Most of the corridor upzoning was decided in Sacramento. Berkeley is arguing about the rest as if it weren't.

Three layers decide what gets built on a Berkeley corridor

Read this bottom-up. The state sets floors the Council cannot go below. The region assigns the numbers. The City decides everything above the floor — and that is a shorter list than most people think.

How tall, corridor by corridor

“The density bonus can go higher than the height limit” is true but useless on its own. The question a resident actually has is: on Solano, North Shattuck or College, under the zoning the Council is about to adopt, how tall does a bonus project get? That is answerable, because the density bonus works off a number the plans state explicitly — the maximum residential density in units per acre. Height follows from density; it is not the other way round.

Central case: 1,050 sq ft of building per unit, 65% coverage, 11.2 ft floor-to-floor. Push to the unfavorable end of both — 1,200 sq ft per unit on a 60%-coverage plate, which is a big-unit mix on a shallow lot — and the 100% column becomes 13 stories on a North Shattuck opportunity site, 11 on Solano and 8 on College. Push the other way and it drops about two stories. That spread is the answer: an envelope, not a number.

Four things the table says

1. Finding 01

   The density cap governs, not the height cap. Berkeley writes both, and on every corridor the density cap is the binding one once a bonus applies. Arguing about 58 ft versus 68 ft while leaving 115 and 140 du/ac untouched is arguing about the wrong number.

2. Finding 02

   The proposed base heights are already known to be non-binding — by the City. Staff’s own May 2026 table publishes, next to a 68-ft North Shattuck base, an estimated 8–9 stories at a 50% bonus and 12 stories at 100%. Solano: 58-ft base, 10 stories at 100%. College: 48-ft base, 7–8 stories. The City is not hiding this. It is printing a height limit it expects to waive, in the same table.

3. Finding 03

   “At least 14 stories, and likely much higher” is not supported for College, Solano or North Shattuck. Two independent methods — the City’s feasibility model and the density arithmetic above — land at 6–12 stories at a full 100% bonus, and they disagree by about a story and a half. Fourteen does not appear anywhere in the range. The top of it is 13 stories, on a North Shattuck opportunity site at the unfavorable end of the sensitivity spread, and that is the ceiling of a tail case, not a corridor-wide expectation. SB 79 does not lift it either: its density floors on that stretch — 100–120 units per acre — sit below the 140 the Update proposes, so the local cap is the one that governs.

4. Finding 04

   2029 University is not a corridor template. That tower’s base density was roughly 281 units per acre on a 12,385 sq ft downtown lot — about twice the densest number the Corridors Zoning Update proposes anywhere. Same law, different input. Two parcels with the same height limit do not produce the same building, because the height limit was never what produced it — the density under it was.

What the rezoning is actually worth

Height is the argument everyone is having. It is not the question that decides whether this was worth doing. The right question is counterfactual: thirty years from now, what is different because the Council adopted this, compared with a Berkeley that did not? Not what the plan permits — what changes. Below is that model, run on the City’s own capacity numbers and its own budget.

The 1,814 figure is the City’s seven-story-everywhere scenario (Solano 617, North Shattuck 1,036, College 161), not the Opportunity Sites option the Planning Commission actually advanced, which models 1,518. The base case then assumes half of that capacity is completed within thirty years and that a fifth of it would have been built anyway under existing zoning and state law. That leaves 726 incremental units — about 30 a year across all three corridors.

What the model says, plainly

1. Finding 01

   Realization is the whole ballgame. Every output scales with the share actually completed, and nothing in the zoning controls that. Berkeley entitled 1,495 units in 2025, permitted 463 and completed 492. The corridors’ entire thirty-year capacity is under four years of citywide completions at that rate. A zoning map is a permission slip, not a construction schedule.

2. Finding 02

   The rent effect is real, and it is small. Even if every permitted unit gets built, modeled rents land about half a percent below where they would otherwise have been. That is relative to the counterfactual — rents do not fall by half a percent in cash terms; they rise slightly less than they would have. And it is a different quantity from the 1–2% and 5–7% declines in the research below: those are rents next to a new building, measured against trend. This is the citywide average, and it is small for an arithmetical reason — 726 units is 1.3% of Berkeley’s 55,031-unit stock, and at the model’s elasticity a 1.3% bigger stock moves citywide rent about a quarter of a percent. Both numbers can be true at once because they measure different things. Anyone promising corridor upzoning will make Berkeley affordable is not describing this arithmetic, and neither is anyone claiming it will make rents spike.

3. Finding 03

   The fiscal case is real, and it is small. About $2.4M a year at year 30 — roughly 2.6% of this year’s Police budget — and about $27M in present value over thirty years, against a city with $30.8B in assessed valuation. The present value of the renter benefit is about $29M. Both are worth having. Neither rescues a budget.

4. Finding 04

   The cost side is not zero. At half realization the model displaces roughly 50,900 sq ft of existing commercial space on the three corridors — the storefronts the Planning Commission spent its May motion worrying about. The service-cost side uses a marginal fraction of current per-capita Police, Fire, Public Works and Parks spending and does not price step-change capital needs. Those are the two places this model is most likely to be too generous.

Two different questions that get answered as one

Question 1 — does upzoning produce buildings? Sometimes, slowly, where demand is strong and parcels are underbuilt. The Terner Center's first-year count of SB 9 — California's statewide duplex-and-lot-split upzoning — found Berkeley received 5 unit applications and 1 lot-split application, none approved, through November 2022. Statewide the pattern was the same: entitlement without construction, because owner-occupancy rules, design limits and costs made projects not pencil. Yonah Freemark's 2023 review of roughly 30 studies puts the short-run land-value jump from upzoning at 3–23% and finds construction follows only when the capacity increase is large, the parcel is underbuilt, and the market is strong. Portland's 15-year parcel study found an average of eight years from upzoning to completion. Chicago's transit upzoning produced higher prices and no additional units in five years.

Question 2 — when a building opens, what happens next door? Here the evidence has converged, and much of it is Bay Area work. Kate Pennington (UC Berkeley economics) used San Francisco fires as a natural experiment for new construction, 2003–2017: rents within 500 meters fell 1.2–2.3%, and displacement risk for existing residents fell 17%. Asquith, Mast and Reed looked at 50-plus-unit buildings in 11 cities including San Francisco, Los Angeles, Seattle and Portland: rents within two blocks fell 5–7% relative to trend, and in-movers came from slightly lower-income neighborhoods than before. Zuk and Chapple at UC Berkeley's Urban Displacement Project found that at the regional scale both market-rate and subsidized production reduce displacement, with subsidized units having roughly double the effect — but at the block level in San Francisco, neither did. The City's own commissioned study (Anti-Eviction Mapping Project) reached the narrower conclusion that affordable production measurably lowers displacement risk and market-rate production's effect "is not statistically significant."

The exception that matters for D1. Where existing rents are far below the new building's, the "amenity effect" can win: a Minneapolis study found rents fell 3.2% near new buildings in high-income areas and rose 6.6% in low-income ones. The blocks west of San Pablo and along Adeline are Berkeley's version of that case. That is the argument for Ask 04, not an argument against building.

Sources: City of Berkeley Housing Element Annual Progress Report (March 2025); Building Berkeley pipeline tally from City permit data. Berkeley's total housing stock is 55,031 units; the City needs roughly 1,460 permits a year for the rest of the cycle. Staff's corridor capacity figures are ceilings, not forecasts.

Any honest demographic claim has to start with the University

Berkeley is a city of roughly 124,000 people at night and about 150,000 by day. UC Berkeley enrolled more than 46,000 students in fall 2025 — about 33,000 undergraduates — and houses roughly 9,700 of those undergraduates, 29%. A decade ago it was 22%, the lowest share in the UC system. The rest live in the city's rental stock, which is why any statement about who moves in and out of Berkeley neighborhoods that doesn't account for students is not a statement about Berkeley.

The long arc. Berkeley was home to more than 27,000 Black residents in 1970 — roughly a quarter of the city. By the 2020 Census the share was 8%, under 10,000 people. The 2020 Census counts Berkeley as 50% white (non-Hispanic), 20% Asian, 14% Hispanic or Latino, 8% Black, 8% multiracial or other. Most of that Black population loss predates the current building cycle and tracks South and West Berkeley — the same blocks the Anti-Eviction Mapping Project flagged in 2021–22: West Berkeley's BIPOC population fell 31% and its white population rose 27% between 2009 and 2019, while citywide the BIPOC share rose 13%; West Berkeley's median rent rose $1,814 a month against a citywide $600.

Two things follow. First, the displacement pressure in D1's western blocks is documented and predates the corridor plans; it is a reason to get the protections right, not a reason to pretend the state law isn't there. Second, short-run "neighborhood got whiter after upzoning" findings from other cities (notably New York in the early 2000s, which had no inclusionary requirement) do not transfer cleanly to a city where a third of the population is students and 20% of every market-rate project is deed-restricted. I'd rather measure it here than import it.

What each side says, and what the record supports

Both sides cite studies. Neither side has run the numbers on Berkeley's own corridors, with Berkeley's own Rent Board data, on the buildings Berkeley already approved. That is the gap, and it is a cheap one to close.

Don't take my word for it

Everything on this page is sourced to a public document or dataset. If you think I've read one of them wrong, tell me and I'll correct the page. A longer, position-free version of this research — with confidence ratings on every figure — is on Learn About Berkeley.

Raw data — go get it yourself

• City of Berkeley Open Data Portal — building permits, zoning permits, inspections, certificates of occupancy. The source for every pipeline count on this page.
• Berkeley permit lookup (Accela) — status of any individual planning or building application by address or case number.
• Berkeley parcel and zoning GIS viewer — every one of the city's 29,024 parcels with zoning, planning and building layers. One of the better municipal GIS systems in the country.
• Berkeley Municipal Code Title 23 (Zoning Ordinance) and the Official Zoning Map.
• Housing Element Sites Inventory tables — the parcel-by-parcel list the City told the state it would build on.
• Corridors Existing Conditions Report (March 2025) — parcels, businesses, rent-controlled units and ownership on each corridor.
• Housing Element Annual Progress Report (March 2025) — the 1,654-of-8,934 figures, by income band.
• California HCD Annual Progress Report data — the same permit data for every California city, downloadable.
• Berkeley Rent Stabilization Board — the rent registry behind any Berkeley-specific rent analysis; annual reports and data requests.
• Terner Center Policy Gauge dashboard — tracks California housing law implementation by city, including Berkeley.
• blockXblock / berkeley-housing-analysis (GitHub) and the master analysis notebook (Colab) — open-source code that pulls the permit portal and tracks each project from application to occupancy. Built by Building Berkeley (buildberkeley.online); anyone can run it.
• Building Berkeley permit pipeline site and Datasette SQL queries — the ~6,000-unit pipeline count and per-project status.
• Bay Area Census — Berkeley — 2020 Census population, race/ethnicity, age. data.census.gov for American Community Survey tenure, income and rent burden by tract.
• Urban Displacement Project maps (UC Berkeley) — displacement and gentrification typology by tract.

Berkeley primary documents

• San Pablo Avenue Specific Plan — Public Review Draft (Oct 2025) and Appendix E, Objective Design Standards.
• Planning Commission Item 10A, Corridors Zoning Update (May 6, 2026) — heights, densities, the 1,518-unit estimate, small-business measures.
• Council appeal packet, 2029 University Ave (Feb 23, 2026) and the staff presentation — the 256-ft approval in the C-DMU Outer Core (75-ft base): 100% density bonus, 3 concessions, 6 waivers including height.
• Housing Element and General Plan Annual Progress Reports (27 March 2026) — the 2025 funnel: 1,495 units entitled, 463 permitted, 492 completed, and 84% of projects over five units using the State Density Bonus.
• FY 2025–26 Proposed Biennial Budget (Police $89.849M, Fire $48.628M, Public Works $7.210M, Parks $9.720M) and the 2025–26 assessed valuation of $30.831B — the denominators in the thirty-year model.
• San Pablo Avenue Specific Plan, Final Draft — Planning Commission recommendation (June 3, 2026), Appendix C, Table C.1 — the current San Pablo standards: 7 stories / 75 ft / FAR 4.25, 150 du/ac residential-only and 145 du/ac mixed-use, against an existing C-W base of 3–4 stories and 40–50 ft. The same table estimates corridor buildout at 6,900 units without the density bonus and 7,000–8,000 with it.
• Design Review Committee staff report and plan set, 2109 Virginia St (May 15, 2025) — the calibration case: 22,342 sq ft lot, 113,000 sq ft, 110 units, 89 ft 4 in against a 35-ft limit (65 ft with a use permit), four waivers, and the plan-set note claiming base density from the General Plan because it exceeds the zoning.
• Berkeleyside, “City Council approves 8-story North Shattuck housing proposal” (Feb 2026) and SF YIMBY — 2109 Virginia St: 110 units, 18 below-market (9 very-low-income, 9 moderate), 8 stories and just over 89 ft on a C-NS parcel with a three-story base limit, approved under the State Density Bonus Law.
• Corridors Zoning Update project page — current status; the May 6, 2026 Planning Commission 7–2 vote and Councilmember Tregub's August 2026 position are from his District 4 constituent newsletter (not web-linkable; quoted as received).
• Planning Commission Item 10A, SB 79 and AB 893 (April 15, 2026) — the state minimums, affected stations, and the City's options.
• Planning Commission Item 10B, Objective Design Standards (April 15, 2026) and the 2021 Residential Objective Standards background report.
• North Berkeley BART Station Area Objective Design Standards (adopted Dec 12, 2023) — the one-tree-per-25-feet precedent.
• Anti-Eviction Mapping Project, Densifying Berkeley: Potential Impacts on Displacement and Equity (executive summary, March 2022) and the November 2021 Council presentation — the West Berkeley figures.
• Middle Housing Zoning; Ashby and North Berkeley BART planning; General Plan and Area Plans index.

Research — UC Berkeley and California first

• Pennington, Does Building New Housing Cause Displacement? The Supply and Demand Effects of Construction in San Francisco (UC Berkeley, 2021).
• Zuk & Chapple, Housing Production, Filtering and Displacement (UC Berkeley Urban Displacement Project, 2016).
• Terner Center, SB 9 Turns One (2023) — the Berkeley application counts; Terner Center, Making It Pencil (2023 update) — the feasibility method behind Ask 02.
• Asquith, Mast & Reed, Supply Shock Versus Demand Shock (Review of Economics and Statistics, 2023) — 11 cities including SF, LA, Seattle, Portland.
• UCLA Lewis Center research roundup on market-rate development (Phillips, Manville & Lens, 2021) and the Planetizen summary.
• Freemark, Zoning Change literature review (Journal of Planning Literature, 2023) and Upzoning Chicago (2020). Effect-size compilation (Barr) for Li and Damiano & Frenier. Out-of-state; cited for the land-value and lag findings only.

Population and the University

• UC Berkeley News, housing milestone (March 2026) — 9,700 beds, 29%, the 2027–28 projects; Daily Cal, fall 2025 enrollment.
• Daily Californian, Black exodus from Berkeley (2022) — the 1970 figure; Bay Area Census — 2020 shares.
• Building Berkeley — housing stock, pipeline, daytime population and per-unit cost figures. Independent civic-data site; I've checked its RHNA numbers against the City's APR and they match.