摘要：The traditional view of cities as monocentric conglomerates of people clustered around an employment center, driving economic growth in cities that subsequently trickles down to the hinterland, is increasingly being challenged. In particular, the role of space, technological leadership, human capital, increasing returns to scale and industrial clustering as well as hierarchical organization principles have been emphasized in the more recent literature. This paper utilizes exploratory and spatial econometric data analysis techniques to investigate these issues for U.S. counties using data from 1969 through 2003. Ultimately, contiguous and hierarchical organization and interaction patterns are captured using an endogenous growth model allowing for spatial effects, inspired by earlier work on human capital and technology gaps. We investigate a neoclassical growth model and compare it to a spatial version of an endogenous growth model allowing for "domestic" investment in human capital and catch-up to the technology leader, and find that human capital strongly contributes to growth in a neoclassical setting, but much less so in an endogenous setting. In the endogenous model the catch-up term dominates in comparison to "domestic" human capital effects.
