The Best and Worst of Dispositions: An Analysis of 30 Years of Juvenile Justice Data in La Porte County, Indiana

Background Both observed studies and experimental studies have identified programs, practices, and treatments that are considered effective, ineffective, or harmful for juvenile offenders. Objective Identify which dispositions were best and worse for each offense in the La Porte County, Indiana, juvenile justice database of over 8500 juvenile offenders from 1983 to 2012. Methods Recidivism time, offense count, and incident count data were each fit separately to a generalized linear model with an independent offense-disposition variable for 27 individual offenses with the covariates of age, gender, and race. The generalized linear model belonged to a quasi-likelihood family with an identity link where variance was assumed to equal mean-squared (mean 2 ). This appears to be the first time that offense and disposition have been paired into an independent variable to assess the effectiveness of juvenile delinquency programs and services. Results These analyses revealed that in 17 of the 27 most frequently occurring offenses, the “catch and release” disposition occurred more often than all other dispositions as the best way to deal with the offense. However, there was no one clearly worst disposition identified for the 27 most frequently occurring offenses. Conclusion The novel offense-disposition pairing technique used in this study supports the notion that the majority of juvenile offenders require relatively limited intervention to change their delinquent trajectory.