Spillovers from FDI and their determinants: the case of China

The thesis examines technology and export spillovers from foreign direct investment (FDI) and their determinants in China, both at an aggregate industry level and disaggregated firm level, using both cross-sectional and panel data.\ud \ud At the industry level, we examine technology spillovers by estimating industry value-added as a function of the interaction terms of the technology transfer of FDI with the technology gap, relative capital intensity and relative labour supply over an eight-year panel data set from 1995 to 2003. It is found that technology spillovers vary across both industries and time, and the technology gap plays a negative role in technology spillovers. At the firm level, using a simultaneous equation model estimated over a comprehensive cross-sectional micro-data set that covers over 180,000 firms in the manufacturing sector in China for 2003, we find positive and substantial technology spillovers from FDI. However technology spillovers do not exist across all industries. In the cultural, educational, and sporting product manufacturing industry, we find no evidence of technology spillovers. In an analysis of technology spillovers in Gansu province, we find positive spillovers, with the scale depending on firm's R&D.\ud \ud On export spillovers, the thesis first set up a partial equilibrium model to show that if the presence of FDI reduces domestic firm's export costs, for example by spillovers of knowledge about foreign markets and through labour movement, then domestic firms' export intensity, which is equal to the share of firms' exports in their total sales, will be unambiguously promoted. This hypothesis is tested using a Heckman sample selection model, estimated over both the comprehensive cross-sectional micro-data in China' manufacturing sector in 2003 and the four year panel data in the cultural, educational, and sporting product manufacturing industry. We find that FDI generates export spillovers, with the scale of spillovers depending on firm characteristics, such as firm size, age, and geographical location