⌂ BERKELEY CITY COUNCIL SCORECARDS
Independent analysis. Informed voters. Better government.
HUMBERT
District 8  ·  Berkeley City Council
2024-12-10 – 2026-07-28 · 61 meetings
Voter Alignment
D
as of 2026-07-28
Performance Pillars
Fiscal Stewardship 55%
Votes, bills & rhetoric on fiscal discipline
D
25%
Civic Focus 25%
Share of speech on core city business
A−
82%
Legislative Skill 10%
Can this member move the needle?
B−
60%
Character 10%
Collegiality, humility & restraint
D+
33%
Why this grade — D

This council votes together: 86% of recorded votes were unanimous, and the two least-aligned members still agree 91% of the time. Comparing votes alone will not tell them apart. What follows is how this member got there.

Asks the right question, then votes past the answer. Humbert is attentive, highly present, and generally focused on municipal business. On two of the period's most consequential decisions his engagement exposed the central problem without producing a decision responsive to it.

On Hopkins he asked Fire Chief Sprague directly whether the newly proposed configuration had been reviewed. The Chief answered that it had not, that no complete design set yet existed, and that review would take two to four weeks once one did. Humbert then voted against both substitutes and for the original recommendation, without correcting its claim that Fire Department review supported the configuration and without waiting for that review.

On the FY2027-28 budget he moved the successful motion, including the $200,000 revenue transfer that restored nine fire positions. That was a real act of prioritization, and it demonstrated that the proposed reductions were not immutable. The same budget eliminated 39 authorized police and dispatch positions and the last 2.6 city-funded crisis-response positions, against the City's own written warnings about 9-1-1 answering capacity. Restoring all of them would have cost far more than $200,000, and nothing here claims otherwise. What is missing is the comparative standard: no part of the motion or the discussion around it explains why those acknowledged risks did not warrant the same reconsideration fire received.

The problem is not inattention. He can identify the decisive question and build a motion capable of changing an outcome. The grade is absolute rather than curved, and it measures what happened next: the answer he elicited, and the tool he had just demonstrated, did not govern the choice he made.

What Humbert does well
What went wrong

This is built from the public record and it is not complete. Absence of a finding here means we have not documented one, not that none exists. Corrections and additions are welcome.

Material governance failures

These are decisions with consequences the city will carry for years. How a member showed up on these matters more than any number of routine votes.

How a governance failure is defined and tested →

Incident Record Full incident log →
-6
Hopkins Item 27 — Fire Review Claim Not Supported by Its Own Source
Character & Conduct · Fiscal Stewardship
→
-8
FY2027–2031 CIP — Infrastructure Backlog Reaches $2.1B With No Structural Fix
Fiscal Stewardship
→
-5
May 19 Budget Hearing — Revenue-Only Lockstep, No Structural Fix
Fiscal Stewardship
→
What the Record Shows
Objective measures from official records and transcript attribution — no weighting or judgment applied.
Attendance — 74 Sessions (Annotated Agenda Record)Official record
Source: post-meeting annotated agenda PDFs — the authoritative attendance record. "Fully absent" = listed absent at roll call and never arrived.
73/74
Sessions present
(attended at some point)
1
Sessions fully absent
(never arrived)
0
Sessions late
(absent at roll, arrived later)
1
Sessions absent
at roll call
Attendance rate
99%
On-time rate
99%
Major Fiscal Votes — 7 Binding DecisionsOfficial record
Curated list of binding fiscal votes Dec 2024–Jul 2025. Source: annotated agenda PDFs.
0/7
Binding fiscal votes
missed
7
Endorsed status quo
(voted YES)
Major Fiscal Votes — 7 Binding Decisionsshow detail
Date Type Item Amount Position
2025-01-21TEFRA BondCMFA Bond – 2001 Ashby Ave$45.0MYES
2025-05-20Revenue BondLease Revenue Notes – Fire HQ$11.0MYES
2025-05-20GO Bond$35M GO Bonds – Measure O Housing$35.0MYES
2025-05-20Budget AmendmentFY2025 Budget Amendment 1st Reading$85.7MYES
2025-06-03Budget AmendmentFY2025 Budget Amendment 2nd Reading$144.0MYES
2025-06-24Budget AdoptionFY2026 Budget Adoption 1st Reading$1.45BYES
2025-07-08Budget AdoptionFY2026 Budget Adoption 2nd Reading$1.52BYES
YES = endorsed the spending authorization (budget adoptions = chose not to cut or reprioritize). ABSENT = missed a binding vote — items are only on the agenda because the Mayor expects passage.
Voting Record — 307 Items (Annotated Agenda)Official record
Source: annotated agenda PDFs across all 307 items with a parseable vote. Council block-vote rate: 92% — individual deviations are high-signal precisely because they are rare.
6
NO votes
(dissent from bloc)
1
Abstentions
(chose not to vote)
2
Absent during vote
(item-level absences)
1
Contested abstentions
(others voted no)
Agenda Impact & EfficiencyOfficial record
5
Discretionary items authored
(from the annotated agenda record)
2
Staff Direction Volume
(40–80 hrs each, no opp-cost review)
30w
Avg words per turn
Rank #7 of 8
2%
Turns >200 words
(monologue territory)
Spending Votes — Roll-Call RecordOfficial record
Dollar amounts from agenda items matched to extracted roll-calls. Coverage is partial: unanimous voice votes and some formats not captured.
$7115.2M
Voted YES on
(53 spending votes)
$8K
Voted NO on
(dissent from block)
81%
YES rate on
spending items
Largest YES vote: $1717.4M  ·  2026-07-07
FY 2027 Annual Appropriations Ordinance
Speech Content — What Was SaidText analysis
81%
Core agenda share
(budget, infrastructure, safety, housing)
17%
Non-core speech share
(foreign policy, ideological resolutions)
5
Fiscal concern mentions
(0.9 per 10k words)
1
New revenue mentions
(0.2 per 10k words; tax/bond advocacy)
Counts are decay-weighted (recent meetings count more than older ones). Source: attributed transcript speech.
Fiscal Stewardship Assessment
The following sections apply this scorecard's fiscal-stewardship framework to the descriptive metrics above. Readers who dispute the weights or philosophy can engage at that level without disputing the factual record above.
Homeless Services Status-Quo AlignmentText analysis
According to the City Manager’s Office, Berkeley spent $35M on homelessness-related activities in fiscal year 2024 (operating costs only; capital expenditures directed by the council add substantially to the total). Score measures investment in the prevailing homeless services apparatus vs. accountability and reform orientation. Lower is better. Based on attributed transcript speech; coverage partial. Full homeless spending analysis →
51
Moderate Status-Quo Alignment
0 = reform-oriented  ·  100 = status-quo aligned
Rhetoric: 3 status-quo-aligned signals (housing first, trauma-informed, unhoused neighbors…)   0 accountability signals (Grants Pass, cost-per-client, metrics…)
Cosponsored/authored 1 homeless-services agenda item(s).
Agenda Behavior — Consent & Action CalendarsOfficial record
Berkeley’s council agenda is divided into two calendars. The consent calendar is a list of items voted on together as a single block, with no individual debate — a routine mechanism for non-controversial matters. The action calendar contains items debated and voted on individually. Any councilmember can request that a consent item be moved to the action calendar, forcing a real discussion before a vote. This score tracks whether members use that power — and whether they sponsor items on the consent calendar that deserve more scrutiny than a block vote allows.
Consent Calendar
0
Off-mission items
buried in consent (authored)
3
Off-mission consent
items co-sponsored
0
False “None” fiscal
claims (authored/cospon.)
$7,000
Discretionary relinquishments
(5 items from council budget)
Action Calendar
1
Off-mission items
brought to action calendar
0
Off-mission action
items co-sponsored
Fiscal Consistency Flag — 5.46 fiscal-concern speeches vs. 1 non-core items on action calendar
5.46 fiscal-concern mentions; 1 off-mission items brought to action calendar
Fiscal Stewardship Alignment — Score BreakdownText analysis + official record
Final score +0.245
Fiscal Stewardship Alignment — Score Breakdownshow calculation
How the Fiscal Stewardship score is built. Each row below adds or subtracts from a running total. The result is kept within a 0–100% range because the score is a percentage-style rating: below zero or above 100% would be meaningless. Individual component values are shown in the table; the grade shown at the top of the scorecard reflects the final result.
Base — from voting record and agenda behavior
HSA alignment HSA score 50.6/100 → inverse 49% → quadratic → 24.4%  (75% weight) +0.567
Scope discipline Off-mission penalty 7.0% → discipline 93.0%  (25% weight) +0.232
Raw score +0.415
Adjustments — documented behaviors and penalties
Incident record3 incidents, tier-weighted, capped ±0.30-0.028
Audit silencestreets rocky road 2025-0.040
New revenue preference penaltyRevenue advocacy without companion reprioritization analysis—
Fiscal referral penaltyBond/tax campaign direction (capped −0.09)-0.050
Rhetoric penaltyFiscal concern rhetoric with no dissent votes (high HSA or serial fiscal-vote absence)—
Final score (kept within 0–100% range) +0.245
P1 speech share: 0.3% of words spoken (174 of 59,083 words) were in turns engaging with a documented structural failure (structural deficit, infrastructure backlog, pension liability, investment non-compliance, reserve policy). This is a diagnostic — not yet incorporated into the composite score.
Ownership of Outcome — 29%

Whether this member behaved as the decision-maker on consequential matters where they had standing to act. Losing an argument counts as ownership; declining to have it does not.

Took ownership
Declined to engage
Official record  Attendance, votes, and authorship drawn from annotated agenda PDFs and agenda item JSONs — primary public record, not subject to interpretation.    Text analysis  Rhetoric signals derived from keyword classification of attributed meeting transcripts, constituent communications, and member statements. Full methodology at METHODOLOGY.md.
Full Incident Record Incident log →