Berkeley consults the public often. But a survey can only measure the choices it offers, and by the time most of these reach residents the contours of the question have already been settled in the room where it happened. What comes back is not a mandate. It is a preference among variations of a decision already made.
This is the companion to The Crayon Box, one level out. That page is about the options a council will consider. This one is about the options a resident is permitted to express — and about what happens when the resulting percentage is later cited as public support.
Public participation is only as meaningful as the choices participants are allowed to make. When every option points in roughly the same direction, a consultation stops measuring what the public wants and starts measuring which version of the City’s preferred approach people dislike least.
Could a reasonable participant reject the premise, keep the status quo, end the current activity, or introduce a materially different alternative — and could that choice genuinely affect what happens next? If not, the process is not exploring public preference. It is collecting reactions inside a frame chosen in advance.
Public engagement is most valuable at three stages: defining the problem, generating alternatives, and choosing among them. If the first two happen internally, the third risks becoming ratification rather than consultation.
A complete frame for a new proposal lets a resident say start it, do not start it, or address this a different way. For something the city already does, it lets them say continue, change it, stop, or again another approach entirely. Those last options are the ones that tend to be absent, and they are the ones the Start, Stop, Continue discipline exists to force.
The recurring forms are easy to recognize once named. The premise is treated as settled, so residents are asked which design or which eligibility rule rather than whether the program is justified. Every option differs in degree but not in kind — a larger program, a smaller one, a phased one, but never no program; several sites, but never “do not build it.” The preferred option carries the value-laden language while alternatives are described only by their mechanics. Independent decisions get bundled, so one answer has to cover policy, location, funding, and staffing at once. And open-ended responses are collected but not analyzed with the seriousness given to the prepared menu — acknowledged as comment, absent from the comparative table.
The 2025–2030 Consolidated Plan sets Berkeley’s priorities for federal housing and community development funding for five years. Staff described the public process to Council:
“In 2024, staff released a survey via the city community message platform to request response to the identified priorities for the HUD funding… 77% of the 590 respondents strongly agreed or agreed with the priority areas you see on the screen.”
Staff presentation to Council, April 29, 2025
Read the construction rather than the number. The priorities were identified first. The survey requested a response to them. Respondents registered agreement or disagreement with a list they had no hand in composing, and 77% of 590 people agreed.
That figure is accurate and it is not evidence of what it will be taken to mean. It does not show that these were the right priorities, or that residents preferred them to priorities nobody proposed. It shows that when a city asks whether its own stated priorities sound reasonable, most people who answer say yes. The reported result establishes broad agreement with the priorities presented; it cannot establish preference over priorities that were never offered.
The Council adopted the plan unanimously. Two speakers.
The Consolidated Plan survey was not an isolated instrument. It went out through Berkeley Considers, the City’s standing public-engagement platform — an Open Town Hall forum operated by OpenGov on Berkeley’s behalf, promoted through recurring “Announcements from the City of Berkeley” emails and posted on the City’s website. It is how Berkeley routinely asks residents what they think.
The City is careful to say the platform is a civic-engagement tool rather than scientific polling. That disclaimer addresses self-selection, which is a real limitation. It does not address the one this page is about: whether the instrument lets a respondent express the choice that actually matters.
Read the City’s own solicitations and a pattern emerges. Each one names a specific, near-term decision the input will feed:
| What residents were asked | What the City said it would do with the answer |
|---|---|
| Telegraph Avenue mobility Jun 2025 — how to improve safety for pedestrians and cyclists while improving bus speed | “Develop a concept design for City Council approval later this year” |
| Bike share siting May 2025 — weigh in on eleven proposed docking locations | “The finalized locations will begin installation this summer” |
| Fire hazard severity zones Mar 2025 — whether to adopt or expand state-defined zones | Used “as the City weighs whether to expand these boundaries at an April 15 City Council meeting” |
| Federal housing priorities Dec 2024 — “tell us which of these programs you see as the funding priorities” | Informs five years of federal housing and public-service spending |
| Parking requirements Nov 2020 — “this survey will ask you to select among options being considered” | “Your response will help shape future development in Berkeley” |
These are consequential decisions and the City is being transparent about where the input goes. But look at what is being asked in each case. How to improve Telegraph, not whether the corridor is the right place to spend the money. Which eleven locations, not whether the stations are wanted. Whether to adopt or expand the fire zones — a menu with no third position, though in that instance state law genuinely removes one, since cities may not shrink the boundaries. Which of these programs are the priorities. And, most explicitly of all: select among options being considered.
There is one more thing the solicitations share. Every one of them promises the input will “inform,” “guide,” “help develop” or “help shape” the outcome. None states how responses are weighted, tallied, or reconciled when they conflict. The one exception is instructive: for three mosaic proposals at the Willard Park clubhouse, the City spelled out the full chain — respondents to an Artist Review Panel, to the Public Art Subcommittee, to the Civic Arts Commission for a final decision in December 2024. The most complete account of how public input travels to a decision was written for the artwork in a park clubhouse.
The Telegraph survey drew 314 registered responses, 186 with text in the open comment field. Read individually and coded by hand, five of those 186 — 2.7% — challenged the survey’s framing rather than arguing within it. They did so in four distinct ways:
“Your survey never mentioned convenience/efficiency of car travel as a possible priority.”
“This survey fails by not providing responders with the choice of DO NOTHING, ‘no project.’”
“Please fix the imbalance between the concept designs… so the public can evaluate & compare them fairly.”
“I am perplexed by your inclusion of access for people with disabilities as something people can vote on. The ADA requires that you take this into account.”
Four of the five framing challenges, Telegraph Multimodal Project survey, June–July 2025
A missing priority category. A missing no-build option. Options not comparable on their face. And a legal requirement presented as a matter of public preference. The fifth declined to engage with the concepts at all until a prior question was answered, on the view that the outcome was already settled.
Two more flagged the project’s scope boundary without contesting the survey itself. The remaining 179 argued the substance — more bike protection, more bus priority, safer crossings. That is the important feature of the result: almost everyone who engaged with the open-text field responded inside the frame they were given. When you hand people three designs, almost everyone tells you which design they prefer. The frame is nearly invisible from inside it, which is precisely why a consultation cannot be treated as evidence of support for the premise.
Notice, too, who a survey framed this way is likely to reach. The invitation went out under the subject line “Help improve mobility by foot, bike, or bus on Telegraph.” The three modes are named in the ask; driving is not among them. A resident who cares about walking, cycling or transit reads that and recognizes their cause. A resident who drives Telegraph, or who thinks the money belongs elsewhere, has been told before opening the survey that their concern is not what this is about — which is what one respondent said back: “Your survey never mentioned convenience/efficiency of car travel as a possible priority.”
An instrument built that way does not sample a community. It can disproportionately summon the constituency already engaged with the modes named in the invitation. The people most motivated to answer are the ones already committed to the direction the survey assumes, and the resulting percentage then travels as though it were the community’s view. This is the mechanism that makes self-selection worse than a sampling problem: the framing decides which self selects.
Berkeley is capable of the better version, which is why this is a habit rather than an incapacity. In the Corridors zoning consultation, the survey asked residents whether to update or retain the existing zoning on each corridor — a genuine question about whether to act at all, not merely how. It drew 1,644 responses, and staff reported the uncomfortable part to Council: overall a slight majority supported updating, while residents living on or near the corridors leaned toward keeping the existing zoning.
That is a consultation doing its job. It offered a real choice, and it surfaced a split the City then had to explain rather than a percentage it could cite. The contrast with the Consolidated Plan survey is the whole argument of this page: Berkeley has demonstrated that it knows how to ask the premise question.
The danger begins when a constrained survey leaves the instrument that produced it. A staff report can accurately report a percentage while shedding the boundaries of the question; a later retelling can shed them further. Eventually a number that measured agreement with a prepared list can appear to carry the weight of a mandate.
The defense against that is not to consult less. It is to make the instrument capable of returning an inconvenient answer — and to keep the alternatives residents propose in the record, in the comparative analysis and the staff report, with a stated reason where one is rejected. An alternative that can be acknowledged without ever being evaluated has not been included. It has been contained.
Define the problem without embedding the solution. Ask the strategic question — start, do not start, continue, change, stop, another approach — before asking about implementation. Provide a real route for alternatives the City has not thought of, and carry the credible ones into the comparative analysis with a stated reason if they are dropped. Describe the tradeoffs of every option symmetrically, including the preferred one. Separate decisions that a reasonable person could answer differently.
That sequence preserves the difference between deciding what to do and deciding how to do it — and it is the only sequence under which a reported percentage means what a reader will assume it means.