Vehicle-Ownership Model Using Family Structure and Accessibility Application to Honolulu, Hawaii

Abstract

Vehicle ownership plays a key role in determining the number of trips made by a household and the modes that the household will use to make those trips. Predicting vehicle ownership is a challenge due to the complexity of relationships affecting the choice of vehicles for a particular household. The income of the household, the accessibility of activities by alternative modes, and the family makeup of a household all affect the choice of the number of vehicles. The Honolulu metropolitan statistical area, which encompasses the island of Oahu, is the primary study area. On Oahu, vehicle ownership varies widely, from the dense neighborhoods in Waikiki to the semirural areas on the fringe of the region. Consequently, the vehicle-ownership model will be a crucial component of the new travel forecasting models for the island. Using 1995 Oahu household interview survey data, a vehicle-ownership model that takes into account income, accessibility, and household family structure has been estimated and found to predict rather accurately the shares in each vehicle-ownership class by geography, household size, and number of workers. In addition, 1990 census data have been used to test a variation of the vehicle-ownership model on several other American cities such as Atlanta, Kansas City, San Francisco, and New York. The resulting models for these regions parallel the Honolulu vehicle-ownership model very closely in structure and statistical significance, thereby demonstrating the model structures’ transferability between regions.