Social Movements & Collective Action: Why Anyone Turns Up

In April 2000 the people of Cochabamba, Bolivia, shut down their city over a water bill. A private company had taken over the water system and prices had jumped. Within weeks the contract was cancelled. Nobody had a great deal of money or an army, and the grievance was not new: people in many cities have paid too much for water. What was new was that tens of thousands of people, who all had a reason to stay home, did not.

That is the puzzle at the centre of this course. Grievances are everywhere and movements are rare. If a protest succeeds, everyone gets the benefit, including the people who stayed in bed. So why would anyone go? A graduate of this course keeps a set of answers, a healthy suspicion of neat stories about "the power of the people", and a clear sense of what the evidence says about how often protest works.

Olson's puzzle: why a shared goal does not organise itself

Before 1965 many scholars assumed that people with a common interest would act together to pursue it. In The Logic of Collective Action the economist Mancur Olson argued the opposite. If the result is a public good (cleaner air, a higher wage for the whole industry, a regime change), no one can be kept from enjoying it, so each person gains most by letting others pay the cost. In a big group, one person's contribution barely changes the outcome, and so the rational move is to free-ride. Olson's conclusion was that large groups will not organise unless they offer selective incentives: benefits for members only, such as insurance through a union or discounts through an association, or penalties for those who do not take part.

The tool is the standard "public goods game" that economists use as a stand-in for Olson's logic. Each of n people can put in one unit. The pot is multiplied by m and shared equally. If everyone contributes, everyone ends up with m units instead of 1. But your own unit comes back to you only as m/n.

Without JavaScript: with 10 people and m = 3, you get back 0.3 of each unit you give, so free-riding pays, even though everyone giving would triple everyone's holdings. With 1,000 people you get back 0.003 unless a private bonus makes up the difference.

Play with it and the pattern is the one Olson found. Small groups can often cooperate: when n is below m, each person gets back more than they put in. Large groups need something extra. The theory has famous critics. Elinor Ostrom (Governing the Commons, 1990) documented many communities that managed shared resources for generations without a state or a market, using monitoring, reputation and graduated sanctions. Sociologists point out that people also join for identity, solidarity and moral reasons, which Olson's version leaves out. Gerald Marwell and Pamela Oliver (The Critical Mass in Collective Action, 1993) argued that contributions are interdependent, so a committed few can make participation worthwhile for everyone else. Olson's real contribution is the question: whenever a movement exists, ask what solved the problem.

Resource mobilisation: grievances are everywhere, organisations are rare

Through the 1950s and 1960s, theories of protest were mainly about people: frustrated, disconnected or irrational. That did not fit the civil rights movement, whose participants were well-organised church members. In 1977 John McCarthy and Mayer Zald argued in American Journal of Sociology for a new frame: grievances are fairly constant, so what changes is the supply of resources (money, volunteers, meeting places, skills, legitimacy) and the organisations that can turn them into action. Movements, on this view, are shaped as much by who funds and organises as by who is angry. Early movement groups, like churches, unions, student groups or village councils, already hold the networks that carry the message.

The theory has been criticised for under-rating anger and for suggesting that professional organisations drive movements. Frances Fox Piven and Richard Cloward (Poor People's Movements, 1977) argued that the power of the poor lies in disruption, and that formal organisation often dilutes it. Both views have evidence. A third useful addition is Bert Klandermans's work on participation (the "supply and demand" of protest): even if a movement exists, someone becoming a participant has to pass through stages, and each stage loses people.

Without JavaScript: with 100,000 sympathisers and conversion rates of 60%, 50%, 40% and 50%, about 6,000 turn up. Doubling any one stage doubles the turnout, which is why organisers work on the weakest stage.

Opportunities and frames: when the door opens and what people think is at stake

Resources are not enough. Doug McAdam's Political Process and the Development of Black Insurgency (1982) and Charles Tilly's work stressed the political opportunity structure: elites split, allies appear, repression becomes costly, a court case or election creates an opening. Tilly also described each society's repertoire of contention, the familiar menu of ways to protest (petitions, strikes, marches, boycotts, occupations) that changes slowly and differs between places. The third element is meaning. David Snow and colleagues (1986) showed that movements win people by framing: diagnosing a problem, naming who is at fault and offering something people can do. "Grievance" is made, not found.

Scholars of Latin America and Asia added something the American models lacked. Sonia Alvarez and Arturo Escobar (1992) argued that many movements in the region were about identity, culture and everyday survival (neighbourhood kitchens, indigenous territory, human-rights groups) rather than bargaining over resources inside a stable state. Orlando Fals Borda's participatory action research, from Colombia, goes further: the people who suffer a problem are the researchers of it.

Granovetter's thresholds: the riot that needs one more person

Why do crowds turn from nothing into everything? In 1978 Mark Granovetter proposed that each person has a threshold: the number of others already acting that would make them join. A hothead has a threshold of 0. A cautious person might want to see 40 others first. His example: a hundred people, with thresholds 0, 1, 2, 3 and so on up to 99. The person with threshold 0 starts; that triggers the one with threshold 1; then 2; and the cascade takes everyone. Now change one thing: nobody has a threshold of 1 (the 1 becomes a second 2). The first person acts, the second person still wants to see two, and the cascade stops at 1. Two crowds with almost identical average attitudes; one empty street, one full.

Without JavaScript: thresholds 0 through 99 give 100 participants; the same crowd with one person moved from 1 to 2 gives 1. A random crowd with average 10 and spread 5 gives 100; with average 20 and spread 4 gives 0.

Two lessons. First, you cannot read a crowd's behaviour off its average opinion: the spread, and the presence of a few instigators, matter. Second, from outside a cascade looks like a mood. Many sudden revolutions were "unpredicted" in this sense. Timur Kuran's Private Truths, Public Lies (1995) argued that people hide their real preferences under repression, so the thresholds are invisible until the cascade begins; that is his explanation of why 1989 in Eastern Europe caught nearly everyone, including experts, by surprise. The model is a simplification: it ignores who knows whom (see Cities, Communities & Social Networks, 35/81) and assumes everyone sees everyone.

Nonviolent and violent campaigns: what the data show and the critics

In Why Civil Resistance Works (2011), Erica Chenoweth and Maria Stephan collected 323 major campaigns for regime change or self-determination between 1900 and 2006. A campaign was counted as a success only if it achieved its goal within a year of its peak and the result was a direct consequence of the campaign. The headline: major nonviolent campaigns succeeded about 53% of the time, violent ones about 26%. Their explanation is about participation: nonviolent campaigns are easier to join (the barrier is lower and the risk smaller), so they draw larger and more varied crowds, which brings defections from the regime's own police, army and business supporters. A well-known side result is that no campaign in the data that gained the active participation of 3.5% of the population had failed. Chenoweth later called it a rule of thumb, not a law.

Without JavaScript: 1900 to 2006, nonviolent about 53% and violent about 26% success; more recently Chenoweth has said the nonviolent rate fell to roughly a third and the violent to about 10%. 3.5% of 10 million people is 350,000.

The critics, stated fairly. Some argue about selection: nonviolent resistance may be chosen where it has a chance, and violent rebellion where the regime leaves no choice, so comparing success rates compares different situations (Rendel Lehoucq, 2016, and others). Others point to coding and measurement choices: who counts as a campaign, what counts as success, and how to classify campaigns that mix methods, as South Africa's did. A third line is the radical flank effect: Herbert Haines (1984) found that when a militant wing appeared in the US civil rights movement, funding and support for the moderate groups rose, so the two may have helped each other. And the facts have moved. Chenoweth herself reported that nonviolent success has fallen since about 2010, and suggested that governments have learnt, digital organising makes quick mass turnout easier but builds less lasting organisation, and campaigns skip the slow work. A fair summary: in the data we have, nonviolent campaigns have done better than violent ones on average, the reasons are debated, and the advantage was smaller in recent years.

Movements from every continent

Most textbook examples come from the United States and Western Europe, which gives a skewed picture. Here are sixteen movements from five continents. Dates are for the movement's best-known phase, and "result" is brief and incomplete, because causes of big political change are always mixed.

Without JavaScript: the movements are listed by region: Africa (anti-apartheid, Green Belt, Tunisia, Sudan), Asia (Salt March, Chipko, People Power, Tiananmen), the Americas (Montgomery, Plaza de Mayo, Cochabamba, Zapatistas, Chile 1988), Europe (Solidarity), Oceania (New Zealand suffrage), plus Syria 2011.

Look across the list and several patterns appear that fit the theory. Churches, unions, markets and villages did the organising (resources). Many began when something shifted at the top: a split among generals, a court ruling, an election (opportunity). Many used familiar local forms: a march, a boycott, a petition, a strike. And the end of the story is often not the end: Tunisia's and Sudan's openings were later reversed, and Syria's turned into a civil war. Winning the moment is different from winning the settlement.

What a graduate just sees

• When a group of people share a goal and nothing happens, ask what keeps them from acting. When something does happen, ask what solved the problem.
• Anger is common and organisation is rare, so look at who has money, meeting places and networks, not only who is aggrieved.
• A crowd is a distribution of thresholds, not an opinion: the spread and the first few movers decide.
• Sudden revolutions are often surprising because private preferences are hidden until the cascade starts.
• In the 1900 to 2006 data, nonviolent campaigns won more often than violent ones, but the reasons, the comparison and the recent trend are all disputed.
• Winning an uprising is not the same as winning the outcome that follows.

Where it shows up next

• Public Policy (35/16): how a movement turns attention into an agenda, and what organised groups do once inside.
• Elections, Parties & Public Opinion (35/15): why turnout is itself an Olson problem.
• Comparative Politics (35/13) and Political Economy (35/17): when regimes fall, and what follows.
• In the world: union organising, a neighbourhood campaign for a crossing, a consumer boycott, a viral hashtag that does or does not become an organisation.

If you want to go deeper

• 📘 Textbook. Donatella della Porta and Mario Diani, Social Movements: An Introduction (Wiley-Blackwell, 3rd edition 2020), is the standard European text. Mancur Olson, The Logic of Collective Action (Harvard University Press, 1965), is short and still the core.
• 📘 Gentler. Erica Chenoweth, Civil Resistance: What Everyone Needs to Know (Oxford University Press, 2021); Peter Ackerman and Jack DuVall, A Force More Powerful (2000); Timur Kuran, Private Truths, Public Lies (1995) for the surprise of cascades.
• 🌐 Free data. The NAVCO dataset of nonviolent and violent campaigns is publicly available from Chenoweth and colleagues (search "NAVCO Harvard Dataverse").
• 🌐 Free course. I could not confirm a single standard free university course on social movements, so I am not naming one. Open course platforms (MIT OpenCourseWare, Open University's OpenLearn, Coursera audit mode) are worth searching for "social movements" or "civil resistance".