Understanding traveller behaviour under congestion pricing: A survey-based decision tracking approach

This paper examines commuter responses to three congestion pricing (CP) schemes including cordon-based, distance-based, and travel time–based in Calgary, Canada. The study uses an efficiently structured, personalized stated preference survey that integrates sociodemographic information, revealed commuting patterns, and responses to hypothetical CP scenarios, with a focus on joint mode and departure-time decisions. A key contribution of this study lies in the survey design, which records respondents’ interactions with the choice interface and identifies the alternatives actively examined during each choice task. This information is used to estimate a tracked multinomial logit model in which the choice set is restricted to alternatives genuinely considered by each respondent. Comparison with a conventional non-tracked specification shows that assuming full consideration of all displayed alternatives inflates sensitivity to travel time, cost, and toll attributes. By contrast, the tracked model yields more moderate elasticities and parameter estimates that better reflect observed decision-making behaviour. Applied to the congestion pricing context, the results suggest shifts towards transit and ridesharing under certain pricing structures, highlighting how pricing design can influence peak-period travel decisions. Overall, the findings demonstrate that incorporating interaction-tracking data into stated preference surveys enhances the behavioural credibility and policy relevance of model-based evaluations of congestion pricing strategies.