徵收進城擁擠費對私人機動車輛使用者行為之影響分析

Abstract

台灣私人運具的取得與使用成本均較為低廉，使得大眾運輸的使用率無法有效提升，也導致都會區在尖峰時段常產生道路擁塞的問題。藉由擁擠定價的策略，在尖峰時段向私人運具徵收擁擠費，將其產生的外部成本內部化，可使部分尖峰時段私人運具用路人移轉至離峰時段，可平衡尖、離峰的旅運量，以藉此改善尖峰時段的擁擠現象。本研究主要目的為構建都會區徵收私人運具進城擁擠費的方案選擇模式，以瞭解三大都會區(台北、台中及高雄)若實施尖峰擁擠定價，對於家戶持有私人運具使用者之影響及移轉至其他運具的情形。 本研究蒐集敘述性偏好資料，並將個體需求模式應用於都會區旅運行為，以擁擠費率、總旅行成本與總旅行時間等三個屬性，模擬擁擠定價策略。透過不同擁擠費率來探討汽機車持有家戶對於方案選擇的行為，替選方案包括「付費進入市區」、「不進入市區或改於離峰時段再進入市區」、「改搭其他交通工具--大眾運輸」及「改搭其他交通工具--其他替選私人運具」等四個方案。本研究以分層抽樣方法進行問卷調查，以取得顯示性及敘述性偏好之資料，從回收之有效樣本中再擷取及區分出研究對象與範圍部分，合計汽車共772份、機車共631份。為瞭解政府於實施徵收進城擁擠費，對持有汽機車之家戶其使用行為之影響，因此分別構建及校估三大都會區之多項羅吉特模式。 多項羅吉特模式校估結果顯示，每週開(騎)車通勤上班(學)天數、每週開(騎)車旅遊或訪友的天數、家戶持有汽機車及自行車的數量、家中到大眾運輸場站最近的步行距離、主要駕駛人之年齡、性別、職業、教育程度、所得與主要上班(學)的交通工具等解釋變數，均會影響汽、機車持有家戶擁擠收費的方案選擇行為。 透過各模式校估結果，由共生變數之總旅行成本與總旅行時間可計算出時間價值，汽車模式高低依序為：台中($4.93元/分)、高雄($4.78元/分)及台北($4.20元/分)；而機車模式則依序為：台中($2.76元/分)、高雄($2.52元/分)及台北($2.28元/分)。整體而言，汽車之時間價值高於機車，而台中又分別大於高雄及台北。

Due to low capital and operating costs of private vehicles in Taiwan, public transport is not utilized effectively and traffic congestion often occurs during peak hours in urban areas. By implementing a congestion pricing strategy such as paying in-town congestion charges, the external costs of using private vehicles could be internalized, and some demands during peak hours would shift to off-peak hours. Therefore, it is possible to balance travel demands between peak and off-peak hours, and further to mitigate the overcrowded phenomena. The major purpose of this study was to understand how implementation of in-town congestion charges in the three major metropolitans (Taipei, Taichung and Kaohsiung) would potentially influence private vehicle users’ driving behaviors and their motivation to utilize other transport modes. Four alternative scenarios were proposed in this study, including: (1) Paying congestion charges when entering a town during peak hours; (2) not to enter a town during peak hours, or entering a town during off-peak hours; (3) using public transport when entering a town during peak hours; (4) using non-public transport (such as walking or bicycling) when entering a town during peak hours. A stratified random sampling method was then utilized to acquire both the revealed and stated preference data from the surveyed private vehicle owners in the three metropolitans. A total of 772 valid car questionnaires and 631 valid motorcycle questionnaires were collected. Next, three variables including congestion charge, total travel cost, and total travel time were utilized in the Multinomial Logit (MNL) model to estimate impacts of in-town congestion charges in each of the metropolitans. The modeling results indicate that a number of factors would influence choice behaviors of private vehicle owners when paying in-town congestion charges is required. The factors include: Weekly driving frequency for work, weekly driving frequency for non-work activities, ownership of private vehicles, distance between home and a public transport stop, and several attributes of principal drivers (such as age, gender, occupation, education, income and major transport mode to work). Finally, value of time (VOT) was calculated using the variables of total travel cost and total travel time. The calculated VOT value for cars decreases in the following order: Taichung (NT$4.93/min), Kaohsiung (NT$4.78/min) and Taipei (NT$4.20/min). For motorcycles, the VOT value decreases in the following order: Taichung (NT$2.76/min), Kaohsiung (NT$2.52/min) and Taipei (NT$2.28/min). Overall, the VOT values for cars are higher than those for motorcycles, and the VOT values in Taichung are higher than in either Kaohsiung or Taipei.