Assessing crossover in a sample of sexual offenders with multiple victims

Purpose. This study examined the proportion of sexual offenders in England and Wales who exhibited 'crossover' in their choice of victim, as defined by age, gender and relationship to the offender. It subsequently aimed to identify criminal and demographic variables predictive of crossover. Method. The sample comprised 1,345 adult male sexual offenders. All had offended against multiple victims and had been discharged from custody in England and Wales between 1992 and 1996. Offence summaries within police records were used to collect victim details for each sexual conviction for each offender, and criminal conviction histories were obtained from the Offenders Index for the sample. Results. A quarter of the sample (24.5%, n = 330) demonstrated crossover behaviour in regard to at least one of the victim dimensions examined (i.e. age, gender or relationship). Using the risk predictor Static-99, these offenders were found to be significantly riskier than those who did not display crossover behaviour Multinomial regression failed to identify reliable variables predictive of crossover. Conclusions. The findings from this study are likely to represent an underestimate of the prevalence of crossover within sexual offenders in England and Wales. Reasons for this are discussed, and the implications of crossover for offender risk assessment, treatment, community supervision and policing practices are explored.