Predicting Sexual Offending

In most courtrooms around the world, one of the prime considerations in terms of handing down a sentence for sexual offenders is the basic question: Will he reoffend?   While there are a variety of actuarial risk instruments available that draw on relevant factors relating to the offense and the offender's criminal history, prediction of long-term risk of sexual reoffending is always a tricky issue.  A recent study published in the August, 2006 issue of Law and Human Behavior (30:525-541) suggests that issue may be trickier still.  Marnie Rice, Grant Harris, and their associates at the Mental Health Centre at Penetanguishene, Ontario assessed a sample of 177 convicted sex offenders and found that police rapsheets tended to underestimate the actual number of sexual offenses which the offenders committed.  It was found that violent offenses were often not recorded as sexual offenses on rapsheets for various reasons.  Given that actuarial measures of sexual violence risk draw on  arrest records to determine criminal history, the potential for obtaining a skewed estimate of possible reoffending is considerable. For a reprint of the article, contact Marnie Rice at riceme@mcmaster.ca