Erica Chenoweth: Why civil nonviolent resistance works

>> Hello everybody. Good afternoon and welcome. I'm Susan Collins, the Joan and Sanford Weill Dean of the Gerald R. Ford School of Public Policy and I'm delighted to welcome all of you here this afternoon for today's policy talks event, which is also cosponsored by the Ford Schools International Policy Center. We're very honored today to be joined by Dr. Erica Chenoweth, who's traveled all the way from the University of Denver's Josef Korbel School of International Affairs. We're delighted to have you here with us today. I'll introduce her more fully in just a few moments. But first, I'd like to say a few words about the Josh Rosenthal Education Fund. Today's policy talks lecture is named for Joshua Rosenthal. Josh was a 1979 University of Michigan graduate. He spent his last year here at the Policy School and then went on to earn a Master's Degree at Princeton University. He was passionate about world affairs and he worked in the field of international finance. He died in the attacks on the World Trade Center on September 11, 2001. Josh's mother, Marilyn Rosenthal, was a longtime Michigan faculty member, and it was very important to her to shape some meaning from what happened on 9/11 to honor her son's optimism about the world, and also about how mutual understanding, dialogue, and analysis can improve communities both here in the United States but also across our world. Marilyn and others established the Josh Rosenthal Education Fund, which enables the Ford School to encourage new and deeper understandings of international issues, and I know that Josh's aunt, Harriet Berg [assumed spelling], is here with us. Welcome, we're delighted to have you. Thanks for joining us. And we're very grateful to your family and friends for their ongoing support. Marilyn Rosenthal passed away in 2007, but we're very grateful to her and the Rosenthal family for making Erica Chenoweth's visit possible and for their lasting contribution to the Ford Schools experience. We also need your help to sustain our students' opportunities to be able to engage with policy leaders like Erica, and if you're interested in supporting the Ford School and the next generation of public servants and policy leaders, I invite you please to visit the Ford Schools website at fordschool.umich.edugiving. And now it's my great pleasure to introduce today's speaker, who I know that Marilyn Rosenthal if she were here with us would really, really be pleased to have deliver today's lecture. Erica is an internationally renowned authority on political violence and its alternatives. She teaches at the Josef Korbel School and is associate senior researcher at the Peace Research Institute of Oslo. She is co-author with Maria J. Stephan of why civil resistance works, the strategic logic of nonviolent conflict. This landmark study won the Woodrow Wilson Foundation award and the 2013 Grawemeyer Award for ideas improving world order. Foreign Policy Magazine ranked Erica among the top 100 global thinkers in 2013. And last year she was awarded the Karl Deutsch Award, which is given annually to the scholar under 40 who has made the most significant impact in international policy or peace research. And it's also a great pleasure for me to tell you that just this week and after the printing of our program, Erica was named a full professor and also the Josef Korbel School's Associate Dean for Research. Congratulations.

[ Applause ]

Following her remarks Erica will take questions from the audience. So beginning at about 4:40 p.m. we will have staff who will be walking the aisles to collect your questions. Ford School faculty member, Professor Susan Walsh, also a Josef Korbel alum, together with two Ford School students, Duverce and Tanner Cooper will facilitate the Q and A session after the remarks. And for those watching online, please send us your questions via twitter using the hashtag policy talks. And so please join me in welcoming Erica Chenoweth, we're delighted to have you here today.

[ Applause ]

Well thank you so much for that very kind introduction. And I also want to thank Dean Collins, Cliff Martin, Erin Florez, and the students and faculty here at the Ford School for making me feel so welcome so far. I also want to thank Josh Rosenthal's family and friends for setting up this memorial fund in his honor. I was able to read a little bit about his life and his family's life today and came across a story about his mother's search for meaning and his death in 9/11 and a beautiful Washington Post article that was published on September 11, 2006, on the mothers of the people who died. And I was struck so much by a comment she made in that article where she said that for her the search for meaning was about the search for knowledge, because she felt that knowledge was agency and knowledge was power. I can identify with that, and so I thank you for having the honor of delivering this lecture today. I want to talk to you little bit about where I was when I first came to the topic of civil resistance. Because I wasn't somebody that came to the topic easily, willingly, or with a very open-mind. So it was 2006, I was a Ph.D. student finishing up my dissertation at the University of Colorado. And my dissertation was all about why political violence happen so much in democratic contexts. I was really bothered with questions about political violence in general, how to stop it, why it happens, and how to reduce it whenever possible. I was not at all familiar with any work on nonviolent alternatives to political violence, and it had never struck me that one could potentially think about nonviolent action as a functional alternative to violence. But what happened is, one of my colleagues forwarded me an email invitation to a workshop called People, Power, and Pedagogy that was happening down the road at Colorado College. Now this was an educational project that was being funded by the International Center on nonviolent conflict. They're a foundation in DC whose main object is to spread and support the development of knowledge about nonviolent conflict. And because I was a graduate student and they were offering free books and free food, I applied and got in and they sent me the box of books in advanced to read. And these were books by people like Gene Sharp, Peter Ackerman, Jack Duvall, Stephen Zunes, and a variety of others who were making what I found to be a really bothersome claim, that nonviolent resistance, where ordinary civilians use coordinated tactics like strikes, boycotts, protests, stay aways, go slows, and a variety of others, they can actively confront, even very repressive opponents, and create radical social change. I thought that this was a very naive view. And I also felt unconvinced by the types of evidence that were being used to support that claim. So when I went to the workshop I came with three major reservations, which I am going to refer to as common claims because I hear them repeated all the time. The first one is that I thought nonviolent resistance, I mean these are unarmed civilians, right, who are trying to collectively organize in ways that disrupt the status [inaudible] power in their countries. Nonviolent resistance could not be effective when the opponent was very powerful, when the opponent was very repressive or willing to use mass repression, or when the opponent was authoritarian in nature, meaning that there was no real political space for effective organizing. The second set of reservations I had was that nonviolent resistance might be all fine and good if you are pursuing [inaudible] reform goal, which is to say reform goals which is to say, I'm sorry.

[ Inaudible ]

Yeah, sure.

>> Sorry. Test, test. Sorry. It's okay. My apologies. Okay.

>> Reform goals, which is to say goals that are like labor rights, for gender rights, or political reforms that would be short of, you know, the removal of a dictator for example, or the removal of a foreign military occupation or colonial power from one's lands, and goals like this that I would usually associate more with violent insurgency. And then the third reservation I had was mostly based on an article that had come out that summer in the Journal International Security by a guy named Max Abrams. And this article was called Why Terrorism Does Not Work. And in it Max collects data on a whole bunch of different active terrorist groups around the world, or so defined by the U.S. State Department, and he looks at how many of them had actually achieved their stated goals and found that it was just 7%, so a pretty ineffective method if you look at it relative to other types of military coercion that are usually around 30% or so. And so I was coming to this workshop thinking, okay so at 7% effectiveness terrorism is judged to be ineffective. And on the last page of his article Max says, so if it's just 7% effective, then why does it happen so much? And he speculates that maybe it's because even at such a low rate of effectiveness it's still more effective than nonviolent protest. So I was coming to this workshop with a very low bar of expectations about the effectiveness of nonviolent resistance, and yet I was coming up against all kinds of material saying that it could be a powerful technique for social change. So I got into some arguments at the workshop and I wasn't a very popular participant, I must confess. But toward the end I met this woman named Maria Stephan, who was randomly kind of paired with me as a roommate in the dorms, and she and I were staying up late in the night having these very antagonistic conversations about it. And ultimately we decided to honestly kind of make an intellectual beat where we decided to do a study that would provide some nominal data to find out whether nonviolent resistance was actually having the success rates in history that its optimists were claiming it to have. So, you know, she asked what kind of research design would be convincing for the average political scientist. And I pulled out like a napkin and drew out what I thought was the beginnings of a research design, and what it would have to do was instead of cherry picking our favorite cases where we looked at successful nonviolent resistance, what we would need to do is do a really wide ranging search of campaigns of nonviolent civil resistance worldwide from a longer period of history. So we went back to 1900 and then we were -- our time period cut off in 2006. So 106 years of data worldwide. And we were going to have to look at all of the existing data sets on things like protests, strike, and other things to try to develop the list of campaigns. So not just one protest but continuous coordinated campaigns of nonviolent action. We also were looking at case studies and histories written by other people about this subject. And good timing, there was a bibliography that came out in the summer of 2006 by a bunch of specialists on this topic named April Carter, Howard Clark, and Michael Randall, where what they did was they collected an annotated bibliography of everything that had ever been written about nonviolent resistance, and then they provided all the source material. So we went to those materials and started basically collecting a list, what you might call a consensus list, of all known nonviolent campaigns from 1900 to 2006 that featured at least 1000 observed participants and that were maximalists in their claims, meaning the removal of a dictator, the removal of a foreign military occupation, or colonial power, or secession. So why did we go after those? Because I was a skeptic and I thought that probably nonviolent resistance wouldn't succeed in those cases. We then looked at campaigns of violent action and we simply pulled those from the correlates of war data set. And what we found there were campaigns where people were using armed action to try to prosecute those same goals. We then created a metric of their outcomes, whether they succeeded or failed. And we applied a strict definition of success, which means that they had to have achieved the removal of the leader or complete territorial independence, both de facto and de jure within a year of the peak of their events. And they also had to have had a discernible impact on that outcome. So like if a leader died of a heart attack while they were in office we wouldn't count that as a success even though you could make the indirect claim that the movement might've had something to do with it. In large part because we didn't want to contaminate the data with cases that were kind of borderline if you want to think about whether the campaign actually led to the outcome or could be judged to have done so. So basically, once we put this all together we submitted it to a review of about a dozen experts around the world and kind of incorporated their feedback and came up with the data set of 323 campaigns of nonviolent and violent action worldwide from that period. Then we ran the numbers and the results kind of blew my mind. This is actually updated data through 2014 to show you that the results are still relevant. But basically the nonviolent campaigns were outperforming the violent ones by about two to one in terms of full success. We also created a mid-level measure, which for all my related arguments today I'm going to consider failures, but that are coded as limited success, meaning that they might have achieved say de facto autonomy rather than full independence in the case of territorial campaigns, but that the violent campaigns were failing far more often than is commonly thought. Another thing that was really striking is that these trends are really increasing over time. What you can see here is that while violent insurgency onsets were kind of peeking out in the 1960's and 70's, after that there's been a sea change in the sort of method of choice for civilian-based opposition movements, such that nonviolent resistance is becoming far more frequent in its onsets and violent insurgency is declining significantly in terms of new campaigns. Another thing is that the trends of success are changing over time, such that nonviolent resistance in the past 40 years has become increasingly effective and violent insurgency is becoming increasingly ineffective. So for a person like me, this raises a huge question. Why is this happening? And another kind of tangential question is, why didn't I ever hear about this before? So basically Maria and I drilled down into the data. And the thing that really struck out is the fact that basically nonviolent campaigns often feature such large and diverse mobilization potential that they are able to activate a number of different points of political power that are not often available to the average violent insurgency. So the large and diverse nature of nonviolent campaigns allows them to disrupt the status quo and start to change and transform power. Now the idea that the more people that participate in an uprising, the more likely the uprising is to succeed is not something unique to Maria's and my study. There's a colleague of several of us here named Mark Lichbach, who in his book, The Rebels Dilemma, had speculated something he calls the 5% rule, which is to say that no government can withstand a challenge with just 5% of its population. That level of mobilization in any society is going to prove very difficult for a government to manage. And, you know, many other people have basically pointed out that numbers matter, Joe DiNardo and many others. But what Marie and I found is that the average nonviolent campaign tends to field something like 11 times more people as a proportion of the overall population than the average violent campaign. And actually the only campaigns in our data set that surpassed Mark Lichbach's 5% rule were nonviolent ones in terms of people actively participating at the peak of the campaign. The reason why we think that nonviolent campaigns are more large and diverse is basically because there are a number of lower barriers to participation in them. The first lower barrier to participation is simply a lower physical barrier, which is to say that many of the tactics available to people participating in nonviolent campaigns are easy to do and don't require any prior physical training prowess or agility. For example, I used to I take military science classes because I thought in college that I might finance my education through ROTC. And so, you know, I enrolled in the classes and then I was going to go and participate in the physical training. Well, you know, I soon found out that meant I had to get up at 4:30 in the morning and run for like 10 miles. I didn't want to do it so I quit it, right. And this is a very trivial kind of stupid example, but the reality is that organized armed activity requires people who are physically able and physically willing to prepare themselves and dispose themselves in situations that are physically challenging. For nonviolent campaigns there are many alternative options that make participation available to a much wider range in the population, including those with physical disabilities, but also the elderly population, youth, and so forth. For example, one extremely powerful technique of nonviolent action is for everybody to stay home and sit on their couch for three days. Now if everybody in a major city sits at home for three days when they ordinarily would not do that, this is going to shut down that city's economy and it's going to create a change. Anybody can do it. In fact, some people might like to do that and take a little break. So these types of physical -- lower physical barriers allow people like this to engage in methods of nonviolent action that aren't otherwise available to oppositional movements. There's also a lower commitment barrier, and by this I don't mean commitment to the cause, but I mean barriers to entry and barriers to exit. So for armed action it's very difficult to know how to get into the group unless you know somebody who's actively engaged, and then once you're in the group it's hard to get out, especially if you've committed what's equivalent to murder in your society. Once people have joined armed groups the armed group becomes their protection, it's hard to get out. So it's going to be a major life choice. You're not going to have a ton of casual participants in these types of movements. For nonviolent campaigns there's plenty of room for casual participants. And in fact, most people who participate in these campaigns don't have to leave behind their entire lives and their families and their friends in order to do so, although that's sometimes the case for upper echelon leadership or so forth. There are also lower informational barriers. And by this I mean that all oppositional campaigns have something of a communication problem. They all want everybody to think they're huge. Violent insurgencies want people to think they're huge by launching armed actions and allowing them to create demonstration effects that attract more people to the fray. Nonviolent campaigns have less of a problem of clandestine attraction because many of the techniques of nonviolent action are deliberately visible so that people think that they're huge. So just take this as an example, imagine that you are in your apartment, you live in a really repressive country, and your neighbor comes to you, and you're close friends, and your neighbor says to you tonight we're going to have a demonstration down the street at 8 o'clock and I know that we have talked a little bit about how you feel about the current regime and so forth and so I hope that you're there. Just know that there will probably be police, but here's a little map with an escape route and hope to see you there tonight. Now if you're like me, you're not to be the person who shows up at 7:55, right, and is sort of like I'm here, where's everybody else. I'm probably going to be the person who looks out my window at 8:30, 8:45 and sees what's going on. And if I see six people down there in the square I'm going to sit this one out. That's just me. But if I see 6000 people and more coming down the way that looks like a big party and I'm probably going to, and I'll probably stay in the middle because that's where you don't get hurt. The more people there are, the bigger the middle is, right. So what happens is that the more people who join, in my mind I can do the probability calculation of how likely it is that I'm going to be hurt, and this kind of creates this kind of critical mass dynamic that you often see punctuating the end of these nonviolent uprisings. The last lower barrier to participation is a lower cognitive barrier, which is just to say the participants in nonviolent action do not have to override the many cognitive inhibitors there are to offense of violence in the typical human brain. So why do numbers matter so much? They don't just matter because they swell into these huge moments of crisis where all of a sudden they melt the hearts of the opponent. That's not at all what we see when we look in these data. Instead, we tested out a series of hypotheses that were derived from Gene Sharp's work on this topic. And Gene Sharp had drawn on [inaudible] notion that obedience to authority is really the critical source of power in any given society, that there's no such thing as a monolithic power that is kind of permanent, and has a continuous source of authority in society. Instead, the key idea here is that no power is monolithic, and no power is permanent. And in fact, every source of power, whether it be a CEO, or whether it be a dictator, or whether it be a foreign occupation, etcetera, is 100% dependent on the reliance, obedience, and help of people that reside in various pillars of support. So for example, they are relying on the obedience of their security forces on their state media, on religious authorities, on business elites and economic elites, educational elites, those pillars that kind of make up the society and continue to obey the power holder in ways that allow it to sustain its power. Now the idea here is that the more people who participate in a campaign, and the wider ranging their own allegiances are, the more likely they are to have access to and be able to put pressure on people who reside in those pillars of support to create moments of crisis where those people reevaluate what their own long-term interests are. This does not require the movement to have moral authority. It just requires the movement to build political power and relationships. So let me give you a very concrete example of when this has happened. In 2000 there was a mass movement in Serbia, which was later called the Bulldozer Revolution because different construction workers brought bulldozers in to downtown Belgrade to remove the barricades that have been set up to keep protesters out. The protests were in opposition to fraudulent elections that were orchestrated by Slobodan Milosevic, who many people think of as the Butcher of the Balkans. This movement ended up with a key moment of hundreds of thousands of people descending on Belgrade in October of 2000 to make him go out. And what happened is that many of the protesters had seized or stolen police radios and were listening as the orders came through in crowd control and other things. Basically they heard the order come over as the key moment came as the protesters approached Parliament for the police and the internal security forces to shoot live fire into the protesters. They also saw that none of them did it. So there was this moment where the activists knew that the order had come down and that the police were disobeying, and that's when they knew that the game was over. Now, in later days, journalists, activists, and researchers interviewed some of these police and asked them why did you disobey the order. And they said, I saw my kid in the crowd, or I saw my neighbor's kid in the crowd, or I saw the guy who sells me liquor on Saturdays in the crowd. And what happened is, because of the sheer size of the campaign, it became so that the people who were in these pillars of support knew that their own lives would be shaped by the decision they made in that moment. And they decided, all of them at the same time, that they didn't want to be the people who were cast aside in the society that they saw inevitably coming to pass. So there were key defections in these moments, and those defections made all the difference in the world. Sometimes it's not possible to have security forces defect. South Africa is an example where the security forces, for whatever reason, were never going to defect, but economic and business elites did. And so the key is to find the pillars of support that are the most influential and through which the power sustains itself, and all of them can be vulnerable to change. We tested this hypothesis and indeed found that as nonviolent campaigns get larger, they are far more likely to bring about defections or disobedience in security forces, whereas violent campaigns get larger it doesn't make as much of an impact. Now sometimes, of course, security forces do shoot into the crowds, and even in those cases we still find the nonviolent campaigns having a two to one advantage. The question is, why? We think there are two reasons. The first is that we find a generalized pattern that repression against people who are unarmed generally is more likely to produce backfire, which is simply the dynamic where people are so outraged by the violence against unarmed people that they actually swell in numbers and actually try to protest precisely because of the repression that takes place. But the second reason, which I think might be even more powerful, is that because of their size there's a very wide array of methods available to nonviolent movements that allow them to maneuver around repression in key moments. There are generally two groups of methods that we talk about, or at least that I talk about. One is methods of concentration. That's when you have the demonstration or the protest in a particular place. It's very disruptive. It assembles mass. It's useful to draw attention and draw journalists and you get the pictures in the papers. But they're also very vulnerable to repression. It's easy to surround them. And if movements do them day after day it becomes very predictable and perhaps dangerous. But then there are methods of dispersion, and that's when people stay away from places they were expected to go. Remember the sitting in your house for three days example. So here we see a general strike paired with a stay at home demonstration. Now these security forces are probably getting paid overtime to stand out there in the sun. So it's costing the opponent something to have this tactic where nobody's going to get hurt. Very large movements are able to shift to these disruptive tactics and have just as much disruption potential as they would if they assembled mass in a key position. In fact, one of the best examples of this switch in sequence is from the Iranian revolution, the one from '77 to '79. The last hundred days of it, called the hundred days of revolution, featured about 90 days of street protests, demonstrations, and rallies. Thousands of people were killed in those demonstrations in that first 90 days. And in the last 10 days the oil workers decided to go on strike. They stayed at home and the internal security forces went door-to-door, dragged them back out on the streets, marched them to the oil fields, and when they got there they worked at half pace. The second day the same thing happen. Oil workers stay home. Security forces go door-to-door, march the people back to the oil fields, they work at half pace. After day three or four of this, security forces start calling in sick. The reason is because it's a very personal kind of violence and they also make the calculation, which was erroneous but they didn't know, that they were not to get their overtime because the oil fields were pumping at half pace. Within a few more days the Shah fled the country and the regime collapsed. So as I wrap up I just want to point out a couple of different longer-term impacts. So I've talked about very discrete moments, where say a leader fleas or a country achieves independence, but we know that not everything is great in every country after these things happen. Iran itself is perhaps a cautionary tale. So Maria and I were very interested in what happens in these countries after success. And what we find is that actually a couple of things, first of all, far fewer people have died in the course of the campaign even if it fails. The second thing is that it's far more likely that the country emerges as a democracy as coded by the polity data set within five years compared with violent insurgencies. And it's also true that these countries are about 15% less likely to relapse into Civil War [inaudible] controlling for other factors that bring about that outcome. And so we're finding that not only are nonviolent resistance campaigns more effective in the short term, but they bring forth different longer-term impacts that make these societies more peaceful and more politically open places to live. So what I would argue is that for people like me, these were very difficult findings to uncover because they fly in the face of many of the types of conventional wisdoms that I had been raised with and that I'd been learned to think about in my field. And so it really points to understanding and unpacking the nature of power. I thought power flowed from the barrel of a gun, but it may flow from consent and obedience, and that place is a very different emphasis on the types of subjects we study. It also places a different emphasis on the role of civilians in conflict. Many people have always, in my field, studied civilians as either just victims that suffer at the whims of the governments that rule over them, or as resources to loot in the context of a Civil War. But few people think of them as active agents making decisions that can drive processes that alter their lives in ways that they actually have some control over. And so it really kind of brings back into my mind a real difficult tension between structure and agency, to what effect our institutions and situations forcing people to use violence for example, or are there realistic alternatives to violence? And I believe that this research suggests that there are. For policy implications I think there are also some key questions. The first question, of course, is whom to support? This one is still open for debate in the United States as our government right now considers revamping its support to Syrian rebels. Now certainly it's the case, as we know from prior research, that supporting rebel movements often gives them a strategic advantage in their conflicts. And we know from our data that when violent rebels receive support from outside states they are about 15% more likely to succeed. However, we also know that when they win they are much less likely to create democratic governments. They are much more likely to commit various human rights abuses, both during the struggle and after. And countries in which rebel groups have been victorious are much more vulnerable to relapse into Civil War. As the people who they depose fight a life-and-death struggle against them and many of these countries end up falling into the conflict trap in ways that are very difficult for them to get out of. And so, you know, if it's true that the United States government stands for democracy, stability, and human rights, then providing guns and money to armed movements is not the best way to spread those things according to our data. That said, it's not the case that providing direct support to nonviolent movements helps them either. In fact, what we found in our data is that when governments provided finances to nonviolent campaigns it had no visible impact on their success rates. First of all, it's pretty -- it was pretty rare and our data, even if it's becoming more common now. But secondly, there was no net impact, positive or negative. And my guess is that because these campaigns are so dependent on the popular legitimacy of their participants that providing support from outside may undermine that level of popular support and create free rider affects. So the question is, okay, if state support isn't the way to go about it then what are the alternatives? And I'm actually undertaking a research study now with Maria, it's our next thing, where we try to evaluate the types of external support that have been provided to nonviolent campaigns over the past 25 years, and the impacts, if we can discern them, on those types of support and types of supporters on the outcomes of these campaigns, both short and long-term. Now my hypothesis on this is that we're going to find that civilian led instruments are much more effective. Because if it's true that agency matters and that people are making choices that alter the course of their conflicts, then things like training, skills, preparation, and knowledge about how to wage nonviolent struggle effectively are going to make a much more difference than whether the movement gets money, or whether it's diplomatic cover from foreign governments. That's my working hypothesis. But we may also find that there aren't any types of external support that are helpful and we may indeed identify some that do great harm. So with that what I'm going to do is just, you know, re-identify the three basic findings here. The first is that nonviolent resistance has been surprisingly effective even in context where we wouldn't expect it to even be possible. The second thing, which we detail much more carefully in our book, is that this is true in a variety of contexts, and even when we control for conditions and factors that might predispose societies to using nonviolent resistance over violent resistance. And then the third major finding is that the way that one fights largely influences the way that one rules when one wins. And so for those that are wishing to change their circumstances in radical ways, it's worth thinking not just about the short-term victories but also the types of societies they're trying to create. We find considerable empirical evidence to support the notion that short-term violent victories lay the groundwork for long-term violent societies. And because of that I'm very excited that this century is actually, despite what you read about in the news today, very much Gandhi's century. And I hope that this generates some exciting discussion, questions are great, insults are fine, and at least, at the very least, I hope I've disturbed you enough to make you think about nonviolent resistance a little more. Thank you.

[ Applause ]

Do you want to use this one and I can use the podium mic?

>> Sure, thank you.

>> Does it work?

>> All right, [inaudible]?

>> Sure. Hello? Can everyone hear me?

>> Yes.

>> All right, great. Thank you again, Dr. Chenoweth, for agreeing to speak to us here at the Ford School. My name is Brenda Duverce and I'm a second year master of public policy student here. And the first question we have for you today is, besides the number of participants, did you discover any other key ingredient to a successful nonviolent movement?

>> Uh-huh. That's a great question. And we did several qualitative case studies and I'm continuing this work. And what I can tell you from knowing the cases well is that generally are four different attributes of successful campaigns. The first is, of course, large and diverse participation. The second is the ability to create loyalty shifts within different pillars of support in the opponents, so whether that's security forces or economic elites, changes place to place, but it can be a powerful moment of leverage for these campaigns. The third thing is whether the campaigns have innovative and creative tactical sequences. That is to say, are they just doing every Friday a demonstration, or are they alternating between demonstrations and other kind of innovative techniques like say strikes or flash mobs or other types of things that are drawing in participants, keeping pressure on the opponent, and also creating those loyalty shifts. And then the fourth element is the degree to which the campaign is able to unify a response to repression. As we know from work by your very own Christian Davenport, my friend here, as governments start to feel more threatened they start to crack down more, right. And so the question is, how does the movement respond? What we find from our data and a forthcoming paper I have with Kurt Schock, is that campaigns that are able to mobilize a unified and disciplined response and maintain nonviolent action in the face of this repression are much more likely to succeed than campaigns that split or divide along the lines of people who want to do violent action and people who want to do nonviolent action or who turn to violence altogether. So those are the four things; participation, loyalty shifts, tactical shifts, and responses to repression that are disciplined.

>> All right, my name is Tanner Cooper and I am a senior in the undergraduate program here at the Ford School, and I've been focused on expanding healthcare for low-income Americans. But our next question to you is, how did your research address movements such as black lives matter or occupy Wall Street that don't have explicit or agreed upon goals? And was there like a measure of success in shifting the conversation or how did your research go about that?

>> Yeah. So we explicitly collected data on maximalist goals. My understanding at the time was that those would be more difficult to achieve. I've since thought more cleanly about that, and I think actually it might be the opposite, that it might be more difficult to achieve some reform goals in highly pluralistic societies. And so occupy Wall Street and -- really more occupy Wall Street is an example of how that can be a real challenge. I think black lives matter actually has a very clear set of goals. I think that despite the fact that they don't enter into our data set because of the nature of these campaigns that the four attributes I just described of successful maximalist campaigns apply just as well to basically any type of reform campaign, so the larger and more diverse the participation, the ability to respond to increasing brutality with discipline, the ability to create changes within the political elites, the economic elites, and with potential third-party supporters and the ability to create incredibly creative and innovative tactics that keep the opponent off balance are all things that would make for successful movement here as anywhere else. And I actually think that black lives matter, as far as I can tell, is performing very well on all of those points.

>> Next question. Your research foc -- your research focused on resistance to dictatorial regimes in a general way. Do you have any evidence that nonviolence resistance works in the face of active genocide or full scale attacks?

>> Yeah, that's a great question. I don't -- I don't have any clear evidence on those cases specifically, although I know many people who have now undertaken studies to find out exactly what's possible in the context of genocide or mass killings of that scale. What I can tell you is that I am working on a project with my colleague, Evan Perkoski, where we're looking at whether nonviolent campaigns are as likely to provoke mass killings and genocide as violent insurgencies, and we find that they are not by a longshot. So, if your question is, what's kind of the safest technique to use in terms of, you know, provoking or creating a moment where mass killing is likely, nonviolent campaigns are clearly much less likely to create that environment than violent insurgencies. In terms of nonviolent campaigns that have occurred in the context of mass killing and genocide, there are cases. There are some famous cases out of Germany, the Rosenstrasse protests by women who were married to Jewish men. German women who were married to Jewish men who successfully protested for their return from concentration camps in front of the SS office in the Rosenstrasse in Berlin. This was late in the stage of the -- of the war. There was also the cases of the Danish resistance to the Nazis in Norwegian resistance, the teachers strike for example, that were moderately, modestly effective methods of resistance. And, you know, my sense is that when regimes are committed to genocidal action against people that really nonviolent or violent resistance are very difficult in those environments. And so I'm not sure that violent resistance gets people further in those environments as when you're talking about collective action. So that's my sense right now. We need to know more about this for sure. And there's at least one dissertation that came out of the school by a guy named Chris Sullivan, who evaluated the archives of Guatemalan Internal Police and found a very sophisticated network of disruption of political opposition in that country among leftist groups, but there was also civil war going on at the times so it's difficult to untangle the impacts of the nonviolent versus the violent resistance there.

>> All right. Our next question is, how can U.S. foreign policy help to promote nonviolent movements over violent rebellions in repressive states?

>> That's a good question that I've been thinking a lot about. I think first of all, it would be useful to develop just less faith in the tool of promoting armed resistance. Like first things first. Let's come to terms with the fact that this hasn't worked out very well in U.S. foreign policy. And I think that's necessary before even the idea comes that there are other alternatives that can make sense. Certainly I think that kind of gleefully jumping to back armed reactors in conflict zones is something that the U.S. has been not so great with. The [inaudible] is like a great example. The U.S. was extremely ambivalent about all of the popular movements that emerged during that time, but as soon as armed reactors developed in Libya it was like finally some people who are speaking our language, like let's go and support the TNC and we'll support them with the NATO intervention and the whole nine yards. And I think that what happened then in Syria, for example, was that there were many people in the Syrian national Council and in other parts of the opposition in the summer of 2011 that saw what was happening in Libya and thought we can have that here too. And so what that did was it created much more risky behavior on the behalf of defectors who basically took their weapons with them to Turkey, and thought that if there was an armed campaign in Syria that the Western countries would likewise support them with an armed intervention and they miscalculated, big time right. But it was -- but I think that when we do that we send the message that we reward armed action and we don't really know what to do with an armed action. I was in Kosovo doing research on the nonviolent resistance to the war between 1989 and 1998. And when I was talking to a few of the interview subjects they mentioned that the leader of the LDK, a guy named Ibrahim Rugova, was very excited when the Dayton Peace Accords were sort of put on the international agenda and thought this is our moment to bring up Kosovo. And he was shocked and dismayed when he discovered that the Kosovo Albanians were not allowed to attend and were not invited to attend. And he asked a French diplomat, what's the deal? Why aren't we, as a nonviolent Democratic movement, coming to the Dayton Peace Accords? And the diplomat said, take a look around the table. You're going to see a bunch of guys in uniforms who have been blowing stuff up. That's who gets to come. Right. And so the message over and over as we reward armed action and we ignore nonviolent action. So I think the first thing first is to stop rewarding armed action, especially in negotiations and make sure that civil society groups and leaders get their rightful place at the table.

>> Great. This is our first question from Twitter, and the question is, how does the study distinguish violence and nonviolence elements in a movement?

>> That's a great question. So the way that we -- this is a question that has a lot of detail to it, as you might imagine. And so we have an online appendix. It's free for our book that you can get on by going to my website ericachenoweth.com and go into research and you can download it. It's about 300 pages. It's a PDF. And it lists all of the decisions that we made around coding something as nonviolent versus violent. But the short answer is that we looked at the primary techniques of action that were used and whether the primary participants in the campaign were unarmed civilians. And if that was true then we coded it as primarily nonviolent campaign. As I mentioned, we drew the violent campaigns from the correlates of war data set and there the definition is simply that armed actions have been taken by a non-state group of -- that resulted in least 100 deaths on both sides. And basically what we did with kind of middle-of-the-road cases where we used those international experts that I mentioned to help us adjudicate between whether we should call something nonviolent or nonviolent with violent flank, because there are some cases, like say South Africa, where it was primarily nonviolent during the final phase, but there was also some armed action going on. And so we often coded those cases as nonviolent with a violent flank. They're about 29% of our cases of nonviolent action that fit that description. And when that was the case what we did was we often estimated our models, both with kind of the ideal types and then we also controlled for whether there was kind of violent flank element to it. Kurt Schock, in my paper, actually unpacks that much more. And what we find, as I mentioned, is that the nonviolent campaigns that feature some violent flanks are generally less effective than the nonviolent campaigns that are nonviolent primarily. So that's how we did it. There is some judgment calls involved but I encourage you to look at the numbers and the data and see what you think.

>> All right. So our next question is basically asking if you mind discussing the role of leadership in the success of nonviolent versus violent resistance, because leaders clearly matter, but do they matter in different ways or are there characteristics that are most important, just how does leadership play a role in nonviolent resistance?

>> Yeah, I think this is an area for continual study. Our NAVCO 2.0 data set actually does have a variable that looks at whether leadership is kind of coordinated and centralized, or whether it's diffused, more like a leaderless resistance. We didn't require that there had to be a single figurehead, which my guess is that that's actually fairly dangerous for nonviolent movements, to have a single leader. But we did look at whether they had a specific leadership structure, such that there were authorities within the campaign that could essentially spell out the rules of participation and move the campaign and maneuver it in key ways so that it wasn't just willy-nilly, everybody's doing what they want. And what we found just in a very simple correlation is that the campaigns that are able to manage more of a discernible leadership structure of that type generally have a higher success rate than those that are leaderless in nature. And my sense is it's related a little bit to what I mentioned about the kind of key factors for success, things like creating loyalty shifts and being able to maneuver around repression when things get very bloody. Those are things that require coordination, that require legitimacy from within the movement leadership, and those two things are very difficult to have when you just have a leaderless resistance, because then it's sort of everybody fending for themselves. But like I said, this requires much more inquiry. There's a lot of work in sociology on leadership and movements. And I think what I just mentioned accords somewhat with what the general, conventional wisdom is now there, but there's obviously remaining controversies.

>> Continuing with the discussion on leadership, do you feel that policymakers have been interested in and receptive to your work in the U.S. and at the U.N.?

>> I think it's all over the map, and you know, Gene Sharp was right. There's no monolithic government. It sort of depends on where they sit. It depends on what their own experience has been. I'd say if I have to put people into general categories, civilian defense officials are the most dubious. But to tell you the truth, uniform military are extremely receptive. It sort of -- it sort of depends on I think what -- and to some extent what their own prior beliefs are about the sources of power. And, yeah, I think that that has made a difference. I think actually though the largest kind real-world impact that we've had relates much more to people that are involved in struggles. They're the ones I hear from the most and who wish to talk about or learn more about what our findings really say for their contexts. That creates a lot of really interesting kind of ethical dilemmas and implications that is another kind of unexpected result of this research. I've gotten to really look carefully about the meaning of research for real-world implications, the limits of research for real-world implications, the ways that people sort of interpret things or misinterpret things, and what my own responsibility is in spelling out what this work might mean for people. So, you know, it's really -- it's been kind of an interesting ride with looking at the real-world implications and applications of this work. And, you know, my hope is that it least, as I said, bothers people to think a little bit about the alternatives to violence, of which there are countless alternatives to violence as they approach their various conflicts.

>> Is it common for violent campaigns to go nonviolent or vice versa? If so, is there any reason why and like what is the nature of these campaigns?

>> Yeah, so this is getting me back to my roots, which is why do -- why do these campaigns go violent in the first place? So here's the interesting thing. There are many examples of violent campaigns that have abandoned violence, but have not fully demobilized in the sense that they just stopped. There are actually a number of campaigns that turn to nonviolent civil resistance seeing that they had run out of their violent options and wanted to try something that they thought at the time might be more effective. One of the most recent examples of this is the Maoist insurgency in Nepal, which in 2006 abandoned non --violent struggle in the Civil War to basically ride the coattails of an urban popular uprising that was using nonviolent resistance to force the monarch out on. But there are other cases like South Africa, El Salvador, Palestine during the first intifada, and many other cases where primarily violent resistance has been supplanted by civil resistance, and often to great effect. East Timor is another very good example. On the reverse, the question of whether nonviolent campaigns turned to violence, well certainly most violent campaigns will tell you they started nonviolently. Many of them will say, well we tried nonviolent action and it didn't work so we went to violence. Now in looking at the cases of violent action in our data set, only 2% of them, 2% used civil resistance, as we define it, for more than nine months before they turned to violence. Now this is an important figure because what we find is that the average nonviolent campaign is about three years to run its course to either fail or succeed. And so to some extent they might be jumping the gun a little bit. But another thing that we find about them is that most violent campaigns, including those that say they tried nonviolence, actually try something like organizing a political party or, you know, writing letters or having one protest that goes badly. And it's kind of interesting the double standard that's there. Because if there's a protest where say 30 people get killed by the police, many people say nonviolence doesn't work so we have to do violence. But if there's a violent episode were 30 people are killed, they would say our tactic isn't working so we need more and better violence, right. What if when 30 people die we need more and better nonviolent action, and maybe a sense that, you know, nonviolent resistance as we find here doesn't work because it's morally right? It works when people are able to wield real political power. And so I think that, you know, the sense that the moral dimension of nonviolence is what distinguishes it from violence has actually made people falsely believe that it could achieve things that it can't in very short terms. And so the more we can do to debunk this and to disseminate knowledge that this is about politics and not about right and wrong per se, I think the easier it will be for people to understand that the [inaudible] protest where 30 people died doesn't mean the end of the potential of nonviolent resistance in that case.

>> Great. Continuing off the discussion of violence, the question we have is, how would nonviolence work with Isis and what practical lessons from your research could be applied to combat?

>> Yeah, so I don't know how nonviolent resistance works against Isis right now, in part because Isis is a very amorphous kind of shadowy presence right now in much of the places where it's governing. But I can tell you that the more that Isis starts to look like IS, the more it starts to look like that, right, and so when it starts to look like that, there are more opportunities for there to be sustained nonviolent resistance against the structures that it develops to rule. Isis doesn't just want to win combat victories, it wants to rule, and so that creates longer-term opportunities for thinking about it like that. In the meantime we can think about Isis as having kind of four basic categories of participants. There are the true believers. There are the psychopaths, to be frank. There are opportunists, who are people who are flocking to the scene because they see their own short-term material interests as being benefited by that. And then I would argue now there's a fourth category of what you might call coerced people, and this is people who live in villages in Iraq or in Syria who when Isis conquers the village they come to all the men in the village and assemble in line and say to the first one, are you with us or against us? And he says, I'm against you and they kill him. And they go to the next person and say, are you with us or against us? And he says, I'm with you. And it's made up of a lot of people like that right now. And so even now it's not fully monolithic, and I think that presents an important opportunity to see that people are joining for various different reasons and have various different short and long-term commitments to the -- to the movement as it stands now. So I don't have a very good answer. I suspect that if I did I'd be much busier than I am now. But I would just say that, you know, many of us are used to thinking in very short-term bits. Maybe the new cycle or something else has brought us to thinking about it that way. But I think in terms of three, five, 10 year increments what the political scene might look like, and I don't see Isis as developing a lot of local legitimacy, and I think that makes them quite vulnerable to challenges from below in future years.

>> All right. So our next question is, how does your research translate to a more local level, we're talking unions, factory workers, community activists, how does it translate to the local level?

>> Yeah, so again I think that the sort of four features of success apply no matter what the scale is really. There was a student that came to a workshop that I was participating in who was a student at Swarthmore University. And, you know, he thought, okay, so what will do is get 30% of the student body to sit in at the board of trustees meeting and force them to divest from fossil fuels. And they had like a whole long conflict and my understanding is that it turned out pretty well for them. So, you know, the sense is like you, you know, any campaign that mobilizes a sustained number of people over a particular threshold, say 3 1/2% or so, is probably going to make a pretty big impact on whatever their environment, even if it's in the very local level. So I think that that's something that translates no matter what.

>> This question focuses on North Korea and countries similar to North Korea, how does nonviolence begin in countries with oppressive regimes and the control of information is so great that a civil resistance movement cannot effectively start organically?

>> Yeah, that's a great question. So I've become increasingly dubious that countries that look from the outside, like they have total information control, really do, just because there are so many incidents. Like even East Germany, I mean people sort of look back on East Germany now knowing what we know, and they're like oh, they didn't have a lot of information control. They were totally vulnerable the whole time. It was just a matter of time until they collapsed totally. But I think in the mid-80's and the -- and the late 70's, I mean people thought about East Germany as like kind of similar to how they think about North Korea now, you know. And so it's -- you know, we have hindsight bias with a lot of the cases where nonviolent resistance succeeds, and because it succeeds we rewrite the way we thought about it before as if it was like an easy place for nonviolent resistance to emerge. And so I think, you know, I don't know what's going happen in North Korea. But generally what we see in many countries is, as processes and practices what James Scott, the anthropologist, calls every day forms of resistance, which is things like foot dragging, or the refusal to obey and kind of typical social conventions, or refusing to buy from a particular fruit seller that you know engages in corrupt practices, or whatever, and these things kind of happen, and the more they're routinized and practiced, the more people approach a level of what Doug Adam calls cognitive liberation, which is just to say this whole system is just kind of bogus and now that I've realized this and I can see that some of my fellows also realize this, I can't un-realize that, you know. And so -- and then that makes the moment inevitable where people will rise up and challenge what they think now to be a totally fabricated system. And one of the best books on that process of cognitive liberation is one by Doug McAdam called, Political Process in the Development of Black Insurgency, about the civil rights movement. And another one he wrote call, Freedom Summer, which is about the process of just minds being opened and changed and then never being able to put the genie back in the bottle, right, when it comes to understanding what's available to people and their own agency and achieving it.

>> All right. So this is going to be our last question and I just wanted to say thank you again so much for coming and talking with us today. It's been an absolute pleasure. But for our final question it's, what is the threshold of violence and oppression that a nonviolent resistance group can endure before they start transitioning towards violent resistance or breakdown?

>> Uh-huh. We don't know precisely what the number is. I think there are good reasons for us not to find out. I'm not up for like noble lies or anything but there are reasons why there are certain types of questions that one wouldn't necessarily want to inquire about and publish about, right. But my sense is that the Iranian Revolution was pretty bad. I mean, there were thousands of people killed in the streets at a pace that was much more intense even then what was happening in Syria in the first nine months, if you can imagine. I think that, you know, cases like Syria approach kind of the threshold level. I'm not sure if it was inevitable that that campaign turned to violence, but the nonviolent moment in that campaign's lifecycle endured quite a lot before the campaign sort of basically fell apart and was undermined by regional geopolitics, and so forth. So I think -- I think that, you know, a lot matters on who's part of the conflict, like is it an internationalized conflict, who is backing the regime, and also, you know, who is backing the movement? Has it become so that it's a proxy fight between different international players? If so, the people on the ground are going to be subject to a lot more things outside their control I think than if it remains kind of localized in international struggle. So I don't know. I mean, East Timor, about 50% of that country was killed in the course of the Civil War, which is wild to think about. Yet they were able to effectively mobilize a nonviolent campaign in the face of continual violence by the Indonesian occupation. So I don't know. I mean, I think sometimes there are situations where it's unimaginable. And then there's a case that pops out where it was fine, not fine but where they won, right, in the end. And so what I would say is that Kenneth Boulding, the Quaker economist, used to say what exists is possible. And we know it exists even in contexts where we wouldn't expect it to and; therefore, it's possible.

[ Applause ]

>> I just -- I wanted to give a special thanks to Erica for a very appropriate and important conversation for our 2015 Rosenthal lecture. I'd also like to thank all of you for all of your questions and for joining us today. I hope you'll stay and continue the conversation. We do have a reception just outside in our Great Hall and I hope that you'll come back and join us for future policy talks. Actually our next one is on Friday. We have Ambassador Thomas Miller, who while will be our speaker and his topic is on the nexus of diplomacy and development. But if you go to our website we have a full calendar and I encourage you to do so. And so please join me in a final round of thank you to our speaker.

[ Applause ]