Transit

The Books Are Closed

SFMTA stopped publishing route-level on-time performance data in December 2018. No mayor, supervisor, or SFMTA Board member has demanded it back. Seven years of silence — and counting.

The J-Church ran on time 43.5% in 2019. By 2023, it had climbed to 63.5%, a genuine recovery that riders could feel. Then it collapsed. The most recent twelve-month measurement puts the J-Church at 37.8% on-time. A 26-point decline from that peak, and worse than where it started. The improvement and the collapse were both invisible to anyone who doesn't manually dig into the data. SFMTA didn't announce the recovery. It didn't flag the decline. It just stopped telling anyone.

Here's why the silence matters more than any single number. SFMTA shifted roughly 70% of routes to headway-based measurement, which tracks whether vehicles are evenly spaced rather than whether they arrive when the schedule says they will. Under headway metrics, the system looks like it's running at about 90%. Under the schedule-based metric that San Francisco's own Charter requires, it's roughly 57%. Both numbers are real. One makes the agency look like it's nearly meeting its 85% standard. The other shows a system failing that standard by 28 points. SFMTA reports the first one.

Nobody knows how fast Muni actually runs. The most commonly cited figure, 8.1 mph, comes from a 2007 GPS study. Nineteen years without an update. That matters because speed is the single most direct driver of operating cost: if a route that should take 45 minutes takes 60, three hours of operator time produces three runs instead of four. SFMTA estimates that existing transit speed improvements save $5–10 million annually. Speed is the only transit investment that pays for itself, and the agency doesn't measure the thing that would prove it.

The 22 Fillmore recovered 124% of its pre-pandemic ridership, the strongest comeback of any D8 route. Its on-time rate dropped. A route that got more popular got less reliable. That's the kind of story route-level data tells. Nobody's been telling it.

The Evidence Is Already Here

Transit speed investment works. San Francisco has the proof from its own corridors.

The 38R Geary's quick-build transit lanes cut travel time by 20% and improved reliability by 37%. Bus collisions on the corridor dropped by two-thirds. The Van Ness BRT, after a decade of construction and a $346 million investment, runs up to 36% faster with travel time variability down 45% — and corridor ridership has recovered to 103% of pre-pandemic levels. The 14 Mission moves 31% faster through downtown transit lanes, and the 14R has reached 106% of pre-pandemic ridership. The 49 Van Ness/Mission is running at 138–140%.

The pattern holds beyond San Francisco. New York launched congestion pricing in its central business district, and the first-year evaluation was definitive: 27 million fewer vehicles entering the zone, morning rush commute speeds up 23%, traffic fatalities down 40% inside the zone and 19% citywide — the fewest in New York's recorded history. Revenue exceeded the $500 million target. A federal court upheld the program in March 2026. Every city that has tested congestion pricing saw public opposition flip to support within one year. Stockholm held a referendum after its trial; two-thirds had opposed the idea, but 53% voted to keep it.

These results raise a question the transit data blackout makes it impossible to answer: what's actually working on Muni's corridors, and where should the next investment go? SFMTA doesn't publish the route-level data that would tell us. The agency has the evidence that transit investment works. It just hasn't built the measurement system that would prove it to the public.

The Fiscal Cliff Has a Date

SFMTA's $307 million annual structural deficit begins in July 2026, when the last of the federal COVID relief funding runs out. The agency has already cut $133 million internally from an original $440 million gap. A $590 million emergency state loan buys roughly one more year of runway. The remaining shortfall cannot be closed through efficiency alone, and by 2030, the projected deficit grows to $434 million.

Two ballot measures on the November 2026 ballot offer the pathway. SB 63, the Connect Bay Area regional sales tax, would provide roughly $148 million annually to SFMTA. The Stronger Muni for All parcel tax would generate $160–187 million per year over fifteen years. Most single-family homeowners would pay $129 annually, scaled progressively for larger and commercial properties. Combined, both measures would produce roughly $331 million for SFMTA, a $24 million surplus over the immediate deficit. That cushion erodes within one to two years at 4% annual cost growth. It's a viable pathway with a thin margin, not a permanent fix.

If either measure fails, SFMTA has published scenarios that include eliminating 20 or more routes, ending regular service at 9 PM, suspending cable car operations, and removing crossing guards at 97 schools. BART's parallel doomsday scenario includes closing 10–15 stations and potentially shutting down entirely by winter 2028.

For D8, the consequences are concentrated on its hillsides. Routes 35, 36, and 37 carry roughly 2,300 daily riders across Diamond Heights and Twin Peaks. Route 35 Eureka has recovered just 40.5% of its pre-pandemic ridership — 462 riders a day. Route 36 Teresita: 688 riders, 50.3% recovery. Route 37 Corbett: 1,154 riders, 52.9%. All three serve a neighborhood where 23.6% of residents are over 65 and the Transportation Vulnerability Index is the highest in D8. When hillside routes disappear, residents who can't reach a grocery store, a pharmacy, or a doctor's appointment aren't just stuck — they're cut off.

The system as a whole remains roughly 25% below 2019 absolute ridership, even as individual routes show strong relative recovery. A route at 124% of its pre-pandemic baseline is a success story within a system that hasn't recovered. That distinction matters when evaluating where cuts fall and who bears them.

The transit fiscal cliff ripples across every other issue in D8. Seventy-eight percent of the district's housing pipeline is within a quarter-mile of Muni service; transit cuts undermine the rationale for transit-oriented development. A 9 PM service cutoff hits nightlife, service workers, and the restaurant economy along Castro, Church, and 24th Street corridors. Outreach workers depend on transit to reach clients. Diamond Heights seniors losing route access compounds the isolation that already makes aging in place harder in D8's hillside neighborhoods.