Estimating the Effect of State Zero Tolerance Laws on Exclusionary Discipline, Racial Discipline Gaps, and Student Behavior
Curran FC · 2016
grade Cquasi-experimentindependentunreplicated
Sample
Two nationally representative data sources; state and year fixed-effects models
Population
US school districts and principals.
Design
State-and-year fixed-effects estimation of the effect of enacting state zero-tolerance statutes. Graded C because identification rests on the timing of state law adoption without a sharper discontinuity, and the behaviour outcome is principals' perceptions rather than an incident count.
Key findings
State zero-tolerance laws predict a 0.5 percentage-point INCREASE in district suspension rates and no consistent decrease in principals' perceptions of problem behaviour. The increases are larger for Black than for White students, so the laws plausibly widened the Black-White suspension gap. Zero tolerance bought more punishment and no more order.
Genetic confound
Low. The variation is legislative.
Replication notes
Directionally consistent with the APA Zero Tolerance Task Force's 2008 evidentiary review, which reached the same conclusion from a different evidence base, and with Hoffman's district natural experiment. No direct replication of this specification exists.
DOI / URL
10.3102/0162373716652728
Effects
| Outcome | Metric | Value | Measure | Timing | Vs | Horizon | Class |
|---|---|---|---|---|---|---|---|
| District suspension rate | percentage points | +0.5 pp after state zero-tolerance law | standardized | post-enactment | business-as-usual | over-2yr | behaviour |
| Principals' perceptions of problem behaviour | direction | no consistent decrease | researcher-designed | post-enactment | business-as-usual | over-2yr | behaviour |
| Black-White suspension gap | direction | larger increases for Black than White students | standardized | post-enactment | business-as-usual | over-2yr | behaviour |
Cited by
- How should a school run its classrooms and its discipline system?mixedconf: mediumgc: low