that the San Francisco County Transportation Authority (SFCTA) is proposing two options for introducing congestion pricing into San Francisco, both of which would represent the first genuine cordon based congestion charging scheme for the USA.
The SFCTA presentation
as part of its public outreach strategy explains the options in moderate detail. If it comes to pass, it would be a revolutionary step forward in the implementation of more sophisticated road pricing in the United States.
Road pricing in the USA to date has been most commonly seen in traditional toll roads, almost all with a mix of manual toll booths and electronically controlled toll booths with barriers. More recently, some toll roads have moved to electronic free flow. However the shift to pricing roads to manage congestion has to date only really been seen in either toll lanes/HOT lanes (where the user can choose to use untolled parallel lanes) or the implementation of peak time charges on existing tolled crossings (e.g. Golden Gate Bridge). No US city to date has dared put a cordon around the CBD to force charges for all vehicle entry beyond that point at fixed times.
What are the proposals?
Peak time charges of US$3 between 0600 and 0900, and 1500-1900 across the proposed cordon options, although it is unclear whether the charges apply in one direction only (reversing AM/PM) or both directions at both times.
A tight Northeast cordon around downtown San Francisco would allegedly result in 12% less car trips and annual net revenue of US$60-US$80 million p.a. A far larger "southern gateway" at the county's southern border would reduce trips by 5%, and generate net revenue of US$60 million p.a.
Now there are a few issues with all of this. Firstly, SFCTA has no legislative authority to impose any charges anywhere, so would need a mandate to do so, but secondly, there will be a real need to do some due diligence on the financial assessment.
Having had some experience in these matters, I'd question the costs (not just lost revenue) for the long list of discounted and exempt vehicles. While taxis, buses and emergency vehicles are not difficult to exempt, discounts for "low income" drivers will be fraught with administrative costs because of the potential for fraud. "Disabled drivers" have a similar, though smaller scale issue. There are also no costs for changes to the road network which are necessary when a cordon is placed, and motorists will approach it and want to divert around it. This is a risky business, and it is important to get the costings right, particularly in the set up phase, because costs will decline significantly after a few years. An interesting dimension completely ignored is revenue from enforcement, because it would be politically unpalatable to suggest that maybe 1 in 20 users will be fined. The claim that collisions will drop tends to ignore that higher speed collisions usually cost more in terms of injuries and deaths, a key error in some very basic evaluations of road pricing.
Another point is whether any free through-routes are needed, given the proposal includes the Bay Bridge and presumably through traffic using the Bay Bridge and associated freeways to bypass downtown San Francisco. A case could be argued either way about that.
Still, notwithstanding these and some other issues (the claim the London Congestion Charge has reduced congestion in London by 33% is a wild misuse of a statistic. The actual statistic is that it reduced car traffic in the charged area by 33% when it opened, which does not correspond to a reduction in congestion across London, as the charged area is small), the proposals are radical for the US.
If I was to choose one of them, I'd pick the Northeast one, if only because cordons are very blunt tools and really only suitable for CBDs where the public transport (and active mode) options are realistic for most of the users. A county boundary charge with a residents' discount would nullify the impacts and impose a far more artificial boundary than one which largely corresponds to central San Francisco commercial activity. Consideration should also be made about the charging periods and perhaps shoulder periods to avoid pre/post peak rushes.
However, a key element to any success will be how the revenue is used, and how the implementation is managed. No US city has implemented anything remotely like this, so experience from outside the US will be essential. Also, whilst the temptation will be to use the money mostly for public transport, it is critically important for motorists to see that they get something in return - which in this case ought to be, at least, far better maintained road surfaces, signage and improvements in traffic signal phasing. Motorists ought to see some of the financial benefit from the charge, which can also include small scale works on improving intersections and improving pedestrian facilities downtown (which motorists also use).
Good luck to San Francisco. A lot of detailed work will be needed to determine exact system design, product design, procurement approach, governance (who sets and reviews charges, on what basis, how is the revenue managed), enforcement, test phasing and the communications strategy for implementation. Where New York once thought it might pioneer, San Francisco may actually get there if it does the work thoroughly, (peer reviewed) and gets buy in from motorists and businesses.
There are probably few US cities that could implement such a cordon and have a significant positive impact on traffic and mode shift, without also encouraging businesses to relocate outside the cordon. It is, after all, a fairly blunt implementation of congestion pricing. However, the potential is there to have more disaggregated pricing at different times and directions.
Then it can be said that the first real congestion pricing scheme (not just lane pricing) in the US will have been implemented in San Francisco, and other cities and states will be watching carefully. If it succeeds, it wont fix congestion throughout the Bay Area, it isn't a magic bullet, but it should make a worthwhile difference.