Measuring the Localization of Economic Activity: A Random Utility Approach

The recent index proposed by Ellison & Glaeser (1997) is now well established as the preferred method for measuring the localization of economic activity. We build on McFadden's Random Utility (Profit) Maximization framework, to develop a parametric version of this measure that is more consistent with the theory originally proposed by Ellison and Glaeser (EG). Given that our method is regression based, it goes beyond the descriptive nature of the EG index, allowing us to evaluate how the localization measure behaves with changes in the determinants that drive firms' location decisions.