Gandhi’s and King’s emphases on nonviolent resistance — in which unarmed people use a coordinated set of strikes, protests, boycotts or other actions to confront an opponent — are not without critics. Some critiques are based on a misunderstanding about what civil resistance is, while others doubt the ability of unarmed and suppressed people to organize and challenge a powerful opponent. With each new movement comes the same set of challenges, including questions about the efficacy of nonviolent action in the face of entrenched power and systemic oppression. In 2011, we published a book exploring these questions and found unexpectedly that campaigns of nonviolent resistance had succeeded more than twice as often as their violent counterparts when seeking to remove incumbent national leaders or gain territorial independence.
To many people, this conclusion may seem naive, but when we drilled into the data, we found that nonviolent resistance campaigns don’t succeed by melting the hearts of their opponents. Instead, they tend to succeed because nonviolent methods have a greater potential for eliciting mass participation — on average, they elicit about 11 times more participants than the average armed uprising — and because this is the source of major power shifts within the opponent regime. Mass participation that draws on diverse segments of society tends to empower and co-opt reformers while cutting off hard-liners from sources of support. When such participation is nonviolent, it increases the chances of pulling the regime’s support from the leadership, allowing security forces, economic elites and civilian bureaucrats to shift their loyalties with less fear of bloody retribution.
In other words, we found that nonviolent resistance is effective not necessarily because of its conversion potential but rather because of its creative, co-optive and coercive potential — a theory that Albert Einstein Institution founder Gene Sharp has articulated for decades. Naturally, not all nonviolent campaigns succeed. But in cases where they failed, there was no good systematic evidence to suggest that violent uprisings would have performed any better.
That was 2011. Now it’s 2016. What have we learned about nonviolent resistance in the past five years? Below we sketch some of the key empirical takeaways from political science, some of which have rather surprising implications for skeptics of nonviolent action.
Nonviolent campaigns have become increasingly common.
If you feel as though we live in a particularly disruptive time in history, you’re right. But it’s the kind of disruption that is unique to our time. The Major Episodes of Contention project (a data project run by Professor Erica Chenoweth at the University of Denver) suggests that nonviolent resistance campaigns have become the modal category of contentious action worldwide. The NAVCO Data Project, a separate data collection project using different source material and inclusion criteria, shows similar patterns, as do a variety of other protest data sets. Whereas the frequency of violent insurgencies — defined with a 1,000-battle death threshold — has declined since the 1970s, campaigns relying primarily on nonviolent resistance have skyrocketed. Note that these figures refer specifically to maximalist campaigns, meaning their goals are to remove the incumbent national leadership from power or to create territorial independence through secession or expulsion of a foreign military occupation or colonial power.
In the first five years of the current decade alone, we have seen more onsets of new nonviolent campaigns than during the entire 1990s, and almost as many as were observed during the 2000s. Our current decade is on track to be the most contentious decade on record.
Although they are more common, the absolute success rates of nonviolent resistance campaigns have declined.
With this precipitous rise of nonviolent campaigns, we also have seen a steep learning curve. The success rates of nonviolent resistance peaked in the 1990s, but the current decade has seen a sharp decline in the success rates of nonviolent resistance.
There may be a few reasons for this. First, state opponents may be learning and adapting to challenges from below. Although several decades ago, they may have underestimated the potential of people power to pose significant threats to their rule, they may now see mass nonviolent campaigns as truly threatening, devoting more resources to preventing them — perhaps following the implications of Bruce Bueno de Mesquita and Alastair Smith’s “Dictator’s Handbook” — or deploying “smart repression” to subvert them when they arise. This phenomenon of learned adaptation, or what Steven Heydemann, the Ketcham Chair in Middle East Studies at Smith College, calls “authoritarianism 2.0,” is a central focus of the “Future of Authoritarianism” project at the Atlantic Council.
Second, activists employing methods of nonviolent action may be learning the wrong lessons from their contemporaries around the globe. For instance, one may be tempted to think, based on the news coverage of the mass demonstrations and strikes in Tunisia in 2010 and 2011, that three weeks’ worth of demonstrations could unseat a dictator. Yet such understandings completely miss the fact that Tunisia had a unique recent history of robust organized labor activity, which lent its support to the uprising, and that general strikes threatened to cripple the Tunisian economy, such that economic and business elites began to withdraw support from President Zine el-Abidine Ben Ali as much as the security forces who defied his order to strafe the demonstrators with automatic weapons.
It is natural for activists to draw inspiration from others in similar situations, but this can often result in failure. For example, Kurt Weyland of the University of Texas points out that during the global wave of mostly violent revolutions in 1848, dissidents sought to replicate the strategy of the initial uprising against the French crown, only to be thwarted by better-prepared, better-resourced monarchs who were, of course, different kinds of opponents. Coming later in the wave, these sovereigns were able to anticipate the revolutionaries’ moves to crush the uprisings and divide the oppositions to their favor. We may be seeing a similar dynamic today, particularly in the later stages of regional waves of uprisings.
However, believe it or not, nonviolent campaigns are still succeeding more often than violence.
Violent campaigns have fared much worse, in terms of absolute rates of success, than nonviolent campaigns since 1960. In fact, in the aggregate, from 1900 to 2015, nonviolent campaigns succeeded 51 percent of the time, whereas violent campaigns succeeded 27 percent of the time. So far this decade, 30 percent of nonviolent campaigns have succeeded, whereas 12 percent of violent campaigns have succeeded — meaning that in fact, the proportional success gap between them is now actually wider than average.
Violent flanks are typically disadvantageous for nonviolent mass movements.
One of the hot topics since 2011 has been the question of whether employing a little bit of violence alongside a primarily unarmed campaign helps or hurts a nonviolent campaign. This question was often represented in the “diversity of tactics” debate here in the United States. But the question of nonviolent, violent or mixed methods of contention is common in many movements seeking radical change worldwide. Despite numerous claims, pro and con, by observers, pundits and activists alike, this question received surprisingly little serious empirical evaluation until fairly recently.
In a recent article in “Mobilization,” Chenoweth and Kurt Schock of Rutgers University use comparative data to study the limited use of violence. They found that violent flanks may achieve some short-term process goals such as media attention, the perception of self-defense, the diffusion of an oppositional culture that builds the commitment of more radical members, or catharsis around the ability to “blow off steam.” But violent flanks typically undermine longer-term strategic goals such as maintaining an increasingly large and diverse participation base, expanding support among third parties and eliciting loyalty shifts among security forces. They find evidence that violent flanks are typically associated with smaller participation rates and more homogenous participation, undermining the main advantage of using nonviolent resistance in the first place. Another study similarly finds that violent flanks tend to increase repression by the state, which tends to be associated with lower participation rates. Thus, on average, violent flanks definitely do not help nonviolent campaigns succeed.
Omar Wasow of Princeton University provides further evidence regarding the political effects of nonviolent vs. “violent” protests. Leveraging data on urban protests by black Americans during the 1960s, Wasow convincingly shows that a higher frequency of nonviolent protests led to higher support for “civil rights” as the primary issue of public concern in the United States, whereas a higher frequency of violent protests led to greater support for “law and order” as the primary issue. After 1965, as violent protests became more common, public opinion shifted away from support for civil rights and toward support for the police response, showing how the movement had ceased to expand its appeal among crucial pillars of support. Strikingly, public opinion mattered not only in the short-term, but also in the long-term: Wasow finds that support for “law and order” was highly correlated with votes for Republican leadership, suggesting that the effects of different protest types have had lasting political effects in the United States.
Nonviolent conflicts are exceedingly difficult to predict.
The entire field of sociology has long been concerned with the question of when social movements or protest movements occur. Maximalist nonviolent resistance campaigns are a slightly different animal, since they presuppose a highly disruptive, contentious series of coordinated actions concentrated against a state opponent with the aim of fundamentally changing the status quo at the national level. Studies evaluating the causes of nonviolent resistance have identified numerous correlates, such as the density of the manufacturing sector (Butcher & Svensson 2014), emotions (Pearlman 2013), geographic proximity (Gleditsch & Rivera 2015), and protest history (Braithwaite, Braithwaite, & Kubik 2015).
In 2015, Chenoweth and Jay Ulfelder evaluated numerous general theories of mass uprisings to find that few of them accurately predict where nonviolent campaigns will occur. Unlike armed campaigns, coups or state collapse — all of which scholars are fairly good at predicting — nonviolent mass campaigns can happen almost anywhere for any reason. They often happen in places where scholars would expect it to be very difficult to mobilize dissent, much less to mobilize dissent effectively. And it’s not at all clear what might trigger them or make them stick. Chenoweth and Ulfelder conclude that people power movements are simply so contextual and contingent that typical forecasting tools and data structures can’t quite pin down their causes. Another way to interpret this finding is that people who organize nonviolent uprisings often overcome adverse conditions in creative ways that defy expectations, which brings us to our final point.
Repression challenges all dissident campaigns but does not necessarily predetermine the choice of nonviolent resistance or its outcome
One popular argument about nonviolent resistance is that it can happen and maybe succeed as long as the opponent plays nice. But as soon as the opponent takes off the gloves, nonviolent resistance is impossible or futile. We dealt with this argument a bit in our 2011 book, but some more recent work also speaks to this important question.
In terms of whether brutal repression influences the possibility of nonviolent resistance, Wendy Pearlman argues in her excellent book on the Palestinian national movement that repression alone cannot explain the reasons why the movement has turned from nonviolent action to violence. She argues that, in fact, repression was just as intense during the nonviolent phase of the First Intifada as it was during several of the movement’s violent phases. Instead, she argues, the level of cohesion can best explain the turn to violence. When the movement possessed collective vision, leadership and a clear set of internal norms and rules, the movement was able to rely on nonviolent resistance despite continued repression by the Israeli government.
Researchers Jonathan Sutton, Charles Butcher and Isak Svensson likewise point to movement structure and organization as a critical determinant of campaign viability in the face of repression. They use quantitative data to argue that when the state uses one-sided violence or mass killings against unarmed demonstrations, the demonstrators can go on to succeed in the long run only when they are part of a larger, coordinated campaign.
Of course, some research casts doubt on the ability of nonviolent opposition to contend with highly sophisticated repressive regimes — particularly those with genocidal or politicidal ambitions. Christopher Sullivan’s recent work on the Guatemalan security forces’ systematic dismantling of the leftist opposition between 1975 and 1985 is a cautionary tale regarding the sophistication and commitment that some regimes pose. As was the brutal, calculated killing of nonviolent protesters by the Bashar al-Assad regime in Syria after the protests in Deraa in March 2011 — a chilling reminder of why nonviolent mass campaigns fail almost as often as they succeed.
But again, it is difficult to predict when such repressive bureaucracies will be able to compel the full loyalty of their subordinates in the face of a mass uprising — even in a seemingly impossible case such as in Syria. Moreover, In forthcoming work, Lee Smithey, Lester Kurtz, and collaborators find that regime repression against unarmed demonstrators can often backfire, by creating moral outrage, drawing in more participation, creating third-party support for the movement and accelerating security force defections. In fact, repressive episodes can often be the cause of a nonviolent campaign rather than its conclusion. The murder of Emmett Till comes to mind as an example of a horrific episode of violence that ultimately generated an upswell of support, sympathy and participation for the U.S. civil rights movement.
In light of Martin Luther King Jr. Day, we thought we’d leave our readers with this insightful passage from his “Letter from a Birmingham Jail,” the full text of which you can find here:
“My friends, I must say to you that we have not made a single gain in civil rights without determined legal and nonviolent pressure. Lamentably, it is an historical fact that privileged groups seldom give up their privileges voluntarily. Individuals may see the moral light and voluntarily give up their unjust posture; but, as Reinhold Niebuhr has reminded us, groups tend to be more immoral than individuals. We know through painful experience that freedom is never voluntarily given by the oppressor; it must be demanded by the oppressed.”
Of course, King was concerned with both the moral and pragmatic dimensions of nonviolent resistance. But his pragmatism should not be underestimated, as Jonathan Rieder’s book on the Birmingham letter strikes home.
Clearly, there is much more to learn about nonviolent resistance: It is an emerging phenomenon, and research on the topic is likewise emerging within the social sciences. People seeking to confront oppression would benefit from more systematic research as to when and how to wage nonviolent struggle in various contexts. Policymakers grappling with challenges ranging from authoritarian resurgence to state fragility to violent extremism would benefit from a deeper understanding of when and why nonviolent movements succeed – and what it means to effectively support them.
In this decade — in which more people are using nonviolent resistance than ever before — scholars and practitioners alike would do well to consult the pragmatic and principled wisdom of Gandhi and King in building a way forward.
Erica Chenoweth is a professor at the Josef Korbel School of International Studies at the University of Denver. She co-hosts the blog Political Violence @ a Glance and is an occasional blogger at The Monkey Cage. Maria J. Stephan is a senior fellow at the U.S. Institute of Peace and a nonresident senior fellow at the Atlantic Council.