Bruce, M. A. (2004). Contextual Complexity And Violent Delinquency Among Black And White Males. Journal of Black Studies, 35(1), 65-98.

This analysis introduces a comparative framework illustrating how context leads to racial differences in violent activity. Data from the National Longitudinal Study of Adolescent Health (N = 4,620) are then used to examine the extent to which the conditions contributing to criminal behavior among Black adolescents differ from White adolescents. The results indicate that communities do influence behavior, although the effect is sometimes conditioned by race. White adolescent behavior appears to be more sensitive to community characteristics than Black behavior. Families are also an important part of adolescent life. Family resources generally can encourage adolescents to engage in conventional behaviors, and depressed resources often lead to unconventional or dangerous behaviors. Results associated with parents’ education show that belonging to families headed by highly educated parents can mean greater access to and awareness of institutions where pro-social activities take place. Finally, both family size and household parental arrangements have implications for behavior, although their individual effects radically differ from one another. Whereas, methodologically, the model represents a fairly conservative depiction of the social environment, more refined measures need to be introduced to flesh out the manner in which some factors affect behavioral outcomes. Nevertheless, this research lays the foundations for future research in two important ways. First, the successful integration of macro- and micro-level factors in one empirical framework establishes the plausibility of multilevel modeling, thereby facilitating a tighter connection between conceptual frames and the empirical models emerging from them. Second, this research also provides a glimpse into the complex relationship between stratification and crime. As such, rather than debating which axis of stratification is more relevant for violence, it may be more fruitful to consider how race, gender, and class independently and interactively influence behavioral outcomes.

Full article can be found here: https://www.jstor.org/stable/4129291