The victim-offender overlap in late adulthood.

This study contributes to the general knowledge of the victim-offender overlap by determining whether the phenomenon exists among older adults and whether known correlates of crime and victimization explain the relationship. Cross-sectional survey data from telephone interviews conducted with individuals 60 years and older (N = 2,000) residing in Arizona and Florida are used to estimate confirmatory factor models for both victimization and criminal offending. The results from a series of multivariate regression models show that victimization is associated with criminal offending. While factors such as low self-control, depression, and spending time in commercial drinking establishments partially attenuate the victimization-crime link, the statistically significant relationship persists in a multivariate context. Further testing indicates that the observed findings are robust across measurement and modeling strategies. Coupled with prior research, the results support the argument that the victim-offender overlap exists (and is difficult to explain) over the life course.