Association for Public Policy Analysis & Management

With the highest incarceration rate in the world, reducing crime and making the criminal justice system more equitable are key policy priorities. For decades there has been widespread concern about the “school-to-prison” pipeline in which students, especially impoverished and minority students, are suspended for relatively minor infractions of school rules. These enforced absences from school can then increase the likelihood of failing courses, failing to graduate from high school, becoming involved in criminal activity, and eventually becoming incarcerated in a juvenile justice or adult facility. Studies of exclusionary discipline (ED) in the education system – defined as suspensions and expulsions – have found that ED increases a student’s juvenile justice involvement and dropping out of high school (e.g., Mittleman 2018; Monahan et al 2014; Bernburg & Krohn 2003). Studies of the effects of ED on adult criminal convictions suggest a connection but face significant limitations. Our work uses a statewide population of public school students to examine the relationship between ED in high school and adult criminal convictions, which, unlike in most states, covers offenders who are 16 and older. The data set includes administrative data on individual students merged with public criminal records from middle school to young adulthood on five complete cohorts (N=~110,000 in each cohort) of entering ninth graders in North Carolina public schools. Our data set includes measures of ED experienced in middle school, test scores, family poverty, school mobility between 6-8^th grade, absences in middle school, and other standard demographic controls. The study has gathered and successfully merged all data needed for the study. We plan to present analytic results on the three aims of our paper: 1) what is the empirical relationship between ED and adult criminal involvement in a large and diverse state, 2) how much variation is there in the ED-incarceration relationship across student subgroups, schools, and districts, and 3) whether ED affects crime directly or through further educational attainment (graduation, college enrollment).