摘要
本文应用Meta回归分析方法对FDI溢出效应的文献进行数量化再分析,以考察FDI溢出效应的实证结论是否受方法与结构异质性因素的影响。结果发现:方法异质性与结构异质性均会影响FDI溢出效应的实证结论。方法异质性因素中指标的构建、采用的数据、遗漏变量问题均会影响FDI溢出效应的结论,特别是文献研究中采用差分广义矩/系统广义矩估计方法会更有可能得到负向FDI水平溢出效应的结论,同时降低了FDI垂直溢出效应的显著性;采用加总的数据不仅会高估中西部地区组、国有企业组以及中国台湾、香港和澳门地区组等组别正向FDI水平溢出效应的可能性,而且还会高估了这些组别得到FDI溢出效应是显著的结论的可能性;另外,与FDI垂直溢出效应相比,文献更有可能得到负向FDI水平溢出效应和FDI水平溢出效应是不显著的结论,在FDI垂直溢出效应中,前向FDI溢出效应的显著性明显低于后向FDI溢出效应的显著性。
This paper presents a meta-analysis of the empirical literature on FDI spillovers to indentify whether the findings of FDI spillovers are affected by method and structure heterogeneity factors.The meta-analysis results suggest that the findings of FDI spillovers are influenced by the indices construction,data characteristic and omitted variables problem;the probabilities of negative FDI horizontal spillovers are higher but the significances of FDI vertical spillovers are lower when the DGMM/sys-GMM econometric method is used in the empirical literature;the probabilities of positive FDI horizontal spillovers and significance of FDI spillovers of Midwest region grouping,state owned enterprise grouping,and Taiwan,Hong Kong,Macao China grouping are overestimated when the aggregate data are used in the empirical literature;the probabilities of negative and insignificant FDI horizontal spillovers are higher than ones of FDI vertical spillovers;the probabilities of significant FDI forward spillovers are lower than ones of FDI backward spillovers.
出处
《金融评论》
2012年第6期34-45,121-122,共12页
Chinese Review of Financial Studies
关键词
Meta回归分析
FDI水平溢出效应
FDI垂直溢出效应
方法异质性
结构异质性
Meta-regression Analysis
FDI Horizontal Spillovers Effects
FDI Vertical Spillovers Effects
Method Heterogeneity
Structure Heterogeneity
