Spatial Autocorrelation in Housing Prices: Analysis of the Brno Metropolitan Area

Abstract

This study examines the persistence of spatial autocorrelation in housing prices within the Brno metropolitan area, focusing on the dynamics between the core city and its peripheral municipalities. The aim of the study is to test whether spatial autocorrelation exists and the extent to which it can be explained by traditional variables. Using a log-linear specification, the analysis employs a hedonic price model estimated separately for the city’s central and peripheral areas. The dataset comprises 19,286 residential transactions from the period 2020–2023. Explanatory variables include structural characteristics, land use, crime rates, and accessibility to services. Spatial autocorrelation is measured using Moran’s I and Local Indicators of Spatial Association, both on raw data and model residuals. The results reveal that price clustering is substantially stronger in peripheral areas, where lower market liquidity and limited substitutability of dwellings amplify the transmission of price signals across municipal boundaries. The findings suggest that localized factors such as planning agreements and limited market liquidity affect price transmission across municipal borders. The study’s findings highlight that spatial autocorrelation is not merely a statistical pattern but a mechanism shaping housing affordability, market efficiency, and the distribution of price shocks. In peripheral areas, prices are strongly affected by proximity to the city center and access to public infrastructure, indicating the need for coordinated metropolitan policies, particularly in rapidly growing suburban areas where spillover effects are most pronounced. The study contributes to housing economics by quantifying the structural sources of spatial dependence and demonstrating their relevance for metropolitan policy design.