3 Urban Travel Forecasting in the USA and UK

Nearly five decades ago in the United States, urban transport was first subjected to a systematic analysis that became the forerunner of modern transport planning. The reasons for the emergence of this new kind of planning activity are well known and will not concern us here. Within ten years both the approach and the embedded travel forecasting models had been implemented at many major urban centres of the world including several in Europe. Although subject to refinements and reinterpretation, the models in widespread contemporary use in urban, regional, national and international transport studies are recognisably the descendants of those four-stage procedures defined in the late 1950s. There have always been criticisms of travel forecasting models. In the early 1960s criticism was muted, but by the late 1970s voices in opposition had turned into a clamour. Following the celebrated San Francisco Bay Area lawsuit and the subsequent Clean Air Act Amendments of 1990 (Garrett and Wachs 1996) the critique of “conventional” models took on new urgency. And yet, nearly 50 years after its inception, the four-stage travel forecasting procedure is still in widespread use in large metropolitan areas. The purpose of this chapter is to examine some of the tensions and conflicts between the multiple states of the art and the states of practice. We interpret the former as comprising the theory of models, conceptions of model systems, model prototypes, and possibly advanced model systems being applied in practical studies which may be on the point of wider acceptance. The latter are characterised by widely available practical applications. We shall structure our discussion of state of the art developments around two related themes: firstly, those evolutionary approaches in which the multi-stage procedure has been significantly refined; and secondly, those attempts beginning in the early 1970s to model travel behaviour “from the ground up.” In Sect. 3.2 we