Now that it seems Mayor Bloomberg is getting really serious with the Manhattan congestion charge, to be about US$8 for every incoming vehicle, it is opportune to assess the London experiment. The following lines are largely based on a piece by Jonathan Leape, published in the Journal of Economic Perspectives, a peer-refereed publication of the American Economic Association, unfortunately not available online for recent publication years.
The premise in London was a desperate situation with an inescapable gridlock in the central district: day-time average speed was 14.3 km/h compared to 32 km/h at night, public transportation and congestion were viewed by Londoners as more important problems than crime. Charging for entering Central London had been discussed for a long time but was thought impossible to implement. The 2000 election of Ken Livingstone to Mayor changed this. Within three years, an impressive system of cameras that recognize license plates was in place and a £5 daily charge was implemented in February 2003, increased to £8 in July 2005. Within a year, day-time average speed increased to 16.7 km/h, the traffic of cars decreased by 34%, trucks 7% and vans 5%, however, taxi traffic went up 22%, and 6% for motorcycles. Other measured of congestion also showed significant relief.
In view of this success, and the welcome net revenue it brings to London, many other cities in Europe as well as New York City are considering a similar congestion charge. One should keep in mind, however, that not everything is rosy in London, though. Circumstantial evidence shows that traffic has increased close to the tariffed zone, presumably from cars driving around it or looking for parking. Also, it appears that cars that have bought a daily pass tend to take more advantage of it than before. Finally, an increasing number of vehicles that are exempt from the charge undermines its impact.
Again and again, it has been shown that congestion cannot be solved by building roads, as the latter just encourage more cars to use them. Besides, this was not an option in London. But incentives can work, making people pay makes them take notice. The availability of a good public transportation system is a requirement, to serve as an alternative to private transportation, and the ability of having a well-defined tariff area is a plus. All criteria apply to London, except maybe the tariff area: it is well-defined by a ring road, but the neighborhoods just outside clearly suffered.
In the case of New York City, the congestion charge has all but one reason to work well: Manhattan is very well serviced by public transportation, it is a well-delimited island that should not lead to a negative externality on neighboring boroughs and towns. But will car drivers take notice? I am not sure about this, considering the number of them who are willing to spend that much time caught in traffic every day while public transportation is available. At the going wage in Manhattan, the opportunity cost of their lost time is much higher than $8, and they seem to think sitting in gridlock is worth it. And looking at their car, $8 a day will not hurt them. But even if it does not reduce congestion, it will bring revenue from the city from those using Manhattan but living outside.
And I do not want to hear about those that do not live close to public transportation and need to work in Manhattan. They chose to live where they live. It is less expensive there for a reason. Take housing closer to rail/metro/bus/boat/work, sell the car, and they will be better off despite higher rent or house costs.
Update (January 24, 2008): This post has been translated to Italian on InterBlog.