Homicide Victimization Among Serious Youthful Offenders: The Victim-Offender Overlap in Three Samples of California Youth Authority Wards

Michael E. Ezell, Duke Universty

Although much research in criminology focuses on either offenders or victims, attention is turning toward the considerable "overlap" existing between the populations of criminal offenders and victims. Using mortality data obtained from the California Vital Statistics, this paper examines the relationship between criminal offending and the risk of homicide victimization among three samples of offenders released from the California Youth Authority in 1981-82, 1986-87, and 1991-92. As of December 31, 1999, 331 of the total 5,101 cases were found to be deceased, with 180 of these deaths classified as homicides (54.4% of the deaths). Significant differences in the risk of homicide victimization were found among the three samples, with the 1991-92 sample accumulating nearly as many homicides as the 1981-82 sample. After separating the offenders into latent classes (on the basis of their prior offending trajectories) using the mixed Poisson methods of Nagin and Land, the Cox proportional hazards survival models were applied to test whether high-rate offenders were at a higher risk of homicide victimization. Results indicate that while the high-rate offenders had marginally higher risks, variables representing the frequency of violent offending/drug sales only, gang membership, ethnicity, and region of residence were much better predictors of risk.

(Return to Program Resources)

Updated 05/20/2006