Foreign direct investment, externalities and geography : an analysis of the effects of geographic proximity on the externalities from FDI in Mexican manufacturing industries
Contemporary research on externalities from FDI suffers from two central problems. First, estimates of such externalities may be biased when FDI is endogenous to the empirical model. Second, there is an important lack in empirical research regarding the identification of structural factors influencing the type and level of FDI-induced externalities. The goal of the thesis is to address both issues. It starts with a theoretical discussion of FDI and externalities, followed by an overview of contemporary empirical research, highlighting the main estimation problems. Next, theories of agglomeration economies are discussed, in an attempt to identify a determinant of FDI-induced externalities in the form of geographical proximity of manufacturing activities. This discussion, supported by an overview of the limited related available empirical evidence, indicates that this concept is a likely candidate to be such a determinant. The next two chapters use unpublished and thus far unexplored data from the 1993 Mexican economic census to estimate FDI-induced externalities in Mexican manufacturing industries. In this part, the main empirical model is developed and estimated. In addition, the robustness of the initial findings of this empirical model is assessed. Furthermore, the estimation issues that are identified in the first part of the thesis are addressed. Most importantly, I introduce an instrumental variable estimation that controls for the problem of endogenous FDI. This instrumental variable estimation functions satisfactorily; as such, it represents the first successful empirical unbiased estimation of FDI-induced externalities in a cross-sectional setting. Finally, the last part of the thesis offers empirical evidence of the effects of geographical proximity on FDI-induced externalities. The findings indicate that geographical proximity does influence such externalities in a multi-faceted fashion. First, the level of geographical concentration of an industry enhances the occurrence of positive externalities within an industry. Second, from a regional point of view, geographical proximity enhances the occurrence of externalities that arise within and between industries in a region. Third, FDI-induced externalities that arise between industries also appear to spill over between neighbouring regions.