Using Network Analytics to Improve Targeted Disruption of Police Misconduct

• Author: Timothy I. C. Cubitt
• DOI: http://doi.org/10.1177/10986111211057212
• Published date: 01 March 2023
• Date: 01 March 2023
• Subject Matter: Articles

Article

Police Quarterly

2023, Vol. 26(1) 24–53

© The Author(s) 2021

Article reuse guidelines:

sagepub.com/journals-permissions

DOI: 10.1177/10986111211057212

journals.sagepub.com/home/pqx

Using Network Analytics to

Improve Targeted Disruption

of Police Misconduct

Timothy I. C. Cubitt

1



Abstract

Research into police misconduct traditionally considers the correlates and antecedents

of misconduct among individual off‌icers, as a means of disruption or prevention.

However, more recently, deviance among police has been considered through net-

work perspectives. This study considered 7755 allegations of misconduct accrued

by 1495 off‌icers from the Baltimore Police Department between January 2015 to

January 2020. A social network analysis was employed to consider the charac-

teristics and differences of misconduct networks between assignments and to

identify key off‌icers within these networks. Findings suggested that the misconduct

networks of patrol assignments functioned marginally different to investigations or

specialist duties. Discrete communities of misconduct were identif‌ied within each

assignment, including a small number of off‌icers that were particularly important

to supporting these networks. This study holds practical implications for the

identif‌ication and disruption of misconduct networks among law enforcement

agencies.

Keywords

police misconduct, social network analysis, broker, complaints, accountability

1

Tasmanian Institute of Law Enforcement Studies, University of Tasmania, Hobart, Tasmania, Australia

Corresponding Author:

Timothy I. C. Cubitt, Tasmanian Institute of Law Enforcement Studies, University of Tasmania, Hobart,

Tasmania 7005, Australia.

Email: Timothy.cubitt@aic.gov.au

Introduction

The prevalence of deviance among police is subject to notable conjecture (Porter &

Warrender, 2009). However, it has recently received a considerable, and justif‌iable,

increase in attention. The ability of law enforcement agencies to maintain legitimacy,

both among the general public and their own staff (Walker & Archbold, 2005), relies on

robust methodologies for detection, investigation, and management of instances of

misconduct. In considering the phenomenon of police misconduct, there appear to be

two approaches under which an array of analyses may fall. The f‌irst approach considers

the individual characteristics of off‌icers who are prone to deviance (Kane & White,

2009), and the second approach favors the view that misconduct is a systemic phe-

nomenon and should be analyzed at the group level (Punch, 2003). At present, the

majority of research in this f‌ield considers the former, with the intention of identifying

characteristics of off‌icers, and antecedent behaviors to misconduct, that may aid

agencies in its disruption. The present research therefore seeks to consider the latter. In

particular, consideration is given to groupings of misconduct, termed communities, and

the role of off‌icers that may not be detected by prevalence of misconduct alone, in the

construction of these communities. The legitimacy of policing agencies relies upon the

ability to demonstrate accountability to the community, through detection and dis-

ruption of misconduct. This research considers the ability to identify groups of

misconduct and attempts to offer an additional detection methodology through the

application of network analytics.

Literature Review

Network Analysis

Social Network Analysis (SNA) is a methodology employed to understand human

behavior through interactions. These methods have been used to consider the function

of criminal networks, on the understanding that organized criminal enterprise is often

nested among a discrete group of individuals, rather than a series of disparate indi-

viduals (Kadushin, 2005). This is an important feature of crime analysis, as strategies

employed for disruption are often contingent upon whether the crime type is committed

by singular individuals or facilitated and committed by an embedded network (Johnson

& Reitzel, 2011). In this sense, the methodology employed for disruption of misconduct

among police is contingent upon an understanding of whether the misconduct is

embedded in a network or relates to an individual.

While network analyses are frequently attributed as originating in Stanley Milgram,

1967 “small world problem,”over time these methods have undergone substantial

analytical advancement, prior to application to crime. Employing SNA using crime data

may consist of the development of a sociogram, or a descriptive graph, and the ap-

plication of network measures to understand the function of the network and the roles of

individuals in that network (McGloin & Kirk, 2010). While the application of SNA to

Cubitt 25