Empirically, travel time averaged across a population is invariant at close to one hour per person per day. Nevertheless, standard transport analysis treats travel time as a variable, scoring the value of time savings in cost-benefit analysis of proposed new infrastructure, and minimising generalised travel time in transport modelling. This paper argues that the in variance of travel time needs to be explained, and that such an explanation might involve a marginal approach to travel time, as well as the possibility of there being an intrinsic utility to travel. These concepts are consistent with economic principles but not with the current practice of transport economics. It is further argued that the benefits of transport system improvement are to be found in the additional access afforded, not in notional travel time savings; and that conventional transport models that pay no regard to the in variance of aggregate travel time may mislead. It is suggested that in transport analysis, aggregate travel time should be treated as a constant, not a variable. (Author/publisher).
Abstract