Quick answer
The outcome data is lopsided: a first out-of-school suspension predicts a 6% lower chance of graduating, 24% lower chance of a bachelor's degree, and 51% higher likelihood of future arrest — while high-restorative schools in a 265,000-student California study showed less exclusion, less misbehavior, and higher GPAs. But 'restorative vs. suspension' is a false binary; the real decision is what fills the middle of your discipline pyramid, and whether every removal ends in structured re-entry.
- Exclusion's long-term costs are documented: −6% diploma, −24% bachelor's, +51% arrest likelihood after a first suspension.
- Black students are suspended at 3.9× the rate of White students — an exclusionary system compounds inequity.
- High-restorative-use schools show less misbehavior and higher GPAs, not chaos (PACE, 265k students).
- Banning suspension without resourcing alternatives produces hidden 'informal exclusion' (Stanford, 2025).
- Keep removal for genuine safety; replace it for the large middle; add re-entry for everyone.
What does suspension actually accomplish?
Suspension does reliably deliver two things: an immediate pause for a volatile situation, and a signal to staff and families that a line exists. Those are real, and the honest comparison starts by admitting it.
What suspension does not deliver, per the accumulated research, is changed behavior or safer schools. The American Psychological Association's Zero Tolerance Task Force found no evidence that exclusion-heavy policies improve safety. Meanwhile the longitudinal costs are severe: tracking 9,000+ students, researchers found a first out-of-school suspension predicted a 6% lower chance of a high-school diploma, a 24% lower chance of a bachelor's degree, and a 51% higher chance of later arrest. The consequence lands hardest and most often on Black students, who are suspended at 3.9 times the White student rate in federal civil-rights data.
What does restorative accountability deliver instead?
Restorative accountability keeps the student inside the community and runs the harder process: name the harm, face the person affected, make it right, and re-enter with support. Students consistently describe it as more demanding than removal — you can serve a suspension playing video games; you cannot outsource facing the classmate you hurt.
The outcome evidence is encouraging where implementation is real. The RAND randomized trial in Pittsburgh measured improved school climate and reduced suspension days. The PACE analysis of 265,000 California middle schoolers associated high restorative use with reductions in exclusionary discipline, self-reported misbehavior, gang involvement, and substance use — alongside higher schoolwide GPAs. For the practice-level detail, see our restorative practices primer.
What are the failure modes on each path?
Exclusion-first fails slowly and invisibly: the same students cycle out and fall further behind, disparities compound, and the office becomes a revolving door that never touches causes.
Restorative-first fails fast and visibly when it's under-resourced. The two documented traps:
- Informal exclusion. A 2025 Stanford study found schools that banned suspensions without building alternatives quietly shifted to hallway time-outs, unsupervised rooms, and undocumented early pickups — the metrics improved while marginalized students kept losing instructional time.
- Unsupported staff. Removing the only tool teachers had, without training and coaching in the replacement, produces chaos, burnout, and a staff revolt that kills the initiative inside a year.
How should a leadership team actually decide?
Reframe the question. You are not choosing between suspension and circles; you're deciding what fills the middle of the discipline pyramid:
- Top (rare): genuine safety threats — removal remains available, but every removal now ends with a structured, supported re-entry.
- Middle (most incidents): defiance, disruption, conflict — this is where restorative conversations and conferences replace removal, and where the outcome data favors repair overwhelmingly.
- Base (everything): proactive Tier 1 systems and community circles that shrink the pyramid from the bottom.
Sequence adult capacity first, policy second. If your team is weighing where to start, that's exactly the conversation our Relationship-Centered Discipline training opens with.
About this guide
Written by Jennifer Awesome, Founder & Educational Consultant at Awesome Consulting & Educational Services — a 501(c)(3) nonprofit with 25+ years in schools and education. Guidance reflects the trainings and coaching we deliver in schools every week. Last updated July 5, 2026.
Common questions
Should we ban suspensions outright? +
The research says no — not as a first move. Bans without resourced alternatives produce hidden informal exclusion (Stanford, 2025). Build adult capacity and restorative structures first; suspension rates then fall as a byproduct rather than a mandate.
Won't keeping disruptive students in class hurt everyone else's learning? +
The data suggests the opposite over time: high-restorative schools report less overall misbehavior and higher GPAs, because repair addresses causes and proactive circles prevent repeats. The transition year does demand real support for teachers — that's where most initiatives are won or lost.
How do we answer parents who want 'real consequences'? +
Show them what restorative accountability requires: facing the person harmed, owning the impact, making it right, and re-entering with commitments. Then share the outcome data on what suspension predicts. Most parents want consequences that change behavior — that's exactly what repair does better.