Youth Violent Offending in School and Out: Reporting, Arrest, and the School-to-Prison Pipeline

Abstract

This study used violent victimization data from the National Crime Victimization Survey (1994–2018) to examine whether criminal justice interventions (i.e. reporting to the police and arrest) for youth-perpetrated violence were more likely to occur in school than outside school. On average, violence at school was 8.4 percentage points less likely than violence outside school to be reported to the police, but if there was a police report, violence in school was 8.0 percentage points more likely to involve an arrest. These statistical differences remained stable throughout the study period. Further analyses of the pooled sample by the offender’s gender and race found that school violence was associated with an increased likelihood of arrest only for Black youth, not White youth, and only for boys, not girls. Implications of these results for the school-to-prison pipeline argument are discussed.

Keywords:

• School-to-prison
• punishment
• youth
• violence
• arrest
• pipeline

Acknowledgment

The analyses, results, and conclusions presented here represent those of the authors and do not necessarily reflect those of the OJJDP or BJS.

Disclosure statement

No potential conflict of interest was reported by the authors.

Notes

1 Von Hippel (2007) argues that, after imputation, including cases that were originally missing on the dependent variable may produce less efficient estimates and may bias estimates. In sensitivity analyses, we deleted cases that were originally missing on both of our dependent variables (e.g. reporting to the police and arrest). We found that our results did not significantly change (Appendix B).

2 In sensitivity analyses, we used a proxy measure of an incident occurring during school by including the location of the incident (school vs. outside school), information about the time of day the victimization occurred (i.e. 6 am to 6 pm vs. other times), the month the victimization occurred (i.e. August to May vs. June and July), and the activity the victim was engaged in before being attacked (i.e. going to work/school, working, or attending school vs. other activities). This alternative measure did not significantly change our results (Appendix C).

3 The NCVS did not start measuring the Hispanic origin of the offender(s) until 2012. Given our focus on the prevalence and trends of criminal justice involvement in youth-perpetrated violence, we chose to omit this measure from our analysis instead of omitting 18 years of data.

4 NCVS records the victim’s household income using only categories (e.g. $40,000–$49,999). We used the mid-points of these categories to code a continuous income variable. With these continuous values, we then used the consumer price index from 2018 to standardize the victim’s household income to its 2018 value.

5 We conducted trend analyses separately for serious violence (i.e. rape, sexual assault, robbery, and aggravated assault) because simple assaults accounted for 73% of violent victimizations and differed in important ways from serious violence (e.g. presence of serious injury or weapon). Arrest trends for serious violence (at school and outside school) were not different from the total sample. For police reporting trends, we found that the probability of reporting violence outside school increased over time, whereas the probability of reporting violence in school followed a curvilinear trend similar to that presented in Figure 1.

6 Average marginal effects represent the average difference between the probabilities of reporting to the police and arrest between youth-perpetrated violence at school and outside school, given the coefficients of the Heckman selection model.

7 Two reviewers suggested that the increase in reporting might relate to increases both in the hiring of school resource officers and in specific security practices (e.g. security cameras, metal detectors, locker checks). Using NCVS SCS data, we tested this suggestion. Although these trends paralleled the increase in reporting to the police from 1994 to 2007, they are inconsistent with the decline in reporting since then (Appendix G).

8 Studies suggest that “other authorities” typically refer to school officials, such as teachers and administrators (Finkelhor & Ormrod, 2001)

Additional information

Funding

This bulletin was prepared under grant number 2017–JF–FX–K029 from the Office of Juvenile Justice and Delinquency Prevention (OJJDP), U.S. Department of Justice, through the Bureau of Justice Statistics (BJS) Visiting Fellows program.

Notes on contributors

Keith L. Hullenaar

Keith Hullenaar is a T32 Fellow Trainee in the Harborview Injury Prevention and Research Center at University of Washington. His research interests include violence and health, victim help-seeking, and youth victimization and offending.

Allison Kurpiel

Allison Kurpiel is a Graduate student in the Department of Sociology and Criminology at Penn State University. Her research interests include youth victimization and offending, fear of crime, and school crime.

R. Barry Ruback

R. Barry Ruback is a Professor of Criminology and Sociology at Penn State University. He is interested in criminal victimization and the reporting of crime.