Spatiotemporal patterns and distributions of harm within street segments

Abstract

PurposeVirtually all analyses of hotspots have been devoted to a crude counting system, i.e. tallying the number of occurrences that take place in pre-specified units of space and time. Recent research shows that while usually half of all criminal events are concentrated in about 3 percent of places commonly referred to as “hotspots” of crime, similar proportions of harm concentrate in only 1 percent of places. These are “harmspots.” Identifying that harm is a more concentrated issue suggests wide policy and research implications, but what are the dynamics of these harmspots? The paper aims to discuss this issue.Design/methodology/approachThis paper provides a descriptive framework for measuring, as well as evidence about, these patterns and concentrations, harmspots in Sussex, England.FindingsThere are four discrete offense categories that account for 80 percent of all the harm within harmspots. These categories include: sexual offenses, violence against the person, robbery and theft. Within these high harmspots, crime counts and harm are strongly correlated (r=0.82, p=0.001). Temporal analyses show that harmspots are not evenly spread across time and place, with night time and weekends becoming substantially more susceptible to harm – more than count-based models. Harmspot trajectory analysis suggests evidence of stability over time within the high harmspots; most harmspots remain chronically inflicted with harm. Violence and sexual offenses are random in their spatial distribution between the harmspots, but robberies and theft are more closely coupled to particular harmspots than others.Originality/valueImplications of these findings are discussed in terms of future research avenues and crime policy.