Psychological research using mostly cross-sectional methods calls into question the presumed function of shame as an inhibitor of immoral or illegal behavior. In a longitudinal study of 476 jail inmates, we assessed shame proneness, guilt proneness, and externalization of blame shortly after incarceration. We interviewed participants (N = 332) 1 year after release into the community, and we accessed official arrest records (N = 446). Guilt proneness negatively and directly predicted reoffense in the 1st year after release; shame proneness did not. Further mediational modeling showed that shame proneness positively predicted recidivism via its robust link to externalization of blame. There remained a direct effect of shame on recidivism: Unimpeded by defensive externalization of blame, shame inhibited recidivism. Items assessing a motivation to hide were primarily responsible for this pattern. Overall, our results suggest that the pain of shame may have two faces-one with destructive potential and the other with constructive potential.
Keywords: antisocial behavior; emotions.