Measuring Agglomeration: Which Estimator Should We Use?

Abstract

With the tremendous growth of research in industrial agglomeration has come increased complexity in the construction of a `better' measure of agglomeration. The research presented here tests the properties of commonly used agglomeration indices through a series of simulations. Results suggest that applied researchers benefit from continuous measures of agglomeration like the Duranton-Overman index or a continuous version of the Gini coefficient, especially for industries with fewer establishments. Conclusions suggest that commonly used estimators of industrial agglomeration vary in how they quantify spatial distributions and researchers need to use the agglomeration index best suited for their specific research question.

Keywords: agglomeration; spatial statistics

JEL Classification: R