The independent and joint effects of race, crime, and social location on the dispositional decisions of juvenile girls

Abstract

The nexus of race, gender, and social location remain a habitual line of inquiry for many criminologists. However, quantitative studies of intersectionality are rare and especially rare as it relates to those studying serious and violent girls sentenced in the juvenile court. Particularly, this study seeks to explore how racialized gender expectations and forms of double/multiplicative jeopardy influence the back-end outcomes for serious and violent girls. Findings reveal the black girls, compared to their white counterparts, were sentenced with less leniency, and instead, punished in a more punitive matter. Importantly, these findings vary across the gamut of legal and extralegal variables. Implications for theory, practice, and future research are discussed, as well as the limitations of the present study.

Keyterms:

• Juveniles
• black feminism and intersectionality
• criminological theory
• girls in the justice system

Notes

1 For a full review of race, girls, historical trends, processing and treatment in the juvenile court, see: Ward (2012) and Chesney-Lind & Sheldon (2013).

2 Purposefully, this analysis excludes juveniles sentenced to DJJ supervision for non-violent offenses and/or status offenses, and includes only serious and/or violent offenses. Specifically, the central focus was in juvenile who fell under South Carolina Code of Law, which would potentially expose them to General Sessions (adult court).Specifically, "DJJ supervision" means that every juvenile in the sample has been formally adjudicated.

3 N = 7. For those appearing in multiple cases, only the most recent case was.

4 As suggested by Browning, Dietz, & Feinberg (2004), a one standard deviation increase in the rate of violent crime is not a non-trivial increase. Browning et al. (20004) note: a one standard deviation or greater increase in neighborhood violent crime above the median substantially reduces the ameliorative effects of collective efficacy on violence, as well as other forms of pro-social capital.

5 Vittinghoff & McCulloch (2007) argue, based on their own logistic regression simulations, that the 10 cases per predictor variable rule is a conservative figure. The authors concluded that 5-9 outcomes per variable is the liberal figure. These models meet and even go beyond the liberal and conservative figures.

6 Model I reflects this code of law. From a courtroom standpoint, this model should account for the highest amount of the variance in sentencing, given these are the factors of relevancy to juvenile court sanctions in South Carolina. However, Model I is not the strongest of the estimated models. Thus, Model I was estimated to determine how well SC Code of Law, 1976 §20-7-760, explains the sentencing outcome.