A behavioral model of work-trip mode choice in Shanghai

This paper analyzes travelers' choice behavior by using combined revealed preference/stated preference (RP/SP) survey data on work-trip mode choice in Shanghai, taking into consideration different scales across the two data sources. Several versions of a multinomial choice model are specified and estimated. The estimation results suggest that the utility function with money cost divided by income by equivalence scale be chosen as the preferred model. Based on the preferred model, values of time and sample aggregate elasticities of choice probabilities are calculated. The results show that values of “in-vehicle time” and “out-of-vehicle time” are 64 and 82% of the sample mean wage, and that in terms of elasticities, “in-vehicle time”, “out-of-vehicle time” and “money cost” are the most important attributes for bus, subway and taxi passengers, respectively. The conditional elasticities given low, middle and high income levels indicate that bicycles seem to be an inferior good for all income levels. Bus and subway transport are inferior goods for people at middle and high income levels but normal goods for those at a low income level. Taxis are a normal good only for low income levels; for middle and high income levels, they are a luxury good. The results obtained may be used for transportation policymaking in Shanghai.