Knowledge spillovers, spatial dependence, and regional economic growth in U.S. metropolitan areas

The main purpose of this research is to investigate three basic issues central to the explanation of knowledge spillovers and regional economic growth: (i) the spatial distribution of innovative activity across U.S. regions, (ii) agglomeration economies and the geography of innovative activity, and (iii) spatial dependence and regional economic growth. To establish a basis of the analysis, theoretical underpinnings on the role of knowledge externalities and increasing returns in the process of economic growth are summarized in Chapter 2. Chapter 3 provides the rationale for employing spatial econometric methods in this research, both from a theoretical context of geographic knowledge spillovers and from a methodological concern for spatial data. Chapter 4 is devoted to an exploratory spatial data analysis in order to better assess differences in the spatial distribution of innovative activity across U.S. metropolitan areas. The analysis of metropolitan innovative intensity provides strong evidence of positive spatial autocorrelation for the period 1990–1999. In Chapter 5, the impact of geographic knowledge spillovers on metropolitan innovative activity is discussed. The main result is that innovative activity in a metropolitan area is positively affected by both specialization and diversity externalities in high technology industries. The spatial econometric analysis shows that geographic knowledge spillovers exist across metropolitan boundaries and high technology specialization externalities are more localized than high technology diversity externalities. In Chapter 6, this research examines the extent to which metropolitan economic growth is related to the three different channels of knowledge externalities in high technology industries. High technology specialization has a positive and significant effect on metropolitan economic growth. However, there are no substantive spillover effects of the three different channels of knowledge spillovers that flow across metropolitan boundaries. From the convergence point of view, the result does not find significant evidence of absolute convergence for the period 1990–1999. However, after controlling for structural characteristics determining metropolitan economic growth, significant evidence of convergence for the period is found. In Chapter 7, this research concludes with a summary of key findings and concluding remarks and explores the implications for regional policy.