Has the Historical Growth in Car Use Come to an End in Great Britain

This paper presents an analysis of recent trends in automobile usage in Great Britain. The authors note significant changes in license-holding, car ownership, and driving levels, and related socio-demographic patterns over the last two decades. Of particular interest is the stabilization in driving mileage per capita around the year 2,000, where economic output grew by more than 15%. The authors’ results suggest that the traditional strong correlation between gross domestic product (GDP) and traffic may have shifted since the recession of the early 1990s to a closer relationship between the size of the total workforce and traffic, rather than between GDP and traffic levels. Research into automobile use patterns in other industrialized nations has identified similar trends as those that are reported in this paper. Such results have led to a debate about the underlying causes of this stabilization, each of which has different implications for traffic forecasting and future transport policy. For example, if the stabilization in driving is occurring because of a saturation of personal travel demand, with most people driving about as much as they please, then current personal levels of automobile use will remain unchanged in the future, and not be affected by changes in network conditions. On the other hand, if demand is reflecting current network performance and congestion levels, then it will be reduced by any substantial deterioration in average traffic speeds, or increased if network performance improves. In either case, it appears that there is now only a weak relationship between automobile ownership and average car travel per capita. Historically, this relationship has been a major component in GB traffic forecasts, and so needs to be urgently reassessed. The conclusions reported in this paper are largely based on simple univariate descriptive analyzes and on a comparative statics framework for the analysis of change. Both these features are open to criticism on the grounds that they may amount to a misspecification of either the cross sectional or dynamic properties of the data (or both). The next stage in the research will therefore involve the application of more sophisticated statistical techniques to address these potential concerns.