– Can parking be managed better?
Posted: May 12, 2011 Filed under: Cars & traffic | Tags: Donald Shoup, Melbourne City Council, parking, pricing, San Francisco, SFpark 10 Comments
SFpark, San Francisco's experiment with dynamic pricing of parking spaces
I’m disappointed by the discussion of parking in Melbourne City Council’s draft Transport Strategy Update 2011-2030 (note – it’s a big download). There’s an opportunity to improve the efficiency of parking space allocation through using technology and pricing in combination, but Council seems content to pass on it.
The broad thrust of the discussion in the report is that the number of on-street parking spaces will decline over the next 20 years to enable public transport and amenity improvements to be implemented. Council is mindful of the impact this will have on its own revenues and those of local businesses, but is persuaded by the social and environmental benefits.
A key recommendation in relation to on-street parking is that Council “will implement new parking technology systems that allow payment without requiring parking machines or meters (and) will remotely sense and assess parking occupancy”. Surprisingly, this recommendation is entirely unsupported by any explanation or discussion. As far as it goes, it nevertheless sounds good – it’ll lower costs by eliminating the need for parking inspectors and it’ll give drivers more flexible payment options.
What seems to be missing, though, is the opportunity to provide drivers with real-time information about parking availability. More importantly, it squibs the opportunity to improve efficiency in allocating parking spaces by setting a price that’s responsive to demand.
The current pilot project just introduced in San Francisco, SFpark, gives a sense of what can be done. Like Melbourne City Council’s plan, it involves sensors that automatically sense if a parking space is empty. SFpark however will convey that information to drivers electronically via a smartphone app. That should reduce the time drivers spend cruising for parking. According to Donald Shoup, a Professor at UCLA and the author of The High Cost of Free Parking, several studies have found that cruising for curb parking generates about 30 percent of the traffic in CBDs in the US. He cites a study he did of a 15 block district in Los Angeles where cruising for on-street parking created 950,000 miles of excess vehicle travel per annum, in the process consuming 47,000 gallons of petrol and producing 730 tons of carbon dioxide.
But the real innovation of SFpark is that prices are adjusted in real time in response to rises and falls in demand. The objective is to ensure that, on average, there is at least one vacant space in each city block:
SFpark will adjust meter prices based on demand to encourage drivers to make trips in off-peak hours and to use parking lots and garages. While high-demand spaces will gradually go up in price, other spaces will decrease in cost……Once a space is found, longer time limits and new meters that accept credit and debit cards will make it easier to avoid parking tickets.
I’m not convinced eliminating cruising would necessarily deliver the same benefits in Melbourne as in LA. The extent of cruising varies between places, depending on variables like the supply and cost of off-street parking. A new study of cruising behaviour in The Netherlands by researchers from the Tinbergen Institute, using a nation-wide sample, found that 30% of car trips involve some cruising. The average trip duration is 20 minutes, however the average cruise time is only 36 seconds. I’d expect cruise time to be higher in Melbourne’s CBD although probably not as high as the average 8 minute cruise time that Shoup found in the US.
However that’s not the key issue. The principal benefits of variable parking pricing would be in prioritising higher value trips and in promoting turn-over of spaces. Those with the strongest reasons for driving would be prepared to pay more – others would park further away. The objection that pricing is inequitable fails to see that people on all income levels will at various times have powerful reasons to pay more for parking – using a dynamic price mechanism means on average there is a greater likelihood a space will be available.
There are also other potential benefits from the way pricing is implemented. For example, rather than incur a hefty fine for over-staying the specified maximum time by a few minutes, drivers could elect to pay (by smartphone) a sharply escalating tariff that gives them, say, an extra 15 minutes at triple the standard rate. Smart meters and smartphones open up a host of possibilities.
Melbourne City Council’s reluctance to use pricing of parking space as an efficiency tool is not surprising – governments usually prefer expensive capital works programs over measures that use existing infrastructure more efficiently but are politically difficult. Council’s Transport Strategy Update 2011-2030 also discusses road pricing but, after pointing out the downsides of congestion and the strong rationale for pricing on efficiency grounds, elects to do nothing about it, solely on the basis of these lame words: “A key issue for the City of Melbourne in considering changes to road pricing is maintaining and enhancing access to the city for a wide variety of trip purposes”.
The Melbourne Urbanist 
Alan, I’ve been following your blog for over a month now, and really appreciate your effort and erudition. Many thanks.
I’m delighted to learn about SFpark through today’s post. I read Shoup a few years back, and he makes so much sense. I’m sure the City of Melbourne will be watching San Fran. I expect by 2016 we’ll see a similar system in operation here.
Variable pricing: Obvious, isn’t it? Just like the vegetable market marks things down in the hour before closing to clear the stock.
But listen to the squeals about ‘profiteering’ if a public authority tries to do it in relation to something that people are used to thinking of as a public resource. An attitude which is reinforced confusingly because they get free parking in many private settings like the shopping mall.
In the example, can the rate you’re paying change while you are away from the car?
Agree that the report sounds weak where it describes congestion charging then conspicuously omits to reach any conclusion or opinion.
My interpretation is the rate is fixed once the flag falls — any newcomers would pay a higher/lower rate depending on movements in
the pricedemand, which I think would only change in fixed chunks of time.Wouldn’t it be great if you could use the SFpark smart phone app to book the parking spot for you as you approach, to avoid a glut of cars arriving at the same time? You’d start paying as you booked, and save other drivers from needlessly traipsing to a spot that was already filled.
Great post Alan.
Quick note; a smaller version of Council’s draft transport strategy is available here – http://www.melbourne.vic.gov.au/AboutCouncil/PlansandPublications/strategies/Pages/transportstrategy.aspx . It is 15mb instead of the full Committee report which is 61mb.
Cheers
My understanding was that pricing was reviewed and changed monthly or quarterly. Pricing would still vary by time and duration of stay. This allows some certainty in the medium term; you’d not likely to leave the house, turn up at a meter and it’s double the price it was when you left your house. It’s still a lot more dynamic than ‘typical’ pricing which tends to updated annually for the new financial year and a flat rate for the metered hours.
To attract drivers and to increase the availability of spaces, rates at garages will vary by the time of day and will be adjusted quarterly in response to demand. For the first two SFpark garages to undergo the demand-responsive pricing, hourly rates will decrease or stay the same, depending on time of day, beginning this month. The rate changes were based on evaluation of the occupancy in these garages.
Street pricing is more dynamic than the garages, however:
Parking meter rates will vary based on time of day, and rates will be adjusted over time in response to demand. Rates will be adjusted on a block-by-block basis using the occupancy data provided by the parking sensors that have been installed in metered on-street parking spaces in the SFpark pilot areas. Meter pricing includes:
• Distinct rate periods: In order to improve parking availability in metered parking spaces, SFpark meters may charge different rates based on the time of day.
• Rates will respond to demand over time: Starting at traditional rates for parking meters (between $2 and $3.50), rates will decrease or increase by 25 cents no more than once per month.
With the gradual changes you end up having smoothed data for occupancy, so it seems like a good starting strategy to me. As smart phones become more prevalent, perhaps more rapid changes could be used. I think though the SFPark poliy is an excellent method for introducing more demand-responsive pricing and I’m sure there will be a lot of interested parties watching the results.
I’m just surprised Melbourne City Council hasn’t mentioned this at all (pricing) as part of their ‘occupancy sensor’ rollout… hopefully the pricing technology can be integrated later.
I agree regarding the Melbourne City Council’s lack of vision with the use of the parking occupancy sensors. Perhaps smart meters can be added in to the occupancy sensors at a later stage.
My understanding of the SFPark system is that the pricing isn’t adjusted in real time. Street meters are updated monthly, and garages quarterly. Street meters, at least, won’t move by more than 25c per hour each update, allowing time to transition to higher/lower prices (ie no sudden shocks if you’re not a regular visitor to the metered areas). Prices still vary by time of day and duration. The proposed set up in my mind represents a sensible transition to perhaps real time, or more fluid, price adjustments at a later stage, without scaring people off variable pricing.
In SFpark pilot areas, meter rates will vary based on time of day, and rates will be adjusted over time in response to demand. Rates will be adjusted on a block-by-block basis, using the occupancy data provided by the parking sensors that have been installed in all on-street parking spaces in the SFpark pilot areas.
Adjusting Rates 1. Meter operational hours will be split into distinct rate periods
In order to help ensure that parking is available in metered parking spaces, SFpark meters may charge different rates based on the time of day in which a car is parked. To facilitate this demand responsive time-of-day pricing, the meter operational hours will be split into distinct rate periods throughout the day.
Rates for parking meters will change gradually and periodically based on demand. Changes to the rates will be made no more often than once per month. When prices are updated, the update will be made on or near the first day of the month. At the outset of the SFpark program, rates at meters will respond to demand as shown by occupancy in the previous month. As SFpark continues to collect occupancy data from the parking sensors, however, SFpark staff will consider including occupancy data from earlier months and years to assist in making pricing determinations
So over time the occupancy data can be refined and smoothed, and probably adjust for monthly/seasonal demands too (even specific weekends, eg holiday/shopping). Be interesting to watch; hopefully the implementation goes well.
Sorry, link for policy:
Click to access FINAL-On-Street-Pricing-Policy.pdf
How about we just eliminate street parking all together in the CBD? There are car parks all over the CBD if people really MUST drive, but no street parking might mean some people would reconsider driving into the CBD all together. And that combined with more road space previously taken up by parked cars could be a good thing.
[…] – Can parking be managed better? […]