This study examines how violent co-offenders shape future offending trajectories, distinguishing between committing violence with a violent co-offender (direct exposure) versus associating with violent accomplices (indirect exposure). Using police data, we applied binary logistic regressions to a sample of Belgian offenders (N = 20,203) to examine the likelihood of general and violent reoffending. Our results show that having a violent co-offender significantly increases the likelihood of future offending, whereas non-violent co-offenders have no pronounced impact. When the co-offense with a violent co-offender involved violence, the heightened likelihood of future offending was mitigated. Among reoffenders, prior violence is associated with future violent offenses, particularly for those with less severe violent offenses. Moreover, committing violence with a violent co-offender reinforces future violence for a large part of the sample (61% of offenders in T2). Overall, violent co-offenders influence future offending both directly and indirectly, with direct participation in violence emerging as a critical predictor of future violent behavior. These findings emphasize the need for further research and targeted interventions at key moments, such as after a violent co-offense.
Geeraert et al. (Thu,) studied this question.