Experiences with transportation models: An international survey of planning practices

- There are lively and debates about the exact role of transport models.
- We report experiences and perceptions of practitioners in different countries.
- A larger distance to transport models correlates with lower trust on model outputs.
- Respondents state they understand assumptions, but other stakeholders do not.
- We propose bringing transport models closer to those who use their outputs.

Transport planning practice is experiencing rapid transitions. This shifting professional environment is prompting lively and sometimes bitter debates about how transportation models should be used. While these models and their outputs play an increasingly more important function in transport-related decision-making processes, growing concerns emerge about their limitations, assumptions, biases, and usability. This paper addresses the question of how different professionals involved in transportation planning perceive and experience these tensions. For that purpose, we developed an online survey which was completed by 229 European transport planning practitioners, primarily working in the Netherlands, Denmark and Germany. Our findings support the following key conclusions. First, and contrary to popular notions on the matter, practitioners are relatively satisfied with the models they use. Second, most respondents are confident that they understand the assumptions and uncertainties associated with transport models, but that other important stakeholders do not. However, third, the larger the distance that respondents have to hands-on working experience with transportation models, the lower is their trust on model outputs. Respondents who are not directly involved in the operation of the models a) report more negative experiences associated with model use in decision-making processes and b) identify more usability barriers. The overall picture revealed a lack of trust amongst transport planning professionals, which is a problem needing to be addressed. We propose bringing models closer to those who use their outputs as a constructive solution to this trust deficit.