Car Traffic Elasticities<BR> A Spatial Panel Data Analysis of French Regions

This paper revisits the elasticities of car traffic using a panel dataset of the twenty-one French regions over the period 1973-99 considering spatial interdependences. Two spatial specifications including random or fixed effects are retained. These models have been estimated using the Maximum Likelihood Estimator (MLE) as an iterative two-stage procedure. We find that significant spatial interdependences characterise the regional car traffic per vehicle. Following the spatial specifications, direct elasticities with respect to kilometres driven for the price of motor fuels fall in the range 0.14 to 0.09. The other elasticities vary respectively from 0.30 to 0.57 for household income per capita, and from 0.54 to 0.28 for the number of cars per capita, confirming that there are more and more cars, but that each car is less and less used. © 2011 LSE and the University of Bath