Inconsistency in the Measurement of Juvenile Reoffending and Its Impact on Assessments of Community Variation

• In Event: Substantive Implications of Variation in Official Measures of Juvenile Recidivism

Abstract

Understanding variation in juvenile reoffending is of great importance to practitioners, policymakers and researchers. However, research on this topic is hampered by inconsistencies in the ways reoffending is measured across states and across studies. Measures are alternately defined in terms of arrest, adjudication or conviction, or commitment or incarceration, and using follow-up periods ranging from six months to four years. This prevents researchers from knowing whether findings using a particular outcome are relevant for studies or jurisdictions that use a different outcome. In this paper, we are particularly interested in the capacity for community-level characteristics to affect juvenile reoffending, and whether conclusions about community variation in reoffending differ depending on the outcome being examined. Using data for 2012-2017 from the Florida Department of Juvenile Justice and from the American Community Survey, we estimate multilevel models with youth nested within census tracts nested within counties. For each reoffending measure, we estimate the proportion of variability that is accounted for by each of these higher-order units, and then determine whether those proportions differ significantly between the various reoffending measures. Variation across measures indicates that the choice of reoffending measure can affect conclusions about how much reoffending varies across geographic contexts.

Authors

• Brian Stults, Florida State University

• Dequan Jordan Cowell, Florida State University

• Sonja Siennick, Florida State University