Berkeley chose the size of Measure U before it finished choosing the projects. When $300 million polled no worse than $200 million, Councilmembers treated that as a reason to take the higher amount.
A sound capital plan starts with the work. Name the projects. Rank them. Price them. Count grants and other available resources. Calculate the remaining gap. Then decide how much to borrow.
Berkeley’s record shows a different sequence.
The City had a large infrastructure backlog and plenty of candidate projects. But the record does not show those projects and resources producing a bottom-up calculation that happened to equal $300 million. Instead, $300 million became the working envelope first. Staff tested whether voters would accept it, Council saw no polling penalty compared with $200 million, and the project list continued to be refined afterward.
The process began much more narrowly.
In June 2025, Councilmember Terry Taplin and colleagues asked staff to examine a revenue measure for Fire Department facilities. The referral contemplated polling for a bond that would “exclusively pay for remodeling or replacement of fire facilities.”
By March 2026, fire had become one of three broad categories:
The City’s March 17 presentation explained that projects were being evaluated under six criteria, then added an important constraint:
The projects were being sorted inside a $300 million envelope. The completed project list did not first produce the $300 million figure.
June 17, 2025 Fire Facility Revenue Measures
March 17, 2026 presentation and survey materials
“What’s the… how far can we take this?”
The February survey did not ask voters to choose between a $200 million bond and a $300 million bond.
It used a split sample. One group heard a broad $300 million bond description. Another group heard a broad $200 million version. Both included fire and emergency facilities, parks and recreation, climate resilience, public buildings, and other infrastructure.
The results were nearly identical:
| Version tested | Estimated annual rate per $100,000 assessed value | Support |
|---|---|---|
| $300 million | $22.14 | 71% |
| $200 million | $18.47 | 70% |
Pollster David Mermin told Council the difference was within the margin of error: there was essentially no measurable difference in support.
That finding had a narrow meaning. The poll could not detect a support penalty for the larger amount at the tax rates presented.
It did not show that voters preferred $300 million. No voter was asked to make that choice.
The survey did ask voters which projects they cared about. That information could help prioritize a capital program. But Council did not take the highest-ranked projects, price a priority package, subtract existing resources, and let that calculation determine the bond amount.
For the amount decision, the striking result was simpler: $300 million did not poll worse than $200 million.
The March 17 discussion makes the decision rule unusually visible.
O’Keefe first asked how the proposed bond and sales tax, taken together, would affect Berkeley taxpayers compared with neighboring cities.
Staff replied that it had not completed that analysis.
Staff did provide pieces of the picture: the proposed bond rate, an example based on a median assessed value, and comparisons with nearby sales-tax rates. But Council did not yet have a completed cumulative household-burden analysis.
The conversation then returned to the poll.
O’Keefe asked:
That question gets to the heart of the process. Once $300 million appeared to carry no electoral penalty, the next question became whether the City could go higher.
O’Keefe then asked how staff had arrived at $300 million in the first place.
The answer should stop any reader who thinks the number emerged from a capital-financing formula.
“$300 million was… It felt like a lot.”
“$22 per $100,000 of assessed value, that seems kind of a reasonable amount.”
“There’s… it’s not like a science.”
There is a discipline for sizing long-term debt.
Define the work. Price it. Identify grants and existing resources. Calculate the financing gap. Match the debt term to the useful life of the assets. Then ask voters to approve the result.
“It felt like a lot” and “that seems kind of a reasonable amount” describe judgment about political and taxpayer tolerance. They do not explain why the City needed exactly $300 million of new borrowing.
That matters because Measure U creates a tax burden extending for decades. The final financing assumptions contemplate multiple bond issuances over time, with repayment stretching roughly forty years from the first issuance to the last scheduled payments.
June 16, 2026 final Measure U staff report
Councilmember Rashi Kesarwani put the logic plainly:
Councilmember Mark Humbert independently made the same move.
“I would be in favor of the 300 rather than the 200, since there’s not a statistical significance as between them.”
“No statistically significant difference” means the poll did not establish a meaningful difference between the two support levels.
It does not provide a financial reason to borrow the larger amount.
The preference for $300 million was a policy judgment Councilmembers added to the polling result.
Councilmembers did raise affordability.
O’Keefe asked about the combined tax burden. Mayor Adena Ishii later asked staff to explain what residents would actually see on their property-tax bills after adding the new bond to existing bond charges.
The answers were still provisional.
At the same time, Berkeley was carrying other recently approved or already-authorized obligations. Measure FF had just added a new parcel tax. Council was considering a new sales tax. And the final $20 million tranche of the previous Measure T1 infrastructure bond had not yet been issued; Council would authorize that final T1 series on June 30, 2026, with delivery anticipated in August.
June 30, 2026 final T1 bond issuance
The record gives far more developed attention to how much bonding capacity the City could use than to a single cumulative household calculation showing how the new obligation sat on top of the others.
The chronology after March is especially telling.
On March 17, staff asked Council for direction to prepare ballot language for a $300 million bond.
By May 19, the agenda asked Council to do two things:
The order is right there in the City’s own agenda: affirm the amount, then keep working on the projects.
On June 16, Council placed the $300 million measure on the ballot.
Berkeley has more infrastructure need than $300 million could ever satisfy. That makes prioritization more important, not less.
A capital plan should be able to answer one basic question:
The public record shows something else: the City found an amount voters appeared willing to tolerate, chose the larger amount when it polled no worse, and continued refining what the money would buy.
Measure U Is T1 Again · The Crayon Box · The Berkeley Budget Framework · Berkeley Decision Project home