ABSTRACT
Typically, it takes only a modest reduction in the number of drivers to unclog a congested road and get the traffic moving faster. Charging people for driving on busy roads at peak periods is the best way to achieve this; such charges encourage some people to drive earlier or later to avoid the rush hour peak, to take less congested routes into town, to car pool, to use mass transit, or to reduce the number of trips, such as by working at home or by combining several errands into one trip. Other policy approaches are far less effective at reducing traffic congestion. Increased subsidies for mass transit may help lure some people away from driving on busy roads. But this policy can be partially selfdefeating: if roads become less congested at peak period because more people are using transit, this attracts onto the roads some people who were not previously driving at peak period because of high congestion. In short, the roads may just fill up with traffic again; this is not the case under road pricing, however, as the charges discourage people from getting back into the car as congestion falls. The same phenomenon tends to undermine other approaches that do not raise the cost of driving, such as expanding cycle access or promoting telecommuting. And higher fuel taxes, which raise the costs of all driving, whether it is in urban or rural areas or occurs during peak or off-peak periods, are an extremely blunt way to reduce traffic jams; before the recent introduction of road pricing, driving in central London was not much faster than walking, despite gasoline taxes seven times as large as those in the United States.
