Panel data analysis of drivers under an evolving cordon tolling system

This paper analyzes a panel dataset of 4011 anonymized car owners in Norway. We observe where they live (on neighborhood-level) and how much they drive. We combine this with data on tolling costs. Over the sample period 2017–2020 we observe four different tolling regimes, with changing tolling costs for peak and off-peak driving in Oslo for different types of cars. We employ both fixed-effects regression and production analysis combined with difference-in-difference methods to analyze effects of cordon policies on driving. The empirical results show that peak and off-peak driving are complementary goods, and consequently that increasing peak tolls can decrease off-peak-driving. Responsiveness varies significantly by geographical location and age, with more responsiveness to the tolls closer to the city center and from elderly drivers. Also, the most noticeable response from the drivers under consideration is shifting more of their driving to the hour before the morning peak charge starts.