The (In)Stability of Residential Burglary Patterns on Street Segments: The Case of Antwerp, Belgium 2005–2016

Abstract

Objectives: To examine the spatial concentration and spatial stability of residential burglary at micro places in the context of a substantial city-level burglary drop in Antwerp, Belgium. Methods: 51,337 police recorded home burglary incidents for the period 2005–2016 are geo-referenced to 26,875 street segments. Longitudinal trends in spatial concentrations of burglary are considered using descriptive statistics, generalized Gini coefficients, and local Getis–Ord statistics. Andresen’s (Appl Geogr 29(3):333–345, 2009) non-parametric spatial point pattern test (SPPT) is used to identify spatial stability in burglary point patterns and evaluate the ubiquity of a city-level burglary drop across street segments. A longitudinal extension of the SPPT is implemented. Results: Residential burglary is substantially concentrated in street segments. Burglary point patterns exhibit a moderate to high degree of spatial stability over time. Local analyses show that 91% of street segments with burglary experienced a net decrease in crime and under 1% of street segments with burglary experienced a net increase. Absolute spatial stability over time is found for just 1.43% of street segments with burglary and minor increases are consistently observed for as few as 11 street segments with burglary. Conclusions: The citywide home burglary drop manifested itself uniformly across street segments with burglary and the majority of street segments that experienced burglary contributed relatively equally to the crime drop. In other words, we find no strong evidence that the city-level crime drop can be tied to substantial decreases in a few specific places, nor do we find that the reductions in burglary are spatially concentrated.