Why Security Studies Should Take Nonviolent Conflict Seriously

The following guest post comes from Professor Erica Chenoweth of Wesleyan University, where she directs Wesleyan’s Program on Terrorism and Insurgency Research, which she established in 2008.

*****

Recent events in the Middle East have revived interest in the theory and practice on nonviolent resistance. Pundits have rediscovered theoretical works by Gene Sharp and other theorists of nonviolent resistance, and have begun to pay more attention to ways that knowledge-sharing among practitioners of civil resistance has influenced these events. At the same time, political scientists have been criticized for failing to predict the onset of these revolutions. Although such criticisms are not entirely fair, even those who did anticipate that change was coming in the Middle East did not anticipate that these dictators would fall to overwhelmingly nonviolent uprisings.

I would argue that part of this failure to grasp the power of nonviolent mass resistance emerges from a tunnel vision within political science—and within security studies in particular—that privileges the study of violence and often neglects civil resistance as a viable form of political contestation. Even those who study revolutions do not typically distinguish between nonviolent, violent, and “mixed” methods of resistance, despite the fact that the type of resistance method these movements select may have discernable effects on their outcomes.

It is time for security studies to take nonviolent conflict seriously, and to incorporate such episodes and their dynamics into the canonical literature.

The following guest post comes from Professor Erica Chenoweth of Wesleyan University, where she directs Wesleyan’s Program on Terrorism and Insurgency Research, which she established in 2008.

*****

Recent events in the Middle East have revived interest in the theory and practice on nonviolent resistance. Pundits have rediscovered theoretical works by Gene Sharp and other theorists of nonviolent resistance, and have begun to “pay more attention to ways that knowledge-sharing among practitioners of civil resistance has influenced these events”:http://www.foreignpolicy.com/articles/2011/02/16/revolution_u.  At the same time, political scientists have been criticized for failing to predict the onset of these revolutions. Although such criticisms are not entirely fair, even those who did anticipate that change was coming in the Middle East did not anticipate that these dictators would fall to overwhelmingly nonviolent uprisings.

I would argue that part of this failure to grasp the power of nonviolent mass resistance emerges from a tunnel vision within political science—and within security studies in particular—that privileges the study of violence and often neglects civil resistance as a viable form of political contestation. Even those who study revolutions do not typically distinguish between nonviolent, violent, and “mixed” methods of resistance, despite the fact that the type of resistance method these movements select may have discernable effects on their outcomes.

It is time for security studies to take nonviolent conflict seriously, and to incorporate such episodes and their dynamics into the canonical literature.

In fact, many of the concepts and strategic dynamics that dominate in the security studies literature are perfectly compatible with those discussed in the literature on nonviolent conflict. For instance, a key for any actor in an asymmetric conflict is to attack the opponent at its weakest point and, if possible, to create divisions within the opponent. Ivan Arreguin-Toft argues weaker powers can do this by adopting indirect (or “guerrilla”) strategies against stronger opponents that use direct (or “conventional”) strategies. Gil Merom and Robert Pape have made the argument that democracies are particularly susceptible to challenges by violent non-state actors, because they create unsustainable divisions within democracies. But few have directly compared how nonviolent challenges would line up compared with violent ones, because most of the security studies literature assumes that the most forceful, effective means of waging political struggle entails the threat or use of violence.

The same concept of attacking the opponent at its weakest point applies to nonviolent conflicts, except that I would argue that nonviolent mass movements are actually superior at undermining regime opponents through asymmetric approaches. This is not because of the “moral high ground,” but rather because their reliance on nonviolent resistance confounds their opponents, whose usual response to internal challenge is to use force. As Qaddafi’s response to the Libyan uprising shows, many dictators are willing to use force against nonviolent protestors; however, this is seldom costless for these dictators. They usually pay a major price in the form of loyalty shifts among security forces or civilian bureaucrats, who are more likely to defect to a nonviolent opposition—especially one that appears to represent diverse constituencies within the country—than to a violent campaign, where their survival is not assured. This may be because violence causes the opponent to cohere and unite. Thus, violent movements play to the regime’s strengths, whether they use indirect or direct approaches.

Another good reason for security studies scholars to pay attention to nonviolent resistance is that dictators themselves seem to fear mass nonviolent uprisings far more than violent insurgencies, as evidenced by this fascinating propaganda video released by the Iranian Interior Ministry several years ago. Authoritarian regimes view nonviolent resistance movements as threatening subversive, precisely because they have fewer tools with which to deal with them without provoking backfire.

Ironically, violent insurgencies may be much easier for dictators to deal with, given that the insurgents confront repressive regimes using methods in which such regimes have a decided resource advantage. The ideal situation for Mubarak would have been for him to face a violent pro-democracy rebellion. He was quite experienced in putting down violent uprisings. This is why he took such pains to employ agents provocateurs to force nonviolent protestors to react with violence—and why we should expect authoritarians in other Middle Eastern regimes to attempt the same tactic. But in the Egyptian case, even when Mubarak’s regime unleashed a wave of armed agents provocateurs, the protestors were prepared to maintain discipline and refuse to escalate their actions, which would have undermined their legitimacy and given Mubarak’s security forces the pretext to repress them. Instead, that repression backfired, inspiring near-universal condemnation, resulting in ever more committed mobilization by pro-democracy protestors, and leading to the total refusal of the Egyptian security forces to comply with Mubarak’s orders.

Given the historical potency of nonviolent resistance, why have security studies scholars avoided studying it? I think that there are a few reasons.

  • The first challenge is that people hear the word “nonviolent,” and they assume that these movements are “passive,” “weak,” “pacifist,” or “activist.”
  • Second, it’s difficult to overcome skepticism about whether nonviolent resistance can work against brutal regimes, or whether it can effectively confront extremely “powerful” countries.
  • Others may think that nonviolent campaigns only succeed when they have help from foreign powers or the international community.
  • Some may think the study of nonviolent conflict must be constrained by interminable endogeneity issues, because nonviolent campaigns only emerge in situations where change is already on the horizon—where victory is already assured, so it is “safe” to organize a nonviolent revolution. According to this view, mass nonviolent mobilization is a consequence of regime transitions or self-determination rather than the cause.
  • And finally, nonviolent resistance may be viewed as extraordinarily difficult to measure—a preconception that has no doubt stymied interest and efforts to collect such data.

Most of these concerns are misplaced.

  • First, well-organized nonviolent resistance campaigns are anything but passive. They are active, coercive campaigns prosecuted by unarmed civilians, often with a great deal of training, planning, and strategic forethought similar to what one might observe on the battlefield. This is why I prefer the term “civil resistance” to terms like “nonviolent resistance” (which defines the activity as the mere absence of violence), or “nonviolence” (which implies a moral or philosophical point of view that opposes the use of violence on moral grounds).
  • Second, civil resistance campaigns are highly effective. In a forthcoming book, Maria Stephan and I find that among major nonviolent resistance campaigns from 1900-2006, over half succeeded, almost all in excessively brutal regimes where victory was certainly not assured at the outset. Compared with a success rate of only about 25% for violent insurgencies, this figure is especially striking.
  • Next, in these nonviolent campaigns, people weren’t necessarily organizing because they saw weakness in their opponent regimes; they organized to create that weakness by attacking the regime at its most vulnerable points (ranging from specific economic sectors to forcing security force defections within the regime).
  • Among the cases we study, foreign aid tended to have no effect on the outcomes of the campaigns—in fact, sometimes foreign aid undermined the legitimacy of the campaign to its own participants.
  • And finally, nonviolent resistance is possible to conceptualize in a way that lends itself to robust inquiry using a variety of empirical methods for causal inference. Projects under the headship of Doug Bond, Doug McAdam and John McCarthy, J. Craig Jenkins, and others make this clear. Whether these efforts have been convincing is open to debate, but similar attempts to measure violent events have been no less problematic.

Those interested in pursuing research related to this topic have many reasons to be encouraged. First, there is considerable potential for fruitful inquiry on the causes, dynamics, outcomes, and consequences of civil resistance as a potent force in confronting powerful state interests. Possible areas for further research include explaining the onset of specifically nonviolent mass campaigns (as opposed to their violent counterparts), explaining variation in the success and failure of different nonviolent campaigns, researching the causes and effects of shifts from nonviolent to violent methods (or vice versa), exploring diffusion effects among movements, measuring and assessing the effects of different tactical choices, evaluating the international dimensions of domestic civil resistance, and exploring the long-term political, social, and economic consequences of civil resistance in post-conflict societies.

Second, there is already a wealth of case-based information that can inform such future studies. A number of scholars and practitioners have unwaveringly advocated and carried out the study of nonviolent conflict for decades, resulting in a voluminous body of work that is just waiting for mainstream security studies scholars to dive into. Moreover, some groups, such as the International Center on Nonviolent Conflict, exist for the expressed purpose of spreading knowledge about the dynamics of nonviolent resistance, and are extremely generous in providing scholars with opportunities to do so.

Nonviolent resistance is here to stay as a powerful force for change in the world. I hope security studies scholars will pursue research that affords civil resistance the prominence it deserves in the field.

5 Responses to Why Security Studies Should Take Nonviolent Conflict Seriously

  1. Realist Writer March 1, 2011 at 12:43 pm #

    I was curious to see how the author looked at nonviolence only to be shocked that I would not be able to determine that.

    First, I clicked on the link ” violence causes the opponent to cohere and unite”, because I have heard of that theory before, as a way to explain why the Russian Revolution of 1905 failed (because the revolutionaries used violence, thereby alienating the Army and causing it to side with the Tsar). I disagree with this theory (why did the Russian Revolution of 1917 succeed then?), but hey, I could be wrong, and I wanted empirical evidence.

    Imagine to my horror to find that the author linked to “Why Terrorism Does
    Not Work”. The whole point of the study was to argue, essentially, that terrorism does not work in achieving its aims, one of the major reasons being that terrorists’ aims are…pie in the sky (“Establish Marxism in Turkey”? “Sever US-Greek Relations”?), and thus could not actually accomplish such difficult goals, either as violent OR nonviolent movements. Furthermore, terrorism is NOT the only sort of violent action that can be directed against States, and even the study itself claimed that insurgencies (that is, violence directed against military units as opposed to civilians) are more effective than terrorism, which undermines the contention of the article.

    The second link I clicked on was “discernable effects on their outcomes”, because I wanted to see how the author ‘coded’ for non-violent conflicts. Unfortunately, it led to an advertising for a book, which I would not be able to afford at the moment. Yet the author has stated that violence-nonviolence is not actually dichotomous, but a spectrum, which I feel may be a lot more believable (or at least easier for me to shallow), but also undermines the point of the article in advocating for nonviolence. Certain forms of violence are much more effective than others, could it be that certain forms of “nonviolence” may also be effective? And what is the distinction between nonviolence and violence blurs (as in the case of civil disobedience)?

    Perhaps the goal of the author was to bring to attention the power of nonviolence as a form of coercion, and in which case, I think it might be interesting to view in this fashion. However, I am still skeptical. After all, formally speaking, the Egyptian and Tunisian regime did not fall because of protesters. It fell because of the military, who did not side with the regime and ultimately forced the regime to resign.

  2. Realist Writer March 1, 2011 at 12:49 pm #

    …And I have to add that even though I’m still skeptical, I still think this is a good line of research to concentrate on. There just needs to be more convincing and nuanced arguments though to convince an already state-centric, “violent coercion
    -centric population.

  3. Realist Writer March 1, 2011 at 12:58 pm #

    Sorry for the triple post, but I found that the article actually buried the study I was looking for under the word “others”:

    http://www.mitpressjournals.org/doi/pdf/10.1162/isec.2008.33.1.7

    This study indeed shows the proof I’m looking for. I’m trying to find the Nonviolent and
    Violent Confict Outcomes data set myself, but I’d be interested in looking at this myself.

  4. Andrew Green March 1, 2011 at 2:49 pm #

    If Chenoweth and Stephan want nonviolent resistance to be taken seriously as a research agenda in security studies, then they should stop promoting and incorporating findings resulting from demonstrably poor-quality research. Specifically, the study by Freedom House and ICNC, linked in the post as an example of research “exploring the long-term political, social, and economic consequences of civil resistance in post-conflict societies,” is deeply, deeply flawed in design, terminology, and analysis; the same study was cited frequently in the International Security article that is the core of Chenoweth and Stephan’s forthcoming book.

    The first and biggest red flag is that the study is based in part on a statistical analysis…but no statistical tables or details are ever provided. This should be enough to make any knowledgable consumer wary of the quality of research, and anyone with Chenoweth’s and Stephan’s deep ties to ICNC would surely be knowledgable consumers of such work.

    I was wary enough to re-analyze the report and data myself. This more rigorous analysis of the study’s data showed that:

    * The combination of civic forces plus powerholders (regime elites) is generally as successful as civic forces alone;

    * Moderate strength of civic forces is quite effective; and

    * Violence is not really a factor.

    In brief…the study design only looked at successful transitions, but is based on factors from the pre-transition period, a selection bias problem that was also readily apparent. The terms used in the report’s narrative are myriad and confusing, leading readers to conclude that civic forces are the prime driver of transition, not the combination of civic forces and powerholders. The analysis is simplistic, e.g., arguing that “people power movements matter” because 50 out of 67 transitions involved civic forces, yet 46 out of 67 involved powerholders. Despite touting results from statistical analysis, no details of that analysis were ever provided as would be de rigeur in scholarly studies.

    The connection between scholarship and policy has not always been strong – one of this blog’s goals is to strengthen that connection – but the goal of conducting quality research that has clear application to policy is a worthy one to pursue. The Freedom House and ICNC study as whole, though, is a cautionary tale for consumers of applied research.

    The re-analysis, “Is This Really How Freedom Is Won?”, can be found at http://www.DGMetrics.net/DGMetrics/FH-StudyCritq.pdf , and the dataset will be provided upon inquiry.

Trackbacks/Pingbacks

  1. The Myth of the Rational Insurgent « rationalinsurgent - February 2, 2012

    […] catalyze new and improved research on the topic of civil resistance–a field I’ve been encouraging security studies scholars to take seriously. One of the ways I’ve been hoping to attract greater attention to the topic of civil […]