A new article on NPR features the work of Professor Erica Chenoweth.
"Chenoweth trawled through 323 campaigns for regime change or self-determination worldwide from 1900 to 2006.
Then Stephan and Chenoweth teamed up to write a paper and a book based on the data. They found that major nonviolent campaigns are successful 53% of the time, while violent campaigns are successful only 26% of the time.
But does nonviolence cause the higher success rate? Or are you more likely to choose sit-ins over shootouts when you know you're already likely to succeed? When Stephan and Chenoweth factored in additional data on why the campaigns turned violent, peaceful resistance still prevailed. The likelihood of success was not a factor in whether a campaign became violent.
"I was surprised," Chenoweth says. "I expected for there to be, basically, at best no significant difference between armed and unarmed action."