SF Merchants Say New Bike Lanes Kill Business. Studies Show Otherwise 

All it took was one month. San Francisco transit officials this summer began rolling out draft plans to dramatically expand the city’s bike lane network. In early July, they convened a meeting for District 3, which includes Chinatown. In early August, they scrapped Chinatown plans due to community pressure. 

There’s no sign that the new network, known as the Biking and Rolling Plan, is in danger. The San Francisco Municipal Transportation Agency is still collecting feedback and will finalize it in early 2025. 

But the anger, especially from small business owners, is part of a pattern. Merchants in neighborhoods including Chinatown, the Mission, Potrero Hill, and West Portal have protested changes that remove parking spots and slow traffic, all in the name of safer streets that encourage bikers and protect pedestrians.

Some changes are a reaction to tragedy. In March, a driver killed a family of four on a West Portal sidewalk, spurring new traffic patterns and calming measures. 

Other changes aim to improve existing infrastructure. Bicyclists long considered Valencia’s curbside lanes, first installed in the 1990s, to be dangerous, so SFMTA tried a one-year experiment last year: center-running bike lanes along an eight-block stretch. 

Update, 9/10/24: SFMTA announced Sept. 9 that it has listened to merchants and crafted a new design for the Valencia bike lanes between 15th and 23rd Streets. The lanes will shift from the center of the road to the curb. But because of restaurant parklets along the curb, the lanes will often have to “slalom” around with more exposure to traffic. The public can view the design and give feedback on Sept. 23 and 25. If the SFMTA board approves the design in November, work should start in January.

With a decade or more of competition from online shopping, then the extra gut punch of the pandemic, shop owners feel like their operating margins are thinner than ever.

“It doesn’t take a lot to push a business over the edge,” says Carol Yenne, who owns a children’s clothing store in Noe Valley and is transit committee chair of the SF Council of District Merchants Associations, which represents a couple dozen neighborhood business groups. 

Many merchants feel left out of conversations, and they are convinced that anything that discourages customers who drive will hurt their businesses. 

There’s a small but growing body of research on bike lanes and small business, and a pattern is emerging. There’s no evidence that bike-friendly changes are bad for entire neighborhoods. 

Polk, Valencia, Columbus, and more 

While there are numerous surveys asking people how they get around town and spend money, The Frisc limited its analysis to four studies that dug into local economic data. All were partially or completely focused on San Francisco. All acknowledged limitations and caveats, which we’ll note.

Here’s what they say. 

* A 2018 study examined Columbus Avenue, Polk Street, and Valencia Street, all of which had bike lanes painted onto the asphalt between 1999 and 2005. The researchers didn’t know if parking spots were removed. The study measured sales revenues two years before the bike lanes, then again three years after installation. On each corridor, sales went up. 

The study noted that in “local serving” shops where a customer could walk (or bike) up — cafes, bookstores, and clothing stores, for example — “sales increased markedly” for Valencia and Polk and slightly on Columbus, “suggesting that bike lanes can have a positive effect.” 

San Francisco Has Many Ways to Make Bike Lanes. Here’s What They Feel Like On Two Wheels

When shops selling larger items were added to the total – such as furniture stores — the data still showed overall sales improvement on Polk and Valencia, but not on Columbus. “Bike lanes do not seem to have the catastrophic negative effect that some merchants claim,” wrote the study’s author, Joseph Poirier of Nelson/Nygaard Consulting. But Poirier also noted “it is unlikely that a blanket ‘bicycle lanes are good for all businesses’ statement is accurate.” 

* In 2019, Poirier and two Bay Area researchers came back with another study, this time to account for the removal of parking spots. 

They looked at two types of situations. First, streets with “sharrows” – pavement markings that encourage drivers to share the road – but no changes along the curb had little effect on local business. Second, adding curbside bike lanes and removing parking spots also had an insignificant effect, but there were notable exceptions, both for better and for worse. 

When curbside lanes appeared, new restaurants and grocery stores had “significantly” higher sales than their existing counterparts, “suggesting bicycle infrastructure may attract more upmarket businesses in those industries,” wrote the authors. On the flip side, businesses that sell appliances, carpets, and other home goods, or auto-related goods and services, saw a significant decline. 

The authors acknowledged that “different types of businesses experience bicycle lane interventions differently, with the implication being that transportation planners need to understand and carefully account for the mix of businesses in implementation sites.”  

* In a 2020 study, Portland State University researchers examined the impact of bike lanes on businesses in San Francisco and three other cities. The SF data focused on Polk Street and 17th Street in the Mission. For comparison, they used parallel streets with similar business density but no bike lanes (Van Ness Avenue and 18th Street). 

The researchers measured Polk from 2002 to 2008, two years after a bike lane went in and a traffic lane was removed. The data consistently showed improvement for small businesses. On 17th Street, a curbside bike lane added in 2011 had a neutral to positive effect on retail sales, and a neutral to slightly negative effect on food service. 

The authors do add a caveat that many factors influence economic performance. 

(The other three cities showed either positive or insignificant impacts on business performance.) 

* This summer, after months of protest on Valencia – including a restaurant owner’s threat of a hunger strike – SF officials issued a study of the corridor’s sales tax revenues between 2019 and 2023. For context, they added the same data from 28 other commercial stretches. (It did not include downtown or tourist-heavy neighborhoods.) 

Valencia had the weakest recovery – 75 percent of pre-pandemic levels – save for upper Fillmore Street. But that’s not enough to pin blame on the center-running bike lane experiment, which began in mid-2023.

The analysis showed no correlation between the new bike lane and sales figures in the second half of 2023: “The challenges facing the corridor pre-date the construction of the bike improvements, and there is no statistical basis for linking the two.”

“This finding does not mean that no business was adversely affected,” the report concluded. “It simply means that any negative impacts on individual businesses were offset by positive impacts on others, and there is no net effect on the corridor as a whole.” 

Taken together, these four studies show that bike lanes don’t depress the economic performance of entire business corridors, and might overall be beneficial, depending on the business mix. 

Valencia Street Bike Lanes to Stay Centered For Now, Despite Merchant Anger and Other Pressure

Kelley Clifton, a professor of urban and transportation planning at the University of British Columbia, has studied spending patterns of customers who drive, bike, and walk to stores. (But not specifically in San Francisco.) Drivers might buy more in a single trip, Clifton says, but bikers and walkers make up the gap by visiting businesses more often.  

However, mildly positive results sprinkled heavily with caveats often don’t soothe merchants’ fears when the lost parking spots and other changes are right outside their doors.

‘Blindsided’ – or not?

Through all of 2023, after announcing its plan to extend the bike network, SFMTA reached out to community groups, gathered input at open houses, and led walk and bike tours around neighborhoods – often with district supervisors pedaling along. 

SFMTA prioritized feedback from communities historically neglected by this kind of planning, such as the Tenderloin, Mission, and Bayview. It hired five community groups to do neighborhood outreach and the SF Bicycle Coalition for citywide outreach to bicyclists.

But there was no specific outreach to merchants, which SFBC executive director Christopher White says was an oversight. White is now working with the Valencia Merchants Association on the transition from the experimental center-running bike lane to new curbside versions. “We had a much stronger relationship with businesses, but that was something that decayed during the pandemic,” says White.

SF’s New Bike Plan Is Coming. Neighborhoods That Feel Left Behind Are the First Stop

Now merchants say last month’s rollout of more specific plans for each neighborhood “blindsided” them. “SFMTA never reached out to us until they came with the news that we’d have bike lanes,” says Edward Siu, chairman of the Chinatown United Merchants of San Francisco. 

SFMTA officials say they were not offering a done deal. In fact, the agency showed up with three scenarios, and one was a Chinatown without additional bike lanes. In that meeting, part of an ongoing series of neighborhood meetings, none of the scenarios are “set in stone,” says Christy Osorio, SFMTA’s Biking and Rolling project manager. 

If someone can’t make a meeting, “I’ll just make time to show them the plans over phone or video chat for 30 minutes,” she says. 

SFMTA also says it has listened to businesses. Not everyone gets what they want. One example is on 17th Street, where the upgraded bike lanes run next to Potrero Hill Montessori School. SFMTA asked for feedback, and the school obliged, says school director Jessica Yang. It requested a “floating” loading zone: one hour for morning drop-offs, and one hour for afternoon pickups. “It was a small ask,” Yang says. “But there was no discussion.”

SFMTA spokesperson Michael Roccaforte says a loading zone on 17th would require complicated and potentially dangerous logistics – moving dedicated police parking, or having cars cross the bike lane. Instead, the agency moved the loading zone around the corner to Carolina Street. Roccaforte says the agency heard that “some parents” preferred the Carolina Street loading zone.

There are positive testimonies too. Jeremy Chan of the Japantown Task Force says SFMTA helped them to find the best spot for bike share parking, even though it would remove a few parking spots. On Tuesday night, Chan was browsing SFMTA’s three mapped scenarios for District 5 in the Park Branch public library on Page Street. 

Also at the meeting, Hayes Valley Neighborhood Association vice president Andrew Seigner gave feedback often heard from bike advocates who want more protected lanes. “No more sharrows, or ‘murder lanes,’” said Seigner, who wants to close Hayes Street to cars. “If it’s not separated, it’s not safe.”

These debates tend to paint participants one-dimensionally, defined by how they get around — cyclists, pedestrians, drivers — when in fact many people use multiple ways to get around, says Clifton of the University of British Columbia. Identity politics, and the demonizing that comes with it, have seeped into these discussions, and “this is unhelpful.” 

 The importance of better bike infrastructure is to encourage as many people as possible to use more sustainable modes of travel. Getting more people out of cars is good for cities and for a warming planet, but working toward that greater good means building trust. 

Correction, 9/4/24: This story has been changed to clarify that for the Biking and Rolling Plan, SFMTA hired community groups to do neighborhood outreach, not the SF Bicycle Coalition. The SF Bicycle Coalition conducted citywide outreach to bicyclists.

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.