Studies Show Car Traffic in San Francisco is Dropping

Car traffic at Mission and Third Streets has declined by 7 percent over the last few years, according to SFCTA counts. Photo: Sergio Ruiz/Flickr

Car traffic has dropped in San Francisco in recent years, despite an economic boom and a growing population, according to studies by the SF County Transportation Authority.

A newly updated study (reported by SF Weekly) by the SFCTA counted fewer cars at 11 of 15 intersections during evening peak hours this year, compared to earlier counts taken between 2009 and 2012. Driving speeds, meanwhile, are “increasing moderately.”

As SF Weekly’s Joe Eskenazi pointed out, the data fly in the face of anecdotes from drivers — who almost universally feel that car congestion is always getting worse. And given the city’s booming economy, population, and construction in recent years, that’s one scenario that certainly would have been plausible had the 20th-century status quo continued.

“Anecdotal evidence is hard to counter,” Eskenazi wrote. “But what statistical evidence does exist flies in the face of your well-worn anecdotes.”

SFCTA transportation planner Dan Tischler acknowledged that, despite the somewhat limited scope of the study, all of the evidence available indicates that San Francisco commuters are driving less, and likely switching to other modes to get around.

“We are not really sure if traffic conditions are worse now than they were a few years ago, but we do have strong evidence that transit is playing an increasingly significant role in handling growth in travel demand,” said Tischler.

Most importantly, Tischler noted, driving speeds have largely remained flat, or even increased slightly, from 2011 to 2013. That contradicts any notion that fewer cars were counted because congestion is causing them to moving through more slowly. (Slower speeds would actually increase throughput, since cars follow one another more closely at slow speeds.)

“Lower traffic volumes, combined with higher speeds, indicates that lower traffic volumes may be due to less demand rather than too much demand,” Tischler said.

The SF Chronicle reported on the increased driving speeds in April, and said the findings were confirmed by a separate study by Inrix, a traffic information company, conducted at the Chronicle’s request.

The 15 intersections studied by the SFCTA were all in and around downtown, mainly along Mission Street (six intersections) and Van Ness Avenue (five intersections). Two intersections were in the Tenderloin: Turk and Hyde streets and McAllister and Leavenworth streets. The intersection of Columbus Avenue and Broadway was the only one to see an increase in vehicle traffic, by 5 percent.

One outlier in the data was the 31 percent traffic drop seen at Mission and Fourth streets, but Tischler noted that it was likely due to the closure of Stockton Street to vehicle traffic due to construction on the Central Subway.

Elsewhere, a traffic drop of 22 percent was recorded at Van Ness and Geary Street, a 20 percent drop at Van Ness and Broadway, and a 16 percent drop at McAllister and Leavenworth. The rest of the intersections saw smaller declines, while three saw a similar amount of traffic.

As the Chronicle reported in its April article, the SFCTA’s studies showed average driving speeds have increased in downtown, by 3 to 5 percent between 2011 and 2013. “Speeds through 95 percent of the intersections monitored in the city’s core remained the same (58 percent) or decreased (37 percent),” wrote the Chronicle’s Michael Cabanatuan. “Freeway speeds showed even bigger increases, rising 15 percent during both the morning and evening rush hours. That implies that fewer people are driving into San Francisco, a supposition other statistics support.”

On the Bay Bridge, car traffic decreased by about 2.8 percent from 2013 to 2014, the Chronicle reported. Golden Gate Bridge traffic reportedly increased by 3.7 percent, though Caltrans counts from 2011 to 2012 at the Highway 1 and 101 entrances showed traffic holding steady and declining 2 percent, respectively.

One major takeaway from all these numbers: The tired narrative of forever-worsening car congestion is often repeated by those fighting efforts to re-purpose space from cars, and make streets safer or make transit more reliable. But that narrative just doesn’t have any grounding in empirical evidence — only anecdotes, and often from the windshield perspective.

But evidence of the trend away from driving is mounting. As we reported, car-free households are also booming in SF — 88 percent of net household growth since 2000 has been car-free.

Even if traffic was getting worse, saving drivers a few seconds still wouldn’t be a good reason to abandon measures that save lives and provide better transportation options. Unfortunately, that was the approach SF took for too many decades, and at last we may be starting to see the fruits of efforts based on a more holistic and multi-modal approach.

Of course, if we really want to accelerate this virtuous trend, SF could finally implement congestion pricing, which is a proven strategy to make everyone’s trip faster, safer, and more reliable.

  • Mario Tanev

    It’s EskEnazi, not EskAnazi, though the latter would be a mildly amusingly ironic last name if segmented around the A.

  • Oops! Fixed.

  • Mario Tanev

    There are two spots with the misspelling. Only one has been fixed.

  • This is even more interesting given the fact that the DMV shows something on the order of 20,000 more cars _registered_ in SF in the last decade, if my memory serves (which it probably does not). That would suggest that even though ownership is up in straight numbers, people are driving those vehicles less.

    Given the anecdotal aspect, though, one has to wonder whether that has to be chalked up to the fact that we’re just fed up with sitting in traffic as a general concept, so it seems more annoying even though it’s not as onerous as it once was.

  • murphstahoe

    Thousands (?) of SF residents now get carted to work in a private bus, Caltrain ridership has doubled. Caltrain median income is over 6 figures, shuttle bus median income is probably higher than Caltrain.

    Own a car, and use it for long distance pleasure trips only. That was my model while living in Noe Valley, we walked to the grocery store and mostly took the car out if we left town to some far flung place – adding little to local congestion.

    If we didn’t own the car to begin with we might have just rented cars, but given we had established parking at home, there was little incentive to sell the car. Even after moving to the netherlands we didn’t add a second car, and we rent a car 2-3 times a year when situations arise, for less than the cost of insuring a second car, let alone depreciation.

  • I’d like to see studies of traffic at intersections around Octavia Blvd. It’s definitely been getting worse there over the past few years, at almost any time during the day.

  • Gezellig

    Btw, even moderate drops in car modeshare can lead to disproportionate drops in congestion for various reasons. One might be the fact that up to 90% of car traffic in some neighborhoods is simply due to drivers endlessly cruising around and around for on-street parking. This kind of heightened concentrated repetitive searching around the block creates a *lot* of congestion per car. So going by another mode not only takes away a moving car on the way but takes away a lot of that endless circling at the end.

    I was actually just thinking how noticeable even slight traffic drops are. Yesterday was Veteran’s Day, which is a day some offices take off but probably a majority don’t. Yet the difference during rush-hour was palpable on Van Ness. There were still plenty of cars, but everything was very free-flowing instead of the normal constant stop-and-go backup. So the moderate (5%? 10%?) drop in traffic made a world of a difference, especially being on a bus.

  • Ask and ye shall receive:

    This is two years old, but it shows that 70K+ vehicles per day clog that area, and the vast majority of them have neither final destinations, nor do they even stop in, SF.

    They are simply driving through, rather than around.

  • Flubert

    How exactly would drivers go “around” Octavia if they do not take it? They are presumably going to 101 from the north or west of the city, or the reverse. What’s the alternate route that is better?

    19th Avenue to 280 might be an alternative if you happen to start and finish on the west side, but that’s hardly less congested anyway.

    Octavia was built to replace the elevated section of 101 that really didn’t cause congestion to Octavia nor congestion trying to get across Market Street. We all voted (eventually, after three ballots) to instead provide local access to 101 via Octavia, and that is exactly what we’ve got, with all the congestion there that that implies.

  • I’m not saying there’s a _better_ route, although 280 to Embarcadero is also an option.

    What I’m pointing out is that this is a stub of the Central Freeway (not part of 101 — that’s Van Ness) that was intended to cut through SF, including through GG Park. It was stopped as part of the Freeway Revolts in 1955, but this legacy route remains despite how deleterious it is to the communities through which these cars travel…and idle.

    If it hadn’t been for the Central Freeway, this wouldn’t be an issue at all. But that’s what we’ve been left with so I suppose we just deal.

  • yermom72

    “Of course, if we really want to accelerate this virtuous trend, SF could finally implement congestion pricing, which is a proven strategy to make everyone’s trip faster, safer, and more reliable.”

    While of course putting the last gentrifying nail into the city’s coffin. No thanks.

  • murphstahoe

    How exactly would congestion pricing impact the lower income portion of the citizenry which doesn’t own a car, or drive it downtown because of the cost of parking?

    I guess your point is that by implementing congestion pricing, San Francisco would become so much more pleasant that the demand for living here would go even higher, drawing more higher income people into the city, thus driving out lower income people who can’t afford to live in such a desireable place?

  • J

    Also, all these new bike lanes are killing the manufacturing jobs! (sarcasm)

  • EastBayer

    Probably particularly noticeable on Van Ness because of all the government workers there.

  • HuckieCA

    The perception of traffic and the reality of traffic don’t always coincide. Traffic for me the few months has been awful due to constant construction in my neighborhood, but in reality, there were probably few cars overall. People who could divert, did. That being said, I think the major reason is more related to the shuttle busses removing cars from the roads, and the fact that many of the tech companies and workers who have recently moved to the city have done so to avoid commuting by car. It would be really interesting to see traffic stats, public transit ridership, and economic growth in the city all plotted together. I suspect that there are probably some areas where traffic got better, and some areas where traffic got worse.

  • patrick_sf

    Through the neighborhood, not the street. As opposed to driving around it, or stopping in it.

  • Gezellig

    So, basically you’re saying we shouldn’t take too many cars off the road or take too much pollution out of the air or reduce too many car-related deaths because Gentrification.

    Wow, that’s some really cynical fatalism.

    By that logic we could *really* hamper gentrification by going ahead and suckifying as many neighborhoods as much as possible. Maybe reintroducing that old Freeways Everywhere plan again?

    Or how about demolishing 50% of all buildings to replace with parking lots to really kill the cool urban vibe?

    That’ll show those gentrifiers! And will *definitely* help the socioeconomically disadvantaged, for sure. Especially by making snooty transit and snobby walking less viable and totally egalitarian driving easier. Because people of lower socioeconomic status in SF are the ones driving everywhere, of course, unlike those selfish transit-taking, elitist walking gentrifiers.

    Whatever we do, let’s *definitely* make sure this never changes!

    Whoa, too nice, Van Ness–no thanks!

    Let’s keep it like this:

    Take THAT, gentrifier 47-takers! Let the *REAL SAN FRANCISCANS*–the selfless drivers in their cars–go FIRST!

  • Gezellig

    Yeah, I was thinking that was a part of it, too.

    Though with how well Civic Center is connected via transit I wonder how many workers there actually drive to work? I mean, this dude certainly does:

    Except for his once-yearly photo-op:

    “Whoa, our bike strategy on Bosworth is a sharrow?! This is terrible. Can’t wait to get back in my Volt tomorrow.”

  • Flubert

    Oh, sure, I supported pulling down 101 back to Bayshore 15 years ago but the political support wasn’t there, I guess.

    But right now access from the NW to the 101 spur is fairly limited – Octavia or Van Ness, unless you want to thread your way through SOMA to 280 or 80.

  • Gezellig
  • Mario Tanev

    Congestion pricing in fact limits the driving to a constant rate. Which means any population growth would be for the pedestrian and transit benefits, not because of driving ease.

  • yermom72

    Congestion pricing is what I was criticizing. It’s the point where this blog veers away from ecological consciousness, into a fetishization of market solutions.

    I.e, the same kind of thinking that got us into this crisis in the first place. And yes, imposing regressive taxes that privilege the wealthy would certainly do nothing to impede gentrification.

  • gneiss

    How is charging car drivers that create congestion on city streets “regressive”? Let’s say this all together again – people who own cars are not the poorest among us. By and large, the people who don’t own cars and take public transit are, and they get shafted by every car driver that blocks a bus during rush hour. If you really want to be “progressive”, level the playing field by taxing the people who slow down transit.

  • Gezellig

    This reminds me of the concern trolling some of the Columbus Ave merchants did against a proposed road diet there–that removing some parking spaces would somehow hurt the working class.

    Despite the fact nearly 90% of people already arrive at Columbus Ave via a non-car means anyway and those 10% that do are disproportionately of privilege. Yet the public right of way devotes far more than 10% of space to the free movement and free/cheap storage of private autos for people mostly of above-average socioeconomic status.

    That’s your *actual* regressive economic model. And it’s everywhere in this city.

  • yermom72

    Yeah, I argue against congestion pricing and that makes me a “troll?” Smooth. I’m sorry but I’m not alone in the belief that these market-driven approaches have no real place in green thinking.

    This is not about parking at all — I’d be happy to see less street parking.

    Yes, working people often do need cars, especially those who now have to commute into the city, from areas without good public transit. Building more public transit is the solution, not making working people’s lives more difficult than it already is.

    Congestion pricing also does nothing to get away from car-centric design, and the unequal access this provides to the city — to the contrary, it ENSHRINES this inequality as a “solution” to congestion.

    That is not a “progressive” answer to the problem.

  • Anandakos

    (Slower speeds would actually increase throughput, since cars follow one another more closely at slow speeds.)

    This has been shown in practice on freeways to be true only above 42 miles per hour. At speeds slower than that cars cluster closely enough together that the “fewer vehicles can pass in a given amount of time” effect begins to exceed and eventually overwhelms the “cars stay farther apart at higher speeds” effect.

    Just think about it. If one slows traffic down to 1 mile per hour (88 feet per minute or 1.47 feet per second) it will take almost exactly ten seconds for one fifteen foot long car to pass, If you observe for a minute, you’ll count six average cars. At ten miles per hour (880 feet per minute or 14.7 feet per second) you’ll observe one car every second. They won’t be exactly nose to tail, so you won’t count 60 of them in a minute, but even using the very conservative-at-ten-miles-per-hour two second rule, you’d count twenty, three times as many as you would at 1 mph.

    But that’s absurd on its face. If traffic is moving ten miles per hour people are only slightly worried about banging into the person ahead of them. Instead they’re fending off the jerk in the next lane trying to squeeze in on them, so they’re as close to the bumper of the car ahead as possible. They’re not “maintaining a two second following distance”.

    At twenty miles per hour (1760 feet per minute or 29.3 feet per second, if they were nose to tail you’d count two per second. Continuing with the two second rule, you’d count a car every two and a half seconds, instead of every three seconds. You’d count 24 of them.

    At thirty miles per hour (2640 feet per minute or 44 feet per second and still using the two second rule, you’d count a car every two and a third seconds, or 25.7. You can see we’re approaching an asymptote. No matter how fast the cars go, with the two second rule they can’t pass more frequently than every two seconds.

    At forty miles per hour (3520 feet per minute or 58.7 feet per second) and still using the two second rule, you’d count a car every two and a quarter seconds or 26.7 of them per minute.

    But the truth is that as speeds increase, so does adherence to the two second rule. Most drivers file down a 30 mile per hour single-lane street with less than two seconds of clearance between their car and the one preceding it.

  • Anandakos

    You pay for the buses, you know. Use them and avoid the “congestion pricing”.

  • Anandakos


    This is the model of the future. Good work, sir.

  • yermom72

    See what I mean? What kind of progressive approach embraces stratification?

  • Did you guys miss a memo? Because I read in SFGate comments that traffic is definitely getting worse, and it’s all because of those dam dirty hippies. No, wait. The other one. Bicyclists.

  • gneiss

    Actually, congestion pricing does help get away from car-centric design, because it allows a city to reduce the total amount of street space that needs to be devoted to cars, since fewer cars will be using the roadways. In addition, by reducing congestion, you allow transit to flow more freely – how can that not be a progressive goal?

    As to saying “build more public transit” that’s the absolutely most expensive option, and won’t work if the roads are already clogged with cars. A more cost effective approach is to better optimize our existing street grid with modes of transport that don’t require as much space as cars. To do that, you need to reduce the total number of cars on the road. Congestion pricing is one way you can do that.

    Honestly, what’s the difference between congestion pricing and paying for tolls on bridges? Are you going to argue that the Bay Bridge toll is also a regressive tax on poor people?

  • yermom72

    I don’t like tolls or toll roads either, but at least those are limited in space. Congestion pricing spreads that logic.

  • SF4SF

    It would be really cool to have an annual aerial 24 hour video survey of the city to really see what’s changing and happening. We all have our own opinions created by our experiences and bias. But in recent block the box hearings, supervisors were saying congestion now is worse than they have ever seen or expected. My own the street experience and friends agree it’s much worse. My bias says it’s due to construction, SFMTA’s engineered congestion changes to roads, and just more people in the city (pedestrians, Ubers, big buses, etc). Also, part of the perception of congestion is from drivers that extend their drive time and complexity by rerouting through other streets and neighborhoods to avoid pinch points. Avoiding congestion feels like congestion too. Increased traffic on my own formerly quiet street adds to the perception. All of this activity is not captured by a few survey points in the city. So I can easily accept all this data and agree that congestion is getting worse.

  • Ryan K

    SFMTA has a vested interest in providing statistics that prove their decisions are having positive impacts. When *every single driver* reports that their commutes are longer, slower, and more dangerous, the ethics of the SFMTA Is called into question.

  • coolbabybookworm

    The study was conducted but the SFCTA, not the SFMTA.

  • EastBayer

    Please, most Americans in the last election also thought that the economy was SO MUCH WORSE in 2014 than in 2008, hence the midterm election results. That, and it was a TA study…

  • 94103er

    As SF Weekly’s Joe Eskenazi pointed out, the data fly in the face of anecdotes from drivers — who almost universally feel that car congestion is always getting worse. And given the city’s booming economy, population, and construction in recent years, that’s one scenario that certainly would have been plausible had the 20th-century status quo continued.

    “Anecdotal evidence is hard to counter,” Eskenazi wrote. “But what statistical evidence does exist flies in the face of your well-worn anecdotes.”

    So I’m figuring your reading comprehension is especially poor, or you’re rooting out some deep, dark conspiracy theory, or your definition of ‘ethics’ is particularly creative. Hmm, which is it?

  • dandyhighwayman

    Anecdotally my commute has been getting easier. Other than the Bay Bridge. And other people I know say the same thing.

    The only slowdowns I hit are from building or road construction. I wonder if those are what people are thinking of.

  • This is great Wonderful post.. Awesome information Thanks..

    Buy Cycles Online

  • Mike Sicard


  • Mike Sicard

    Have you been in the Mission in the past two months?

  • Mike Sicard

    They could require a $5 million fee to move into or stay in the city. That would solve most of the problems
    Of course, homeless would be exempt.

  • Mike Sicard

    I’m all for getting cars off the road, but not everyone can or will walk or ride a bike. The buses and trains are filthy, poorly maintained, unreliable, and are decades behind in expansion. You can’t build Manhattan on the infrastructure of Modesto and expect smooth commutes with little traffic or pollution.


Today’s Headlines

SFCTA to Vote Tomorrow on Extending Congestion Pricing Study (SF Examiner, City Insider) San Mateo County Leaders “Threaten to Enact Toll” for San Francisco Residents (Mercury News) Muni Management Seeks to End Year-End Payouts for Muni Operators (SF Gate) Mayor Newsom’s Office Orders Department Heads to Compile Info for Transition (SF Examiner) Residents of Rent-Controlled […]

SFCTA to Test Variable Road Pricing on Treasure Island

Treasure Island will serve as San Francisco’s proving grounds for road pricing that adjusts in response to traffic conditions, as the city looks to minimize Bay Bridge car congestion generated by residents expected to move to the development site. When the first housing units on the artificial mid-bay island, formerly owned by the Navy, are occupied in 2019, the SF County Transportation […]

SFCTA: Geary BRT Will Take Hundreds of Cars Off the Street Every Hour

Once bus rapid transit is finally up and running, Geary Boulevard will carry thousands fewer cars every day in 2020, compared to a scenario where it doesn’t get built. That’s according to a preliminary analysis [PDF] presented by the SF County Transportation Authority. The traffic counts vary, depending on which of several design alternatives are […]