This brief was developed using ChatGPT and edited by Charlotte Sutcliffe, Duke undergraduate research assistant; for full text and references see
Shi, Y., & Zhu, M. (2022). Equal Time for equal crime? racial bias in school discipline. Retrieved from https://doi.org/10.1016/j.econedurev.2022.102256
Background:
This study investigates the presence of racial bias in school disciplinary actions using administrative K-12 data from North Carolina spanning 2008–2018. Disparities in exclusionary discipline (e.g., suspensions) are well-documented, with Black students disproportionately affected. While such disparities could stem from unobservable behavioral differences, this paper explores whether students of different races receive different disciplinary outcomes for the same behavior - indicative of racial bias. The study aims to isolate the role of bias by controlling for variables such as prior disciplinary history, socioeconomic status, and incident-level characteristics.
Findings:
Two strategies were employed. The first used within-incident variation to compare disciplinary outcomes among students of different races involved in the same incident. The second utilized student fixed effects to compare how a given student’s outcomes varied depending on the race of the peer they were disciplined alongside. Both approaches found that Black students are more likely to be suspended and for longer durations than their White or Hispanic peers, even when involved in the same disciplinary event and accounting for personal history and attributes.
Specifically, Black students are 0.4 percentage points more likely than White peers to be suspended and receive suspensions that are on average 0.05 days longer. These differences persist after adjusting for prior infractions and socioeconomic factors. Hispanic-White disparities were not observed, while Black-Hispanic gaps were seen with longer suspension times but not in suspension probability.
Importantly, these disparities were concentrated in “subjective” infractions (e.g., insubordination), where administrators have more discretion, as opposed to “objective” infractions (e.g., skipping school), suggesting that bias is more likely to manifest in discretionary contexts. The fixed effects model further confirmed that Black students are more harshly punished when incidents involve peers of a different race, particularly White students.
Takeaways:
The evidence indicates that racial disparities in school suspensions - especially for subjective offenses - are not solely attributable to behavioral differences or socioeconomic status. Instead, the consistent patterns across robust methodological designs strongly suggest the presence of racial bias in school disciplinary practices. These findings highlight the need for policy reforms that address discretion in disciplinary decision-making and promote equitable treatment across racial and ethnic lines in education systems.