This paper investigated road traffic congestion caused by speed differences using both analytical and numerical simulation models. Especially outside peak hours, speed differences are probably one of the most important reasons for congestion. Some main conclusions are that optimal tolls for slow vehicles are higher than those for fast drivers, that the marginal external costs and the optimal tolls for slow drivers are actually decreasing in the equilibrium number of slow drivers, and that "platooning" may become an attractive option especially when the desire for allow speed is caused by a lower value of time. (A)
Abstract