Vehicle Miles Traveled in Multivehicle Households

Vehicle miles traveled (VMT) is an important factor in the development of transportation plans, emission mitigation measures, and energy conservation policies. Therefore, estimation of VMT is a crucial task supporting such plans and policies. This research addresses the estimation of VMT in households owning multiple vehicles. This sector is expected to use vehicles differently from single-vehicle households because usage of any vehicle may depend on usage of other vehicles. Previous studies concluded that there is a substitution effect between usages of two vehicles (i.e., greater usage of one vehicle lessens usage of the other). In view of more recent changes in sociodemographic structure, the problem was revisited with the 2001 National Household Travel Survey database. The proposed VMT model is a system of simultaneous equations. Each equation explains the VMT for one of the household's vehicles. The three-stage least-squares method was used to estimate the coefficients. A case study of two-vehicle households was investigated. The resulting model shows that VMT can be explained by variables such as the vehicle's newness, number of potential car users in a household, and household income. Surprisingly, the results show not a substitution effect but a spilling effect. The VMT of the first vehicle does not depend on how much the second vehicle is driven. However, increased use of the first vehicle tends to spill over and increase the use of the second one. Some explanation of this behavior shift is provided.