Parking is the kind of thing most people don't think about much. Except when they can't find a spot, causing them to circle the block, their temper rising as they muck up traffic.
So what if there was a better way? A three-year pilot project in San Francisco – where they kept adjusting the prices, aiming to keep some spots free in every garage and block – found that availability went up, even as the average cost to park went down.
Jay Primus, who ran the SFpark experiment, was in Toronto this month for meetings with the city's transportation department. He sat down with The Globe and Mail to explain the San Francisco approach.
An oft-cited statistic in urbanist circles has it that nearly one-third of downtown traffic is people looking for a parking spot. This is somewhat misleading as it lumps together studies – some nearly a century old, in a variety of cities – but few would dispute that traffic flow is hurt by people looking for a spot. There's also a safety factor: By searching for parking as they cruise along, these are, by nature, distracted drivers. So if giving up valuable urban real estate to create more parking spots is not an attractive option, what about pricing based on supply and demand? Many cities already acknowledge this basic economic theory by charging less to park at off-peak times, but San Francisco went much further.
Helped by a federal funding grant, SFpark set out to gather a wealth of data about exactly where and when people were parking. The pilot, which covered the central business district and five other neighbourhoods, used in-road sensors to monitor street parking spots and swing-arm data to track paid-garage use. Every few months, this data was used to shift prices, based on area or time of day, with the goal of keeping parking-spot occupancy in the 60- to 80-per-cent range. At the same time, they worked to improve the driver experience, making it possible to pay by credit card and smoothing out the price differential between street and garage parking. They also spruced up the garages, making them less dingy, and improved wayfinding so it was easier for people to locate them.
Under the new model, the average time spent looking for a parking spot was slashed to six minutes from 11.5 minutes. There were significant reductions in double-parking on some roads, on which Mr. Primus said there were "very noticeable improvements" in transit speed and reliability. Another difference: The amount of sales tax generated in the pilot area went up compared with the rest of the city. The specific reasons for this haven't been proved but one possibility is that the perception that parking would be less onerous made these areas more attractive. And people seemed to roll with the cost changes, with Mr. Primus saying zero complaints were received over 13 price adjustments.
Parking prices dropped – an average of 12 per cent – but so did the amount the city raised by issuing tickets as, amid efforts to make it easier to pay for parking, the scofflaw rate went down. Over all, parking revenue went up 3 per cent. As for the argument that charging more for desirable parking is regressive, that has less traction in San Francisco, where revenue goes to transit, which is more likely to be used by lower-income residents.
The program itself cost about $30-million (U.S.) to operate, although Mr. Primus said that the price for this kind of technology is coming down fast. And there are even cheaper ways to do it. "Cities don't need perfect data in order to perfectly manage parking," he said. "Some data about occupancy and availability will be enough to start managing to get it about right, or certainly more right than most cities do now."