What drives the utility of shared transport services for urban travellers? A stated preference survey in German cities

The supply of shared mobility solutions has been increasing during the last years, so has the popularity of Mobility-as-a-Service. Both promise an easy access to and usage of shared vehicles or shared rides. Nevertheless, usage of these services remains low in German cities. Hence, the question arises: what determines the utility of travellers regarding shared modes and how is this different to conventional modes? To answer this, we conduct a stated preference experiment amongst 1,445 respondents (8,670 observations). The sample is drawn from residents of the 83 largest cities in Germany. We consider four shared (e-scooter-, bike-, carsharing, and ridepooling) and three conventional modes (walking, private car, and public transport). We estimate a mixed logit model and calculate the respective value of travel time (VoT) as well as the value of access, egress, and parking search time. The importance of the individual attributes is analysed drawing on a part-worth analysis. Further, we calculate average treatment effects to show simulated mode-choice probability changes. We find that costs are more important than travel time for carsharing and ridepooling whilst they are equally important for the remaining modes. For shared services, access is more important than egress. Moreover, among the shared services, e-scootersharing shows the highest VoT (23.73 EUR/h), followed by bikesharing (18.53 EUR/h). Finally, cost changes to private cars and public transport show the highest simulated shifting potential with carsharing profiting most from cost increases in these two modes.