Peace Science Made Accessible, Understandable, and Useful.

How Do “Violent Flanks” Affect the Outcomes of Nonviolent Campaigns?

It is a common belief: however “nice” the use of nonviolence may be, in the real world violence is necessary—and ultimately more effective, the thinking goes—for challenging a brutal regime, fighting injustice, or defending against an armed opponent. But what are the actual effects of adding violence to a movement’s repertoire of resistance strategies? Previous scholarship has been inconclusive on this question of so-called “radical flank effects,” as studies tend to focus on individual cases and also reflect collective confusion over what is meant by “radical”—does it, for instance, refer to the means used or the ends sought? Focusing, therefore, on violent—as opposed to “radical”—flanks, the authors set out to bring clarity and systematic analysis to bear on the question of positive versus negative violent flank effects. Examining the full universe of nonviolent campaigns with radical (“maximalist”) goals (“removal of an incumbent national government, self-determination, secession, or the expulsion of foreign occupation”) from 1900 until 2006, they ask: how does the presence of armed resistance at the same time as nonviolent resistance affect the success of nonviolent campaigns?

The authors generate three hypotheses, as follows:

  1. Nonviolent campaigns with violent flanks are more likely to succeed than nonviolent campaigns without violent flanks.
  2. Nonviolent campaigns without violent flanks are more likely to succeed than nonviolent campaigns with violent flanks.
  3. Violent flanks have no impact on the success rates of nonviolent campaigns.

The authors use both quantitative and qualitative research methods to answer their question. To test their primary hypotheses, they search for any significant statistical relationships that might exist between the presence of violent flanks and the success/failure of nonviolent campaigns. They find none, thus providing no support for either Hypothesis 1 or 2. As the authors note, this could mean either that the presence of violent flanks has no discernible effect on outcomes or that it has mixed positive and negative effects that cancel each other out when taken together. When they compare the effects of violent flanks that emerge from inside a nonviolent movement to those of violent flanks that develop parallel to a nonviolent movement, they find that the former are associated with failure, suggesting that negative violent flank effects are more pronounced when a nonviolent campaign cannot distance itself from its armed counterpart. Moreover, they find that mass participation is the strongest determinant of nonviolent campaign success and that the presence of violent flanks has a negative effect on participation levels, suggesting that violent flanks may indirectly decrease the likelihood of success.

To flesh out how violent flanks operate within individual cases, the authors examine four cases where violent flanks were present: Burma 1988, Philippines 1983-1986, South Africa 1952-1961, and South Africa 1983-1994. Two campaigns were successful (Philippines and South Africa 1983-1994), and two were not (Burma and South Africa 1952-1961); two had violent flanks outside of the nonviolent movement (Burma and Philippines), and two had violent flanks associated with the nonviolent movement (the two South Africa cases). After examining the histories of these nonviolent campaigns—and the ways they interacted with armed resistance—the authors found mixed results. Violent flanks had negative effects in the two unsuccessful cases, no net impact in one of the successful cases (the Philippines), and a weak positive effect in the other (the later South African case). Overall there was greater evidence for negative violent flank effect mechanisms than for positive ones. In the one case where a violent flank had a weak positive effect (South Africa 1983-1994), that effect was mostly symbolic—energizing activists around the revolutionary mystique of violent resistance—rather than instrumental to gaining power over the apartheid regime. However, in the two cases where violent flanks had negative effects, these effects were seriously detrimental: the presence of an armed movement diminished “chances of success for otherwise nonviolent campaigns by legitimating repression, demobilizing participants, shifting to violent strategies where the state [wa]s superior, and discrediting regime opponents.” Notably, the armed movements were consistently shown not to protect nonviolent activists but rather to put them at greater risk, as authorities used the presence of armed actors to justify widespread repression against all resistance movements, violent and nonviolent alike.

The case studies show that violent flanks do actually influence the outcomes of nonviolent campaigns, despite the earlier quantitative findings suggesting otherwise. Negative and positive effects simply appear to cancel each other out when taken together over a large number of cases, with negative violent flank effects being somewhat more prominent than positive ones. The authors argue, therefore, that “on average, maximalist nonviolent campaigns often succeed despite violent flanks—rarely because of them.”

Contemporary Relevance:

Despite recent scholarship demonstrating the greater effectiveness of nonviolent resistance (see Chenoweth & Stephan 2011), assumptions about the effectiveness of violence—along with its supposedly radical and/or revolutionary nature—stubbornly persist. When faced with a brutal or blatantly unjust opponent, many people are inclined to believe that only violence will bring about needed change or be able to protect/defend one’s community or fellow activists. We have seen this recent thinking everywhere from Syria to Venezuela, but for those of us in the U.S. struggling against the Trump administration and the white supremacist forces it has unleashed, we need look no further than the presence of Antifa (anti-fascist groups who do not rule out engaging in violent confrontations) in our own protests to see this same logic at work—as well as its counterproductive effects. Such groups see themselves as a necessary counterpart to white supremacist or neo-Nazi groups who come armed to demonstrations, ready to engage in street battles with left-wing activists. Although this logic of needing to use violence to defend against violence is so widespread and deeply engrained as to be almost intuitive, the problem is that such moves feed into and reinforce narratives on the Right that inspire—and provide cover for—their own claims to self-defense. Just as the presence of a violent flank in an anti-regime nonviolent movement can provide necessary and/or further justification for government security forces to fire on protesters, so too can it create a similar dynamic among non-state groups, including neo-Nazis and white supremacists, mobilizing more recruits and ultimately increasing the vulnerability of anti-racist/anti-fascist activists and the marginalized/targeted communities whom they wish to defend.

Talking Points:

  • There is no significant statistical relationship between the presence of violent flanks and either nonviolent campaign success or failure, the result of violent flanks having both negative and positive effects that cancel each other out when taken together.
  • Violent flanks that emerge from within otherwise nonviolent campaigns appear to decrease these campaigns’ likelihood of success.
  • Mass participation is the strongest determinant of nonviolent campaign success, and violent flanks have a negative effect on participation levels, suggesting that violent flanks can indirectly contribute to campaign failure.
  • In case studies, armed movements were consistently shown not to protect nonviolent activists but rather to put them at greater risk, as authorities used the presence of armed actors to justify widespread repression against all resistance movements, violent and nonviolent alike.
  • Research shows that, “on average, maximalist nonviolent campaigns often succeed despite violent flanks—rarely because of them.”

Practical Implications:

This research highlights an area where collaboration between academic and activist communities is urgently needed: the dissemination and discussion of research that directly contradicts persistent assumptions about the necessity of violence to achieve movement goals.

In the wake of recent events in Charlottesville, outrage has rightly focused on the neo-Nazi and white supremacist groups who came armed and even killed one of the counter-protesters. Their goals of racial supremacy and purity, fueled by hate and fear, have no place in a country that values equality, pluralism, and human dignity, and their ascendency now is nothing short of terrifying. But, for the sake of effectively challenging these groups and their repulsive vision, we must also engage in critical reflection inward, especially with regards to the strategic implications of the presence of Antifa affiliates who also came to Charlottesville armed, among otherwise nonviolent counter-protesters. Although their work to expose and tirelessly organize against fascism is admirable and necessary, those who identify with Antifa and its full range of tactics appear to endorse at least two flawed assumptions: 1) that truly radical action to effectively challenge fascism must include violence—often termed “physical confrontation”—and that nonviolence equals “dialogue” or “normal politics” and therefore implies acquiescence, submission, or cooptation, and 2) that violence/“physical confrontation” is also necessary to protect activists and targeted communities. But, in fact, here is what we know from recent social scientific research: Nonviolent resistance is twice as effective as violent resistance when used for radical goals such as the removal of an authoritarian regime or national liberation, cases with no shortage of brutal, unreasonable opponents (see Chenoweth & Stephan 2011). Furthermore, nonviolent resistance strategy is all about analyzing and dismantling an opponent’s sources of power, including through direct action. Finally, as noted above, instead of protecting nonviolent activists, the presence of a violent flank creates justification for further repression against them, making them more vulnerable to violence.

Therefore, it is time that we un-tether violence from its “radical” and “protective/defensive” associations. Not doing so—hanging on, as Antifa does, to these tired old assertions that violence is a necessary response—is, quite simply, poor strategy. It gives white supremacists and neo-Nazis exactly what they want, reinforcing their “we’re embattled” narratives and thereby strengthening their movement. It muddies the waters by giving commentators on the Right something to point to when they try to create (ludicrous) moral equivalencies between white supremacists/neo-Nazis and anti-fascist activists. And, in doing so, it does nothing to diminish the strength of white supremacism.

Furthermore, the continued presence of armed elements like Antifa has negative effects within the resistance: it makes many fellow activists feel more vulnerable to violence and therefore less likely to show up to demonstrations, especially with children in tow, diminishing mass participation in the resistance and thereby decreasing its power and effectiveness.

For all these reasons, if Antifa activists care—as they no doubt do—about effectively challenging fascist, white supremacist forces, they must think more strategically, considering the short- and long-term effects of their actions.  And fellow activists must engage in discussions with them about the strategic value and radical credentials of completely nonviolent resistance, together strategizing about those actions that will best diminish the power of the opponent to realize its fascist agenda and raising the following points:

Despite common-sense associations of armed/violent action with defense and protection, nonviolent discipline has a better chance of keeping activists safe than armed resistance does, even—counter-intuitively—in the face of a violent adversary. There is no guarantee of complete safety with either type of resistance, but armed resistance is much more likely to elicit further—not less—violence from the other side.

There is a strategic logic to nonviolent resistance of which Antifa seems unaware. Far from being synonymous with “dialogue” or “debate,” nonviolent resistance involves the dismantling of an opponent’s sources of power through a range of methods, including various forms of disruption and direct action, and is twice as effective as violent resistance in achieving radical goals (see Chenoweth & Stephan 2011). Contrary to mainstream belief, there is a historical record of successful nonviolent resistance against fascism in countries under Nazi control (e.g., women saving their Jewish husbands in Berlin’s Rosenstrasse demonstrations, Denmark’s rescue of a sizable number of its Jewish community, etc.).

Only by maintaining nonviolent discipline can the resistance dramatize and capitalize on the clear contrast between its activists and the white supremacists or neo-Nazis they confront.

Finally, far from embodying a radical challenge to fascism, Antifa affiliates are doing exactly what neo-Nazis and white supremacists are hoping they will do—this is precisely the reaction that will energize the very fascists they are hoping to shut down. Only by disassociating one’s radical credentials from participation in violence will we ultimately move away from these knee-jerk responses to racist violence that do nothing to minimize the draw and strength of white supremacy—and instead move towards more strategic, effective action that actually has a chance of advancing the cause of a diverse, affirming, just society.

Violent flank effects: the effects of armed resistance that is carried out at the same time as an otherwise nonviolent resistance campaign (with this armed resistance having developed either from elements within the nonviolent movement or separate from it).

Continued Reading:


Chenoweth, E. & Schock, K. (2015). Do contemporaneous armed challenges affect the outcomes of mass nonviolent campaigns? Mobilization: An International Quarterly 2(4), 427-451.

Next article Rethinking Security, Violence, and Development from the Margins
Previous article Counterproductive Effects of Military Counterterrorism Strategies