Estimation and specification of the effects of parking costs on urban transport mode choice

The effect of changes in parking fees on urban transport mode choice is investigated to evaluate the claim that parking taxes are an effective substitute for road pricing in influencing congestion. It is shown that previous analyses of the modal choice decision with respect to parking costs have misspecified the model, resulting in biased predictions. Binary logit analysis is used to estimate the traditional and correctly specified models. Elasticities for four policy oriented variables are calculated. The elasticities provide a measure of the bias from misspecification and indicate the most effective policy variables to reduce auto use.