An electric sign warns drivers about speed cameras on San Francisco's Gough Street at night. Red taillights are visible on the street, and a city park under lights is to the right.
Not so fast: A sign warns drivers about a speed camera on San Francisco's Gough Street. (Photo: Lisa Plachy)

Last week, street safety advocates threw a little party to celebrate San Francisco’s first speed camera at the corner of Geary Boulevard and 7th Avenue. They blew up balloons, kids held signs imploring drivers to slow down, and local news crews showed up to capture the moment. 

This camera and thirty-two others are about to switch on across the city in the latest attempt to curb dangerous driving. Despite a 2014 pledge to eliminate traffic deaths within a decade, city streets are deadlier than ever, with 42 traffic-related fatalities in 2024. 

Another safety milestone on the other side of town, however, didn’t get nearly as much attention. (Or balloons.) In December and January, neighbors on Hearst Avenue, near City College, watched city workers lay down humble rows of asphalt to create speed humps on their street.

It was more relief than celebration: It took nearly two years for their request to become reality, even though Hearst is an official “slow street,” meaning the city wants fewer cars and dramatically slower speeds there. 

Turns out the bureaucratic molasses of the Hearst process is no outlier. The city has no proactive program to lay down speed bumps. Residents must petition for them, and then wait years, not months, for action on a single block.

Speed cameras show promise, as data from New York City and Philadelphia demonstrate. But low-tech speed bumps (as well as humps, which are wider) are cheaper and faster to install. Or they could be — if San Francisco learns a lesson from another large city across the country. 

In 2023, Boston decided to ditch its residential request system, like the one in SF, and get up to speed on speed bumps. Boston plans to install 500 a year, a pledge that advocates would love to see in SF as part of a comprehensive approach to calming traffic. “San Francisco has to stop debating some things,” says Walk SF communications director Marta Lindsey. 

The first speed camera in SF’s new program drew a small celebration last week as a city billboard warned drivers of the change. (Photo: Kristi Coale)

It’s not clear how many bumps and humps are already on SF streets. A 2022 report from WalkSF said the city had installed 900 the previous two decades, but The Frisc could not confirm these numbers. One SFMTA map puts San Francisco’s count at a little over 500.

When asked to confirm this number, agency spokesperson Michael Roccaforte referred The Frisc to a DataSF set, which tallied 563. According to the data set, San Francisco installed only 22 last year.  

By any measure, the city is making slow progress.

Speed killed in 2024

In SF’s 42 traffic fatalities last year, 36 percent of drivers were speeding. Statewide, excessive speed is behind more than one-third of traffic deaths, according to the California Office of Traffic Safety. Efforts to even allow speed cameras in California have been in the works for six years.

Proponents like Walk SF’s Lindsey traveled to Sacramento to push for the legislation, which was finally signed into law in 2023. SF’s five-year pilot program, with cameras at 33 intersections, is just now underway. San Francisco Municipal Transportation Agency spokesperson Michael Roccaforte said it will cost up to $7.4 million for the program.

In New York City, which began using the tech in 2022, a study showed streets with cameras had fewer injuries and fatalities than streets without them.  

Shoes and name placards laid out on the steps of SF City Hall as people line up near the building's front door and a sign that memorializes people who have died in traffic-related deaths.
A vigil on the steps of SF City Hall for victims of traffic crashes, Nov. 17, 2024. Forty-two deaths were among the highest for a year in recent memory. (Photo: Lisa Plachy)

But there is another tool that SFMTA describes as “the most effective” for slowing traffic that is “less resource intensive” than cameras. That would be the speed hump. 

A Federal Highway Administration survey found that speed humps reduce vehicle speed by nearly 10 mph. A study in Oakland found that children’s odds of injury or death as a result of an automobile collision were cut by more than 50 percent when they lived near speed humps.

The SF traffic calming program has a lot of room for improvement. 

robin pam, kidsafe sf co-founder

Though data in SF are limited, SFMTA reports from 2019 and 2021 show that speed bumps have slowed down drivers across the city.  

Yet the process for getting even one speed hump in SF is  incredibly slow. So slow that in 2021, Vimeo co-founder Zach Klein decided to install his own after SFMTA told him it could take 30 months to get one on his street — if it was even approved.

Wait and see

Want a speed bump on your street? Get in line. SFMTA accepts applications from residents on a quarterly basis. This is an improvement; it used to accept applications once a year.

KidSafe SF co-founder Robin Pam says one of her neighbors applied in June 2022 and have been told the speed hump won’t be installed until at least the end of this year. “The SF traffic calming program has a lot of room for improvement,” Pam said. 

SFMTA waits until the end of the quarter to begin evaluations on applications, going in the order they are received. The evaluation process is time-consuming. A consultant must visit each individual street to gather traffic data and then report back to SFMTA, which determines eligibility by a couple of factors: traffic speed and volume and whether there is a school, bus stop, or hospital on the block or nearby. Most non-residential and heavily-trafficked streets are not eligible.

Applicants can end up waiting eight months or more to hear back from SFMTA on whether their street is approved. That’s just phase one.

If approved, agency staff go through an “engineering and design” phase, where they decide which specific traffic-calming measure a street will receive. Most often, they choose speed humps — they’re simple to install, inexpensive, and the most effective way to slow traffic. SFMTA doesn’t give a time frame for this step.

Once a traffic-calming measure is selected, the agency holds a public hearing run by agency staff where residents are invited to comment. After this, final approval comes from the city traffic engineer. 

Only then can construction begin — likely “within 18 months of the public hearing date,” according to the SFMTA website.

Fresh asphalt: “Slow” Hearst Ave. got a new speed bump recently after waiting nearly two years. (Photo: Kristi Coale)

Slow Streets aren’t moving any faster. When SFMTA turned the pandemic experiment into a formal program in December 2022, it pledged to add speed humps and other calming measures to help streets lower vehicle volumes and speeds. The targets are 15 mph and 1,000 or fewer cars per day on average. 

Slow Hearst, which only received a set of Slow Street signs and plastic posts at the ends of blocks, never met the speed target. So the advocacy group Friends of Slow Hearst turned to their district’s supervisor, Myrna Melgar. She helped secure $50,000 in July 2023 to add three speed humps to Hearst and Flood Streets, adjacent to an elementary school. 

Slow Hearst steward and KidSafeSF organizer Sara Barz thought that securing money through Sup. Melgar’s office would streamline the process. But SFMTA still had to go through its evaluation, engineering, and design steps. The agency then elected to add speed tables — wider than speed humps — and to add two per block, for a total of six.

The city installed the first two in December, and the remaining four in January. All told, the process from writing a proposal to installation took 21 months. 

Boston’s bumps

In May 2023, Boston mayor Michelle Wu announced a plan to install more than 1,500 speed humps on residential streets over three years. The point was to provide a 24-7 method of speed enforcement, said WalkMassachusetts co-executive director Brendan Kearney. “You can’t have a police presence everywhere,” he told The Frisc

The plan prioritizes areas with a history of crashes and socioeconomically disadvantaged residents. (According to recent studies, these areas suffer the most traffic injuries and fatalities.) Transportation officials divided Boston into zones, where speed humps are added to groups of streets in a series — a speed hump every 150 to 250 feet. 

A map of Boston with blue, black and purple highlights showing the city's speed bump installation progress.
Boston’s speed bump program began in 2023. Officials have pledged to install 500 a year. (Boston Transportation Dept.; The Frisc)

Wu said that the city’s old, application-based process often got bogged down in public comment. Plus, the disjointed approach of putting single humps on individual streets didn’t have much effect. “Drivers just go one street over where there isn’t a speed hump,” Wu said at a May 2023 press conference. 

Another problem was that safety measures were not distributed with equity in mind, said Kearney. Only people who were “plugged in” would take steps to get speed bumps, he told The Frisc. 

The program has not been perfect. On one street in the Jamaica Plain neighborhood, the city removed 13 speed humps — too many and too high. They’ll add back four when the weather improves.

Still, Boston’s chief of streets, Jascha Franklin-Hodge, believes the speed humps are making a difference. In a Bluesky post on Nov. 8, Franklin-Hodge provided early data from the program and said, “Speed humps reclaim neighborhood streets from aggressive drivers.” (The Frisc reached out to Franklin-Hodge for comment and for the “full report” promised in November, but got no response.)

We’re doing some before/after analysis of recent non-separated bike lane projects on residential streets which included speed humps. Full report to come, but this stood out. Speed humps reclaim neighborhood streets from aggressive drivers.

Jascha Franklin-Hodge (@jfh01.bsky.social) 2024-11-09T01:42:29.665Z

Kearney points to Boston’s Vision Zero numbers as more evidence that the speed humps are working. Pedestrian injury collisions went down from 587 in 2023, when the speed hump program began, to 190 last year. 

Boston expects to have at least 1,500 speed humps installed by 2026, with expansion plans beyond.

Smile, you’re on camera

Meanwhile, San Francisco will spend five years evaluating the 33 cameras spread around its most dangerous streets. SF has distributed them based on a few criteria. They’re on the high-injury network, the 12 percent of streets where most traffic-related injuries and fatalities happen. (Fulton Street, Geary Boulevard, and Cesar Chavez Street will each receive two cameras.)

SFMTA has also chosen streets where other safety measures, such as speed bumps or lower speed limits, aren’t practical — because they’re a major thoroughfare or on a bus route — or where those measures haven’t been effective. 

SFMTA says the speed cameras aren’t meant to punish drivers or be a significant source of revenue. But there are financial penalties. Fines start at $50 and can go as high as $500, depending on how fast the driver is going. Low-income drivers and people on public assistance will receive discounts. (The cameras are triggered when a driver exceeds the posted speed limit by 11 mph.) The agency said the citations are civil penalties, not moving violations, because they are issued to the registered vehicle’s owner. They won’t count as points against a driver’s license. 

Mainly the cameras are intended to change driver behavior, which has worsened since the pandemic. Nationwide, pedestrian deaths are at a 40-year high.

SF’s traffic engineers report unsafe speed among the most common causes of collisions. At the same time, the number of SFPD citations have dropped dramatically compared to pre-pandemic years. The department says one reason for the decline is a 2015 law, meant to curb racial profiling, that requires officers to list reasons for traffic stops. SFPD has also blamed staffing shortages.

New York is already changing driver behavior. In the study issued in January, 74 percent of vehicles that receive violations get no more than one or two a year; there are few repeat offenders. New York’s program expires this year unless the city council renews it. 

Back in San Francisco, cameras are part of a broader arsenal of safety tools that includes lower speed limits, “daylighting” of intersections, and more, efforts that took several years to accelerate. Right now, speed bumps are stuck in a lower gear. The potential is there, but the city needs to make its move.

Correction, 3/3/25: This story originally said that the speed camera at 7th Avenue and Geary Boulevard has been activated. It has not yet been activated, but will be soon.

Update, 3/9/25: A previous version of this story said SF had installed 10 speed bumps in 2024. SFMTA has since updated its map, and the revised total is 22.

Kristi Coale covers streets, transit, and the environment for The Frisc.

Join the Conversation

2 Comments

  1. Thanks for reporting on this important issue! One comment and one question:

    Based on my experience requesting traffic calming measures in my neighborhood, the MTA seems to prioritize ensuring the flow of traffic over pedestrian safety. As long as that remains the case, the city has no hope of reaching Vision Zero.

    Question: I’ve heard MTA refer to some speed bumps as “cushions.” I think those are the bumps that have flat spaces that I believe are designed to allow buses to avoid the bumps. Is that correct? Unfortunate plenty of cars use them as well, which defeats their purpose.

    1. Kristi Coale responds: There are several varieties of this type of traffic calming. You are correct: speed cushions are the ones with the slots for wheels. These are intended to allow first responders through, particularly fire and ambulance. I’ve spoken to EMTs, and going over a speed hump, bump or whatever you call them, with a patient is not great. I’ve been in cities where these cuts in the cushion straddle the two lanes of traffic. So for a car to avoid the cushion by using the slots in these instances, the driver would need to move half-way into the oncoming traffic. That serves as a good deterrent to avoidance moves.

      SF’s speed cushions don’t do this, at least none that I’ve seen. (One example of cuts making the cushion useless is on Masonic between Frederick and Waller.)

      So you are correct, the cuts as deployed defeat the purpose.

Leave a comment