An analysis on household demand for transportation in Taiwan

[[abstract]]隨著經濟成長，民眾私人運具持有數提高，導致石油需求量增加，以進而影響到大眾運輸之搭乘。過去多數的研究，多將私人或大眾運輸需求分開推估，然而在運輸使用選擇上，卻是在同個時間點做決策。因此，石油使用量與大眾運輸搭乘人次存在著替代關係，使得兩運輸方程式所產生的殘差項有關。故本研究將透過近似無相關迴歸分析，同步推估模型中的參數。此外，機車的使用在台灣運輸型態中，為普遍使用的交通工具，故本研究的私人運具持有數，將同時考量汽車與機車於模型中。並將家庭人口結構做更深入的分析，藉此可進一步了解各層面的人口，對運輸需求影響大小之關係。 實證分析結果顯示，各變數在兩運輸需求方程式中，呈現顯著的影響。從油價分析的係數得知，大眾運輸搭乘人次與私人運輸石油使用量，兩者呈現替代關係。家戶擁有較多的私人運具，會有較多的石油使用量，進而減少大眾運輸之搭乘人次。而所得的改變對於運輸需求之影響，呈現較為不敏感的結果。在家庭人口特性中，有較多0-6歲小孩人數的家戶，石油使用量較高，而大眾運輸搭乘人次為相對較低。最後透過模擬油價上漲情境得知，家戶將減少石油需求量，並增加對大眾運輸的使用。以居住在台北市的家戶而言，大眾運輸消費最高，推測較健全的大眾運輸基礎設施，會使民眾較為依賴大眾運輸的使用。[[abstract]]Due to growth of the economy, the number of private vehicles is increasing and so is the gasoline demand. In return, this will also affect the number of public transport passengers. However, when analyzing transport demand, most of the previous researches estimate private and public transport individually. Nevertheless, the decision of choosing private or public transport is at the same time. As a result, the gasoline consumption and the number of public transport passengers exists substitution relationships; consequently the residual of the two transport equations are correlative. Therefore, different from the literature, this research estimating private and public transport by Seemingly Unrelated Regression. Also, our model will make further modification and discussion. Motorcycles are very popular vehicles in Taiwan. So we modify our model to close to real situation by including number of motorcycle in our vehicle number. Besides, this study will further discuss the effects of household structure to transport demand. The empirical results show that estimated parameters for all variables included in our model are significant. Since the signs of estimated parameters for fuel price in both equations are opposite, it implies that the number of public transport passengers and the gasoline consumption are substitutes. Households owning more vehicles have more gasoline consumption and less passengers taking public transport. However, demand of transport is insensitive to the household income. The impacts of household demographic characteristics show that households with 0-6 year children use more gasoline but less public transport. Finally, by the simulation of increasing oil price scenarios, the rise of gasoline price will result in less gasoline consumption but more public transport for households. Households in Taipei have the highest consumption for public transport which may contribute to its most completed public transport infrastructures in Taiwan