The basic theory on urban transport pricing and public transport optimisation has been reviewed, highlighting the main methodological contributions found in the literature. The review indicates that there are sufficient theoretical grounds to set public transport fares below average operator costs, and therefore an optimal subsidy seems justified on first best grounds, second best grounds, or both. Nevertheless, the answer as to which is the appropriate level of fare and subsidy does depend first on the modelling approach (first or second best, what externalities are included, possibility of day of time substitution), and second, on the actual context or city. As a result, estimated optimal bus fares vary from negative figures to values that actually cover operating cost (without considering capital investment). The optimisation of bus frequency and size is also extensively discussed, with reference to elements that influence their optimal level such as active capacity constraints and the setting of road pricing.
This thesis adds to this body of literature by analysing the influence of non-motorised transport on public transport pricing, introducing new decision variables like the choice of a fare collection technique and level of infrastructure investment for bus corridors, and by providing a more comprehensive methodological framework for the introduction of bus congestion and passengers crowding in the optimisation of bus services and setting of fare and road pricing. These elements are presented in the next chapters.
47
Chapter 3
3 Multimodal transport pricing: the influence of non-motorised modes
Multimodal Transport Pricing: the Influence of Non-
motorised Modes
143.1 Introduction
Second best pricing models that take into consideration only two modes - cars and public transport (bus or rail) - have found that subsidies for public transport are desirable, with fares below marginal cost due to the underpricing of cars. However, as put forward by Kerin (1992), this approach neglects the existence of other modes, notably walking and cycling, that play an important and growing role in urban transport systems, especially for short trips. Disregarding non-motorised transport is a growing concern because low bus fares not only deter some drivers from using their cars, but also divert walkers and cyclists onto trains or buses, which is not necessarily a desirable outcome. As such, a pricing model that also includes non-motorised transport seems desirable in order to estimate the impact of these modes on (possibly decreasing) optimal subsidies for public transport. Even though there are no analytical models that address the issue of the influence of non- motorised transport on urban transport pricing policy, we do find that walking and cycling are considered as travelling alternatives in applied models (Safirova et al., 2006; Proost
48 and Van Dender, 2008), but no attempt is made to identify how the design of the pricing instrument would change by considering or ignoring walking and cycling.
In this chapter, a multimodal pricing model is developed, including three modes - automobile, public transport (either bus or rail) and non-motorised transport (either walking or cycling), with the objective of maximising social welfare. This model extends the previous literature by identifying the role that non-motorised transport can play in the optimal setting of fares for public transport. Emphasis is given to the effect of bus demand on car congestion when both modes share the right of way, and the way in which the optimal fare, frequency and vehicle size should be determined when the capacity constraint is binding for a public transport service, i.e., when demand meets the capacity offered by the operator (see Section 2.4.3). We also include in the framework the cost of externalities other than congestion, such as accidents, pollution and noise, and the toll collection cost, all of which increase the marginal cost of motorised transport compared with walking and cycling alternatives. The emphasis of this chapter is not on the determination of the empirical value for optimal fares and subsidies (where applicable) but with the economic principles behind them15.
With reference to the outcomes of this multimodal pricing framework, it is shown that the effect of considering non-motorised transport alternatives on optimal public transport fares depends on the demand substitution between modes; the stronger is the demand substitution between public transport and non-motorised modes, relative to the substitution between car and public transport, and car and non-motorised modes, the more likely it is that a higher public transport fare would result from the allowance for the role of walking or cycling on fare setting. On the other hand, a capacity constraint on public transport plays a role in optimal pricing only when the transport capacity cannot be set at its optimal level. Finally, the internalisation of externalities other than congestion is
15 For numerical comparisons on fares and subsidies among several studies, see Proost and Van Dender
49 likely to increase optimal fares and road charges, therefore increasing the generalised cost of motorised transport modes relative to a non-motorised alternative.
The remainder of the chapter is organised as follows. Definitions, assumptions and formulation of the social welfare maximisation model are presented in Section 3.2. The first best and second best problems are solved and analysed in Sections 3.3 and 3.4, respectively. Section 3.5 extends the model by including external costs other than congestion and toll collection costs into the social welfare objective function. Finally, a summary and the main conclusions are given in Section 3.6.