Housing
What does it actually take to turn a housing approval into a building?
San Francisco approved 56,000 housing units that have never secured construction permits. Nobody in city government can tell you how many are stalled, how many belong to megaprojects on 20-year timelines, and how many are financially dead. The city built an elaborate system for deciding whether to allow housing. It built almost nothing for actually delivering it.
District 8 approved 88% of housing applications — well above the citywide 79% — and processed them 17% faster. The approval bottleneck isn't here. The construction bottleneck is. And a supervisor controls roughly 15–25% of what determines whether housing gets built: fees, permitting timelines, land use oversight. That's not nothing. It's also not everything. This platform focuses on the 15–25% where a supervisor can make a measurable difference.
The Approval Isn't the Problem
Here's the number that should reframe every housing conversation in this city: 56,000 approved units have never received construction permitsThe permit that authorizes actual building work. Distinct from a planning approval, which only says a project is allowed. A project can be "approved" for years without ever getting a construction permit.. That figure includes megaprojects with 20-year build-outs, smaller projects waiting for financing, and some that are functionally dead. The mix matters. But the fact that no public system distinguishes between them is the finding. San Francisco tracks whether a project gets approved. It does not reliably track what happens next.
In D8, the pattern is smaller-scale but the same. 257 projects producing 960 net units — an average of 3.7 units per project. This isn't a megaproject district. It's an incremental-housing district: backyard cottages, in-law units, three-story additions on commercial corridors. That shapes which policy tools actually work here, and it's why generic "build more housing" platforms don't translate. D8 needs a permitting system that works at the scale this district actually builds.
Where the Pipeline Narrows
Citywide pipeline. The bottleneck isn't approval — it's everything after. Source: SF Planning, DataSF.
Of the permits D8 issued between 2018 and 2022, 42.9% extended beyond five years without completion. Of the 15 incomplete, only 2 were formally cancelled. The other 13 show zero DBIDepartment of Building Inspection — the SF agency that issues construction permits and inspects building work. inspection activity in over a year. Median dormancy among those 13: 29.3 months. As far as anyone can tell, nobody's been keeping track.
Ezra Klein and Derek Thompson call this pattern "chosen scarcity": the policy decisions and institutional failures that produce artificial shortages even when abundance is technically possible. San Francisco has the land, the demand, and now the zoning. What it lacks is the administrative infrastructure to convert a permit into a set of keys.
Same District, 4x the Wait
D8's median permit processing time is 487 days. The citywide median is 587. That's 17% faster — credit where it's due. But the district-level number hides something more interesting.
Permit Processing by Corridor
Median days from application to permit issuance. Same city, same rules, a 4x gap. Source: DataSF permit records.
Dolores Street: 85-day median. Diamond Street: 349 days. Same city. Same regulations. Same DBIDepartment of Building Inspection — the SF agency that issues construction permits and inspects building work.. A fourfold gap within a single district. That variation is the interesting finding, because it means the problem isn't a uniform policy failure. Something about specific corridors, specific project types, or specific review pathways produces wildly different timelines. A pipeline tracker that makes corridor-level data public would expose those patterns — and that exposure is the first step toward fixing them.
The 2023 permit surge tells a related story. Applications jumped from roughly 150 to 3,397, coinciding with Family Zoning Plan-related activity. That's application volume, not construction starts — an important distinction. The surge suggests developers respond to regulatory signals. The question is whether the processing infrastructure can absorb the demand when macroeconomic conditions eventually improve and applications rise again.
A Supervisor Controls 15–25%. Here's How to Use It.
A supervisor controls roughly 15–25% of what determines whether housing gets built. That's fees, permitting timelines, land use oversight, and budget votes on affordable housing funding. What's not in scope: construction costs ($500–$800 per square foot in San Francisco), interest rates, land prices, and lender risk appetite. Most candidates don't say this. I think you should hear it, because knowing the limits is how you focus on what actually works.
San Francisco already ran the experiment. In 2023, the city cut housing fees by a third and slashed affordable housing requirements from 22% to 12%. Applications still didn't rebound. D8 permit applications dropped roughly 80% from their 2018–2022 average. If regulation were the binding constraint, reducing it should have helped. It didn't. The macroeconomic barriers — interest rates, construction costs, lender reluctance toward San Francisco projects — are real, and they're outside any supervisor's reach.
That makes the tools a supervisor can reach more important, not less. Enforceable permitting timelines. A public pipeline tracker so residents can see where projects stall. Affordable site diversification so the district isn't one delay away from losing 74% of its affordable production. A visitability standard that costs $100–$600 per unit now instead of $10,000 later. These are the levers within the 15–25%. This platform uses every one of them.
One Project, 74% of the Plan
1939 Market Street is 187 units of LGBTQ+ senior housing, the most significant affordable development in D8's pipeline. A $47.6 million state AHSCAffordable Housing and Sustainable Communities — a California program that funds affordable housing near transit corridors. grant awarded in December 2025 resolved the primary funding risk. The project still needs roughly $100 million more through 4% LIHTCLow-Income Housing Tax Credit — a federal program giving tax incentives to developers who build affordable housing. The "4%" version typically finances about 30% of a project's cost. tax credits, with an application planned for May 2026. Construction is projected to begin late 2026 or early 2027, with completion expected in 2029. This project is a genuine achievement worth defending.
D8's Affordable Housing Pipeline
Dedicated 100% affordable units in D8 pipeline. One project delay = nearly all affordable production stalls. Source: MOHCD pipeline data.
It's also 74% of all dedicated 100% affordable housing in D8's pipeline. The remaining 66 units are spread across 11 other projects. That concentration isn't an accident — it reflects a district that has historically produced very little affordable housing and relies on a single large project to close the gap. But concentration is risk. One delay, one financing complication, one LIHTC cycle that doesn't go D8's way, and nearly all of this district's affordable production stalls.
1939 Market is also where housing meets LGBTQ+ preservation. At $3,800/month average rent in the Castro, affordability determines whether the community that built this neighborhood can still live in it. This project isn't just housing policy. It's the district's primary investment in LGBTQ+ senior housing. Those are the same conversation, and the backup site strategy acknowledges that both depend on the same timeline.
Every New Home Should Work at 35 or 75
A visitability standard means zero-step entries, wider doorways, and one accessible bathroom on the ground floor. The cost during new construction: $100–$600 per unit. The cost to retrofit later: $10,000 or more. That's a 20:1 penalty for doing it in the wrong order.
The Cost of Doing It in the Wrong Order
Sources: National Association of Home Builders; HUD visitability cost studies. Multiple jurisdictions have adopted comparable standards.
The Family Zoning Plan created capacity for roughly 36,200 new units. Every one built without visitability features is a unit the city pays 20 times as much to fix when the resident ages, is injured, or needs a wheelchair. D8's 65-to-74 population is 63% of all D8 seniors — the bubble that hits 75–84 in the next decade. And 77% of San Francisco's housing was built before 1950: stairs, narrow doorways, no elevators. New construction is the only realistic place to get this right at scale.
Caregiver ADUsAccessory Dwelling Units — secondary homes on single-family lots. In D8, usually a converted garage or backyard cottage. are the second piece. D8's typical project averages 3.7 units, and the ADU goes in the Victorian garage — exactly where accessibility is most needed. One permit creates a new housing unit, a caregiving solution, and an accessible ground-level home. When housing production is also aging-in-place infrastructure, the cost math shifts: $500 per person per year for community-based senior services versus $159,000 for a nursing home. The 318:1 ratio makes visitability the single highest-return investment in the housing pipeline.
78.1% of D8's housing pipeline sits within a quarter-mile of MUNI transit. Housing and transit are the same policy in this district. But that connection only works if the transit works — and the $307 million SFMTASan Francisco Municipal Transportation Agency — runs Muni buses, light rail, streetcars, and cable cars. structural deficit threatens the routes that make density livable. Routes 35, 36, and 37 serve Diamond Heights, where the housing pipeline is thinnest and the senior population is highest. Cutting those routes doesn't just affect transit. It undermines the rationale for building housing there at all.
What I'll Do
Publish a Housing Pipeline Tracker
Within my first month, I'll begin building a public dashboard tracking every permitted project's status, queue time, and blocking issue. The 15 incomplete permits with zero recent inspection activity become visible — and accountable. No new laws. Existing data. Day one.
Set Enforceable Permitting Timelines
Replace generic 'cut red tape' with size-specific targets: ADUs toward 180 days, small projects toward 150 days, large projects toward 365 days. When deadlines pass, automatic escalation — not another email chain. The Dolores-to-Diamond gap proves the system can move faster. The question is whether anyone requires it to.
Build Every New Home to Work at 35 or 75
A visitability standard — zero-step entries, wider doorways, one accessible bathroom — adds $100–$600 per unit during construction. Retrofitting later costs $10,000 or more. That's a 20:1 cost penalty for doing it in the wrong order. Every unit built under the Family Zoning Plan's new capacity without this standard is a unit the city pays to fix later.
Build a Second Affordable Site in D8
1939 Market Street is 74% of D8's dedicated affordable housing pipeline. 187 units of LGBTQ+ senior housing, de-risked by a $47.6M state grant, expected 2029. It matters enormously. It's also a single point of failure. One delay stalls nearly all of D8's affordable production. I'll begin a site identification process so the district isn't one setback away from zero.
Fast-Track Caregiver ADUs
One permit, three problems solved: a new housing unit, a caregiving solution, and an accessible ground-level home. D8's typical project is 3.7 units — this district's housing future is backyard cottages and in-law units, not towers. Pre-approved designs, reduced fees, expedited review. The ADU goes in the Victorian garage, exactly where accessibility is most needed.
The Zoning Worked. Now Comes the Hard Part.
The Family Zoning Plan is the most significant housing reform San Francisco has adopted in decades. It passed the Board 7-4 in December 2025, met the January 2026 state deadline, and reversed a 1978 downzoning that had frozen more than half the city into low-rise residential. It created capacity for roughly 36,200 additional units. The Chief Economist projects realistic production of 8,500–14,600 over 20 years — meaningful, but a fraction of the 82,069 the state requires by 2031.
It faces lawsuits from both sides. It's still a genuine achievement.
But it passed during a hostile construction market: interest rates elevated, construction costs among the highest in the world, lenders reluctant to finance San Francisco projects. D8 permitted only 751 units over the past six years despite holding about 10% of the city's population. The Lurie administration's PermitSFMayor Lurie's permit reform initiative, launched February 2025. Aims to consolidate and speed up the city's building permit process. initiative aims to streamline the post-approval process, and early legislative wins on ADU rules, fee reductions, and over-the-counter permits are real. What's still missing is enforceable timeline targets, a public pipeline tracker so residents can see where projects stall, and financing tools that survive when federal support shrinks.
That last point is the one everyone's avoiding. The Trump administration proposed a 44% cut to HUDU.S. Department of Housing and Urban Development — the federal agency that funds affordable housing, homelessness programs, and rental vouchers.. Roughly 1,000 San Francisco households face loss of Emergency Housing VouchersFederal rent subsidies that help people exiting homelessness afford stable housing. Funded through HUD.. The state's January budget contained zero funding for affordable housing. The $10 billion state housing bond on the November 2026 ballot is a lifeline for the entire affordable development model. Without it — and with federal dollars disappearing — projects like 1939 Market face compounding risk. The question isn't whether San Francisco should build local financing infrastructure. It's whether the city starts before the money runs out.
The plan created possibility. Whether that becomes housing depends on whether the city treats permitting, financing, and affordable site strategy with the same urgency it brought to the rezoning itself. That's the work ahead.
Other Issue Areas
Sources
SF Chief Economist Ted Egan, Family Zoning Plan economic analysis (October 2025). SF Planning Department, PRO Housing Application (56,123 approved units). DataSF permit records, DBI inspection activity logs. RHNA allocation: HCD 6th Cycle (82,069 units). BLA budget analysis. MOHCD affordable housing pipeline data. AHSC grant award, 1939 Market Street ($47.6M, December 2025). Turner & Townsend International Construction Market Survey (2022). Klein & Thompson, Abundance (2025). Full citations in deep dives.
Go deeper
Explore the full Research Desk
Every position Darshini takes is backed by evidence. See the full list.