本研究主要針對國防支出對經濟成長正負影響不一，過去的實證研究使用橫斷面資料分析多國現象，但橫斷面資料只適合分析多國的平均整體現象，無法反映出個別國家的異質現象，再加上過去在討探國防支出與經濟成長的實證研究大多使用二元共整合與 Granger 因果關係為實證方法，但是二元共整合可能無法指出變數間是否存在長期關係，且 Granger 因果關係只能檢定短期因果關係。 本文針對過去文獻之不足，加入資本、勞動與出口等變數，建立一組多變量向量自我迴歸模型，並加入影響台灣經濟成長之出口因素，以重新檢驗台灣1955年至2006年間，國防支出對經濟成長的影響。在時間序列的方法上，採用單根檢定、多變量共整合、誤差修正、變異數分析、脈衝反應等方法進行實證分析。實證結果顯示，國防支出對經濟成長在長期上有顯著的正向影響，在短期上國防支出對經濟成長影響不顯著，而國防支出自發性衝擊反應對經濟成長影響效果不大。 The purpose of this study is to reinvestigate the influence of defense expenditures on economic growth. In the past, most research use cross-sectional data analysis to analyze the phenomenon ; however, the cross-sectional data analysis is only suitable for analyzing the average holistic phenomenon of multi-countries, unable to reflect the heterogeneous phenomenon of individual countries. In addition, Bivariate Co-integration and Granger Causality have been used for empirical methods mostly in the previous empirical researches exploring defense expenditures and economic growth. However, Bivariate Co-integration may be unable to point out whether long-term relation exists among variables; also, Granger Causality can only test short-term causal relation. As the past studies may have problem of omission of relevant variables, the variables such as capital, labor and export have been added in this study, in order to establish a Multivariate Vector Autoregressive Model. Also, the factor like export that influence Taiwan’s economic growth, have been added in order to review the influence of defense expenditures on economic growth in Taiwan from 1955 to 2006. The econometric methods such as Unit Root Test, Multivariate Co-integration, Error Correction, Analysis of Variance (ANOVA), impulse Pulse function and so on have been adopted to conduct empirical analyses. The results of empirical analyses have shown: Defense expenditures have significant and positive influence on economic growth in long-term while defense expenditures have no significant influence on economic growth in short-term. Besides, the spontaneous impulse response of defense expenditures shows an insignificant influence on economic growth.
