| 摘要: | 外人直接投資(foreign direct investment, FDI)被視為經濟成長的引擎之一, 主要原因是FDI會帶給接受國新的技術以及外溢效果, 進而促使其經濟成長加速。 但文獻上對FDI帶給總體經濟環境的影響, 看法並不一致。因此, 本論文將對以下三個議題進行討論, 第一, FDI是否真的會使經濟成長加速; 第二,FDI是否會讓所得不均度下降; 第三, 雙邊FDI在景氣共移中所扮演的角色為何? 第二章主要探討FDI是否對經濟成長存在不對稱的門檻效果, 本章利用Caner and Hansen (2004)所提出的門檻回歸方法並運用期初GDP、人力資本與貿易開放程度為門檻變數,係以1975年至2000年的62個國家為研究對象。在未考慮門檻效果時, FDI對於經濟成長的效果是不一致的,但納入門檻迴歸時, 實證結果證實期初GDP和人力資本的程度對FDI與經濟成長具有重要的解釋能力。若FDI的接受國擁有較佳的期初GDP與人力資本, FDI對於經濟成長是正向顯著的影響。 第三章欲重新驗證金融中介機構在FDI對經濟成長的角色, 本章首先建立一個簡單的理論模型說明金融中介與FDI的外溢效果之間的關聯。 在實證分析上, 我們利用工具變數法來處理有可能的內生性問題, 但我們也發現在這個議題上存在弱工具變數的問題, 此問題會讓原先的估計結果產生偏誤。 因此, 本章運用較可信賴的聯合檢定, 如: Anderson and Rubin (1949), Kleibergen (2002)與Moreira (2003)再次檢驗金融中介的角色。我們也考慮四種在有弱工具變數問題時較佳表現的估計方法進行分析, 如: limited information maximum likelihood (LIML)、Fuller (1977)及Hausman et al. (2008)所提出的兩種heteroskedasticity-robust估計方法(HLIML和HFUL)。本章與Hermes and Lensink (2003)和Alfaro et al. (2004)不同之處, 我們發現擁有較佳的金融發展國家並不會讓FDI產生外溢效果, 使得經濟快速成長。 在第四章, 運用1980年至2005年56個國家分析在不同基礎建設的水準之下, FDI與所得不均度之間的關係。本章利用Hansen (2000)所提出的門檻迴歸方法發現, 在基礎建設略差的國家,FDI的引進會讓接受國的所得不均度惡化; 相反地, 擁有較佳的基礎建設時, FDI對所得不均的影響不大。此外,本章發現國際貿易會使得所得分配更加平均, 此外, 實證結果也支持Kuzents (1955)年所提出的Kuzents inverted-U hypothesis。 第五章利用77個國家組合分析FDI、貿易、產業相似度與景氣共移性的關係, 在過往文獻探討中大部份都忽略FDI的重要性。本章採用Baltagi (1981)提出的error component three stage least square (EC3SLS)進行實證探討,我們發現FDI關係越緊密的國家會讓提高兩國的景氣共移性, 而貿易及產業相似度對景氣的影響則不顯著,這代表FDI是解釋景氣相關性重要的變數之一。此外, 本章發現FDI是屬於水平性質, 因為FDI與貿易存在替代關係,FDI取代貿易在景氣共移中的角色, 這也是FDI的重要性提升的原因之一;實證結果亦顯示兩國產業差異性越大, 越會透過貿易來獲取各自所需的商品。Foreign direct investment (FDI) is usually viewed as an engine of growth. An influential economic rationale for offering inventive to attract FDI is based on the belief that FDI enhance host countries' economic growth through technology transfers and spillovers. However, the literature provides debatable evidence concerning the effects of FDI. In this dissertation, we want to discuss the following issues: Does FDI promote economic growth? Does FDI make the income distribution more equal? Does bilateral FDI contribute to the business cycle synchronization? In Chapter 2, we use the threshold regression technique developed by Caner and Hansen(2004) to examine whether the effect of FDI on economic growth is dependent upon different absorptive capacities. There are three absorptive capacities, namely, initial GDP, human capital and the volume of trade, that are used as threshold variables in this chapter. The empirical analysis shows that FDI alone plays an ambiguous role in contributing to economic growth based on a sample of 62 countries covering the period from 1975 through 2000. Under the threshold regression, we find that initial GDP and human capital are important factors in explaining FDI. FDI is found to have a positive and significant impact on growth when host countries have better levels of initial GDP and human capital. Chapter 3 revisits the role of financial intermediation and the effects of foreign direct investment (FDI) on economic growth using cross-country data between 1975 and 2005. To motivate our empirical investigation, we set up a simple model illustrating how local financial markets create linkages in the process of technology spillover. We apply instrumental variable regressions to avoid the endogeneity problem and find that weak instruments may lead to bias in estimating the effects of FDI on output through local financial markets. Some fully robust tests, such as the Anderson and Rubin (1949), Kleibergen (2002) and Moreira (2003) statistics, are used to re-evaluate this issue. We consider the limited information maximum likelihood (LIML) and the Fuller (1977)(FULL) methods to provide more reliable point estimates and inferences under weak instruments. We also use the heteroskedasticity-robust estimators, HLIML and HFuller(HFHUL), suggested by Hausman et al.(2008). In contrast with Hermes and Lensink (2003) and Alfaro et al.~(2004), our empirical results show that economies with better-developed financial markets do not necessarily to benefit more from FDI in accelerating their economic growth. In Chapter 4, we focus on the effect of FDI on income inequality and whether this relationship is dependent on absorptive capacity or not, by using a cross-sectional dataset taken from 56 countries covering the period 1980--2005. Our results, using the endogenous threshold regression model proposed by Hansen (2000), suggest that there is strong evidence of a two-regime split in our sample. That is, FDI is likely to be harmful to the income distribution of those host countries with low levels of absorptive capacity. By contrast, our results support the perspective that FDI has little effect on income inequality in the case of countries with better absorptive capacity. We have also found that international trade can lead to the income distribution being more equal. The Kuznets inverted-U hypothesis is thus found to hold based on our empirical results. Chapter 5 analyzes the relationships that exist among FDI, trade and industry dissimilarity on business cycle co-movements using a panel data set taken from 77 pairs of countries over the period 1988--2002. Most of the previous literature has overlooked the possible channels through which FDI might influence business cycle synchronization. Our results, based on the error component three stage least squares(EC3SLS) approach suggested by Baltagi (1981), show that FDI is likely to lead to more synchronization. By contrast, trade and dissimilarity are not significantly correlated with business cycle correlations. This might indicate that the FDI channels are better able to explain the business cycle linkages. In addition, our findings show that FDI is of the horizontal type in our sample and is demonstrated to be a substitute for trade relations. FDI might thus replace the prominent channel of trade. We also find that economies with different structures enjoy high levels of international trade. |
