Abstract

This study examines the influence of violent co-offenders on suspects’ future offending trajectories, distinguishing between the effects of committing violence with a co-offender (direct exposure) versus merely associating with violent accomplices (indirect exposure). We conducted logistic regressions on a sample of Belgian suspects (N = 20,203) using police data to examine the likelihood of reoffending and future violent offending. Our results demonstrated that having a violent co-offender increases the likelihood of future offending, while non-violent accomplices have no significant impact. However, if the co-offense with a violent accomplice involved violence, suspects are more likely to desist, possibly due to moral boundaries or deterrence. Among reoffenders, prior violence predicts future violent offenses, while more severe past violence slightly reduces the likelihood of future violence, likely due to shock or severe sentencing. Further, direct violent exposure with a violent accomplice increased the likelihood of future violence for reoffenders, suggesting a reinforcing effect. These nuanced findings highlight that violent accomplices influence future offending in complex ways. Overall, having a violent accomplice generally raises the likelihood of continued offending, while direct participation in violent acts with such accomplices plays a critical role in predicting future violent behavior. These findings emphasize the need for further research and targeted interventions at pivotal moments.