透過您的圖書館登入
IP:3.137.180.32
  • 學位論文

政經結構變遷與景氣循環-台灣的實證研究

Democratization, Globalization and Business Cycles: Evidence form Taiwan

指導教授 : 林向愷
共同指導教授 : 蕭文宗
若您是本文的作者,可授權文章由華藝線上圖書館中協助推廣。

摘要


本論文是以景氣循環研究與台灣政經結構變遷為主軸的三篇相關文章。首先,論文的第二章發展一套探討景氣波動是否存在跨時異質現象的統計檢定方法,並發現1987年後台灣主要總體經濟變數景氣波動變異程度明顯下降,而民間投資與每人平均工時於2000年後再度上升。該章再利用EGARCH-X模型以了解1987年後,民主化、全球化以及產業結構變遷對上述景氣波動變異程度的影響,估計結果顯示民主化是導致台灣景氣波動變異程度下降最關鍵的因素,但2000年後台灣資金外流無法解釋民間投資與每人平均工時景氣波動變異程度持續變大的現象。 論文的第三章探討台灣儲蓄與投資相關性之跨時異質性,以及1987年後民主化、貿易自由化以及金融國際化對儲蓄與投資關係的影響。為此,該章發展一套兩階段的處理方法,首先將儲蓄與投資波動幅度中的跨時異質性剔除,再探討影響儲蓄與投資相關性之因素。透過雙變量DCC GARCH-X模型估計結果,在第一階段發現民主化具有抑制景氣波動的效果,而2000年後儲蓄與投資相關性顯著上升係因儲蓄與投資的波動程度受民主化影響而下降。第二階段估計結果則顯示,已剔除波動異質性的儲蓄與投資仍為正向相關且相關係數具跨時異質性,其中金融開放程度對儲蓄與投資相關性有顯著的正向影響。 論文的第四章在探討台灣國民所得帳中主要總體變數以及勞動市場相關變數長期成長趨勢與短期景氣波動的認定是否受到政經結構變遷的影響。該章以成長理論強調的線性時間趨勢加上政經結構因素如民主化、全球化等,做為總體變數去除長期趨勢的方法,並與景氣循環研究常用的HP filter相比較。實證結果顯示,去除掉包含政經結構因素的長期趨勢後,所歸納之台灣景氣循環典型現象相較於HP filter更能呈現如Danthine and Girardin(1989)與Blackburn and Ravn(1992)等學者所強調之景氣循環「時間獨立」性質。

並列摘要


This dissertation consists of three related articles which contribute to the business cycles and institutional changes in Taiwan. The first essay in Chapter 2 investigates the time-varying volatility of business cycles in Taiwan from an institutional perspective. Test results reveal that the volatility has substantially declined for most macroeconomic variables since 1987, and private investment and hours worked per capita rose again after 2000. Then, we use an EGARCH-X model to identify the empirical link between democratization, globalization, and other structural changes and time-varying properties of volatility, and find that democratization is the most significant contributing factor. The second essay in Chapter 3 investigates the time-varying correlation of national saving and domestic investment, and the link between the major institutional changes since 1987 in Taiwan. We use a two-step estimation to account for time-varying volatility in saving and investment first, and investigate the correlation property after that. By the multivariate DCC GARCH-X model, we find that democratization is the most contributing factor to suppress saving and investment volatility, which also results the rising correlation after 2000. In the second stage, we find that financial openness has a significant positive impact on the correlation of saving and investment. The third essay in Chapter 4 clarifies whether the long-run trend and short-run cycle identification is related to institutional changes in Taiwan. Given that the most popular Hodrick-Prescott (HP) filter could distort some important stylized facts of business cycles, this essay explicitly incorporates factors such as democratization and globalization in the trend-removing procedure. The resulting empirical regularities of business cycles are more time-invariant than those using HP filter. Hence, by properly taking into account the possible effects of these institutional factors on growth, our approach provides a better integration of new growth theory and business cycles.

參考文獻


王高文、毛維凌 (2004), 「單一方程式共整合-GARCH模型:台灣股市之實證研究」, 《經濟論文叢刊》, 32, 1-24。
陳仕偉 (2006), 「景氣波動變異對景氣轉折點認定之影響:跨國的實證研究」, 《人文及社會科學集刊》, 18, 37-76。
林向愷、岳俊豪 (2009), 「政經結構變遷與景氣波動-台灣的實證研究」, 《經濟論文》, 37, 411-456。
林向愷、賴惠子 (2009), 「預算體制與政府跨期財政行為-台灣的實證研究」, 《經濟論文》, 37, 207-252。
李秀雲 (2005), 「貿易條件與開放經濟失業率之單根-兼論台灣自然率之估計」, 《經濟論文》, 33, 35-66。

延伸閱讀