Geographically Weighted Regression Models to Investigate Urban Infrastructures Impacts

Land transformation, producing effects on the physical and functional system, can generate externalities, modifying the individuals’ perception of a given place. These effects, and in particular the changes in terms of urban quality, influence the success and effectiveness of an intervention in social, economic, environmental terms. To understand the perception of urban quality and define the elements that can contribute to its definition, urban infrastructures are increasingly subject to evaluation and quantitative measures and estimation. In the literature, it is possible to identify different types of approaches for assessing the quality of urban environments, both monetary and non-monetary. This study identified the impacts of urban infrastructure on residential property prices in Singapore. Through the application of the Hedonic Prices Method. Once the significance of the model variables was verified through an Ordinary Least Square (OLS) estimator, Geographically Weighted Regression (GWR) and Multi-scale Weighted Regression (MGWR) models were applied to take under control the spatial heterogeneity of the real estate market. Results show how MGWR model explained better the relationship between the selling price and structural and accessibility variables. This application built a bridge between economic valuation and local planning, supporting policy makers to map and identify weak areas in Singapore.