Regional Underdevelopment: Is FDI the Solution? A Semiparametric Analysis

In this Paper we aim to examine the regional impact of foreign-owned establishments on the performance of domestic establishments in the electronics industry in the UK. We use establishment-level data taken from the UK Census of Production (the ARD) to this end. In the econometric specification, we allow for the time-varying endogeneity of the factors of production function, and correct for the sample-selection bias generated by plants with larger capital stocks surviving in spite of lower productivity realisations. The FDI spillover literature has so far abstracted from the selection problem generated by plant exit. To our knowledge, this is the first Paper that simultaneously attempts to correct the production parameter estimates for selectivity induced by plant exit as well as time-varying endogeneity (the ‘not so fixed’ effect), before identifying the impact of foreign direct investment on domestic plant’s productivity. The results indicate that positive spillovers exist but are mostly confined to the region in which the MNE locates. A number of characteristics influence their level, they are higher from non-US firms (in particular Japanese firms) and in more-developed regions.