Congestion Charging and the Optimal Provision of Public Infrastructure: Theory and Evidence

Abstract

The paper provides a theoretical framework for analysing the effects of public infrastructure provision on private sector productivity using the example of a transport network. Public infrastructure such as a transport network is assumed to be a (congested) public good. When the provision of this good is at the long run equilibrium level, consumers pay a price which reflects the (individually-determined) marginal productivity of the good and the supplier is also recovering all its opportunity costs. In the traditional literature on transport congestion (Walters, 1961; Mohring and Harwitz, 1962), the concept of infrastructure capacity is often defined in term of the maximum level of traffic flow, which is more of a usage concept rather than a ‘capacity’ concept. Congestion is then defined in terms of the difference between the marginal social cost of this traffic flow and the marginal private costs. There has been some debate in the literature on the way travel demand in general, and traffic congestion in particular, has been defined in terms of traffic flow because this will tend to give an ambiguous definition of the concept of ‘congestion’ in some cases. An alternative measure for the concept of traffic demand (and supply), and of ‘congestion’, is in terms of traffic density or volume rather than in terms of traffic flow. In this paper, we explore this alternative definition of ‘capacity utilisation’ and of ‘congestion’ in terms of traffic density. We arrive at an alternative definition for the concept of optimal congestion tax that turns out to be more robust. This is because it can be applied, not only to the situation of ‘low congestion’ but also to the case of bottleneck or ‘hyper-congestion’ which is not well analysed in the traditional literature. The paper also illustrates this new concept with some numerical calculations based on empirical observations on an actual road network.