Research Article: Adolescent offenders’ current whereabouts predict locations of their future crimes

Date Published: January 30, 2019

Publisher: Public Library of Science

Author(s): Wim Bernasco, Hedwig Eisenbarth.


Knowing where crime is likely to happen can help prevent it. Here I investigate whether two basic mechanisms of human mobility—preferential return and spatial exploration—explain and predict where offenders commit future crimes. A sample of 843 adolescents reported their hourly whereabouts during four days. In line with findings from other sources and populations, their locations were concentrated and predictable. During the subsequent four years, 70 of the 843 were apprehended for committing one or more crimes. Compared to others, these 70 future offenders had visited slightly more different locations. However, their action radius and the predictability of their whereabouts had been very similar to those who would not become offenders. The offenders perpetrated most of their crimes around places they had visited before, including places where they previously offended. These findings show that the predictability of human mobility applies to offending and to offenders as well, and helps us understand and forecast where they will commit future crimes. They may prove particularly useful in criminal investigations, as they suggest that police should generally prioritize suspects who are familiar with the location of the crime and its environs, either because of their legal routine activities or because of their prior offences.

Partial Text

The prevention of crime heavily depends on our ability to forecast where and when it will happen. Recently, police forces have successfully started to allocate their resources using algorithms that predict where and when crime is most likely to occur next [1, 2]. The algorithms are based on well-corroborated evidence that crime risk is temporarily elevated within a few weeks and within a few hundred meters from a previous crime [3]. This space-time pattern of crime (but not its spatial and temporal scale) is similar to the patterns that characterize the diffusion of infectious diseases and the propagation of earthquakes. In fact, it has first been demonstrated with methods originally developed in epidemiology [4, 5] and in seismology [6].

The statistical analyses are informed and structured by the crime pattern theory [7]. The theory asserts that offenders perpetrate crimes in and nearby their activity spaces, locations they are already familiar with through their recurrent mobility. In other words, it proposes that preferential return and spatial exploration not only drive routine activities, but also crime location choices.

The research was approved by the Ethics Committee for Legal and Criminological Research of the Faculty of Law of the Vrije Universiteit Amsterdam (CERCO).

The findings reported here suggest that all of us, law-breakers and law-abiders alike, may be less innovative and more predictable than we think we are, or would like to be. Although crime may be an exceptional activity, its location can be predicted by preferential return and spatial exploration, the same two mechanisms that generate our routine mobility. In line with the findings from prior studies, but based on direct measures of individual activity spaces, I demonstrated that offenders not only tend to commit crimes in and around the places they have visited before when pursuing either their legal daily activities, but also have a tendency to commit crimes in and around the locations of their previous crimes. Although low crime reporting rates and low clearances rates require cautious conclusions, the findings do seem to support the general assumption of environmental criminology that legal and illegal conduct are not fundamentally different, and follow the same behavioral laws.




Leave a Reply

Your email address will not be published.