Simple indicators of crime and police: How big data can be used to reveal temporal patterns
| Published date | 01 May 2023 |
| DOI | http://doi.org/10.1177/14773708221120754 |
| Author | Philipp M. Dau,Maite Dewinter,Frank Witlox,Tom Vander Beken,Christophe Vandeviver |
| Date | 01 May 2023 |
| Subject Matter | Articles |
Simple indicators of crime and
police: How big data can be
used to reveal temporal
patterns
Philipp M. Dau
Department of Criminology, Criminal Law and Social Law, Ghent
University, Belgium
Maite Dewinter
Department of Geography, Ghent University, Belgium
Frank Witlox
Department of Geography, Ghent University, Belgium
Department of Geography, University of Tartu, Estonia
College of Civil Aviation, Nanjing University of Aeronautics and
Astronautics, China
Tom Vander Beken
Department of Criminology, Criminal Law and Social Law, Ghent
University, Belgium
Christophe Vandeviver
Department of Criminology, Criminal Law and Social Law, Ghent
University, Belgium
Abstract
This study demonstrates how temporal summary statistics can be a guiding tool for big data ana-
lyses to unravel temporal patterns of crime and police presence. Simple indicator statistics were
used to identify temporal clusters of crimes and police presence, and to investigate potential links
between the two. The methodology was applied on an anonymized police database, including
Corresponding author:
Christophe Vandeviver, Department of Criminology, Criminal Law, & Social Law, Ghent University,
Universiteitstraat 4, 9000 Ghent, Belgium.
Email: christophe.vandeviver@ugent.be
Article
European Journal of Criminology
2023, Vol. 20(3) 1146–1163
© The Author(s) 2022
Article reuse guidelines:
sagepub.com/journals-permissions
DOI: 10.1177/14773708221120754
journals.sagepub.com/home/euc
reported crime events and police presence data, from a medium-sized European police depart-
ment. The results illustrated that certain crime types occurred more during the day (e.g., burglar-
ies), while others were more prevalent at night (e.g., drug crimes, motorbike and car theft). Police
presence showed dispersed temporal patterns and little temporal focus on any type of crime. The
research shows that temporal summary statistics can be used to support an explorative analysis of
big datasets and guide subsequent spatiotemporal analyses of crime and police data. The summary
statistics offer an accessible approach to analysing extensive datasets of policing activity and
improving evidence-based policing strategies.
Keywords
Covid-19, crime, evidence-based policing, global positioning system, police, temporal analysis
Introduction
The rapid development and adoption of novel technologies have led to a new age of big
data. The challenges and benefits of this for the public sector will reshape the criminal
justice system (Elevelt et al., 2019; Hutt et al., 2018; Wain and Ariel, 2014; Wain et al.,
2017). One clear benefit is the growing availability of extensive datasets of reported
crimes and police activity. Continuous improvement and digitization have enabled police
departments to utilize dense digital databases and introduce real-time tracking of police activ-
ity and reported crimes (Cordner, 1979; Elevelt et al., 2019). Crime databases are now able to
store more detailed information on the time and place of offences or linked offender biomet-
rics. This has enabled more detailed investigations to be carried out into the temporal aspects
of crime, such as repeat offending or the seasonality of crime (Felson and Clarke, 1998; van
Sleeuwen et al., 2021b). Real-time tracking of policing activity allows precise measurement
of the location of patrols or the amount of patrol time not assigned to emergency call
response, so-called “officer downtime”(Cordner, 1979; Famega, 2005).
This article has two objectives. First, we describe the context and the use of Felson and
Poulsen’s (2003) summary statistics and demonstrate why and how they can be adapted
and applied to police tracking data. Second, we apply those statistics to a variety of crime
types and the proportion of officer downtime (i.e., unassigned patrol time) to test and
illustrate the usability of this method to analyse big datasets of policing data. In doing
so, we demonstrate how extensive datasets and detailed temporal analysis can be made
more accessible for both crime scientists and practitioners.
First, we highlight the need to include temporality in policing as well as criminology
and introduce the methodology of the statistics and data used, then we present the results
of their application and discuss potential implications and applications.
Temporal aspects of police presence and simple indicators of crime
The data-orientated development has sparked plenty of spatiotemporal research on crime
and policing, and the interplay of the two (e.g., Ariel et al., 2019; Davies and Bowers,
2020; Weisburd et al., 2017). However, extant research has focused more on spatial
than temporal aspects (Ratcliffe, 2002, 2010). This is partly due to the exponential
increase in the units that need to be analysed when the micro levels of time and space
Dau et al. 1147
To continue reading
Request your trial