Offending Characteristics, Risk of Recidivism and Prediction of Reoffence in Sex Offenders, from Adolescence to later Adulthood

The basis of this study was a census of all offenses of sexual assault and sexual abuse that were reported to the Berlin police between 1994 and 1999 (N = 2,446). The purpose was to analyse the particularities of elderly offenders (60 years or older, n = 68) compared to the 273 juvenile and the 2,105 younger adult offenders with regard to prior offenses, index offenses and recidivism rates. Furthermore, the analysis focused on the prediction of sexual and violent reoffence by means of the static-99 within the elderly age group. The analysis showed that the rate of former sexual offence committed by seniors was comparable to the younger adults; however, sexual and non-sexual violent pre-offences were much less frequent. On the other hand, the seniors were more frequently involved in sexual abuse, their modus operandi showed lesser degrees of intensity of the sexual acts and less use of force and violence (even among sexual abusers only) and group-offence was definitely uncommon. Furthermore the rate of sexual reoffence was similar to the juvenile offenders, but lower in comparison to adult offenders younger than 60 years. A more sophisticated analysis of the age-effects on recidivism-rates confirmed the anticipation of growing risks until middle adulthood and decreasing risks of reoffence afterwards. However, an explicit trajectory of first growing and than decreasing risks in the course of aging could only be found among offenders with sexual abuse, whereas the zenith was shifted in elderly stages of age. On the other hand, among offenders with sexual assaults, the risks of sexual reoffending seemed rather similar until the middle adulthood and in-creasingly lower in later stages. In spite of these aging-effects and the particularities of the elderly sexual offenders, the static-99 predicted sexual reoffence among the senior offenders quite well and there was no difference in comparison with the total sample.