Modelling the effects of travel time use on mode choice and the value of travel time savings: tackling self-selection and endogeneity bias

Abstract

Preferences for travel-based activities may lead individuals to choose travel modes that best accommodate activity needs. In theory, “passengerised” modes, such as public transport and automated vehicles (AVs), could become the preferred travel options, as they enable a wider range of travel-based activities. This potential to use in-vehicle time could, in turn, decrease the value of travel time savings (VTTS), leading to lower resistance to longer journeys. Together, these two phenomena create challenges to demand modellers associated, including (i) mode choice self-selection associated with travel-based activities, and (ii) potential endogeneity bias in VTTS estimates. This study employs an attitude-based approach to capture mode choice self-selection while controlling for endogeneity bias in VTTS estimates by considering traveller characteristics and travel context variables that influence travel time use and account for value of time heterogeneity. Results suggest that the construct capturing the causal effect of travel time use on mode choice increases the probability of choosing rail modes in journeys over 5 km and car driving in journeys shorter than 5 km, confirming the presence of mode choice self-selection. Concerning VTTS, accounting for travel-time-use-related preferences and travel context characteristics led to travel-based activities having non-significant effects on VTTS, suggesting this approach may effectively control for endogeneity bias. These findings have implications for road safety, residential location choice, and potential AV-related behavioural changes, all of which have consequences for transport policy formulation. Overall, this study highlights the importance of accounting for travel time use to achieve a more realistic representation of activity-travel behaviour.