voters were told the whole cost. Measure U tells them half of it, and the other half lands in a General Fund already carrying a structural deficit.
Council Special Item 02, 2 December 2025 · Measure C (2010) ballot textMeasure V raises the rate to 10.75% starting April 1, 2027, and in the City Attorney’s words lasts “until ended by the voters.” It is a general tax, which is a term of art with consequences: the money carries no spending obligation, so there is no promise on the record to check the result against. You cannot verify a commitment nobody made. That is a design choice made before the measure reached the ballot, and it is the reason no reporting standard can be bolted onto this one after the fact.
City Attorney’s impartial analysis, November 2026 measuresMeasure Z taxes improvements at $0.06 per square foot residential and $0.09 non-residential — about $9,200,000 in year one, $58,300,000 over six years — for a bank that does not have a charter and needs approval from the California Department of Financial Protection and Innovation. The measure then provides that if a public bank cannot secure authorization to conduct business by June 30, 2033, alternative lending activities may proceed. Read that again: the measure has a plan for the money if the bank never opens. The tax also rises annually with inflation, and I can find no provision letting Council pause that increase.
City Attorney’s impartial analysis, November 2026 measuresNot my characterization — the county Tax Collector’s. In his own words: “Berkeley taxpayers have 29 of these types of taxes, more than any other Alameda County taxpayer.” In the most recent year he found the average Berkeley bill up about 13%, with special assessments alone up about 20%, against roughly 2% on the county portion and about 4.5% on ad valorem taxes overall — because Proposition 13 caps that side at 2% and does not reach this one. The schools measure rose 15.72% in a single year at $0.54 per square foot; the landscaping tax 20%, at 26.5 cents. The fastest-growing part of the bill sits outside the protection most people believe they have, and four more tax measures are on this ballot.
Alameda County Treasurer-Tax Collector, “City of Berkeley Taxes: Why the Increase?”I want to be exact about what this is not. It is not an anti-tax position: I am voting yes on three measures on this same ballot, one of which is a tax. It is not a claim that Berkeley’s needs are exaggerated — the city’s own auditor puts unfunded capital and deferred maintenance at $1.8 billion, which is worse than most residents think. And it is not an accusation against staff, who wrote what they were asked to write.
It is a claim about instruments. A bond that lists thirty-one projects and commits to none, with no operating budget for what it builds. A general tax that makes no promise and never ends. A parcel tax for an institution that has no charter and a written contingency for what happens to the money if it never gets one. In my working life, if a client brought me any of these three and asked whether the savings were real, the answer would be that the question cannot be answered as written. I am not willing to vote yes on something I could not later verify.
Berkeley requires measurement and verification before it pays an energy contractor — measured outcomes, against a defensible baseline, under international protocols. The city applies that standard to a $2 million retrofit. It does not apply it to a $300 million bond.
Berkeley already has a Program Plan — the Council adopted it on July 26, 2022, with $1.802 billion in needs and six targets. What it does not have is a year on any of those targets, or the updated plan the Mayor’s Task Force set for spring 2026, which I have not been able to locate. Publish the update — the asset baseline, the sequence, the funding path for the rest — alongside the thirty-one projects ranked in order, and a $300 million bond stops being a menu and becomes step one of something. I would vote for that bond and say so in writing. I set this out in full on the Vision 2050 page, where my answer is yes to the plan: date the targets and publish the score.
Every capital measure states what the finished thing costs to operate, and where that money comes from. Measure C did this in 2010 and Berkeley has not done it since. A building the city cannot afford to staff is not an asset; it is a liability with a ribbon on it.
Not “street repair.” Lane-miles brought to a stated pavement condition index at a stated cost, by a stated date. A purpose is what a measure is for. A boundary is what it promises. Voters can only hold the second one to account.
Measures U and Z both already require an audit at least every three years and an annual report. I want to be fair about that: it is real, and it is more than nothing. But an audit confirms the money went where it was supposed to go. It does not say whether the streets got smoother or the loans got made. Add one column — what we said we would achieve, what we achieved — and the reports become worth reading.
Twenty-nine assessments rise automatically whether or not they deliver, and Measure Z would add a thirtieth that rises with inflation with no stated way for Council to pause it. Tie the escalator to the annual report: miss the target, the increase waits until the next report shows it met. This is ordinary practice in performance contracting and it is the highest-leverage line on this page.
The structural deficit is real — the City Auditor projects $32 million in FY2027 and $33 million in FY2028 — and voting no does not make it disappear. I am not pretending otherwise. But the escalator asymmetry on the existing bill is a larger revenue story than half a cent of sales tax, and nobody is working on it. That is work I would do.
What did it promise, in the ballot language? What would success look like, as a number? What do we observe, from independent data? What are the escalator and the sunset? Every full text is linked in Look it up — they are shorter than you would expect.
Two-thirds threshold, about $22.14 per $100,000 of assessed value, roughly $15.2 million a year, a projected forty-year term. The purposes are real and I support them: fire stations, sidewalks, seismic work. The instrument asks for a two-thirds vote against an unranked menu of thirty-one projects, with no operating budget for what gets built and no dated plan showing where it sits in the larger job. Berkeley has run this experiment recently: Measure L in November 2022, a $650 million infrastructure and housing bond, took 59.4% and failed against the same two-thirds bar. Publish the plan, date the targets and price the operations, and this becomes a yes.
Majority threshold, 10.75% from April 1, 2027, “until ended by the voters,” $9–10 million a year into the General Fund. A general tax carries no spending obligation, so verification is structurally impossible — there is no promise to check. It is regressive, it is permanent, and it lands on people who cannot move their spending out of town. The deficit it addresses is real; this instrument still cannot be held to anything.
$0.06 per square foot residential, $0.09 non-residential, about $9.2 million in year one and $58.3 million over six years, rising annually with inflation. The bank has no charter and requires approval from the state Department of Financial Protection and Innovation. The measure provides that if authorization is not secured by June 30, 2033, alternative lending activities may proceed — a contingency for spending the money if the bank never opens. I support the purpose. I have said for months that this one needed published capitalization milestones, a governance structure, and a date by which it lends or the tax lapses. It arrived without them, and with an escalator instead.
Among other things it cuts the council filing fee to $25 and raises the public financing administration budget. It lowers the barrier to running for office. This campaign is the argument for it.
District 1 is majority-renter and roughly 45% of renters here are cost-burdened. Pair it with a Rent Board that publishes enforcement outcomes annually — the same standard, applied to a body I am asking voters to strengthen.
The arts are a real public good and the institutions face real risk — I am not arguing otherwise, and I am not asking anyone to vote it down. My position is about sequence. I do not think a new parcel tax is the best use of limited tax resources until we get our budget lined out, and Berkeley’s budget is not lined out: a $32 million structural deficit in FY2027, $33 million in FY2028, $1.8 billion in unfunded capital and deferred maintenance, and twenty-nine assessments already on the bill, which rose about 20% in the most recent year against the 2% Prop 13 allows on the rest of it. Ask me again when the city can show what it already collects is working, and I will have an answer instead of a reservation.
The measure on this ballot with the strongest evidence base behind it. Berkeley’s original soda tax is one of the few local policies anywhere that has been studied properly and shown to work. When a measure can show its results, I vote for it. That is the whole standard, applied in the direction fewer people find convenient to notice.
Affordable Berkeley is the organized opposition to these measures and its research on the bank tax is the most useful public account of that measure’s history. I am linking it below. I am not adopting its slate, and because I am asking you to trust my numbers, here is where I part from theirs.
Affordable Berkeley urges no on the arts tax and no on the soda tax. I am voting yes on the soda tax — we disagree about the strongest-evidenced measure on the ballot, which is not a small disagreement. On the arts tax I am neutral rather than opposed, which is a different thing from agreeing with them. A shared position on three measures is not a slate, and I would rather say that plainly than let a link imply otherwise.
Their Measure U page says Council identified 35 potential projects. The December 2, 2025 council item lists 31, totaling $307,317,742. Their underlying point is right — nothing was prioritized and nothing was committed — and the correct number makes it no weaker.
Their Measure V page describes an increase “by 5%.” Measure V raises the rate by half a percentage point, from 10.25% to 10.75%. As a share of the existing rate that is about 4.9%, so the sentence is defensible arithmetic and misleading English. The honest version — a permanent general tax that promises nothing — does not need the help.
They cite a $29.5 million structural deficit for FY2028, which is the budget book’s number. The City Auditor’s April 2026 report puts it at $32 million in FY2027 and $33 million in FY2028. The larger figure is the one that argues against my own position most strongly, and it is the one I use.
That the bank tax was designed as a four-jurisdiction split with Berkeley’s share at about $2.5 million; that every other city parcel tax lets Council suspend its automatic increase; and that Measure U duplicates sidewalk funding already promised by Measure FF in 2024. All three would strengthen the case if true. I have not found the documents, so they appear here as their claims rather than mine, and if you can point me to the sources I will publish them.
I changed my position, and you should know when. Through August my written position on all seven measures was the same: support the purpose, and ask that the measure carry a baseline, a target, an annual report and a sunset. In September I moved the bond and the sales tax to no, and then the bank tax as well. What changed is that the conditions did not arrive. The ballot language is now fixed and carries no outcome target; the prioritized project list is still unpublished and the spring 2026 Program Plan update has not appeared; a general tax cannot be conditioned at all; and the bank tax arrived with an escalator and a 2033 contingency instead of capitalization milestones. I would rather publish the change than have somebody find it. The old position is in writing, the new one is here, and you can judge the reasoning.
Total repayment on the bond is a projection, not a city figure. Affordable Berkeley puts the total cost of Measure U at $550–600 million at current interest rates. That is plausible for $300 million over forty years, and I have not verified it. What the city has published is principal ($300 million), annual cost (about $22.14 per $100,000 of assessed value), annual collection (about $15.2 million) and term (projected forty years). If you want the real number, the debt service schedule exists in the Finance Department and somebody should ask for it — I intend to.
Measures U and Z do carry accountability provisions. Both require an audit by the City Auditor at least every three years and an annual report from the City Manager. My objection is to what those reports measure, not that they are missing, and I would rather state that precisely than let an impression stand that these measures have nothing attached.
The polling figure is press reporting. The bond was reported at 64% support in April 2026 against a 66.7% passage threshold, from a city poll by Lake Research Partners. I have not read the instrument or the crosstabs. I mention it because a measure polling below its own bar has every reason to earn trust it has not yet earned, but I am not resting the position on it. The harder fact is Measure L: 59.4% in November 2022, against the same two-thirds threshold, and it failed.
Two figures came out of an earlier draft of this page. It carried a dollar-weighted average escalator of 8.51% across the City and school assessments, and a crossover point below which square-footage taxes exceed the property tax itself. Both were my own arithmetic rather than anything published, and the second rested on an assumed ad valorem rate. I took them out rather than ask you to trust a number I could not hand you a document for. The county Tax Collector’s published figures above make the same point and you can check them.
I have not read every word of all seven full texts. I have read U, V and Z, the impartial analyses for all seven, and the summaries. If something in a full text contradicts a characterization above, that is exactly the correction this page exists to publish.
Every claim on this page comes from one of these. If a link breaks, the document title and date are enough to find it on berkeleyca.gov.
If a number above is off, or a document says something different from what I've quoted, I want to know. Accepted and rejected corrections both get published here, with my response.