Understanding Co-Offending Patterns and Criminal Mentorship Using Police Registry Data.

The current study examines registration data from the Dutch police pertaining to 2,757 individuals and 10 different crime categories, at the offense, individual, and network levels of analysis. Moreover, a novel estimate is introduced which identifies individuals’ tendency to operate as mentors or mentees in criminal networks. We find co-offending to be conditional on the type of crime and to typically involve groups of two to three offenders. Moreover, we find that men and those aged 18–25 are most likely to co-offend and that co-offenders typically do not show a preference for particular co-offenders. Individuals with central positions in the co-offending network were found to be older and to offend more frequently with a higher average of co-offenders per offense, than those less central. Lastly, we find the novel estimate to be a promising tool for identifying mentor and mentee roles in large quantitative network data. The results of the current study emphasize the importance of studying co-offending on different analytical levels, and to differentiate co-offending by different crime types, network positions, and individual roles, to better inform theory and policy. [ABSTRACT FROM AUTHOR]