Personal car or shared car? Predicting potential modal shifts from multinomial logit models and bootstrap confidence intervals

Abstract

Households' daily mobility in France is characterized by the preponderance of the automobile. Passenger cars, mainly used by households but not only, are thus responsible for more than a half of fuel consumption in road transport (CGDD/SoES, July 2013) and more than a half of CO 2 emissions in the transport sector (SoES/CDC, December 2012). The main objective of this paper is thus to explain the modal choice of French households for their local daily trips, particularly the importance of the car, and to predict potential shifts from personal car to shared car. A multinomial logit model is estimated and reveals the particular importance of car equipment on modal choices and specifically on car use. Simulations by 2020 are thus conducted under three scenarios depending on the household's motorization (no car, one car, two cars or more) and per different mobility profiles. Personal car should remain the main mode of transportation by 2020 except if households have no car. In that case, modal shares would be more balanced, public transport would become the main transport mode and the shift to shared car would be at a maximum. Modal share of shared car could thus reach 16% for "exclusive motorists". A conditional logit model is also estimated and shows no particular importance of the means of transportation's costs in the modal choices. These results show that the increase in distances between 2010 and 2020 makes motorized modes more necessary. Thus, personal car and public transport should remain the main modes of transportation by 2020. Moreover, expected changes in costs and travel time by 2020 does not seem to have any effect on the deployment of shared car, its modal share being constant (in an average) between 2010 and 2020.

The existing and the emerging: car ownership and car sharing on the road towards sustainable mobility

“…Car sharing is important, because the new use of cars as shared goods through collaborative consumption can be part of a trajectory of changing frontiers in the automotive industry for sustainable mobility. Within this trajectory, we find new business models, new technologies and service innovations emerging from recent developments in share-economy concepts (Chevalier and Lantz, 2015;Tinnilä and Kallio, 2015). Although privately owned cars dominate personal transport today, this may shift as a consequence of changing behaviour and official policies on energy efficiency and pollution reduction (Lanzini, 2018).…”

Factors influencing adoption and usage probability of car sharing in Bangkok

“…On the other hand, Becker, Loder, et al (2017) found that car-sharing membership is less likely for larger households. Chevalier and Lantz (2015) found that marital status also affected the choice between personal and shared car. N. Wang and Yan (2016) found that married people are more willing to use car-sharing than unmarried people.…”

“…The main attributes were identified as trip travel time and cost and the number of cars per household. Dissanayake and Morikawa (2010) Chevalier and Lantz (2015) proposed a multinomial logit model to investigate the mode choice of French households for their local daily trips as well as to estimate the potential shifts from private car to shared car. In addition, a conditional logit model was taken into account the economic rationality (cost and travel time) of individuals in their modal choices.…”

“…On the other hand, Le Vine et al (2014) found the minority of car-sharing users were predicted to be under age 40. Chevalier and Lantz (2015) found that the older the person, the great the opportunity to use a shared car. However, De Luca and Di Pace (2014) found that age did not have any effect on the choice to use a shared car.…”

“…Family structure was also found to influence the travel mode choice decision. Those from larger households were likely to be more willing to join car-sharing schemes (Chevalier et al, 2015;J. Kim, Rasouli and Timmermans, 2017;Nazari et al, 2018).…”