A recent study by Evan Rose, Jonathan Schellenberg, and Yotam Shem-Tov estimates the effect of teacher quality on criminal justice contact. It uses education data and arrest records that cover all students in North Carolina public schools in grades three to twelve from 1996 to 2013 and include information on student test scores, teacher and classroom assignments, demographic characteristics, and disciplinary and attendance records. The criminal justice data include arrests, convictions, and sentences. Combined, the data include almost two million students and 40,000 teachers, allowing the researchers to track individual students, their teachers, and many different student outcomes.
The researchers found that elementary and middle school teachers have large effects on students’ rates of future arrests, convictions, and incarceration. Teachers who are good at improving behavior and attendance and reducing the odds that students will repeat a grade are particularly likely to reduce later criminal justice contact.
Importantly, however, teachers who increase student achievement do not necessarily improve discipline, attendance, or grade repetition—or reduce students’ odds of criminal justice contact. Shifting a student to an elementary or middle school teacher with a 1 standard deviation higher effect on test scores has no significant effect on the likelihood of arrest between the ages of sixteen and twenty-one.
What are those teachers who minimize criminal justice contact doing so well? Recent academic literature links non-cognitive and social-emotional skills to good attendance and discipline in school, as well as long-term outcomes. Teachers who reduce criminal justice contact the most may be those best at teaching those skills.
To gauge how important this is, various authors have looked at hypothetical scenarios examining the effects of removing the worst performing teachers from the classroom. For example, in a seminal (and much-debated) article, Eric Hanushek assessed the academic effects of firing the bottom 5 percent of teachers based on the short-term academic improvements of their students and found that this would sharply increase educational outcomes.
Similarly, the present study simulated the impacts of replacing the bottom 5 percent of teachers based on their effects on long-run outcomes such as criminal justice contact and college attendance. They found that this would likely result in large improvements, including a 6 percent reduction in criminal arrests.
Certain teachers are good at boosting academic outcomes, while some are better at boosting other outcomes. In classrooms where students are at risk of skipping school or being suspended, teachers may choose to focus on developing non-cognitive and social-emotional skills at the expense of academics. If discipline and attendance are proxies for non-cognitive and social-emotional skills, then much of the effects of teachers talented at teaching those skills are not being captured in standardized testing data.
The upshot of all of this is that non-cognitive skills are crucial for long-term success, especially in reducing likelihood of criminal justice contact. The data show that individual elementary and middle school teachers can have large impacts on criminal justice outcomes and college attendance years down the line.
In theory, teacher evaluation could be based partially on non-cognitive/criminal justice value-added data. But such data would only be captured years later, and by that time would likely not be useful for HR decisions.
In practice, principals will not have this kind of information at their fingertips. Principals should see this study as further evidence that test scores do not capture all of the effects that teachers have, and therefore should be cautious in using test scores as an overriding factor in hiring and firing decisions
SOURCE: Evan K. Rose, Jonathan T. Schellenberg, and Yotam Shem-Tov, “The Effects of Teacher Quality on Adult Criminal Justice Contact,” National Bureau of Economic Research, Working Paper 30274, July 2022.
