油價及國內生產毛額對高速公路交通量之影響

Abstract

本研究旨在探討民國九十三年一月至九十八年六月間油價及國內生產毛額對高速公路各收費站通行車輛數之長短期因果關係及跨期動態之衝擊反應，俾提供有關單位交通量管理策略研擬之參考。本研究首先將各收費站通行車輛數通行車輛的月資料依路線（國道1號、3號、5號）、區域（北區、中區、南區、宜蘭地區、全島）及車種（小型車、大型車、聯結車）加以區隔分析。在實證方法上的選用是以 Engle and Granger 兩階段法及Johansen最大概似法進行共整合檢定（cointegration test），來檢定變數間是否存在長期均衡關係。並根據Granger(1969) 所提出變數預測力的方法, 利用Wald 檢定及Toda and Yamamoto (1995)之方法來衡量變數間之短期領先落後的因果關係。最後，再輔以向量自我迴歸模型進行後續的衝擊反應函數分析，以了解變數間之動態、跨期的影響與衝擊，俾檢視我國油價、國內生產毛額及高速公路交通量間是否存在緊密的關係。

實證結果發現，兩種共整合之檢定結果皆顯示油價及國內生產毛額與高速公路交通量不存在長期均衡共整合的關係。此說明了長期的交通量之增減並無法由油價及國內生產毛額來加以判定，而係由其他因素所左右。此外，不同Granger因果關係的檢定方法雖產生部份結果不一致之現象，但就宜蘭地區而言，兩種Granger因果檢定的結果均一致指出該地區之高速公路交通量並不受油價及國內生產毛額的影響。這顯示國道5號交通量仍受其他因素所控制，此與該條國道大部分均為旅遊觀光旅次與雪山隧道的該開通未台灣北部及東北部的交通帶來很大的便利性有關。最後，透過衝擊反應函數進行變數間之動態跨期衝擊影響分析中，發現油價及國內生產毛額對部分地區大型車及聯結車有正向且立即的短暫性衝擊。這說明若有關單位欲透過油價調整來長期且有效抑制高速公路交通量恐難達成。

It is a concept universally acknowledged that when gas price rises, the toll road use will reduce. It is also well-fixed in the minds of many policy makers that they devise energy-related policies based on this premise. With an aim to successfully implement the government’s proposed green tax policy for achieving certain levels of environmental protection by increasing the gas price to reduce the usage of vehicles, the causal effect of gas price and traffic volume plays a significant role here. As a result, this study takes a closer look at the conceptual grounds of the notion of causality in Granger’s sense. In addition, taking into account that GDP may be also a key component of affecting vehicle usage, it is also incorporated into the study.

Cointegration test for the long term equilibrium relationship, Granger causality test for the short run lead or lag relationship and the analysis of impulse response function are employed to unveil and justify the linkage among the interested variables. The main findings are not in line with what we used to take for granted that no co-integration for the long run, not all the gas price or GDP Granger cause traffic volume and as well as some positive impulse responses of traffic volume to gas price shock and GDP innovation. However, as even the simplest descriptive statistics can be deceptive, our findings must be treated with caution especially when extrapolating any anticipated effects on relative policy making.