Composition of Vehicle Occupancy for Journey-to-Work Trips Evidence of Ridesharing from 2009 National Household Travel Survey Vermont Add-On Sample

Ridesharing serves to mitigate pollution and congestion with minimal investment of public capital while also increasing the efficiency of the transportation system. This research addresses the gaps in the literature on the structure and formation of ridesharing by identifying individual, household, and physical-environment characteristics that correspond with an individual's choice to rideshare instead of drive alone. For a fuller understanding of ridesharing behavior, there first must be a better understanding of who is in the vehicle, not just how many. A distinction is made between intrahousehold (internal) and interhousehold (external) ridesharing. With the Vermont add-on sample of the 2009 National Household Travel Survey, a multinomial logit model and nested logit model were developed to examine the determinants of ridesharing. The analysis in this research stresses the importance of the way ridesharing behavior is extracted from survey data. Further, a new method for calculating household vehicle availability is presented; the method places less importance on drivers who are not full-time workers. Results indicate that employment density, longer distance traveled to work, and working in a small urban area have positive influences on the likelihood of ridesharing. Higher vehicle availability in the household, age, sex (male), and a longer time spent per trip on the journey to work were all found to influence the propensity to rideshare negatively. Cost of travel does not significantly affect ridesharing.