Are commuter train timetables consistent with passengers’ valuations of waiting times and in-vehicle crowding?

Many models have been developed and used to analyse the costs and benefits of transport investments. Similar tools can also be used for transport operation planning and capacity allocation. An example of such use is the assessment of commuter train operations and service frequency. In this study, we analyse the societally optimal frequency for commuter train services. The aim is to reveal the implicit valuation (by the public transport agency) of the waiting time and the in-vehicle crowding in the commuting system. We use an analytic CBA model to formulate the societal costs of a certain service frequency and analyse the societally optimal frequencies during peak and off-peak hours. Comparing the optimal and the actual frequencies allows to reveal the implicit valuations of waiting time and crowding. Using relevant data from the commuter train services in Stockholm on a typical working day in September 2015 (e.g., OD matrix, cost parameters), we perform a numerical analysis on certain lines and directions. We find the societally optimal frequency and the implicit valuation of waiting time and crowding. The results suggest that the public transport agency in Stockholm (i.e., SL) adopted service frequencies that are generally slightly higher than societally optimum which can be explained by a higher implicit valuation of waiting time and crowding. We also find that the optimal frequencies are more sensitive to the waiting time valuation rather than that of crowding.