Implications of sex offender classification on reporting demographic characteristics, health, and criminal careers: Results from an Australia jurisdiction

Background: Using cross-sectional and retrospective data to classify sex offenders has implications for epidemiological and survey research. However, there is little empirical evidence regarding the practical implications of these for applied research. This study describes the data linkage process and subsequent classification of sex offenders in a cohort of prisoners recruited as part of an Australian inmate health survey. Methods: Data linkage was used to link the New South Wales (NSW) Inmate Health Surveys to the state’s offending database to identify men with histories of sexual offending. Sex offenders were classified into men who sexually offended against children only (ChildSOs), against adults only (AdultSOs), and men who sexually offended against both children and adults (PolySOs). Results: By using historical offending data rather than the current offence information only, an additional 35.4% of men were identified who had histories of sexual offences. Differences were found between the sex offender subgroups in terms of demographic characteristics, health, and criminal careers. PolySOs were the least likely to report less than high school education, self-employment, being single, and children. Half the ChildSOs reported a mental health issue and, similarly to PolySOs, half reported four or more chronic health conditions. In terms of their criminal careers, PolySOs were the eldest when committing their first: non-sexual, non-violent crime (M=43.2 years, SD=13.8); violent crime (M=39.5 years, SD=11.1); and sexual crime (M=47.8 years, SD=11.2) compared to the other groups. These men also committed more sexual offences (M=5.91, SD=11.2) compared to those who only offended against one victim age group. Conclusion: These findings suggested that whenever possible historical offending records should be used to more accurately identify sex offenders, and that differences in demographic, health, and criminal careers exist for the different sex offender subgroups. Keywords: adult sex offender, child sex offender, classification, data linkage, health, polymorphous, specialisation.

Adult Offender Assessment and Classification in Custodial Settings

This chapter describes the developments that have occurred over the past three decades in the area of offender assessment and classification, including discussion of why offender classification is so vital to correctional agencies. The importance of using actuarial approaches to predicting the risk of reoffending and danger to others is discussed, as well as the inclusion of static and dynamic factors on composite measures of offender risk and need. Particular attention is paid to the application of the principles of Risk, Need, and Responsivity (RNR) to offender assessment, classification, and subsequent work with the offender, often described as “offender case management.” How prison environments (including inmate and officer subcultures) can potentially interfere with the accuracy of risk and needs assessments is also debated.

Risk/Need Assessment, Offender Classification, and the Role of Childhood Abuse

Over the last several decades, the use of risk/need assessments has had a great impact on correctional classification and treatment decisions. One instrument that is now being used by many correctional agencies in the United States is the Level of Service Inventory-Revised (LSI-R). Criticisms of the LSI-R include failure to consider physical and sexual abuse as risk factors and failure to modify the instrument for the assessment of female offenders. Using a sample of 442 offenders (317 male and 125 female), this research investigates (a) the validity of the LSI-R for both females and males and (b) the relationship between childhood abuse, the LSI-R, and recidivism rates. Analyses indicated that the LSI-R is a valid (predictive) instrument for this sample of female offenders and that a history of prior abuse fails to add to the prediction of reincarceration, once risk is controlled for using the LSI-R. Implications for further research are discussed.