• violence;
  • co-offending;
  • peers;
  • longitudinal;
  • crime


Although the idea that youthful offenders are affected by the company they keep is widely accepted, evidence in support of this idea is based primarily on reports provided by offenders and their peers. As an alternative to relying on reports of criminal behavior, a method that may overestimate the role that peers play in criminal behavior, the current research on co-offending uses court records to identify and track over time individuals who are known to commit crimes together. The present investigation is the first co-offending study to track patterns of violent criminal behavior (over an 18-year period) among a sample of urban offenders and their accomplices. The study tests whether violence “spreads” from violent offenders to those inexperienced in violence. Results indicate that nonviolent offenders who commit their first co-offense with violent accomplices are at increased risk for subsequent serious violent crime. Findings suggest that lessons of violence can be learned “on the street,” where knowledge is passed along through impromptu social contexts, including those in which offenders commit crimes together. Aggr. Behav. 28:97–108, 2002. © 2002 Wiley-Liss, Inc.