IMPLICATIONS OF DYNAMIC MODELLING FOR CAR OWNERSHIP AND USE

This paper explores the idea of a dynamic model for car ownership and use, by arguing that the transition from static to dynamic models is of central importance to understanding the market for cars. In theory, dynamic models are more general, and include static models as a special case. Whether their use is worthwhile depends on the interest and importance of their actual results. The authors present some results, based on three sets of analyses: (1) some disaggregated analyses using repeated panel surveys; (2) a summary of recent published results of studies of demand elasticities, using dynamic econometric methods; and (3) a reanalysis, using dynamic specifications, of the type of data used for national road traffic forecasts. The respective findings are as follows. (1) Car ownership is volatile at the individual level, and reductions in car ownership are extensive. (2) The effects of a sustained 10% real increase in fuel price suggest that petrol price manipulation is a more efficient tool for achieving fuel consumption objectives than for controlling congestion. (3) In the long run, car ownership is very sensitive indeed to running costs. The argument is then taken one stage further, by considering the dynamics of car use as well as those of car ownership. This has consequences for traffic congestion, energy consumption, and transport policy in general. There is an appendix on estimation of car ownership and car use models. This paper was presented at the Lex Symposium on Car Marketing Forecasting, held in London on 11 May 1993.