Most scholars believe that new firms concentrate geographically where conditions favor entrepreneurial activity. However, empirical evidence on this point is mixed. For example, while Zucker, et al (1998) find that new biotechnology firms congregate near universities, Dumais, et al (2002) conclude that startups have a “de-agglomerating” effect by locating away from same-industry agglomerations.

Audrestch, et al’s (2005) “knowledge spillover theory of entrepreneurship” posits that a region’s total level of entrepreneurial activity is driven by knowledge creation and local economic growth. The theory suggests that knowledge-driven entrepreneurship is distinct from growth-driven entrepreneurship. Thus, the mixed evidence may reflect the heterogeneity across different types of entrepreneurial activity and their respective causes.

The purpose of this study is to explore this possibility using a multi-industry dataset of firm births encompassing the continental U.S. This study explores two fundamental conjectures—that new firms concentrate geographically and that the determinants of knowledge-based entrepreneurship are distinct from the determinants of growth-based entrepreneurship.