An interesting issue in contemporary travel behaviour research is whether the transportation demand has to be considered purely derived from underlying activity patterns or whether a utility is also associated with travelling per se. In the latter case, substantial amendments of current planning models would be needed to represent this phenomenon adequately. Earlier research consistently gave evidence of the existence of this specific utility, but its quantification is hindered by a specific measurement problem because survey respondents tend to mingle the utility of travelling and the utility of reaching a destination. The present work defines a methodology to quantify the decrement in the specific utility of driving a car due to the presence of difficulties and self-limiting behaviours. This is in turn responsible for an alteration of driving frequency. A structural equation modelling technique is used for the analysis. The structural submodel represents the complex relationships between socioeconomic variables, specific utility, and driving frequency. The measurement submodel defines the specific utility on the basis of reported self-evaluations concerning physical fitness and self-limiting behaviours while driving. An application of the method based on data collected in the 2002 National Transportation Availability and Use Survey is presented. The results show that the decrement of specific utility (which can be seen as a disutility) of driving a car has an important impact on the frequency of performing this activity compared with the derived utility that is customarily modelled through socioeconomic variables.
Abstract