Estimating behavioral parameters for mode choice typically relies on revealed- or stated preference data. However, applying GPS-based revealed preference (GPS-RP) panel data in modeling mode choice, particularly in response to shocks or policy interventions, remains relatively rare and methodologically under-explored. This paper discusses the preparation, processing, filtering, and modeling of (semi-)automated travel diaries collected over 3 months, including the 9-Euro-Ticket fare policy intervention in Germany in 2022. By estimating two multinomial logit models, we investigated how this intervention influenced the value of travel time savings (VTTS) across different modes. Our findings revealed a substantial reduction in VTTS for public transportation during the intervention period, with values approximately half those in the months following the intervention, highlighting the profound impact of this nearly fare-free policy. This study debates the difficulties and complexities of estimating VTTS using GPS-RP data for urban travel behavior. It underscores the importance of robust preprocessing and filtering methodologies when handling complex GPS data, and discusses how the intervention’s effects on VTTS and project appraisal could inform future transportation policy and investment strategies.
Beck et al. (2026) studied this question.