| Summary |
In order to improve traffic congestion and emissions in urban areas, transportation managers unanimously agree that to promote public transportation and to increase usage cost of private vehicles, such as parking fee increase, parking space reduction, fuel fee increase, and in-town congestion charge, are at the heart of urban traffic management. However, without knowing the key factors contributing to the usage of private vehicles, it is rather difficult to propose effective countermeasures. Meanwhile, previous studies regarding the usage of private vehicles mainly rely on questionnaire surveys on road users, the representative of the samples might be questionable.
Based on this, this study aims to investigate the factors contributing to the usage of cars and motorcycles based on the annual mileage traveled of vehicles calculated from the periodical vehicle inspection dataset. Two types of models are then estimated and compared. The first type of models is to regress annual mileage traveled of cars or motorcycles separately on three types of variables, including socio-demographic variables of vehicle owners, characteristics of vehicles and environmental variables of the located cities/counties. Considering the hierarchy of the explanatory variables, the linear hierarchical models are adopted. The second type of models are to simultaneously regress both mileages of cars and motorcycles of the same owners on the three types of variables by using simultaneous equations modeling approach so as to investigate the endogeneity and substitution effect between cars and motorcycles.
The estimation results of the separate motorcycle model show that engine size, male owner, and length of roadways have significantly positive effect on usage while vehicle vantage, owner age, number of motorcycles owned, population ratio of secondary and tertiary industries, operating mileage of city buses, and mass rapid transit (MRT) system (with or without) show significantly negative effect on motorcycle usage. While for the estimated car usage model, engine size, alternative fuel, diesel and density of freeway have significantly positive effect on car usage while car vantage, owner age, number of motorcycles owned, population ratio of secondary and tertiary industries, density of city road and operating mileage of city buses have significantly negative effect on car usage. The estimation results of simultaneous equation model show that the substitution effect between car and motorcycle usage is significantly tested and the substitution effect of motorcycles is stronger than that of cars, suggesting simultaneously control measures towards car and motorcycle usages should be applied. |
