Regional economic effects of university research: A spatial econometric perspective.

A sizable literature has documented the important role of universities in the development of the world\u27s largest high technology concentrations: Stanford University in Silicon Valley, MIT in Boston and Cambridge University in Cambridge (U.K.). It is the fundamental research question of this dissertation to assess the extent to which local university knowledge impacts are unique, non-repeatable phenomena or whether they can be experienced in other regions as well. This dissertation represents the first study in the literature that provides a systematic, US-wide analysis of university-high technology connections at the lowest possible level of spatial aggregation. Its contributions are twofold: (1) it focuses on the regional aspects of the interaction between high technology innovations, university and private R&D at the proper spatial scale, at the level of counties and metropolitan areas; (2) it uses the specialized methodology of spatial econometrics to explicitly deal with potential spatial effects in cross-sectional data. The study is based on a unique data set of high technology innovations, industrial R&D employment in the US in 1982. From a county level exploratory spatial data analysis, it is shown that the concentration of high technology product innovations, private R&D and university research follow a similar spatial pattern across the USA. A formal regression analysis is carried out for 43 US states and 125 Metropolitan Statistical Areas which implements the Griliches-Jaffe knowledge production function framework. This provides strong evidence that university research expenditures have a positive and very significant effect on aggregate high technology innovation. Moreover, this university impact follows a distance decay pattern. Additionally, results for US MSAs indicate that research employment in high technology R&D laboratories is significantly determined by the level of local university research expenditures. A model that explicitly accounts for spatial variation in the coefficients of the knowledge production function indicates that the same amount of university research expenditure yields substantially different number of innovations depending on certain agglomeration factors. The findings in this dissertation strongly indicate that a stimulation of research activities in universities located in existing agglomerations of high technology production and research has significant positive spillover effects on the regional economy