On to the next one? Using social network data to inform police target prioritization

Abstract

Purpose Target prioritization is routinely done among law enforcement agencies, but the criteria to establish which targets will lead to the most crime reduction are neither systematic, nor do they take into account the networks in which offenders are embedded. The purpose of this paper is to propose network capital as a guide for prioritization exercises. The approach simultaneously considers a participant’s network centrality and their crime-affiliated attributes. Design/methodology/approach Data on all police interactions are used to map the social networks of two mutually connected police targets from a mid-size city in British Columbia, Canada. Network capital is captured by combining the extent to which individuals act as brokers between otherwise unconnected individuals (betweenness centrality), their number of contacts in the network (degree centrality), and whether they have a criminal record, gang ties, and a firearm carrier status. Findings The network comprises 101 associates, with nine mutual contacts amongst the two targets, and half of the network having a crime-affiliated attribute. Network capital directed the prioritization process to seven associates who stood out. Targeting strategies from two different investigative outcomes are compared. Research limitations/implications The specific recommendations of the study can only be interpreted within the context of the initial targets around which the network was constructed. As a prioritization approach, however, network capital is generalizable to other contexts with implications for law enforcement officials and, more broadly, the community. Originality/value The study provides insights into the practical application of network analysis with already existing police data. Network capital is data driven, which comes with its own limitations, but which constitutes an improvement over purely informal approaches to target prioritization.

Smart Policing: A Critical Review of the Literature

“…For example, Philadelphia police estimates the harm scores of violent crime offenders by aggregating the weighted scores of static factors like gravity of offence, severity of sentence and time decay [22]. In the case of gangs, network analysis is a better approach to identify influential targets, for instance, using capital scores and other network metrics [23]. However, network analysis suffers on account of missing links arising from untraced and unreported cases.…”

“…Summary of new roles of data in smart policing Network analysis calculates network capital from gang-related data for targeting the most instrumental criminals for their apprehension[23].• Predictive crime mapping applies machine learning tools (near repeats & risk terrain analysis) to directed data to forecast crime patterns for directed patrolling to prevent crime (e.g. PredPol, HunchLab etc)[24][25][26].…”

Human and Social Capital Strategies for Mafia Network Disruption

“…The social capital approach requires the use of SNA which focuses on the computation for network measurements such as density and centralization, analysis of clusters and the measure of the centrality of individuals [18]- [20]. Centrality measures are able to identify critical actors in a criminal group [21]- [24].…”