Abstract
Recently, Paternoster et al. used data from the Cambridge Study in Delinquent Development, a longitudinal study of 411 South London boys mostly born in 1953, to investigate the linkage between adolescent and adult offending and found that variations in adult offending were consistent with a random process after conditioning on adolescent offending. In this paper, we test the robustness of this early study across data sources and genders. Here, we use data from the Dunedin New Zealand 1972 birth cohort study to replicate previous findings regarding stability and change in criminal offending between the adolescent and adult years. In particular, our interest centers on the stochastic properties of the adolescent and adult conviction distribution in the cohort and whether the structure of this distribution is similar for males and females. This replication and extension of prior work is especially important since criminologists have little understanding of the pattern of female adolescent offending or how the patterns are linked to adult offending for women. The analysis reveals that variation in adult offending after conditioning on adolescent offending is consistent with a random (Poisson) process. Furthermore, this pattern is evident for both the males and the females in the Dunedin New Zealand 1972 birth cohort.