Hello, this is Terrence McNally. Hands Off. The first national mobilization of opposition to Trump and Musk is called for Saturday, April 5th in Washington, D.C. and in thousands of communities across the country. Ezra Levin of Indivisible says attending those smaller, more local events will have greater impact.
Sponsored by 40-plus organizations including Indivisible, MoveOn, Third Act, and Our Revolution, and as I said, many more. Need some motivation? Here's my 2019 conversation with Erica Chenoweth, professor at Harvard's Kennedy School and author of Why Civil Resistance Works: The Strategic Logic of Nonviolent Conflict. She's done the research and run the numbers, and the data favors nonviolence in a big way. And when nonviolent mass protests involve 3.5% of the population, regimes nearly always are overthrown.
Chenoweth says, "Could you imagine if 11.5 million people, now 12 million, that's about three times the size of the 2017 Women's March, were doing something like mass noncooperation in a sustained way for 9 to 18 months, things would be totally different in this country." And that lays out the promise and the challenge we face.
Hello, I'm Terrence McNally, and welcome to another newly recorded episode of Free Forum: A World That Just Might Work. I'm going to be speaking today with Erica Chenoweth. She's co-host of the award-winning blog Political Violence at a Glance and co-author of Why Civil Resistance Works. And you can learn more at ericachenoweth.com.
Erica, E-R-I-C-A Chenoweth, C-H-E-N-O-W-E-T-H, one word, ericachenoweth.com. The show streams weekly on the Progressive Voices Network on TuneIn, and podcasts are available anytime, anywhere on Apple Podcasts or iTunes, podomatic.com, Google Play, and at my site, terrencemcnally.net.
Now to today's guest. Let me set this up by saying that in the fall of 2000, I interviewed the creators of a PBS miniseries and a book, both titled A Force More Powerful: A Century of Nonviolent Conflict. In that millennial year, we saw a lot of retrospectives on the 20th century, featuring images of world wars, genocide, nuclear weapons, human destructiveness on a scale unprecedented in history. But A Force More Powerful highlighted an entirely different theme: that before, during, and after the Great Wars, there was a series of successful nonviolent conflicts about racial and social justice and self-determination. And they predicted that in a global information culture, such methods would become even more powerful. Now, their work was meaningful and inspirational, but it was also anecdotal.
Now, nearly 20 years later, I'm joined today by Erica Chenoweth, who, working with Maria Stephan, set out to rigorously study the question of whether violent or nonviolent resistance was more likely to succeed. Over two years, they collected data on all violent and nonviolent campaigns from 1900 to 2006 that resulted in the overthrow of a government or in territorial liberation. They created a data set of 323 mass actions, and they analyzed nearly 160 variables. Their finding: in the aggregate, nonviolent civil resistance was far more effective in producing change.
Now, I look at an article in The New York Times this past June 25th by Amanda Taub and Max Fisher, and it opens with this line: "2019 might be the year of the protest." And they cite examples in Prague, Hong Kong, Kazakhstan, and the UK. We can now add Puerto Rico. In April, the president of Algeria left office after protests, and the president of Sudan did the same a few days later. So, what's going on? Why are nonviolent protests working, and what might that mean for us here in the US?
Erica Chenoweth is a professor of public policy at Harvard Kennedy School and a Susan S. and Kenneth L. Wallach Professor at the Radcliffe Institute for Advanced Study. Foreign Policy magazine ranked her among the top 100 global thinkers of 2013. She blogs at Rational Insurgent and co-hosts the award-winning blog Political Violence at a Glance. Along with Jeremy Pressman, she's founding co-director of the Crowd Counting Consortium that collects data on the size of political crowds protesting within the US since the first Women's March of 2017. She's co-author of Civil Action and the Dynamics of Violence, The Oxford Handbook of Terrorism, and Why Civil Resistance Works: The Strategic Logic of Nonviolent Conflict. Welcome, Erica Chenoweth, to Free Forum: A World That Just Might Work. Let me tell listeners we're recording this conversation Tuesday, July 23rd.
Well, thank you very much. Thank you for having me. And I know when I listened to you being interviewed, you cited one of those folks I referred to, one of the authors of A Force More Powerful, Peter Ackerman, as someone that had informed your thinking. So that's nice that that goes full circle.
I like listeners to get a feel for the people behind the work and the ideas that we talk about on the show. So can you talk briefly about your path to the work that you find yourself doing today, and so that they get a sense for you, you feel free to go way back: mentors, turning points, moments of decision, that kind of thing.
Sure. I think actually the fall of the Berlin Wall and the Eastern European revolutions in 1989 were probably some of the most formative events in my life, in my own political awakening and consciousness. I was a kid at the time. I was born in 1980, so I was just starting to watch the news and notice newspaper front pages and headlines and such.
And then I can recall noticing that there were wars breaking out in southeastern Europe, the Balkans, and in many of the former Soviet republics. And a lot of the conflicts that caught my attention were those in which people were talking about children who had been the victims of violence. There was a girl named Zlata Filipovic who'd written a diary during the siege of Sarajevo, and my mother had bought me the book at a bookstore, and I'd read it, and Zlata and I were exactly the same age.
And so even as a kid growing up, I was really intrigued by and concerned by political violence and the ways that it affected people around the world, and so wanted to have some kind of career in which I could be involved with responses to violence. When I went to college, I figured the way that I could contribute was trying to potentially pursue a career in the military. And so I sort of tried out for Army ROTC. I took a couple of military science classes and went to some of the trainings and found that it wasn't for me, and that I actually took to more of the scientific side of things.
But I was very interested in when and why political violence happens. And when I went to graduate school, I committed to studying terrorism, why terrorism happens in democratic countries so much, and I was really interested in trying to use statistical tools to explain and describe patterns of terrorism and its causes. And so by the time I was finishing my PhD, I was really fully entrenched in kind of security studies and the academic study of violence and how to contain it, prevent it, or reduce it.
But it didn't really occur to me that nonviolent action was a part of that story at all. The fields were very distinct from one another. Most people associated with nonviolence or nonviolent action were considered more normative theorists, from how I could kind of access that information. And so it didn't really come through in the work that I was doing until I was kind of accidentally invited to a workshop that was put on by the International Center on Nonviolent Conflict, which Peter Ackerman and his family had financed.
And there I met Peter and Jack DuVall and Maria Stephan, who was working there at the time. And you know, they were talking about nonviolent action, their book A Force More Powerful, and other examples from around the world where ordinary people had used mass collective nonviolent action to achieve really radical transformations in their societies. And I found myself really curious but skeptical about a few key questions.
The first was really whether there had actually been some kind of systematic comparison in the reasons why people use both nonviolent and violent resistance, and then whether they kind of occur in different types of countries or different types of regions of the world against different types of states. So, for example, one of my main reservations was that nonviolent resistance couldn't work if the opponent was very powerful, brutal, if they had closed off space for civil society, if they had, you know, really sophisticated methods of repression that they were willing to use.
I also didn't think that nonviolent resistance could achieve radical transformations. I thought it could only achieve incremental or reformist changes when it happened. And so these were some of the things that I kept kind of pondering and asking people about. And at the end of the workshop, Maria and I kind of sat down and came to some kind of a research design that would address all of my skepticism.
So, you know, basically, how would we design a study that addressed the key hypotheses I was sort of spelling out about the fact that nonviolent resistance might be less effective against authoritarians or for real radical goals? So that's really what motivated the research. And then in the end, as you mentioned, the research really supports people who are more guardedly optimistic about the promise of nonviolent action as a category or method of conflict.
There are all kinds of different questions and caveats that come along with that, of course, but I was surprised by the results. And from there, Maria and I have both kind of dedicated much of our professional lives to continuing to study different aspects of nonviolent resistance, both occasionally together and in separate collaborations with others, and solo projects too. So I think that that's, you know, much of what I'm doing now is related to that initial kind of discovery that Maria and I came across in 2008, really, and then we wrote the book that came out in 2011.
Yeah. And one of the things that I think people probably heard in that is that you liked the notion of nonviolent resistance, but you assumed that it was not as successful. I gave a few statistics in the intro, how many conflicts you looked at and how many variables and so on, but can you take us through in a little more depth how you go about rigorously seeking a credible, meaningful, and maybe most important, useful conclusion when you undertake that comparison between nonviolent and violent? And I think one of the things that I was surprised by, and you were surprised, was that no one had done it prior to you and Maria's setting out to do it, right?
Yes. So, you know, to the question of how do you even design a study that can credibly deal with this question? Of course, there's lots of different kind of methodological inclinations that people have. But in my case, with my own toolkit at the time, what I was trying to do is come up with a design that would capture a really long-term study. So one that covered a significant period of world history, that also covered basically every country in the world. So every potential candidate for a place where people could wage some kind of struggle, armed or unarmed.
And then basically it came to looking at comparable campaigns of action. So you have to make sure that you're not comparing apples to oranges. So in our case, we used kind of a strict inclusion criteria. First, the campaign had to be seeking maximalist goals. That is, as you mentioned, overthrow of an incumbent national leader or territorial liberation, which could be either a secession campaign, some kind of self-determination campaign, or the expulsion of a foreign military occupation or colonial power.
And then basically we excluded campaigns where people were asking for less than that. So, for example, if they were asking for the expansion of certain rights, if they were asking for policy reforms or even institutional reforms, we were not going to include those because we didn't want to include campaigns where, you know, the goals might be easier, so to speak, to achieve because they were demanding less.
And then we had to come up with a criterion for success. And again, we followed standard practice in political science, which is to essentially try to compare on discrete outcomes that you can observe across many cases that are comparable. So what we chose to do was look at the removal of the incumbent national leader through irregular means. We looked at territorial independence, that is full independence, both de facto and de jure. And we counted like a dictator being voted out through elections as not a success.
We counted autonomy, even significant autonomy, as not a success. So you had to sort of achieve the full basket of that maximalist objective. The campaigns had to do so within a year of their peak mobilization, and they had to have had a discernible impact on that outcome such that we couldn't imagine it happening without the campaign.
So, for example, if there was some kind of international intervention that forced the outcome, we wouldn't necessarily say that that campaign was the only reason it happened. And we wouldn't say that if the dictator died in office of a heart attack that the campaign won, even if they had an indirect impact on that outcome.
So basically, we then circulated our data to a panel of field experts to make sure that we had included all of the campaigns that had been reported from 1900 through 2006 around the world, both through history books, through different existing databases of different protest events. And we use as our unit of analysis campaigns rather than events of protest.
And the reason is because campaigns are a bit easier to observe. That is to say, it's easier for us to make a pretty credible case that we picked up all reported incidents where people had at least a thousand members who were actively confronting the opponent over a period of time, as opposed to like a one-off event that could go unreported and unnoticed.
So, you know, basically our inclusion criteria are pretty strict. And once we compare armed and unarmed and mixed campaigns along those criteria, then we can really make a much more credible argument that we are looking at one-to-one comparisons.
Very good. Let me ask a couple of questions about that one. When the campaign, whether violent or nonviolent or mixed, as you say, some where there was a nonviolent component but there was violence, when that is not sort of the sole determinant, or you can't say that was what did the trick because there was, as you suggested, foreign intervention or something like that, do those go as failures or do they fall out of the count?
Yeah, so they stay in the count, but we would mark them down as partial successes. So that means the campaign might have had an impact on the outcome but wasn't the main driver necessarily, or it was really ambiguous and hard to untangle. So, for example, the Indian independence movement, Gandhi's movement and the Quit India campaign, we put as like a partial success.
Now, functionally what this means in the actual analysis that we've done in the book and in the data set is that those campaigns that didn't succeed outright are counted as failures. And this again is for a really conservative reason, which is that we don't want skeptics to say that we overcounted successes or biased the results in favor of nonviolent campaigns, which also have a higher success, a higher partial success rate, and that in the sense that they get more concessions short of full success than armed campaigns. So we wanted to exclude those. So the only campaigns that we coded as successes are those that met those strictest criteria.
Okay. So that's one thing. And the other is, this just occurs to me, what about those less than maximal conflicts? In other words, when I think of the US, you know, we're in need of some resistance, we're in need of some change, but it's not going to fall into, you know, your overthrow sort of model. Has anyone done research on that, or is that just too loose? You get where I'm going. In other words, do we assume that the same preference for nonviolence shows up in those, and so that would be sort of guidance for us going forward.
Yeah. So, we definitely couldn't with our resources collect those data because, as you can imagine, if you just were looking around the world at any campaign where there are at least a thousand people engaged in mobilization, it's a major undertaking. And I'm actually doing that with a research team right now that kind of covers the whole world from 1955 until 2017.
So we're collecting data on all kinds of different movements that are, you know, basically the only criterion is that there are at least a thousand people engaged for longer than a week in some kind of collective action. So that really opens up the opportunity to do some more thorough analysis about what we can extrapolate. But I will say that once you look at the campaigns that way, it's very clear that there are far, far more nonviolent campaigns that happen around the world than violent ones.
So it is the dominant mode of conflict prosecution, so to speak, when people have an issue that they're trying to resolve and their institutions are not able to provide solutions or redress for those grievances. And we'll know a lot more about more generalized findings, I hope, in the next few years.
That's exciting. So for anyone who's, this is civil rights issues, obviously a lot of the stuff that's happening around climate change and will continue to happen around climate change, those sorts of things where you're not overthrowing the government, you just want them to do the work.
Yeah, and Jeremy Pressman and I are collecting events data in the United States on a daily basis. And that's through our Crowd Counting Consortium, where we look at every single political crowd that has assembled in the US since the Women's March. So we have pretty good data just about like patterns of protest in the United States in the last couple of years.
Okay. One thing I'd love you to do is cite some examples, just from your work and so on. And so I'm going to mention three of them, then I'll ask you to talk about the role of nonviolent resistance in those. In 1986 in the Philippines, the end of the Marcos regime. In 2003, the ouster of Eduard Shevardnadze in Georgia. And then the one that I mentioned in the intro earlier this year, the presidents of both the Sudan and Algeria announcing they would step aside. Just give us a little feel for what the resistance was and how it worked.
Sure. So in the Philippines there was about a three-year-long campaign that was really a pro-democracy movement, the sort of iconic People Power movement of the Philippines. And Marcos, Ferdinand Marcos, who was the president of the Philippines at the time, was backed by the United States, and there were significant conflicts that had been ongoing in the Philippines over the previous several decades.
And you know, in the end, there was kind of a political crisis that erupted. There was a kind of bloodless coup, and the protesters had both protested on behalf of the military stepping in and then protested to protect the military once they had tried to step in to remove him.
A couple of the things that are notable about the People Power movement is the role of women and religious communities. In particular, the Catholic Church was pretty important in mobilizing a kind of moral argument both for greater freedoms and for the protection of civilians and unarmed protesters at the time.
And so in the end, not only did the People Power movement assemble huge mass, that is just, you know, hundreds of thousands of people at key moments of the mobilization, but they were able to do it in a way that prevented some mass atrocities from actually taking place. There's one anecdote where there was an order for apparently the air force or some kind of special forces to fly above demonstrators and attack them from the skies.
And some of the nuns that were associated with the Catholic Church in Manila arranged themselves into a cross in the streets so that when the planes or helicopters were flying overhead, they looked down and they saw the crucifix. So it apparently prevented some of them from engaging in the violence.
And in the end, I think it's one of those iconic cases because it shows how civil resistance works. It doesn't necessarily work because it melts the heart of the dictator, per se, but it works because, or when it works, it works because there were huge numbers of people that figured out a way to elicit shifts in the loyalties of the pillars that support the dictator, in this case security forces and even kind of first-violent loyal security forces, and that those defections are key in really dislocating the opponent from its pillars of support.
So of course this all happened in the context of a contested election outcome and a political assassination, which points to another thing that's really interesting about many different civil resistance campaigns and kind of semi-authoritarian regimes. They often will happen in the context of elections that raise questions about the fairness of the process and the dignity of people who are casting their votes in good faith only to have them stolen or manipulated.
And so there's something about elections that seem to galvanize and catalyze movements even in pretty repressive places, that makes those kind of key moments of collective action. So in the Georgian case, this was another case where there was really kind of a mass, a very large-scale uprising against Shevardnadze that was ultimately successful.
Based on the cases you've studied, what are the key elements that are necessary for a successful nonviolent campaign? And my other question as you begin to answer that is, the data study where you set, you know, here's what defines an uprising, here's what defines success and so on, it seems to me once you get into the explanations you're not in quite as definite and clear a territory. How do you come up with those explanations and then what are they?
Yeah. So basically, I guess I can sort of summarize what I see as the four things that seem to be associated with successful nonviolent campaigns, and these are factors that we did collect and code for each of the different campaigns in the initial data set. So the first is the numbers. So the larger the campaign and the more diverse the campaign participants in peak moments, the more likely the campaign is to succeed.
And that's because large-scale participation opens up all kinds of political opportunities for the movement that smaller campaigns don't have. For one thing, they make movements that are nonviolent much more likely to elicit defections from security forces. So we can observe when that happens, and those are coded in the data set when they do, and it's definitely the case that when security forces in particular start to disobey the regime, it's much more likely that the campaigns succeed.
The third factor is large-scale campaigns in particular are much better at innovating new tactics. So they are much better, for example, at being able to shift to methods of dispersion, meaning like strikes and stay-at-home demonstrations that are actually very disruptive as opposed to just engaging in protests and demonstrations.
And then the fourth factor is that nonviolent campaigns that maintain discipline, that is they kind of are able to stick to their plan and remain resilient to repression over time, are much more likely to succeed than campaigns that kind of fall into disarray as soon as repression escalates.
So the way that we can sort of glean those factors is both, as I mentioned, in some of the descriptive data that we get from the large-N study, but Maria had done her doctoral thesis on the topic of nonviolent resistance and had done some really thorough case studies that allowed us to pull out some of these other more qualitative factors.
And then we also did some new research for our book that we wrote together that covers Burma, the Philippines, Iran, and the first Palestinian Intifada. And so we're able to compare basically successful and failed cases of nonviolent action and of violent action across those cases, that helps us to pull out these key factors of success.
Okay. So yeah, that's kind of a mixed methodology, but I can understand, as I said, it's not quite as black and white as does this qualify as a conflict and does it qualify as a success. This is a little more gray, but I get it. I'm just going to repeat them so people get it. Large, diverse, sustained participation, number one, the more people that are involved and stay involved, and then the loyalty shifts. Key one is security forces, but it's also the rest of the elites that are supporting.
And then as you say, mere repetition of demonstrations, mass demonstrations, is not going to be as successful as when you begin to vary it with boycotts, stay-at-home strikes, those sorts of things. And then to finally resist the temptation to respond in kind to violence and repression from the authorities.
Let me just jump off from that last one and say you've actually studied what happens in these mixed situations when there is a block within the protest movement that says, you know, we've got to fight back, we've got to meet them with violence and so on. And I know that that temptation has to exist, and we even see that in the US, it seems to me, where you have a lot of peaceful demonstrations and then you have confrontations between, say, white nationalists and Antifa and those sorts of things. So what are your findings? What are your findings when violence is inserted into a nonviolent campaign?
Yeah. So again, I've only really looked at this in depth myself using these maximalist campaigns. And you know, what I found with Kurt Schock in particular in a 2015 study is that nonviolent campaigns that embrace or adopt violent flanks are less likely to succeed than nonviolent campaigns that don't. The nonviolent campaigns that embrace or adopt violent flanks are still more likely to succeed than outright armed campaigns, but they are less likely to succeed than the ones that avoid that.
And I think that the main thing to look at there is that the nonviolent campaigns in which there is an armed flank, they tend to be smaller in size, and the size actually tends to, people tend to stay away once a campaign starts to adopt arms or be associated with that, and the participants become more homogeneous.
So what that generally means is that there are fewer opportunities to create those different political leverage points that I mentioned in the prior discussion. Now it is the case that sometimes nonviolent campaigns with violent flanks succeed, but I would argue that they typically succeed in spite of the violent flanks rather than because of them.
So this is true, for example, in the South Africa case, in the anti-apartheid movement, where, you know, there's a temptation to say, oh, there was an armed campaign and an unarmed campaign, and that made the unarmed campaign much more kind of desirable as a negotiating partner. And, you know, Mandela was helped by the fact that there was an armed flank waiting in the wings and engaging in kind of credible threats of escalating the conflict.
And you know, reasonable people can disagree about whether that's the way it happened there. I think in the South Africa case it's much more convincing to say that the presence of an armed flank in that context actually created a much higher intensity of repression against even unarmed dissidents and made it a much tougher environment in which they could organize effective nonviolent action for a number of years.
But they nevertheless found a way to engage in mass noncooperation economically, and that forced an economic crisis and galvanized the international community to isolate South Africa economically in a way that really created the opportunity for a negotiated outcome to that conflict that ended legal apartheid.
And yes, there is still a lot of room to go in terms of racial discrimination and all kinds of structural violence that remain in that case, but there's no evidence to suggest that an armed campaign would have set the country on a smoother path than the nonviolent one.
Right. One of the things I've heard you say is that when the authorities can get the situation to become either chaotic or violent, that you're now playing on their turf, and that works for them. One of the things that you've said, which I think people will understand, is that as soon as that begins to happen, then the accusation that the uprising is terrorist or communist or anarchist or whatever, you know, that they are the enemies of the people as well as the state, becomes a much easier case to make.
And one thing you have researched, or someone has researched, but I read it in your notes, is that among Americans that have been polled, I believe that when there's violence involved, people have said I'm much less likely to get involved.
