Spatial Dependence in Entrepreneurship Research

Abstract

Measures used in entrepreneurship research are often subject to spatial dependence. Spatial dependence renders ordinary least squares (OLS) estimation inappropriate because the estimates will be biased, inconsistent, and/or inefficient. The aims of this article are (a) to demonstrate how spatial dependence is especially problematic for entrepreneurship research and (b) to arm researchers with spatial modeling techniques that are more appropriate for such analysis. As such, not only will this article illustrate how to incorporate spatial dependence explicitly into the linear regression model, it also discusses how these techniques make it possible to explore and locate areas with particularly high levels of spatial dependence (i.e., hot spots). These techniques, although new to the management literature, are well known in both the regional science and geography literatures and are rapidly diffusing to economics, sociology, and related social sciences.