高速公路實施計程收費前後對短程用路人行為之影響分析

Abstract

本研究恰逢計程收費制度全面上路之時，以事前、事後比較分析概念，探討國內高速公路實施計程收費後對短程旅次用路人路線移轉行為的影響。研究方法採用結合潛在變數之二元羅吉特個體選擇模式，分析樣本之旅次特性、社經特性、心理認知潛在變數及旅行時間和旅行成本對於用路人路線選擇行為的影響。最後，則以無潛在變數之顯示性偏好模式進行計程收費實施前後之機率分析、彈性分析和福利經濟變動分析。 研究結果顯示，計程收費實施後有98.5%的受訪者留駛高速公路；留駛高速公路的樣本中，97%的人維持和過去相同的起迄點交流道。事前模式的SEM分析中，各假設皆成立：「認知效果與效率」對「政策接受度」有正向的直接影響（H1）；「認知公平」對「政策接受度」有正向的直接影響（H2）；「轉移障礙」對「留駛高速公路的意向」有正向的直接影響（H3）；「政策接受度」對「留駛高速公路的意向」有正向的直接影響（H4）。而事後模式的SEM分析中，僅支持H1、H2、H3，不支持H4。 本研究建立之無潛在變數及結合潛在變數之二元羅吉特模式，所有解釋變數皆達顯著水準，且同一變數於事前和事後模式中的符號相同。模式較估結果發現，影響路線選擇的變數有：方案特定常數、旅行時間、旅行成本、轉移障礙、政策接受度、通勤目的、所得、性別。事前選擇模式較估而得的時間價值與事後選擇模式之時間價值的結果相當一致。其他應用分析方面，總體用路人於計程收費實施後行駛高速公路的機率上升；總體彈性值約為-0.0374；計程收費實施後用路人整體效用有提升。This study happened to be in the year of the distance-based toll scheme being implemented. This study applied the conception of before-and-after analysis to explore how the new toll scheme influences short-trip road users’ choice behavior. As to the research method, binary logit model combining with latent variables was utilized to analyze the effects of basic attributes (e.g., trip attribute, social-economic attribute, cognitive latent variables, travel time, travel cost) on short-trip road users’ choice behavior. At last, the model of revealed preference with no latent variable was employed to analyze the variations in probability, elasticity and economic welfare before and after the distance-based toll scheme being implemented. The effective sample size of this study is 200. Among it, commuters occupy the proportion of 59%, and on the other hand, business trip is in the majority of non-commute purposes. The result shows that after the distance-based toll scheme, there are 98.5% participants remaining using freeway and 97% of them remain the same routes as past. According to the SEM before-model, every hypothesis is supported. The supported hypotheses are as follows: (H1) road users’ cognitive effect and efficiency of the distance-based toll scheme positively and directly affects their policy acceptance; (H2) road users’ cognitive fairness of the distance-based toll scheme positively and directly affects their policy acceptance; (H3) road users’ cognitive transference obstacle positively and directly affects their intention to remain using freeway; (H4) road users’ policy acceptance towards the distance-based toll scheme positively and directly affects their intention to remain using freeway. However, according to the SEM after-model, only H1, H2, H3 are supported while H4 is not supported. All the variables are significant in both binary logit model with and without latent variables, and the signs of same variables are consistent in before-model and after-model. The variables which influence route choice include alternative specific constant, travel time, travel cost, transference obstacle, policy acceptance, commute purpose, income and gender. As for other applied analysis, the result shows that after the distance-based toll scheme was conducted, the entire road users’ probability of using freeway increases; the entire elasticity is about -0.0374; the entire road user’s utility increases.