Understanding the Persistence of the Victimization-Offending Relationship: Modeling Causal Mechanisms Across Place and Time

Violent victimization is one of the most salient predictors of violent offending, and offending reliably predicts risk for victimization. Existing research shows that the relationship between victimization and offending is robust to a wide variety of variables. Some scholars speculate that violent victimization and offending are so intimately connected that it is not possible to fully understand them apart from one another. Despite much effort to identify the etiology of this phenomenon, it remains rather elusive. There is reason to believe that neighborhood disadvantage represents an important piece of the puzzle, and perhaps plays two meaningful roles in the genesis of the victimization-offending relationship. First, one view holds that the relationship is spurious and an artifact of disadvantage. Second, an alternative perspective maintains that the magnitude of the relationship is conditioned by levels of disadvantage. The main purpose of this dissertation is to examine each of these questions. A supplementary component first of the study also focus on the relationship between income-generating offending and victimization. Specifically, the supplementary section first examines whether incomegenerating offending explains the link between violent victimization and offending and second, whether the link between income-generating offending and violent victimization is conditioned by neighborhood context. To address these questions, pooled data from ten waves of the Pittsburgh Youth Study (PYS) are analyzed. Controls for time-stable unobserved heterogeneity are implemented using a hybrid random effects model. The findings from the study provide for main conclusions. First, neighborhood disadvantage does not explain the relationship between victimization and offending. Second, the relationship is conditioned by levels of disadvantage. Third, regarding the supplementary sections, the findings show that the relationship between violent offending and victimization is robust to measures of income-generating offending. Fourth, the relationship between income-generating offending and violent victimization is conditioned by neighborhood disadvantage. The implications of these findings for future theoretical and empirical research on interpersonal violence are discussed.