Violent or nonviolent: which protests succeed?
Do violent protests achieve more than nonviolent ones?
A comparison of 2,255 major movements seeking policy reforms shows:
🔵 Mainly nonviolent: 50.4% achieved full or partial success.
🔴 Mainly violent: 47.6%.
But the sample sizes tell an essential part of the story: 2,234 nonviolent movements versus just 21 violent ones.
What changed over time?
Between 1955–2005 and 2006–2018:
- Nonviolent movements: 51.7% → 46.8%.
- Violent movements: 40.0% → 66.7%.
That apparent rise for violent movements comes from just 4 successes among 6 recent cases. It cannot establish that violence became more effective.
The comparison covers demands about wages, public services, rights and other reforms. “Success” consistently includes both full achievement and partial gains.
These are sustained movements involving at least 1,000 participants—not every demonstration. Classification follows the predominant method, so mainly nonviolent movements can include minor violent elements. Differences in context, reporting and unresolved outcomes also limit comparisons.
The figures describe historical outcomes; they do not prove that choosing a tactic causes success.
Source: Chenoweth & Kang, Major Episodes of Contention (2026), covering 1955–2018. The Data Atlas calculations using the authors’ replication data.
What should count as protest success: full achievement, partial concessions, or lasting influence on policy?