Chenoweth, Erica and Stephan, Maria J. (2011). Why civil resistance works: The strategic logic of nonviolent conflict. New York: Columbia University Press. (320 pp)


Reviewed by Brian Martin

A book review published in Interface: A Journal for and about Social Movements, Volume 4, Issue 1, May 2012, pp. 371-374

Interface pdf


Imagine you live in a country with a repressive government and you want to do something about it. You are ready to take strong and risky action. What's the most promising way to have an effect? Some of your young friends have left university to join an armed guerrilla movement; others, who don't want to use violence, are calling for protests in the streets. Which of these options is more promising?

The debate over how to challenge oppressive regimes and policies has been going on for over a century with little resolution in sight. Armed struggle has a long tradition, including but not restricted to Leninists. Prominent successes include struggles in China, Cuba, Vietnam, and Algeria. Proponents usually assume armed struggle is the only way to overthrow a regime willing to use unlimited force against challengers.

In contrast is another tradition whose most prominent figure is Gandhi, who led major nonviolent struggles in South Africa and India. Gandhi objected to using violence to promote change; his approach was followed in the US civil rights movement in the 1950s and 1960s, led by Martin Luther King, Jr. Less well known than these campaigns are a host of other unarmed struggles against repressive governments in places like Guatemala, Indonesia, Iran, Philippines, and Serbia.

What do researchers say about challenging repressive regimes? Most attention has been on conditions that enable or hinder success using frameworks such as resource mobilisation and political opportunity structures. Scholars have not systematically compared different methods of struggle. Most of them assume peaceful protest can be crushed by a sufficiently ruthless ruler. As a result, researchers have not provided much guidance for activists. After all, if the key is political opportunities and the prospects are not very good right now, then the methods used by challengers should not make that much difference.

The assumption by proponents of armed struggle and by many scholars is that success without armed struggle depends on a regime being soft. In this way of thinking, Gandhi faced a weak opponent, the kind-hearted British. Likewise, the collapse of Eastern European communist governments in 1989 is attributed more to weaknesses of the regimes than to citizen action.

Due in part to these assumptions, there has been no systematic testing of the comparative effectiveness of armed and unarmed struggles against repressive governments. Until now. Erica Chenoweth and Maria Stephan in Why Civil Resistance Works have provided a powerful statistical analysis that undermines claims for armed struggle and, incidentally, the assumptions of most social movement researchers. (In the context of their study, civil resistance means the same as nonviolent action.)

The basis for their analysis is a database of 323 campaigns, between 1900 and 2006, of resistance to regimes or occupations -- or in support of secession. Included in the database are, for example, the 1944 October revolution in Guatemala, the 1955 Naga rebellion in India, the 1960-1975 Pathet Lao campaign in Cambodia, and the 1974 carnation revolution in Portugal. The database has all sorts of information, such as locations, key protagonists, lengths of campaigns, maximum numbers of participants, methods used, and outcomes.

For Chenoweth and Stephan's core argument, the key bits of information are the methods used (either primarily armed struggle or primarily civil resistance) and the success or failure of the campaign. Deciding whether a campaign is successful is sometimes difficult; maybe only some of the goals of the challengers were achieved or maybe the goals changed along the way. This is only one of many difficulties faced in quantifying the elements of resistance struggles. The authors report a careful process for validating the information in the database including checking judgements about campaigns with experts on the countries and events involved.

With such a database, it is possible to test various hypotheses. Their most significant and striking finding is that nonviolent campaigns are far more likely to succeed than violent campaigns.

A sceptic might claim the nonviolent campaigns were against softer targets. Chenoweth and Stephan tested this: one of the elements in the database is how repressive the regime is. The answer: the strength of the regime makes very little difference to the success of the resistance. This is remarkable. It means that civil resistance works against even the most repressive regimes, and with a much greater chance of success than armed resistance.

What happened to the idea, widely used by social movement scholars, that movements succeed because political opportunities are favourable? Chenoweth and Stephan have replaced it with a quite different conclusion: the keys to success are the methods and strategies adopted by the challengers. Conditions such as the level of government repression don't make very much difference to outcomes. This means that success depends far more on what activists do than scholars, political analysts, or governments have ever realised.

The statistics in the book are supplemented with many illustrations, including four detailed case studies: the 1977-1979 Iranian revolution, the first Palestinian intifada (1987-1993), the 1983-1986 people power movement in the Philippines, and the 1988-1990 Burmese uprising. These vivid stories give flesh to, and help validate generalisations from, the statistical findings.

If Chenoweth and Stephan are right, social movement scholars should reconsider their frameworks and focus on agency, namely what activists choose to do. Why haven't scholars done this before? One answer is that it means relinquishing some of their authority to experienced activists.

What are the lessons for activists? The first and foremost is that armed struggle is not a promising option. It is less likely to succeed and, when it does, it is more likely to lead to a society lower in freedom and more likely to lapse back into civil war. Mixing armed struggle and civil resistance is not such a good idea either. The best option, statistically speaking, is to forego any armed resistance and rely entirely on nonviolent methods.

Why are nonviolent methods so much more effective? Chenoweth and Stephan argue that the key is greater participation. Most of those who join an armed struggle are young fit men, a relatively small sector of the population. Methods of civil resistance include sit-ins and public protests which allow involvement by a greater proportion of the population. Methods such as boycotts and banging pots from balconies allow nearly everyone to join in. It turns out that participation is a key factor in success. The maximum number of participants, as a fraction of the population, is highly correlated with success of the campaign -- and a large number of participants is more likely to be achieved with a nonviolent campaign.

Participation is crucial, in part, due to spin-off effects. More participants, especially when they include a wide cross-section of the population, means the resistance builds links to more people with the likelihood of causing shifts in the loyalty of security forces, which are absolutely vital to success. This process can happen in both violent and nonviolent struggles, but high participation is more likely in nonviolent struggles because there are fewer barriers to involvement. Joining a guerrilla movement or a terrorist organisation requires high commitment, especially due to a high risk of death, whereas joining a large rally or participating in a general strike requires less commitment, thereby allowing the movement to grow. The case studies -- each of which involves a primary nonviolent struggle in which there was a parallel armed struggle -- vividly show this.

Why Civil Resistance Works is an academic work published by a university press. It contains statistical data, explanation and justification of database construction, careful analysis of contrary hypotheses, and much else. Unlike some scholarly writing, it is clearly written, logically organised, and provides helpful summaries. Nevertheless, it is unlikely to become bedtime reading for activists. What then are the takeaway messages?

Here is my list.

* Civil resistance works. A well-organised unarmed campaign against a repressive government is much more likely to succeed than a well-organised armed campaign. The message from nonviolent activists to those who advocate armed struggle should be "show us some good evidence that your approach works better, because the best study so far shows civil resistance has better prospects."

* When civil resistance works, the outcomes are likely to be better. Use nonviolent methods if you want a nonviolent society; use armed struggle if you want a militarised successor regime.

* The key is participation. The more people involved in a campaign, and the more diverse the participants, the more likely its success. Beyond this general conclusion, I think it is a plausible extrapolation from the data for activists to say, "let's choose actions that will involve the most people from different sectors of society."

* Winning over the security apparatus is crucial. Undermining the loyalty of those who maintain order should be a central goal.

* Plan, innovate and strategise. The evidence shows that the methods used by challengers are crucial to success. In other words, how a campaign proceeds sensitively depends on the actions by the players, so it is vital to be creative, respond wisely to opponent movements, and be able to survive repression.

Regimes strategise too, so there is no set of steps that guarantees success; campaigns need to innovate against opponent strategies. Struggle against injustice is like a game: to win, it has to be played well. This is why diverse participation is important, because it brings in people with different skills, ideas, and contacts. Running a campaign from a central headquarters, with a fixed ideology, is not a promising approach. Having widespread participation and encouraging experimentation and diversity is.

The more people understand the dynamics of nonviolent action and learn to think strategically, the more likely a campaign is to develop the staying power, strategic innovation, and resilience to succeed. Why Civil Resistance Works is not an activist manual, but its findings should be used by anyone writing one.

About the reviewer

Brian Martin is professor of social sciences at the University of Wollongong, Australia and author of many books and articles on nonviolence and other topics.


Go to

Brian Martin's book reviews

Brian Martin's publications on nonviolent action

Brian Martin's publications

Brian Martin's website