A much used phrase in current transport policy is "reducing the need to travel". Unfortunately this seemingly innocuous concept is rarely defined, and without a clear understanding of what it should mean, it is likely to remain a pious hope for policy-makers while allowing travellers to continue in their ways, convinced that their journeys are "necessary". The aim of this paper is to stimulate some fundamental discussion about destination choice and hence a definition of need. At the same time as politicians hope to reduce travel need, there is a concern that accessibility should in some way be enhanced. Accessibility also remains an elusive concept, though recent efforts have been made to tighten up the definition. If all travellers were homogeneous, and all destinations serving a given travel purpose provided identical facilities, there would be no point in travelling further than the minimum (generalised) cost destination. Even in this extreme case, "unnecessary" travel could result if the costs faced by travellers differed from true resource costs, as is widely considered to be the case. In practice, of course, destinations do differ in terms of "quality", individuals differ in their tastes and perceptions, and modellers dealing with destination choice almost invariably aggregate (to zones, e.g.). All these contribute to the "randomness" in utility-based models of destination choice. In practice, while allowance for size variables is more or less standard, there is very little empirical evidence on the scale of the non-homogeneity effect, nor does the existing theory allow it to be partitioned between different sources. The paper addresses these issues in terms of both their modelling and policy implications. It also discusses questions of evaluation, where current practice would appear to be at fault in making no allowances for changes in the cost of not travelling.
Abstract