Of all nonviolent methods, the ones that withdraw economic cooperation, boycotts and strikes, have one of the longest records of concrete results. They are also among the most accessible: almost anyone can decide where to spend money.

A short history

• Colonial America. In the 1760s and 1770s, American colonists signed non-importation agreements refusing British goods to protest taxes such as the Stamp Act.
• The sugar boycott. In the 1790s, hundreds of thousands of British people, many of them women running household budgets, stopped buying sugar produced by enslaved people in the Caribbean.
• The word itself. In 1880, during Ireland's Land War, local people in County Mayo refused to work for or deal with Captain Charles Boycott, a land agent who had moved to evict tenants. His name became the word.
• Farmworkers. From 1965 to 1970, Filipino and Mexican American farmworkers, soon united as the United Farm Workers under Cesar Chavez, Dolores Huerta and Larry Itliong, struck California grape growers and organized a nationwide consumer boycott of table grapes, winning union contracts in 1970.

What makes a boycott work

The experience of organizers points to a few common features:

1. A clear, specific demand. “Desegregate the buses” or “sign a union contract” is easier to rally around, and to declare won, than a vague call to “do better.”
2. A clear target that can actually grant the demand and is sensitive to the pressure, such as a company that depends on its reputation with customers.
3. Sustainability. Participation must be cheap enough to keep up for as long as needed: easy substitutes, a simple message, and support for people who bear real costs, like Montgomery's carpools.
4. Unity and discipline, so the target cannot dismiss the campaign as small or violent.
5. A path to victory: a way to negotiate, check the result and end the boycott with a clear win.

Boycotts often work less through lost sales than through damage to reputation and the uncertainty they cause.

Does nonviolent resistance work?

Many people long assumed that violence was the more effective way to force major political change. In Why Civil Resistance Works (2011), political scientists Erica Chenoweth and Maria Stephan tested this. They gathered data on 323 major campaigns from 1900 to 2006 that sought big goals: removing a government, ending a foreign occupation or winning independence. Nonviolent campaigns succeeded about 53 percent of the time, violent ones about 26 percent.

Their main explanation is participation. Nonviolent campaigns can include far more people: the old, the young, people unwilling to risk killing. Broad participation makes defections from a regime's pillars of support more likely, especially among security forces. Countries where nonviolent campaigns succeeded were also more likely to become democracies afterward.

Critics raise questions about the comparison. Movements may choose nonviolence where it already has better chances, and classifying a messy campaign as violent or nonviolent involves judgment. The authors tried to test for such problems, and most researchers accept the broad pattern while debating its size and causes.

The 3.5 percent figure

Chenoweth later observed that, in their data, no campaign failed once around 3.5 percent of the population took an active part at its peak. The figure spread widely, often as a “rule.” It needs caveats:

• It is a historical observation, not a law of nature or a guarantee.
• Very few campaigns ever reached that level; 3.5 percent of a large country means millions of people actively taking part.
• It concerns large campaigns against governments, not consumer boycotts of companies.
• Chenoweth has noted possible exceptions, and that success rates for nonviolent campaigns fell noticeably in the 2010s as governments learned to counter them.

Worked example: designing a consumer boycott

Suppose a community group learns that a popular clothing brand buys from a factory where workers were injured because fire exits were locked. The group wants to act. Here is how the five features apply.

• Demand. A weak demand would be “treat workers better.” A strong one would be: “Commit publicly to independent safety inspections at all supplier factories within six months, and publish the results.” It is specific, achievable and checkable.
• Target. The brand, not the factory. The brand has the buying power to change conditions, and its value depends on its reputation with customers.
• Sustainability. Ask supporters to avoid one brand, and offer a list of alternatives, rather than asking them to stop buying clothes altogether. Most people can keep that up for months.
• Unity and discipline. Agree on one message, stay factual, and do not harass store employees, who did not make the decision.
• Path to victory. Decide in advance what counts as a win and who will verify it. Announce openly when the demand is met and end the boycott, so future targets know a deal will be honored.

Then the group should check its expectations. A boycott of one brand by a few thousand people will rarely dent sales much. Its power usually comes from news coverage, from retailers and investors asking questions, and from the brand's wish to make the problem go away. Measuring success by the demand won, not by lost sales, keeps the campaign honest.