Steal that car: A spatial analysis of motor vehicle theft in New South Wales, Australia

This article seeks to better understand the increase in motor vehicle theft (MVT) in New South Wales (NSW), Australia, by answering three research questions: (i) what is the spatial distribution of MVT rates across Local Government Areas (LGAs) in NSW?, (ii) which factors predict MVT?; and (iii) do those factors play a more significant role in particular regions as compared to others? Using generalized least squares (ordinary least squares (with robust standard errors)) and geographically weighted regression, we attempt to identify significant predictors of spatial non-stationarity of three years of average MVT rates across 128 NSW Local Government Areas. Rates of MVT varied widely across NSW LGAs and showed significant spatial clustering. Geographically weighted regression revealed strong spatial non-stationarity, particularly for youth, Indigeneity, migration, and opportunity measures, with effects concentrated in regional and remote areas. The predictors of MVT across NSW are spatially dependent and prevention strategies should target both social and opportunity structures at the local level.