Violence Doesn’t Work (Most of the Time)

By Steven Pinker

People have long assumed that violence is necessary for political change. Rulers never cede power voluntarily, the argument goes, so progressives have no choice but to contemplate the use of force to bring about a better world, mindful of the trade-off between a small amount of violence now and acceptance of an unjust status quo indefinitely. Terrorists invoke this trade-off to justify what would otherwise be wanton murder. Even their most vociferous condemners concede that terrorism, though highly immoral, is often efficacious.

Of course, Mohandas Gandhi, and later Martin Luther King Jr., argued the opposite—that violence, in addition to being morally heinous, is tactically counterproductive. Violent movements attract thugs and firebrands who enjoy the mayhem. Violent tactics provide a pretext for retaliation by the enemy and alienate third parties who might otherwise support the movement.

So how effective is violence? Political scientists have recently tried tallying the successes and failures of violent and nonviolent movements. The evidence is piling up that Gandhi was right—at least on average. In separate analyses, Audrey Cronin and Max Abrahms have shown that terrorist movements almost always fizzle out without achieving any of their strategic aims. Just think of the failed independence movements in Puerto Rico, Ulster, Quebec, Basque Country, Kurdistan, and Tamil Eelam. The success rate of terrorist movements is, at best, in the single digits.

In their recent book, Why Civil Resistance Works, Erica Chenoweth and Maria Stephan found that about three-quarters of nonviolent movements get some or all of what they want, compared with only about a third of the violent ones. The Arab Spring bears this out: consider the more or less nonviolent movements that ousted the leaders of Egypt, Tunisia, and Yemen (together with the violent one needed in Libya). Even more encouraging, the success rate of nonviolent protest movements has steadily climbed since the 1940s, while that of violent movements has fallen since the 1980s.

Next idea: Fix Law Schools

This article available online at:

http://www.theatlantic.com/magazine/archive/2012/07/violence-doesn-t-work-most-of-the-time/309031/