How each gap was closed, FY2023 to FY2028
Each bar is one year's gap, split by how it was closed. Lasting cuts barely appear until FY2027. In FY2026 about a third of the fix lasted, mostly revenue changes.
City of Berkeley · General Fund
Why the City's main budget comes up short, and what the November 3 ballot means.
Berkeley from the hills. Photo by Janet Ganbold on Unsplash
Berkeley's main budget comes up about $30 million short every year. One-time money and savings covered it for years, and most of the savings are spent. The FY2027-28 budget closes it by eliminating 150 jobs and counting on Measure V.
Every budget since at least FY2019 started $10M to $40M short. Until FY2027 the City closed it mostly with short-term fixes, so it came back.
Each bar is one year's gap, split by how it was closed. Lasting cuts barely appear until FY2027. In FY2026 about a third of the fix lasted, mostly revenue changes.
Spending ran $47M (FY2024) and $32M (FY2025) above regular revenue. Cash fell about $55M, because about $23M of one-time transfers came in from other City accounts.
Facts from the City Auditor and the budget, not a grade. Short-term fixes and savings carried the budget for years. The General Fund also pays $2.7M a year to parking funds and the 1947 Center St building, which should pay for themselves. Reserves and retiree health debt are the bright spots.
Top bar: Berkeley. Bottom bar: the goal or an earlier year. Bars start at zero.
Pay, pensions and health care are two-thirds of the General Fund.
Actual pay and benefits grew 8.3% in one year. The FY2026 budget jump comes from funding every job at full cost, plus raises and higher pension and health costs.
Two-thirds of the pension bill is old debt, set by CalPERS.
Police and Fire are about half the General Fund.
General government is the offices that run the City, from the City Manager and Finance to the Council and Auditor. Spending rose from $34.0M to $53.2M, FY2022 to FY2025, mostly from about 35 added positions and one-time spending in FY2025, not raises. The growth was typical for nearby cities; the level was not: $419 per resident against a median of $255.
Dashed line: median of the other nine cities.
Dashed line: median of the other seven cities. Berkeley's bar is from its own books, all funds, not counting IT, so it differs from the 57% above.
Revenue has grown, but costs have grown faster. Most newer voter-approved taxes can only pay for specific programs. Two general taxes, Measures P and U1 ($18M to $20M a year), go to homelessness and housing by Council choice.
Transfer taxes fell by half after FY2022. Transfers in from other City accounts, mostly one-time, swing from $29M to $4M.
About $100M a year of City special taxes can only pay for their stated purposes. Measures P and U1 ($18M to $20M) are general taxes the Council puts toward homelessness and housing. Letters repeat across elections, so each measure shows its year.
Most big Bay Area cities have gaps. Berkeley spends about what Oakland does per resident.
Berkeley's gap is on the high side: below Oakland, close to Palo Alto and San Francisco. Sunnyvale has none.
Berkeley spends about what Oakland does per resident. About $400 of it is the health department.
Berkeley spends about $1,000 more per resident than the typical city here, mostly on public and mental health, housing and homelessness, and ambulances. Most of these cities don't run their own health department or ambulances.
Of these ten cities, Berkeley ties Palo Alto for the most property tax per resident, collects the most in parcel and transfer taxes, and among the least in sales tax.
How Berkeley compares on officers and crime, what the two departments cost per resident, how they perform against the City's own targets, and why costs rose. Options are in Sections 6 and 7.
Berkeley has fewer officers per resident than Oakland, and clears (solves) 4 in 10 violent crimes to Oakland's 1 in 10.
Berkeley's public safety cost per resident is close to Oakland's; the difference is Berkeley's own ambulances. From FY2023 to FY2025 its police spending grew less than these cities' (5% against a median of 21%) and fire and ambulance grew more (31% against 23%), much of it the paramedic program paid by the 2020 Measure FF tax.
Police, dispatch and fire are where the City would cut if Measure V fails. Here is Berkeley against its own targets, and the one public 911 comparison. What the cuts would mean.
Top bar: Berkeley. Bottom bar: its own target, or Oakland for crimes solved. Bars start at zero. For times, shorter is better.
The most urgent calls have held at a 7-minute median since 2021. Lower-priority calls now wait about 50 minutes, up from 38. Meanwhile 911 calls rose from 12,400 to 15,300 a year, the most in seven years, and only 17 of the 911 center's 36 positions were filled in early 2025 (a wider count than the budget's 29 dispatchers).
Berkeley doesn't publish how fast it answers 911 calls. Its last public figure, from 2017, was 89% answered within 15 seconds, below the state's 95% standard. Oakland's auditor found 30% of Oakland's 911 calls waited more than a minute in 2023.
The 2023 pay contracts and higher pension debt payments explain most of the police increase and about three-fifths of fire's. The rest of fire's is 50 ambulance jobs, paid largely by the 2020 Measure FF tax, and overtime.
Top bar: the department now. Bottom bar: FY2022, the department's spending increase since FY2022, or the median of nearby cities. Bars start at zero.
The City has already balanced FY2027 and FY2028 with lasting cuts plus Measure V, on the November 3 ballot, and has named the cuts if V fails. Below: how FY2027 closes, other ways to cover the gap, and what happens if V fails.
$19.8M of lasting savings, $2.1M of one-time money, and $9.4M to hold 33 jobs (one-time money in FY2027, Measure V from FY2028). The City's published gap, $29.2M, comes after the one-time money.
| Positions by department | Cut, empty | Cut, filled | Held for V vote, empty | Held for V vote, filled |
|---|---|---|---|---|
| Health, Housing & Community Services | 37 | 8 | 0 | 0 |
| Police | 18 | 0 | 12 | 9 |
| Public Works | 13 | 2 | 0 | 0 |
| Fire | 2 | 3 | 0 | 9 |
| City Manager | 4 | 5 | 0 | 0 |
| Parks, Recreation & Waterfront | 1 | 2 | 3 | 0 |
| Finance | 4 | 0 | 0 | 0 |
| Information Technology | 4 | 0 | 0 | 0 |
| Human Resources | 2 | 0 | 0 | 0 |
| Total | 85 | 20 | 15 | 18 |
Still empty after these cuts: 128 jobs, most paid from other City accounts. About 28 are General Fund jobs outside police and fire; eliminating them would save about $5M a year, close to the $5.9M cost of the 18 filled jobs on hold for Measure V. What happens to the 33 jobs on hold if V fails.
The bottom two bars are the City's plan: $20.1M of lasting cuts (105 jobs, 85 already empty, plus other savings) and Measure V ($9.6M), which together cover the FY2028 gap. The bars above are alternatives the Council could use instead, or if V fails. All ten add up to about $24.5M.
With V, the FY2028 budget has $0.75M to spare; without it, it is $8.9M short, a $9.6M swing. The budget already names the cuts that would cover it. One-time money keeps all 33 jobs through FY2027; the cuts would come in the FY2028 budget update.
The three cuts, $9.7M, would cover the $8.9M gap.
Engine 4 answered 28 fires and 966 medical calls, including 17 cardiac arrests, in 2025; those calls would go to farther stations. The City says the closure "will significantly degrade response times" and that the effects "cannot be fully mitigated." Fire is already 59 seconds over its 8:30 target; the City hasn't estimated how much the closure would add.
Specialty units shrink or end: traffic, community services, the bicycle team and major detective work. The City says this means less proactive work, fewer investigations and less community engagement. Police beat their 8-minute target for top-priority calls today (7 minutes); the effect on response times and on solving crimes (42% of violent crimes, against Oakland's 10%) hasn't been estimated.
The 911 center has 20 of 32 authorized positions filled and relies on mandatory overtime. Cutting the open jobs stops hiring into them. The City says this "will slow 9-1-1 call-taking and impact reliability." Berkeley hasn't published its 911 answer times since 2017.
Pools may close some weekends, or one pool for the winter. Therapeutic recreation for up to 21 youth with special needs ends. Summer camp and afterschool programs each lose 36 slots, and community centers close some evenings and weekends.
Source: City of Berkeley, May 19, 2026 staff report (Council Item 21) and its Supplement 2 department narratives. Targets, response times and crimes solved: the scorecard in Section 5.
Only Measure V helps close the gap. The others cover repairs (U), elections (W), renters (X), the arts (Y), a public bank (Z) and health programs (AA), each with its own cost for a typical home. This page doesn't recommend a vote.
| Measure | What it would pay for or change | Cost to a typical home | Effect on the City's gap | Lasts |
|---|---|---|---|---|
| U: $300M bond (two-thirds vote) | Public buildings and infrastructure: emergency response facilities, parks, accessibility, seismic and climate upgrades. About 17% of the $1.8B repair backlog. | About $120 a year (property tax; about $190 in the peak year) | None | About 40 years of repayment |
| V: half-cent sales tax | Any City service. The FY2028 budget counts on it to keep Fire Station 4, 15 police officers, 6 dispatcher jobs and 3 recreation jobs. | About $100 a year for every $20,000 of taxable purchases in Berkeley | Closes about $9.6M a year of the gap | Until voters end it |
| W: City Charter changes | Election rules: candidate filing fee from $150 to $25; more public campaign financing money for administration ($250K to $400K); recall and initiative procedures. | None directly | Up to $150K a year more in costs | Permanent |
| X: rent law changes | Caps rent increases at 10% a year; tenants keep coverage if their unit later becomes exempt; tenant associations at 5 or more units; fee waivers for nonprofit subsidized housing. | None for most owners | None (Rent Board runs on its own fees) | Permanent |
| Y: arts parcel tax | Arts grants: 75% performing-arts organizations, 10% buildings and equipment, 5% individual artists, 5% festivals. | About $130 a year | None | 12 years |
| Z: public bank parcel tax | Money to start a public bank lending for affordable housing, green energy and small business. If no bank is licensed by mid-2033, the fund can lend directly. | About $115 a year, rising with inflation | None | 6 years |
| AA: soda tax 1¢ to 2¢ | Health, nutrition and drinking-water programs: at least 42.5% to community grants and 42.5% to Berkeley Unified. | 1¢ more per ounce of sugary drinks (12¢ on a 12-ounce can), if passed on to shoppers | Between −$1.15M and +$0.3M a year, depending on the City's soda grants | Until repealed |
Together, U, Y and Z would add about $370 a year to the property tax bill of a typical home. These are the November 3, 2026 measures; the W, X and Y in "What voters already pay for" are 2024 measures with the same letters.
The City's plan covers FY2028. After that the gap comes back if pay and benefits grow faster than revenue. These options answer the causes above: services Berkeley runs that most cities don't (Section 4), general government's growth and how fast pay grows (Section 2), and short-term budgeting (Section 1). Examples, not recommendations.
At 5% a year growth in pay and benefits, a $12M gap is back by FY2030.
Each changes the budget every year, not just once. Examples, not recommendations.
| Option | Per year | What would change | What isn't known |
|---|---|---|---|
| Change what the City runs | |||
| Bring general government closer to nearby cities | $2.2M per 5% | General government, the offices that run the City, costs $419 per resident against a median of $255 for nearby cities. It grew from $34.0M to $53.2M from FY2022 to FY2025, mostly about 35 added positions and one-time spending. A 5% cut means 10 to 12 fewer positions and slower permits, contracts and records. | Which offices would shrink, and how much of the difference is how cities book costs. |
| Use part of Measures P and U1 for other services | up to $18M | These general taxes go to homelessness and housing by Council choice. Moving money out shrinks those programs by the same amount. | Which programs would shrink. |
| Hand public and mental health to Alameda County | $1.3M to $6.3M | Berkeley is one of three California cities with its own health department; nearby, the county runs it. It costs the General Fund about $7.9M net. City-only programs would end or move: public health in FY2028 at the earliest, mental health around FY2030. Shelters and homeless outreach stay with the City. | Which programs the county keeps. |
| Hand ambulance transport to the County's contractor | likely $3M to $8M | Elsewhere in the county a contractor runs ambulances, paid only by patient fees. It would keep Berkeley's fees ($9.2M in FY2027), and the City would lose a Medi-Cal add-on only public providers get. Fire engines with paramedics would still answer medical calls first. Earliest start about FY2030; full range $2M more cost to $13M saved. | What is inside the $23.8M emergency medical budget; whether the paramedic tax ($4.15M) and Measure FF could pay for paramedics on engines; whether the County would take Berkeley in. |
| Send civilian responders to calls that don't need police | $0 to $2M net | About a fifth of officer call time goes to welfare checks, alarms, noise, parking and similar calls: the work of 4 to 6 officers ($1.6M to $2.5M). The County has run crisis response since 2025, so the City might add only 2 to 3 Community Service Officers: about $1M to $2M net, or $0 to $1M with its own crisis team. | Saves money only if officer jobs or overtime actually go down. |
| Slow pay growth | |||
| Hold pay growth 1 point lower | $2.1M, then $8.6M | Smaller raises for every group: $2.1M in FY2028, $8.6M a year after four years. At 2.4% growth the gap stays within about $2M a year through FY2032; at 5%, a $12M gap is back by FY2030 (chart above). Police and fire contracts are being negotiated now. | Effect on hiring and keeping staff, with 10% to 12% of City jobs already empty. |
| Budget rules other cities use | |||
| Plan 10 years ahead, and require the plan to balance | No direct saving | 20 yrsSunnyvale must balance 10 yrsPalo Alto, Albany, San Leandro 5 yrsOakland, Hayward, Richmond, El Cerrito 2 yrsBerkeley, Alameda | Whether a forecast alone changes decisions: Palo Alto and San Leandro plan 10 years ahead and still face gaps. How much of Sunnyvale's result comes from the rule. |
| Pay into the pension savings trust every year under set rules | Costs money each year | Palo Alto and Sunnyvale do this. Berkeley drew $3M from its trust in FY2025 and $6M in FY2026 to fill gaps, with no rule to refill it. | How large a yearly deposit would be. It adds to spending in the years it is made. |
Berkeley Unified is a separate government with its own budget. This page asks the same questions: how big is the gap and why, is the money managed carefully, what it buys, and what could balance the budget without hurting results.
About half is the main account the district can use as it chooses. Parcel taxes and grants can only pay for set purposes.
Lasting fixes against one-time money, by school year. One-time money covers a year, then the gap comes back. Most came from the retiree-health savings fund ($7.4M down to $2.2M); in 2026-27, from moving regular bills onto a one-time state grant. Hover a bar for details.
The district spends about $26,000 per student, but its flexible money was squeezed from several sides at once.
The third point above: in 2024-25 and 2025-26 the raise outran what the state added.
Facts from the district's own budget documents, not a grade.
Top bar: Berkeley Unified. Bottom bar: a minimum, an earlier year, or the other five districts. Bars start at zero.
Berkeley spends more per student than its neighbors and puts it into smaller classes. Results are high overall, but low-income and Black students trail districts that spend less.
Berkeley spends more per student than Albany, Alameda and San Leandro, mostly from parcel taxes: BSEP ($46.8M in 2026-27) for smaller classes, BERRA ($12.6M) for educator pay. That buys more teachers per student than any neighbor, at about state-average pay.
Top bar: Berkeley Unified. Bottom bar: the state or a nearby district. Bars start at zero. For absence, shorter is better.
More money doesn't line up with better results here. Albany and Alameda spend about $8,000 less per student than Berkeley and get more of their low-income students to the reading standard.
The district must find $7.0M of lasting fixes for 2027-28 and $7.5M by 2028-29. Its own list (board resolution 26-046) isn't posted online, so these are the options its documents, or our analysis, price. They fall into administration, enrollment, attendance, the costs that caused the gap, and smaller options. Examples, not recommendations.
Each bar adds one option to those above it. Priced options come to about $5.8M a year, so more would be needed. Hatched: one-time. Outlined: no price published yet.
Berkeley spends a larger share on administration than similar districts. This is the largest priced option.
Berkeley has 5.5 administrators per 1,000 students, the most of six nearby districts. The clearest outlier is central-office managers in non-teaching jobs, such as finance, HR, IT and facilities. Trimming administration to the typical share of spending would save about $2M to $3M a year.
Most of the main account comes from a state formula that pays for each student on each day they attend: about $13,755 a year. Berkeley Unified's own schools have about 4% fewer students than in 2014-15, and births are down about a quarter since 2008.
Berkeley Unified's own schools are down about 4% since 2014-15. California's public schools are down 8%, and its district-run schools 12%, as charter schools grew. The often-cited drop from 10,462 students includes two charter schools that closed in 2019. Births set kindergarten size five years later, and they are down about a quarter since 2008, so smaller classes are still coming.
Each line starts at 100 in 2014-15. Births since 2008: Alameda County down 23.5%, California down 27.1%, Berkeley-area ZIP codes down about 29%. Berkeley and California include TK, which grew as the state phased it in from 2022-23; grades 1-12 alone fell 12.0% in Berkeley (all public schools) and 9.7% statewide.
Each student who attends every day brings about $13,755 a year. The district estimates 138 more students would bring about $1.6M.
No country has raised births back to two children per woman. Childcare and paid leave have the best evidence of more children, not just earlier ones; one-time baby bonuses mostly move births earlier. Those need state or federal action and take five or more years to reach kindergarten. Within Berkeley's reach: the mix of new housing, staff housing and local preschool.
| Policy | Where | Measured effect | Evidence |
|---|---|---|---|
| Universal childcare | Norway, Germany | More children, not just earlier ones. Norway: about 0.5 to 0.7 more children per woman going from no coverage to 60%. Germany: about 0.12 more per 30-point rise in coverage. | Strong |
| Longer paid parental leave | Austria, Germany | Austria (1 to 2 years of leave): 15% more likely to have another child within 3 years. Germany: more births among university-educated women; no overall effect on second children. | Strong |
| Ongoing child allowances | Israel, France | Israel: about 8% more births, mostly in lower-income families. France spends 3.4% of GDP on families; its birth rate still fell to 1.56 children per woman in 2025. | Moderate |
| One-time baby bonuses | Spain, Quebec, Australia, Singapore | Spain: births up 3%, then down 5.5% when the bonus ended. Quebec (up to C$8,000): about 12% more. Singapore: little or no effect. Mostly moves births earlier. | Mixed |
| Large national packages | Hungary, South Korea, Japan | Hungary: 1.25 children per woman (2010) to 1.59 (2021), back to 1.31 (2025), mostly timing. South Korea: 0.72 (2023) to 0.80 (2025). Japan: a record low 1.15 (2024). | Weak |
| Lower housing costs | U.S., England | When home prices rise, renters have fewer children and owners more. England: renters have 4.9% fewer births per 10% price rise, about 1.3% fewer overall. | Strong |
| Family-sized and affordable homes | Alexandria (Va.), Vancouver, Toronto | In Alexandria, new market-rate apartments add 0.03 to 0.07 public-school students per home; income-restricted affordable homes add 0.60 to 0.76. Vancouver's and Toronto's family-unit rules haven't been evaluated. | Moderate; none for the rules |
| Universal pre-K or free childcare | Washington DC, New Mexico | DC (2008): mothers' work up about 10 points; public enrollment grew by 19,000 by 2019-20, alongside population growth and new charters. New Mexico (2025): too new to evaluate. | Moderate for jobs, weak for enrollment |
| School staff housing | Santa Clara Unified, San Francisco, Berkeley Unified (110 homes, about 2027) | Helps keep staff; no measured effect on enrollment. | Weak |
The state pays about $76 for each day a student attends. Each point of attendance, about 2 more days per student a year, is worth about $1.3M. Absence is highest for homeless students (49% chronically absent), students with disabilities (26%) and low-income students (24%).
Share of students missing 10% or more of school days. Berkeley's overall rate is among the lowest nearby, but its low-income and Black students miss more school than in Alameda, Castro Valley, Fremont, Palo Alto, Piedmont or San Leandro. That is where better attendance would have to come from.
Only a few approaches have been tested well.
| Approach | Evidence | What it did | In Berkeley now |
|---|---|---|---|
| Absence reports mailed or texted to families | Strong: randomized trials in many districts | Chronic absence down 10% to 15%, for about $6 per family. Generic reminders did nothing. | Already doing it: 6,685 families got personalized texts and mailers in 2025-26. |
| A mentor for each chronically absent student | Strong for grades 5-7 | In a Chicago trial, 3.4 fewer days missed (against 17). No effect for younger children. | Case management for homeless students; no mentor program for all chronically absent students. |
| School buses | Moderate: a national study, not a trial | Kindergartners who rode the bus: 12% chronically absent, against 14% who didn't. | A bus fee is on the district's options list. Since 2022 the state repays 60% of district home-to-school transportation costs. |
| Walking school bus: adults walk a group of children to school | Weak: program reports only | Reported drops in chronic absence, such as 15% over two years in Springfield, Mass., alongside other efforts. | None found. |
| Before-school care | Not tested on attendance | No trials found. | BEARS runs from 7 a.m. at five elementary schools, with income rules; LEARNS runs mornings at three. Neither has room for everyone: BEARS keeps a waitlist ranked by need, and LEARNS says demand exceeds space at several schools. |
The district names these two as the main causes of the gap (Section 1).
The district has named options for special education but hasn't priced them. Raises are a choice made in each contract.
$33.5M from the main fund in 2025-26, up from $25.5M in 2022-23
$1.6M a year for each 1% raise for all staff
Each is worth less than $0.6M a year.
The Rent Stabilization Board runs Berkeley's rent control and eviction rules. Voters created the program in 1980 and elect its nine commissioners. It has its own budget, paid mostly by a yearly fee on rental units.
Revenue beat spending in four of the last five years, by as much as $1.47M (FY2026, projected), because spending came in under budget, mostly from empty positions. That is how the reserve grew from $1.35M to a projected $4.63M. The FY2027 budget plans to spend about $1.05M more than it brings in, paid from that reserve.
Numbers above the bars (under the years on a phone, in millions): surplus (+) or deficit (−). Rent Board fund only. FY2021 isn't shown: we didn't find its actual figures. *FY2026 is projected; FY2027 is the adopted budget, not a result.
Actual spending rose from $3.9M in FY2015 to $7.7M in FY2025, and the FY2027 budget is $11.3M, including $2.0M of one-time items. Staff grew from 19.75 to 31 positions. The fee for a rent-controlled unit was $250 from 2018-19 to 2022-23 and is $397 for 2026-27.
Landlords pay a yearly fee for each registered unit, set each March to cover the next year's budget. Because the fee is set to match the budget, the Rent Board doesn't run deficits the way the City does: when its costs rise, the fee rises.
$397 a year per rent-controlled unit; $244 for other covered units
$250 to $397 in four years
Berkeley's fee for a rent-controlled unit is the highest of these ten programs. Fees pay for different things in each city, and some spread costs over far more units: San Francisco bills about 217,500 units, Berkeley about 28,000. Dividing each budget by its units, Berkeley spends about $337 a unit on recurring costs, San Francisco $66 and Oakland about $105.
Of the $11.3M budget, $9.25M is recurring, mostly staff. The other $2.0M is one-time; the largest piece is $750K for outside lawyers, including for a class-action suit over the fee on partially covered units. Spending has come in under budget every year since FY2022, so the uncommitted reserve grew to a projected $4.63M; the Board plans to use about $1M of it in FY2027.
The Rent Board budgets 31 positions for FY2027, up from 19.75 in FY2015; it had 31.5 in FY1986, when rents were also controlled between tenants. Counseling and registration are the largest groups. The work: about 40,400 unit records, about 4,000 new tenancies registered a year, counseling for over 10,000 people a year by the City's count, about 70 petitions and 100 mediations a year, eviction and Ellis Act tracking, and the City's Empty Homes Tax.
Positions by what they do, FY2026 (29 in all). FY2027 adds an office specialist and a three-year assistant planner, for 31. Berkeley has about one position per 900 registered units; Richmond about one per 1,200 and San Francisco about one per 4,350. Programs do different things, so this is a rough guide; the City Auditor's planned audit of the Rent Board may say whether staffing fits the workload.
About 28,100 units are registered and pay a fee: 20,965 fully covered (rent limits and eviction rules) and 7,153 partially covered (eviction rules only). Another 11,410 records are exempt, 7,108 of them because the owner lives there. Registration compliance rose from 56% in 2022-23 to 97% in 2025-26.
Since 1999, state law lets landlords set any rent when a unit turns over; Berkeley's limits apply only to increases during a tenancy. So rents depend mostly on when a tenancy began: the 2,448 tenancies from before 1999 have a median rent ceiling of $906 a month, while tenancies that began in 2025 started at a median of $2,495.
Petitions to the Rent Board peaked at 193 in 2017 and were 69 in 2025, while mediation requests rose to 104. Ellis Act withdrawals, where an owner leaves the rental business, removed 2 to 8 units a year from 2022 to 2025. Owners filed 8 owner move-in eviction notices from July 2022 to June 2025, and 233 eviction cases were filed in court in Berkeley in 2025.
Open questions in the Rent Board's numbers, and things that could change them.
Sources: Berkeley Rent Stabilization Board budget packets, fee hearings, statistical reports and minutes (2021-2026); City of Berkeley budget books (FY2016-17 to FY2027-28); peer cities' rent program fee pages and budgets. Details under each card's Notes and sources.
Sources: City of Berkeley FY2025-26 and FY2027-28 budget documents, FY2026 mid-year revenue report, City Auditor financial condition audit (Apr 2026), City Attorney impartial analyses (Nov 2026 ballot). Full list with links under Sources below.