Modelling Travellers’ Choice Between Park-and-ride and Other Modes of Travel to Work in the Context of Risk and Uncertainty

Modelling Travellers’ Choice Between Park-and-ride and Other Modes of Travel to Work in the Context of Risk and Uncertainty

Nominated for The Planning for Sustainable Land Use and Transport Award


Ying Huang, University of Western Australia, Brett Smith, University of Western Australia, Doina Olaru, University of Western Australia


Using stated preference data in Perth, this paper focuses on the choice between park-and-ride and other modes in terms of uncertainty in the mode choice setting: car travel time variations, parking availability and crowding on public transport


1. Introduction
Park-and-ride (PnR) is seen as an attractive public transport alternative in a city with low densities as it combines the efficiency of a mass transit system with the flexibility of the car (Holguin-Veras, Hart et al. 2012). Perth’s newest rail lines were developed with the expectation that park-and-ride would play a major role. However, the system is already under stress with commuters having trouble securing a bay at the railway stations (Martinovich 2008). Moreover, commuters experience non-trivial day-to-day travel time variation on private motorised vehicles and crowding in public transport vehicles conditional on the departure time and unexpected situations occurred during the commuting trip. This paper focuses on the choice to commute by PnR and the trade-off between departure time and the uncertainty of securing a parking bay. The paper also explores other sources of uncertainty in the mode choice setting: day-to-day travel time variations when travelling by car and crowding on public transport.
2. Methodology
Expected utility theory has been applied to commuter choice under travel time variations or travel time reliability in traffic (Bates, Polak, Jones, & Cook, 2001; Noland & Small, 1995; Senna, 1994). Hensher et al., (2011) extended the theory to apply non-expected utility decision heuristics to travel time reliability and crowding on public transport. Whilst the number of available PnR bays has been shown to influence station choice (Chakour & Eluru, 2014), examining commuters’ risk behaviour in this context will broaden the understanding of the role of uncertainty in the mode choice decision. Based on overall goodness of fit, the results presented here make use of the constant relative risk aversion (CRRA) and Tversky-Kahnman’s (Tversky & Kahneman, 1992) probability weighting function.

Parking availability is modelled as a risk adjusted cost. It is assumed that the commuter would prefer to secure a free bay at the station, but face the possibility of having to pay or being unable to locate a bay. Crowding on public transport is incorporated in the model as a measure of discomfort whereby the expected standing time during the trip is the probability of finding a seat multiplied by the duration of the trip. Car travel time reliability is presented to respondents using three time values and associated probabilities – expressed as how often the trip time occurred in 20 days.
3. Data collection
Two household revealed preference (RP) and stated preference (SP) survey were undertaken between September 2013 and April 2014. The sample was drawn from the catchments of seven train stations with PnR facilities in Perth, WA. The attribute levels for the stated preference were pivoted on the revealed preference responses. An efficient design (Rose & Bliemer, 2008) was developed and optimised using genetic algorithm (Olaru et al 2011). In total, 215 commuters responded to the SP survey. An indicative choice scenario is offered in the Appendix.
4. Preliminary results and analysis
Statistically significant risk aversion parameters indicate that travellers are risk-seeking towards car travel time and chance of securing a parking bay when making decisions for commuting. This is consistent with results given by Hensher, Greene, and Li (2011). Travel time reliability parking availability and crowding play an important role in commuters’ mode choices. Results also show individual heterogeneity in expected car travel time and expected parking cost at railway station.
Travel cost, access and in-vehicle travel time are important factors constraining commuters’ willingness to travel on public transport. Commuters other than PnR users are less prone to shift current departure time, if it results in the loss of punctuality. Moreover, commuters are more reluctant to depart late than early. On the other hand, PnR users prefer to depart early to secure a parking bay at railway station. Commuters with car availability during the working day are inclined to choose car as their travel mode or for access. Driving to the destination is the appealing option for women due to the safety concerns. Compared with the young and the middle, the senior in Perth prefer to use public transport.
5. Conclusion
Results from Perth PnR travel study indicate the importance of including uncertainty and risk attitudes when analysing commuters’ travel behaviours. Given a specific probability weighting approach, a risk-seeking attitude emerges among travellers when they face uncertain travel times and parking costs at railway station. Finally, sufficient parking space at railway stations is more likely to induce mode shifts for car drivers who experience travel time variation.


Association for European Transport