Spatial Econometrics: When Geography Shapes the Economy

Abstract

This article reviews the theoretical foundations, formal tools, and empirical applications of spatial econometrics. Building on Tobler's first law of geography, which holds that geographic proximity systematically structures economic interactions, the article traces the emergence of the discipline from Isard's pioneering work to Anselin's methodological formalization. It presents the properties that distinguish spatial econometrics — spatial dependence, spatial heterogeneity, and spatial weight matrices — and outlines the main econometric specifications (spatial lag, spatial error, spatial Durbin) as well as measures of spatial autocorrelation (Moran's I, LISA indicators). These tools are illustrated through several cases of economic clustering (Silicon Valley, European industrial districts) and spatial contagion of economic decline, before being applied to a synthesis of empirical findings on regional convergence, technological spillovers, and interjurisdictional fiscal competition. The results show that incorporating the spatial dimension substantially alters econometric estimates and the resulting economic policy recommendations. Spatial econometrics thus stands as a methodological necessity for any analysis based on georeferenced data.