Business location and spatial externalities: Tying concepts to measures

Spatial externalities among businesses, though notoriously difficult to measure, are a central concern in urban and regional economics. Traditionally they–along with closely related concepts such as agglomeration economies–have been studied empirically with hedonic models, production and cost functions, and simplified growth models. More recently, researchers have begun using direct measures of proximity among businesses to shed light on the influence of externalities on industrial location, regional growth, and localized technological change. The shift has been aided by an explosion in spatially-referenced economic data, advances in spatial statistics, and the advent of affordable and user-friendly GIS and related software. As existing indicators of concentration and spatial association have been adapted for the economic domain and new ones developed, the pool of measures useful for business location research generally, and externalities more specifically, has expanded. In this paper, we systematically review and compare a set of leading indicators of business co-location using standard data sets and evaluation criteria. Ultimately, our aim is to assess the capabilities and limits of the measures for understanding spatial business externalities. More generally, we discuss a number of common challenges associated with drawing inferences from cross-sectional spatial data.