Valuing (travel) time: Models and data for activity scheduling

Measuring and valuing travel time is a core concern of transport science. Savings in travel time are the central element of assessing infrastructure projects. The major criticisms of current methods are the inclusion of small time savings, the consequences of the postulated fixed travel time budget and the inconsistency of its value in the underlying models and assessment guidelines. The aim of this project is to improve existing assessment methods and models of changes in time use. It will use a new, comprehensive and technologically advanced travel behaviour diary to achieve its methodological goals. The project is divided into the following work packages: Expand and improve existing models for the number, type and sequence of (out-of-home) activities by incorporating the latest scientific understanding of transport economics Estimate the first German values of the components of travel time Improve methods of measuring time use by using GPS-based tracking of times and routes travelled Develop, test and conduct a major stated-response survey to measure the way daily schedules are adjusted in response to new generalised cost structures Evaluate the resulting policy implications using an updated version of an agent-based travel demand simulation (MATSim) that includes the aforementioned improvements The project will produce two main results: 1) the first travel time values by component for Germany which are based on a dedicated database, and b) policy-sensitive and comprehensive choice models of daily activity scheduling.

The aim of the "Valuing (travel) time" project was to further develop the best models currently available for valuing travel times and to apply them for the first time in Austria. Values of time (time cost rates) are a key input in cost-benefit analyzes (RVS 02.01.22, 2010) because they capture the economic benefits of shorter travel times, for example, through new high-level transport connections. Currently, values of time are only estimated on the basis of travel behaviour. Meanwhile, there are better models that allow a more reliable estimation. For these time use models to be estimated, information about time use and consumer expenditures must be included in addition to the travel behavior. However, these data are usually collected from different individuals by means of different surveys: travel surveys, time use surveys, and consumer expenditure surveys. So far, there is no dataset worldwide that collects this information simultaneously from the same individuals. Therefore, the time use model could only be estimated incompletely with missing components. In this project, a novel survey procedure was developed which we call the Mobility Activity Expenditure Diary (MAED). Based on this procedure, information about all three areas was collected simultaneously over a period of one week from 748 individuals, which were randomly selected among Austrian workers. In a stated preference follow-up survey, the subjects` choices of transport modes, routes, and destinations were investigated. With this data, the complete time use model with all components could be estimated for the first time. The value of leisure was found to be 9.50 / hour, which is in the expected range. More surprisingly, the time spent in public transport was valued much less negatively than in a car. It contradicts the popular claim that time in a car is perceived more pleasant. The reason may be the increasing availability of mobile devices that can be better used in public transport because passengers are released from the driving task. In this respect, the shift of car trips to public transport results in a benefit, even if travel time and costs did not change. We recommend to adapt the benefit-cost analysis to reflect this effect. In addition, it is to be expected that the automated driving and the concomitant release of car drivers from the driving task will trigger a similar effect: The driving time is used for other activities and thus perceived more pleasant, making car use less sensitive to longer travel times.