Qiang Xu, Bowling Green State University


Criminological research has identified both stability and change in antisocial behavior across time. Two theoretical perspectives, the population heterogeneity and state dependence models, compete to explain why a small group of offenders persist in antisocial activities throughout the life course, while the vast majority of their contemporaries desist from such behavior. Although prior research has provided support for both perspectives, most of the previous studies are based on samples of “average” or “typical” offenders and thus fail to study serious offenders. In addition, few have examined the concurrent relationship between delinquency and various social domains, which is crucial for our understanding of the stability and change of criminal behavior over time. Meanwhile, we know very little about gender and race differences in the reciprocal relationship between adulthood delinquency and such critical social domains as social capital, self-identity, social networks, and opportunities for prosocial transitions. Based on a longitudinal sample of previously institutionalized youth (T1, n=254; T2, n=210), this project expands the testing of the state dependence perspective on persistent criminal behavior by using LISREL and structural equation models. Utilizing these techniques, this study examines the reciprocal effects between adult criminal behavior and critical social domains; how prior delinquency influences protective and risk factors for adult delinquency; and whether there are gender and race differences in the pathways from juvenile delinquency to adult criminal involvement. Results suggest that besides the effects of prior delinquency, various critical social domains affect adult criminal involvement of this sample of serious offenders. Differential association theory and life opportunity theory are shown to have more explanatory power in sketching the pathways from juvenile delinquency to adult criminal involvement. These findings indicate that the state dependence model can be applied in explaining the continuity of criminal behavior among serious offenders as well as more mundane general population offenders. Gender and race differences in the pathways to adult criminal involvement and the contemporaneous effects of adult crime on critical social domains are also highlighted in this research.