Shoup: SFpark Yields Promising Results, Lessons for Demand-Based Pricing

Drivers were most sensitive to changes in parking prices in the early afternoon, and were more sensitive during the week than the weekend.

Donald Shoup may be known as a guru of smart parking policy, but even he has found a few surprises in the data collected so far from SFpark.

“The biggest surprise I got was that prices went up and down, but overall, they stayed the same. The average price actually declined by 1 percent,” said Shoup, professor of urban planning at UCLA and author of The High Cost of Free Parking, the bible of modern parking policy. “That surprised everybody. People thought it was just a way to jack up prices, but the city specifically said, ‘We are going to set prices according to this principle.'”

[Update: The SFpark final report found the average price dropped by 4 percent, not 1.]

SFpark, which uses “smart meters” and ground sensors to measure parking occupancy and adjust prices accordingly, is providing valuable lessons for San Francisco and cities around the world that want to reduce the amount of time drivers spend cruising the streets for a parking space.

The growing body of data collected from the program is shedding more light on the complexities of parking demand. But overall, Shoup says, it’s providing hard evidence that raising and lowering meter prices is an effective way to keep enough parking spots available for drivers who need them — and to help ensure too many spots don’t sit empty.

Donald Shoup at the launch event for SFpark in 2011. Photo: Bryan Goebel

Keeping, say, one parking spot open on every block “will make the transportation work best — it’ll reduce cruising, speed up buses, reduce air pollution,” said Shoup. “It’s easy to explain [a goal] like that — we’re aiming at what you want to see.”

In a recent report [PDF] published in the Journal of the American Planning Association, Shoup and UCLA doctoral student Gregory Pierce explain that since SFpark managers began adjusting meter prices in August of 2011, the “elasticity” of parking demand — the degree to which price changes affect parking occupancy — has varied across different locations and times of day (due to different trip purposes, they surmise), and that drivers changed their behavior most profoundly after the second price adjustment, possibly due to a spike in awareness of the program. As prices have been refined, elasticity has declined.

Prices appeared to have the lowest impacts in highly residential neighborhoods like the Mission and the Marina, while retail districts like Fisherman’s Wharf and the Fillmore saw the most drastic adjustments to new prices, according to the report.

Parking demand was also affected by temporary events like festivals and construction sites, which Shoup says SFpark should strive to plan for in advance more effectively. SFpark did begin setting special prices at spaces around AT&T Park during Giants games this year. “There are a lot of other things affecting occupancy other than just prices,” said Shoup. “It could be road closures, sales in stores, or Christmas season.”

As occupancy rates get closer to the target of roughly 85 percent, adjustments to SFpark meter rates have become less drastic and less frequent. Under the last adjustment announced in March, 58 percent of meter rates stayed the same, while 20 percent of meter hours increased by 25 cents, and 22 percent dropped by 25 or 50 cents. In June, the latest adjustment for rates at city-owned parking garages affected only 7 percent of parking hours, and garage rates have generally dropped sharply throughout the course of the program (complemented with signs pointing drivers towards the garages). SFpark managers are allowed to make adjustments every two months at the most.

Drivers’ responsiveness to parking price changes varied in different neighborhoods, with drivers in the residential Mission being the least sensitive to adjustments.

But a lack of awareness that prices vary between each hour and block can hold back the program’s effectiveness, since fewer drivers are looking to park in areas with less demand. For this reason, Shoup and Pierce recommend that SFpark increase the visibility of the program beyond current features like the program’s phone app, data-rich website, advertisements, and meter labels. “Many drivers, especially tourists, remain unaware of these price variations and thus miss the opportunity to save money by walking a few blocks from their cars to their destinations,” the report says.

The report also calls upon California to tackle the problem of disabled parking placard abuse, pointing to other states like Michigan and Illinois that have removed incentives for drivers to abuse placards to get free parking:

California allows all drivers with disabled placards to park free for an unlimited time at parking meters, so higher prices increase the temptation to abuse placards. Raising prices on crowded blocks may simply drive out paying parkers and make more spaces available for placard abusers. If so, prices will not reduce occupancy, and the price elasticity of demand will remain artificially low.

The SFMTA should also extend its parking meter hours past 6 p.m., when parking demand remains high in many areas, said Shoup. “I hope San Francisco will ask, ‘Why is the right price at 7 p.m. on Union Square $0?'” he said. “We have the equipment, all the software, and we just put it to sleep at 6 p.m.”

Outdated policies aside, Shoup praised San Francisco for pioneering the most comprehensive system of demand-based parking in the world. “Normally, you expect America to have fallen behind the rest of the world in transportation policy, but this is the one area where the United States has really seized the lead and is way ahead,” he said.

Leaders in cities around the world have expressed interest in launching similar programs, Shoup noted, including Rio de Janeiro, which plans to host the Olympics in 2016 but doesn’t have a single parking meter on its streets. “They were quite entranced by it,” he said.

Shoup calls SFpark a “political success” — insofar as it’s converted existing parking meters (adding new meters is another question) — and he says the program shows how dynamic parking pricing can be an alternative to congestion pricing that requires less infrastructure and fewer political battles.

“Cities can adopt programs like SFpark even if they do not yet have all the resources and political will to adopt congestion pricing,” the report says. “In effect, performance-parking prices are a poor man’s congestion pricing, and they may represent a step toward full congestion pricing.”

Down the road, demand-based parking pricing could help make people aware of the hidden costs of car use, say Shoup and Pierce. In fact, they argue, drivers might come to expect the convenience of available parking, even if they have to pay for it. “What seemed unthinkable in the past may become indispensable in the future,” they wrote.

“With performance-parking prices, drivers will find places to park their cars just as easily as they find places to buy gasoline. But drivers will also have to think about the price of parking just as they now think about the prices of fuel, tires, insurance, registration, repairs, and cars themselves.”

  • Anonymous

    Quite fascinating that the Mission has the lowest elasticities. Kind of gives the lie to the idea that low-income residents there are negatively affected by parking pricing forcing them out. In fact, it appears people who actually drive in the Mission aren’t price-sensitive at all — event compared to the (gasp!) Marina.

  • Brazil doesnt have parking meters…but street parking isnt free. You pay via dashboard booklets.

  • Joel

    Any word on when SFpark will be expanded? I vaguely recall hearing it’d be sometime this year…

  • Ryan Brady

    I got the impression that’s because these people live there. They gotta park because they can’t just turn around and go home. That’s why touristy areas are more elastic – people can just select a new destination.

    I don’t feel that elasticity was explained very well here though, so I’m not sure if I’m drawing the right conclusion.

  • Anonymous

    I’m not sure the inelasticity in the Misison and Marina is due to actual residents parking, since no self-respecting residential parker would actually pay to park on a metered street. That’s what Residential Parking Permits (and the 90%+ of blocks that don’t have meters) are for.

    My hunch is most parkers on streets like Mission and Valencia are probably visitors. The lower elasticity values in residential neighborhoods may be due to the lack of immediate alternatives (such as cheap off-street garage parking or low-occupancy un-priced side streets). Perhaps there’s also an awareness problem about alternatives (including parking down the street on a cheaper SFpark block).

  • Joel

    It’s important to point out that the prices haven’t changed uniformly throughout the Mission pilot area. Prices on Valencia are generally twice as high as those on Mission. I find this odd since it usually seems just as hard to find a parking space on both streets. The two streets have very different demographics and it indicates that people who park on Mission are more price sensitive.

  • Anonymous

    Great point, though I’m not sure it’s all that difficult to find a parking space on those streets during meter hours (certainly it is after 6 p.m.) The two streets do have different demographics, though differences in price do not necessarily directly correlate with differences in price sensitivity. There may also be more people trying to park on Valencia. I also suspect that many visitors who want to park on Valencia are less willing to park on Mission, which may explain why they’re not taking advantage of the cheaper parking a block away — hence the inelastic demand on Valencia.

    Either way, the fact that parking on Mission has remained fairly cheap (basically $2 plus inflation) suggests that there’s not necessarily an equity problem with SFpark, as has often been suggested.

  • Anonymous

    Interesting that the average price was relatively unchanged. Seems kind of like the early results from LA’s HOT lanes on the 110, where it looks like people’s willingness to pay more was a little overestimated.

  • I’ve long supported “Shopian” parking policy, as well as congestion pricing, but I’m finding as the long, slow auto-collapse continues I just don’t care about using market-based policy to promote marginally more efficient auto use. If average prices for parking don’t go up, we aren’t even winning on external costs. His policy is as revolutionary as joining REI.

    As mentioned towards the end of this post, an underlying problem for many cities is that insiders have found ways to avoid paying, without a tinge of guilt or legal consequence. It doesn’t matter how cleverly we calculate a price if sufficient parkers avoid it through political connections. On my street, a transit authority hi-viz vest slung on the front seat of an auto — safety first! — is enough to exempt the bearer from parking law. I don’t even know what parking price these sleazy city employees are avoiding, and neither do they. Raise the prices to free up spaces for their abuse? Whatever.

    I interested in poured concrete that makes it difficult for motorized cretins to invade every square inch of public space. I’m interested in less technology on my streets: remove the speed-maximizing, traffic-clumping signals. Bridge and border tolls are great, we should bring that bronze-age tech back to our silly free crossings. But I’ve walked and cycled in many European cities, and they all had better street environments than London. Whatever the positive effects of computerized street pricing, they pale in comparison to the physical accommodations that some other cities have been refining for decades.

    Instead of trying coax our plainly unsustainable, lethal, and all-around crappy transportation system to spew exhaust for a few extra decades, I’d like to walk calmly away from its impending self-destruction. One sidewalk extension, bike corral, and non-thru street at a time.

  • Steve

    Pricing parking in residential areas is of great importance. My city has free (and scarce) residential parking which encourages people like me to use my parent’s extra car and leave it there for weeks. We need to price it so people like me won’t needlessly bring cars there. Making parking that isn’t priced (market or not) into a pricing scheme (even if $0 is the price at certain hours) is incredibly important. It will help to utilize our supply and make clear that we don’t need to subsidize off-street parking through or zoning codes.

  • Shinji

    Congestion pricing has the effect of reducing car traffic entering downtown. Does smart parking have a similar effect? I’m afraid smart parking can rather increase people driving to downtown, because they can more easily find spaces to park their cars. Is there any research about this possibility?

  • guestguestguest

    One of the problems right now is that people with disabled placards can trump any efforts made towards smart parking solution such as demand based pricing because they are not affected by it. It’s estimated that 25% of the metered spaces in SF are occupied by placard holders.

    One idea is to have a tiered placard system like Michigan where most placard holders still have to pay the meter (I’m not sure how time limits apply though). I think only 1 or 2 percent of placard holders there applied for the tier that permitted free parking.

  • Parking Enthusiast

    SFPark’s sensor system combined with VoicePark’s mobile app technology have reduced the amount of time that it takes to find parking in SF from 6.5 minutes to 45 seconds. When SFPark expands, then the environmental impact of smart parking will truly begin.


San Francisco to Start Smart Parking Management Trial Soon

The central principle of San Francisco’s cutting-edge parking management program, SFPark, comes right from Econ 101. If there are more people looking for parking than there are parking spaces (i.e. demand is greater than supply) adjust the price of parking until there is enough turnover on a given street, or roughly one free parking space […]

Shoup Weighs in on Oakland Parking Controversy

A newly installed SFpark parking meter in San Francisco. The SFpark program was inspired by Donald Shoup’s theories on parking management. Photo: Bryan Goebel If the recent parking battle in Oakland had you thinking of UCLA Professor Donald Shoup, you’re not alone. After the Oakland City Council raised parking fines and extended parking meter hours […]

SFMTA Launches SFPark to Much Fanfare and Political Support

San Francisco launched the world’s most innovative parking pilot today, a federally-funded trial that promises to revolutionize the way cities manage and price metered curb parking. SFPark will make it easier for motorists to find spaces in busy commercial districts, while reducing congestion, speeding Muni, and improving air quality and safety for bicyclists and pedestrians. […]

SF’s Parking Experiment to Test Shoup’s Traffic Theories

SF Park Pilot Areas – Richmond and West Portal control areas not featured The Municipal Transportation Agency’s federally-funded parking experiment, SFPark, is shaping up to be the most powerful tool remaining in the city’s traffic-busting toolbox considering the San Francisco Chamber of Commerce’s criticism of congestion pricing and Mayor Newsom’s recent tempered support for the […]