Private operators and time-of-day tolling on a congested road network

Private-sector involvement in the construction and operation of roads is growing around the world and private toll roads are seen as a useful tool in the battle against congestion. Yet serious concerns remain about exercise of monopoly power if private operators can set tolls freely. A number of theoretical studies have investigated private toll-road pricing strategies, and compared them with first-best and second-best public tolls. But most of the analyses have employed simple road networks and/or used static models that do not capture the temporal dimension of congestion or describe the impacts of tolling schemes that vary by time of day. This paper takes a fresh look at private toll road pricing using METROPOLIS: a dynamic traffic simulator that treats endogenously choices of transport mode, departure time and route at the level of individual travellers. Simulations are performed for the peak-period morning commute on a stylized urban road network with jobs concentrated towards the centre of the city. Tolling scenarios are defined in terms of what is tolled (traffic lanes, whole links, or toll rings) and how tolls are varied over time. Three administration regimes are compared. The first two are the standard polar cases: social surplus maximization by a public-sector operator, and unconstrained profit maximization by a private-sector operator. The third regime entails varying tolls in steps to eliminate queuing on the tolled links. It is a form of third-best tolling that could be implemented either by a public operator or by the private sector under quality-of-service regulation. Amongst the results it is found that the no-queue tolling regime performs favourably compared to public step tolling, and invariably better than private tolling. Another provisional finding is that a private operator has less incentive than does a public operator to implement time-of-day congestion pricing.