Patrick Linehan was backing out of his driveway near City College on the evening of October 19 when he saw a driver make a left turn at the corner of Hearst and Forester streets. The car slammed into a bicyclist.
The rider managed to jump off his bike, Linehan recalls, but the driver continued, dragging the bike six feet before stopping. Linehan measured the scar on the pavement.
Sunnyside Elementary School is at that intersection, with crowded morning drop-offs and afternoon pickups. Linehan often sees annoyed drivers and hurried parents speeding, running stop signs, and driving around the pickup line. “The most dangerous thing for my kids and others is cars,” Linehan says. “We should slow people down.”
Hearst Avenue is one of SF’s 18 official Slow Streets, where drivers are expected to proceed cautiously and create space for pedestrians, bikes, and others. The idea began during the pandemic lockdown to create breathing room for residents. Later, Slow Streets gave more people confidence to ride bikes on city streets, some for the first time.
Now these handful of streets are the lynchpin of an expanded citywide network to entice more San Franciscans to ditch cars for bikes and other transport modes. In the draft plan for the new network, SF’s Municipal Transportation Agency underscored that the 16 miles of Slow Streets help make SF one of the most bikeable cities in the nation.
City officials want even more people to leave cars behind in a not-too-distant future where the population could approach 1 million. (SF is supposed to plan for more than 80,000 new homes by 2031.) “We don’t have the capacity for that many cars,” said Christy Osorio at a November public meeting.
Osorio, who is overseeing the bike plan rollout for the San Francisco Municipal Transportation Agency, introduced the first draft of the plan at the meeting.
SFMTA faces challenges on more than one front. Anywhere drivers feel constrained, or merchants and neighbors feel threatened, street changes have required ballot measures (Golden Gate Park, the Great Highway) or experimentation (Valencia Street), or have been watered down (West Portal and Lake Street) or discarded (Chinatown).
Meanwhile, a steady diet of street improvements haven’t done much to reduce the city’s injury and death totals. A 2014 pledge to eliminate traffic deaths by this year has failed. Through Nov. 24, this has already been the second-worst year of the so-called Vision Zero era with 33 traffic fatalities, 22 of them pedestrians.
In the same ten years, injury collisions increased from 2014 to 2019, before dropping the following year due to COVID. They began rising again in 2021.
To work well, the new plan must convince people that biking is safe. As SFMTA’s “bicycle conditions index” analysis shows, residents don’t feel comfortable throughout much of the city.
As the index map shows, most of the highly-rated streets (green and dark green) are in western, quasi-suburban neighborhoods.
Slow Streets are demonstrating that less traffic and lower speeds can make a big difference in safety, and more could be on the way. Some of the streets are better than others, however, and The Frisc has done its own analysis of all of SF’s Slow Streets to see how well they’ve worked. There are plenty of lessons to learn from the data that, if heeded, could shape the future of our city.
‘Low tech,’ big results
The Frisc started with the city’s injury collision database, which gathers police and Department of Public Health reports.
We looked at six years of incidents for all 18 Slow Streets, from 2018 through 2023. Most became slow in 2020 — albeit in different months. Three were added in 2021, and when the city formalized the program in 2022, Cayuga Avenue and 22nd Street were included.
Distance also varies. Most “slow” stretches are less than one mile, but a few stretch over 1.5 miles or more.
To account for these variables, we measured injury collisions per month — as well as per mile when comparing the streets in total. We broke them down into two groups: before a street got a slow designation, and after.
The picture that emerges is clear: Slower is safer. Injury collisions on SF’s 18 Slow Streets decreased 45 percent compared with their pre-slow states. This is demonstrably better than the citywide rate. Across SF, injury collisions in 2022-23 were down 14 percent from the full two years before.
“This is a huge success story. Low-tech traffic interventions have led to better street safety,” says Marcel Moran, faculty fellow at New York University’s Center for Urban Science + Progress, who reviewed The Frisc’s data.
Slow but not equal
The 18 Slow Streets share similarities — a mix of residential, often with patches of commercial areas, along with schools, parks, and community centers. To create Slow Streets, SFMTA has tried to choose environments conducive to meeting its traffic target averages: fewer than 1,000 vehicles a day driving at less than 15 mph.
That said, developing a Slow Street is as much art as engineering with the different lengths, neighborhood considerations, and histories. Some streets have required —and received — constant adjustments.
For example, Page Street, one of the longest of the 18 at 1.63 miles, has deep roots as a vital bikeway from Golden Gate Park to Hayes Valley. Over time, SFMTA has added diverters to block cars from turning left onto Page from Divisadero Street and to prevent left turns from Stanyan Street.
The agency is now focusing on the eastern end near Octavia Boulevard where a one-way block, designed to funnel cars onto the freeway, is plagued by wrong-way drivers.
But there has been little effort to improve another long Slow Street. At 1.52 miles, Lake Street’s traffic volume often exceeds the target limit of 1,000 cars a day, particularly at the eastern end near Arguello Boulevard. Beyond small signs and a few speed humps, SFMTA has no current plans to install “calming” infrastructure or even barriers to divert traffic, as it has on Page.
The birth of Slow Lake produced one of the most heated battles, as the street has historically served as a cut-through for drivers coming off the Golden Gate Bridge.
These two examples show the role of a street’s history — and perhaps community pressure — in differing approaches.
As with the Lake Street fight, opponents have charged that limiting traffic will only make conditions worse on parallel streets. We examined these adjacent streets to see if these suspicions have merit.
Taken as a whole, parallel streets have not borne an extra burden. Injury collisions on these streets in total decreased 13 percent compared with the pre-Slow era. That’s just about the same drop in citywide incidents over the same period.
But as with the Slow Streets, not all parallel streets have done equally well. California Street, parallel to Lake, had 46 percent fewer injury collisions after Lake became slow. But Alemany Boulevard, which has on- and off-ramps from the 280 freeway, fared much worse after Cayuga became slow one block to the west.
(A note on methodology: For many Slow Streets, the alternate route for cut-through drivers is obvious. Lake, for example, is bounded by the Presidio and only California runs parallel to it. For others, we tried to select streets of similar width lanes and residential and commercial mixes.)
Waiting for roadwork
Injuries are the gold standard for measuring Slow Street success. SFMTA has also set important measurements of congestion and speed goals. Yet it has only done scattered readings.
The most recent was between January and April 2023, when it took the “temperature” of 16 Slow Streets for 48-hour periods. Only 25 percent met the speed target (average of 15 mph), while 75 percent streets stayed under the volume target (average of less than 1,000 cars a day).
Not satisfied with SFMTA’s data collection and lack of response on some streets, advocate group KidSafe SF launched a grassroots monitoring program using sensors that measure traffic activity and vehicle speeds. (The sensors attach to windows in volunteers’ homes and face the street.)
On the whole, the sensors show that most Slow Streets are in compliance with the program’s targets for average vehicle speed and volume.
Still, the most recent data on the group’s dashboard — October 2023 through June 2024 — shows traffic on Lake Street is persistently above the average daily volume, while average vehicle speeds along Cayuga (20 mph) and Hearst (30 mph) frequently exceed the 15 mph target.
SFMTA reviews and compares the data with its own, said agency spokesperson Michael Roccaforte. But they “don’t use it for official decision-making because city staff are not involved with its collection or quality control.”
Safety has long been a concern along Hearst, says Sara Barz, a Slow Hearst steward and KidSafe SF organizer. Sensor data has confirmed residents’ observations. Sup. Myrna Melgar, whose district includes Hearst, helped procure $50,000 in 2023 to add speed tables (a lower version of speed bumps) and crosswalks to help slow traffic. But the city has yet to do the work.
Emma Heiken, an aide to Sup. Melgar, says projects funded in this process take at least a year to complete. “SFMTA is going to take data into account — do studies and all — to see where the most effective placement will do the most good for the money,” says Heiken.
SFMTA has finished its studies on Hearst and four speed tables are “now on the Department of Public Works’ list for installation in the coming months,” said Roccaforte.
‘Drivers flipping you off’
NYU’s Moran, who studied SF’s crosswalks and bus stops as part of his UC Berkeley doctorate, says SFMTA and its QuickBuild program can make a difference, even with small adjustments. “It shows you how profound that little change can be,” he says.
The folks on Cayuga, the newest Slow Street, would love to see more signage and other fixes, but there’s no specific money allocated. The 2.2-mile street runs past or near 10 preschools and K-12 schools, including Balboa High School and James Denman Middle School.
“You get car drivers flipping you off if you’re in the middle of the street, because there’s little to tell them that this is a different street,” says Sheila Chung Hagen, steward for “Slowyuga.” (The good news is there were zero crashes in the street’s first full slow year.)
This raises an important question: How are Slow Streets prioritized for safety upgrades? In SFMTA’s May 2023 snapshot, Slow Noe and Slow Minnesota exceeded the volume target, though no injuries have been reported. Their upgrades – completed earlier this year – included traffic diverters and speed cushions.
In data compiled by The Frisc, Slow 22nd Street in Noe Valley and the Mission and Slow Cabrillo in the Richmond ranked first and fourth on the list of injury crashes in 2022-23. SFMTA said Cabrillo will receive a new type of Slow Street paddle (posts with signs) in the New Year. But the agency’s “small team doesn’t yet have plans to turn to 22nd Street,” said Roccaforte.
Meanwhile, the highly successful Slow Sanchez Street is getting even more safety infrastructure. As we reported, the city had some use-it-or-lose-it funding last year from an expiring sales tax measure. Nearly $300,000 was available for District 8, where Slow Sanchez is located.
Making Sanchez “as awesome as possible,” as SPUR’s Annie Fryman told The Frisc in 2023, is a useful strategy: a real-world demonstration of how good street changes can be.
Will such demonstrations be enough to convince other neighborhoods to embrace new Slow Streets? After more than a year of community meetings and input, SFMTA has said a final plan should come to a vote early next year.
It could include new Slow Streets in Japantown, the Western Addition, and the Excelsior. And the plan could also add more calming measures like those overdue on Hearst Street.
Seeing how there’s no grassroots consensus about bikes, some folks are going to be more inclined than others to roll with the final map.
Correction, 12/16/25: This story originally misspelled the last name of SFMTA’s Biking and Rolling project manager. Her name is Christy Osorio, not Osario.