There are two totally different ways of approaching the measurement of transport subsidies. The first is to compare total costs and total revenues, in order to see how far users pay the total costs, and how far explicit or implicit subsidies are provided. Often only money costs are examined, but we will argue that in a sector such as transport, where externalities are very important, total social cost should be examined. The second approach is to consider the relationship between marginal social cost and price, and to regard the failure of price to cover marginal social cost as a subsidy. The first approach is often seen as important politically and in the consideration of equity issues. It is the second however that is appropriate in the consideration of economic efficiency. In sectors which are broadly characterised by constant returns to scale the two may give a similar picture. Thus it is common to emphasise the first approach to subsidies. In transport, however, both infrastructure and operations are typically characterised by major economies of scale, whilst many externalities, such as congestion and noise, are important and strongly non linear. The result is that marginal and average social cost are highly divergent. This paper considers evidence on both approaches to the level of subsidy. It presents the results of a study of social costs and revenues from transport for all Western and some Central European countries. It then describes the state of the art methodology for quantification and valuation of the environmental costs of transport, and reports on a review of studies that have considered the implications of inefficient pricing in terms of distorted demand for transport.

URL: http://www.thepep.org/ClearingHouse/docfiles/OECD%20transport%20subsidies.pdf