Using a spatial econometric approach to identify the main determinants and spillover effects of residential property prices in La Spezia (Italy)

Abstract

We employ a spatial econometric approach to investigate the factors influencing residential property prices in La Spezia province (Italy). Unlike traditional hedonic models, which often overlook spatial dependencies, our methodology explicitly accounts for spatial autocorrelation, thereby yielding more robust and accurate estimates. Diagnostic spatial tests reveal significant spatial dependence in both property prices and context variables. To address this, we adopt the Spatial Durbin Error Model (SDEM), using a first-order Queen contiguity weight matrix. This model not only enhances explanatory power but also improves predictive accuracy. By incorporating spatial effects, the SDEM enables the disentanglement of direct and spillover influences, offering a more comprehensive understanding of the determinants of property prices. The findings demonstrate the importance of spatially-aware models not only in the formulation of effective housing policies and urban development strategies but also in appraisal practices, where they improve the accuracy of real estate valuation.