Effects of fuel price on individual dynamic travel decisions : binary probit selection model that uses GPS panel data

The goal of this study is to explore the impact of fuel price changes on individual activity travel behavior. Previous studies in transportation research have examined both short- and long-term fuel price elasticities on vehicle miles traveled (VMT) and vehicle stock through aggregated data at the country or subnational level. This study is one of the few to use individual-level panel data. Furthermore, studies that consider context effects such as weather conditions, accessibility, and travel companion are rare. Noninclusion of these context-variable effects may produce spurious results on the influence of fuel price changes on activity travel patterns. On the basis of a unique set of GPS panel data of 122 respondents living in the Eindhoven area of the Netherlands, The authors collected corresponding weather conditions data and daily fuel price data. A two-step estimation approach was applied to estimate (a) individuals' decisions of making car-based trips and (b) the influence of fluctuations in real fuel price on car travel distance. The results show that fuel price has significant negative effects both on an individual's decision to use a car and on travel distance by car if a person uses it. In addition, results indicate that days of the week, weather conditions, and 3-weck lagged and 1-week expected fuel prices have signiflcant negative effects on individuals' car use decisions. However, once people decide to travel by car, weather conditions do not have any signiflcant effects on VMT. The 2-week lagged fuel price has a signiflcant negative effect on only VMT.