Sometimes people who get frustrated by deep injustices in democratic societies seem to give up on their governments doing the right thing and start to consider the possibility of violent revolution as the only way to get any meaningful change. Andrew Marantz writes about an empirical study by Erica Chenoweth, the Berthold Beitz Professor in Human Rights and International Affairs at the Harvard Kennedy School, to try and address empirically the question of which kind of effort, mass civil rights struggle or revolution, is more likely to produce the results sought.
During the next five years, Chenoweth and Stephan built a database called Nonviolent and Violent Campaigns and Outcomes, or navco. It aimed to account for every attempted revolution worldwide, between 1900 and 2006: the Carnation Revolution, in Portugal; the Blancos rebellion, in Uruguay; the Active Voices campaign, in Madagascar; and three hundred and twenty others. “I took for granted, as did all the political scientists I was familiar with, that the serious thing, the thing you do if you’re a rebel group that really wants results, is you take up arms,” Chenoweth told me. “Then I ran the numbers.” Much of Chenoweth’s career since then has consisted of interpreting and explaining what those numbers showed.
In 2011, Chenoweth and Stephan published their findings in a book called “Why Civil Resistance Works: The Strategic Logic of Nonviolent Conflict.” It included detailed narrative case studies in which the authors hypothesized about why, say, the Philippine People Power movement of 1986 achieved its goals whereas the Burmese uprising of 1988 did not… In the database, Chenoweth and Stephan condensed each campaign’s months or years of struggle into a binary line of code: violent or nonviolent, success or failure.
Chenoweth and Stephan selected only “maximalist” resistance campaigns—big movements, with a thousand or more participants, that sought to fundamentally alter a nation’s political order, either by seceding or by overthrowing a foreign occupier or a head of state. The American civil-rights movement of the nineteen-sixties was not included in the navco data; although there were secessionists and insurgents within the movement, its main demands were reformist, not revolutionary. Moreover, campaigns were counted as successful only if their goals were achieved within a year of peak activity, without an unrelated intervention. The Greek resistance to the Nazis was coded as a failure, because although the movement contributed to the Nazis’ retreat from Greece, Allied troops seemed to contribute more. The Indian independence movement, the popular archetype of nonviolent insurrection, was classified as a partial success—for one thing, the British did eventually quit India, but not within a year.
Many of Chenoweth’s articles are quantitative and technical, but the upshot is simple enough: civil-resistance movements prevail far more often than armed movements do (about 1.95 times more often, according to the most recent version of the data). This seems to hold true across decades and continents, in democracies and autocracies, against weak regimes and strong ones.
As with all such statistical results, it can never be totally prescriptive. Even if on average civil resistance movements are more successful in achieving maximal goals, that does not imply that that is the best course of action for any given situation in any given country. But it does mean that one should not give up on the civil rights option too quickly.