Introduction vs. price change of road toll – a panel data analysis of revealed preferences

Abstract

Cordon-based congestion charging systems effectively reduce traffic, with initial implementations often achieving 15–20 % reductions in vehicle volumes, as observed in Singapore, London, and Stockholm. However, subsequent toll increases typically produce much smaller elasticities, a phenomenon known as "Large Elasticity at Introduction" (LEI). This suggests that introducing tolls on previously free roads triggers stronger behavioral responses than adjusting toll rates on already tolled roads.

This study explores LEI using mode-choice data from a panel of 2814 commuters in the Stavanger urban area, collected before and after substantial changes to the toll-cordon system in October 2018. Employing random utility models, we test linear and log-transformed cost specifications to investigate the roles of diminishing sensitivity and the zero-price effect.

Our findings indicate that diminishing sensitivity to cost, captured by a log-transformed specification, is more critical to explaining LEI than a strict zero-price discontinuity. Real-world data confirm strong responses to new tolls at city-center cordons (16–18 % traffic reductions) but weaker responses at previously tolled municipal borders (4 %). These results emphasize that LEI is largely driven by non-linear cost sensitivity, suggesting that introducing low tolls in areas with robust modal alternatives can achieve substantial congestion reductions without requiring steep price hikes.