Impact of Crime on Spatial Analysis of House Prices: evidence from a UK city

David McIlhatton, W. S. McGreal, P. Taltavul de la Paz, A. Adair

Research output: Contribution to journalArticlepeer-review

16 Citations (Scopus)


Purpose There is a lack of understanding in the literature on the spatial relationships between crime and house price. This paper tests the impact of spatial effects in the housing market, how these are related to the incidence of crime and whether effects vary by type of crime. Design/methodology/approach The analysis initially explores univariate and bivariate spatial patterns in crime and house price data for the Belfast Metropolitan Area using Moran’s I and LISA models and secondly utilises spatial auto-regression models to estimate the role of crime on house prices. A spatially weighted two stage least squares model is specified to analyse the joint impact of crime variables. The analysis is cross-sectional based on a panel of data. Findings The paper illustrates that the pricing impact of crime is complex and varies by type of crime, property type and location. It is shown that burglary and theft are associated with higher income neighbourhoods whereas violence against the person, criminal damage and drugs offences are mainly associated with lower priced neighbourhoods. Spatial error effects are reduced in models based on specific crime variables. Originality/value The originality of this paper is the application of spatial analysis in the study of the impact of crime upon house prices. Criticisms of hedonic price models are based on unexplained error effects, the significance of this paper is the reduction of spatial error effects achievable through the analysis of crime data.
Original languageEnglish
Pages (from-to)627-647
JournalInternational Journal of Housing Markets and Analysis
Issue number4
Publication statusPublished - 2016
Externally publishedYes


Dive into the research topics of 'Impact of Crime on Spatial Analysis of House Prices: evidence from a UK city'. Together they form a unique fingerprint.

Cite this