探討GDP成長與汽機車成長之因果關係

Testing Causality between GDP and Motorcycle/Car

吳佳紋(Chia-Wen Wu)

指導教授 ： 許添本

國立臺灣大學/工學院/土木工程學研究所/碩士(2006年)

https://doi.org/10.6342/NTU.2006.02020

摘要

我國歷年機動車輛數持續不斷增加，而此種以機車、小客車為私人運具發展導向的情形，亦在許多亞洲國家出現，例如：印度、越南、泰國…等，然而隨著汽機車數目的增加，其所帶來之負面衝擊也日益嚴重，為解決此問題，已付出許多代價；同時，根據過去的研究指出，機動車輛持有及使用的成長通常被視為是在GDP與所得水準增加過程中不可避免的情況。本研究期望透過日本與台灣之汽機車及經濟間之互動關係研究，作為東南亞國家制定運輸政策之參考。
再者，由於傳統之計量分析方法，是根據先驗理論基礎所建立的結構化計量模型，在認定上非常困難且不具說服力，且根據先驗理論基礎建立之模型所估計的實證結果，無法確切反應真正的資料特性。基於上述的原因，故本研究以日本及台灣為研究對象，盡可能選取1952年至2005年之年資料，採用機車登記數、小客車登記數及以2000年為基期之各國國內生產毛額為變數，係利用ADF單根檢定、VAR向量自我迴歸模型、共整合檢定、Granger因果關係檢定及預測誤差變異數分解…等方法，分別探討「機車登記數與GDP之因果關係」及「小客車登記數與GDP之因果關係」。
從共整合檢定的結果顯示，臺灣資料的變數均具有共整合關係，意即隱含了這些變數長期而言，是具有往「均衡方向調整」的特性；相對的，就日本而言，則是不具共整合關係。而在Granger因果關係的檢定結果發現，台灣之機車登記數與GDP間並無雙向的回饋關係，僅存在單向「機車登記數成長造成GDP成長」之關係，另一方面，在小客車登記數與GDP間也僅存單向「GDP成長帶動小客車登記數成長」之關係；就日本資料而言，僅存在單向「機車登記數的成長造成GDP成長」的關係，但在小客車登記數與GDP間卻是獨立關係。進一步利用預測誤差變異數分解，觀察各變數之總預測誤差變異數，藉由模型中所有變數之衝擊所解釋的百分比，以了解各總體變數的相互解釋能力，此實證結果與Granger因果檢定的結果相符合。

關鍵字

單根檢定 ； 共整合 ； Granger因果關係 ； GDP ； 小客車登記數、機車登記數

並列摘要

The number of mobile vehicles in Taiwan has been increasing for many years especially motorcycles. The trend of using motorcycles for private communication also has been discovered in many other Asian countries like India, Thailand, Vietnam
… and so on. However, with increasing numbers of mobile vehicles, the negative impacts which took are more serious. In order to solve these problems, we have already paid a lot of costs. At the same time, according to the past studies, the ownership of mobile vehicles and growth of the usage are related to the increasing of GDP. The main objective of this study is to establish transportation policies by researching the relationship between mobile vehicles and ecnomy of Japan and Taiwan.
On the other hand, the model according to the traditional econometric analytic method which is established based on prior theory is very difficult and not has convincingness. Since the model is unable to reflect to the real characteristic of the data. On the basis of above-mentioned reason, the Granger causality and forecast error variance decomposition techniques were applied on 1952-2005 data for Taiwan and Japan to examine cointegration and causality between “GDP and the number of registered motorcycles” and “GDP and the number of registered cars”.

The results show that the causality between GDP and mobile vehicles（motorcycle and car）is “from the number of registered motorcycle to GDP” and “from GDP to the number to registered car” in Taiwan. As for Japan, the results are “from the number of registered motorcycle to GDP” and “the independence relationship between GDP and the number of registered car”. Further, utilize the way of impulse response function and forecast error variance decomposition to find out the mutual explanation abilities between each variable.

並列關鍵字

ADF test ； Cointegration ； Causality ； GDP ； car ； motorcycle

參考文獻

18. 蔡佳佳，機車持有預測模式之研究，國立台灣大學土木研究所碩士論文，民

1. Walter Hook and Michael Replogle (1996)：”Motorization and nonmotorized transport in Asia ”. Land Use Policy, Vol. 13, No. 1, pp. 69-84.

2. Joyce Dargay and Dermot Gately(1999): “Income’s effect on car and vehicle ownership, worldwide:1960-2015”. Transportation Research Part A 33, pp. 101-138.

3. Joyce M. Dargay (2001): “The effect of income on car ownership: evidence of asymmetry”. Transportation Research Part A 35, pp. 807-821.

5. R. Ramanathan：” The long-run behaviour of transport performance in India: a cointegration approach”. Transportation Research Part A 35 (2001) 309-320