Charles Lindsey

Tradable permits have been proposed in the literature as a tool for controlling road traffic congestion. Unlike with tolls, travelers in aggregate do not incur net out- of-pocket costs if permits are distributed free. Moreover, if travel conditions do not change, tradable permits can be designed to support the same traffic flows as tolls. However, the two instruments are no longer equivalent if conditions fluctuate and neither permit quantities nor toll levels can be adjusted accordingly. In this thesis, we evaluate the efficiency of tradable permits and road tolls in controlling traffic congestion during a morning peak travel period on a stylized urban road network with daily commuters who travel from their homes to their workplaces. We compare the economic efficiency of the two instruments when road link capacities are susceptible to reductions due to incidents, bad weather, road repairs or other shocks. We apply the dynamic traffic network simulator METROPOLIS in which travelers make mode, departure time and route choice decisions. Fixed permit quantities and toll levels are chosen to control the number of vehicle trips crossing a cordon into the city centre with the goal of maximizing welfare defined by expected social surplus. Simulations are performed for a range of scenarios in which road capacities are correlated or uncorrelated, and suffer reductions of varying magnitude. Most of our analysis suggests that road tolls outperform tradable permits. However, in some instances tradable permits do better; in particular, when the road links that bypass the centre of the network experience capacity fluctuations.

View record