How Popular Movements Achieve Regime Change: Chenoweth, the 80/20 Rule, the Network Effect, and What Predicts Success
The mechanics of removing entrenched power, and why they favor the people who persist.
Note from the author: This article is around an 8–12 minute read. It’s the first in a series highlighting the academic and empirical research illustrating successful social movements. Follow-up pieces will include the following subjects: the catalog of nonviolent tactics assembled by Gene Sharp; Steven Levitsky and Kim Lane Scheppele on how authoritarian consolidation proceeds and how to counter it; selectorate theory, which identifies the coalition whose defection removes a ruler and tells us where to most effectively apply pressure; my own academic contributions on ideological recruitment and consensus building in social movements, “The Multiplication Engine”; and my work on state-level opposition to federal authoritarian consolidation, “Federalism for Anti-Fascists.”
Nearly every nonviolent campaign that reached the active, sustained participation of roughly three and a half percent of a population achieved its objective, a pattern Chenoweth and Stephan found across 323 campaigns between 1900 and 2006. The exception, Bahrain, exceeded the threshold and failed, and most successful campaigns never reached it at all.
Reaching three and a half percent, then, is not what determines whether a campaign succeeds. In later work, Chenoweth found two things. Nonviolent campaigns succeed less often than they once did: their success rate was highest in the 1990s and has fallen most sharply since 2010. And they have shrunk, drawing in a smaller share of the population than they used to.
Two of my favorite ideas come from outside Chenoweth’s work, and laying them over those findings is the part I find fascinating. The first is the Pareto principle: across many settings, a small share of participants does most of the work. The second is the network effect: each person who joins makes joining safer and more worthwhile for the next, because larger numbers lower the risk to everyone and raise the odds that the movement succeeds.
Together they identify two factors: a committed minority, and the relationships through which its actions reach the wider population.
The network effect explains why people join a movement already under way, but not how one starts. Lower risk and better odds come only as more people join, so the first to act have neither. In a 1978 model of collective action, Mark Granovetter showed why overwhelming public support for change can still fail to produce a movement. If each person waits for enough others to act before acting, and no one is willing to go first, that support stays latent, no matter how large. A few people willing to act before anyone else are what allow the rest to follow. Their presence matters more than the eventual size of the movement.
For those of us who want change, the lesson is to act, and to act where others can see it. The people who would join are waiting to see someone go first. Every visible early action lowers what the next person needs to see before they move. Going first, in the open, is itself what brings the next person in.
This is the pattern behind the fall of East Germany. Susanne Lohmann, in “The Dynamics of Informational Cascades: The Monday Demonstrations in Leipzig, East Germany, 1989 to 1991,” World Politics 1994, tracked 42 Monday demonstrations across five cycles and showed the cascade in real time. The earliest protesters were high-grievance radicals who turned out when repression was most likely. Their visible turnout showed everyone else how much opposition existed, which lowered the cost for moderates to join, until the regime lost its viability.
The formal collective-action backbone is Marwell and Oliver, The Critical Mass in Collective Action, 1993, the standard work showing a critical mass of early contributors triggers wider participation when the payoff curve accelerates. And for a modern empirical version on real protest data, Barberá et al, “The Critical Periphery in the Growth of Social Protests,” PLOS One 2015, shows the committed core generates the signal and the periphery spreads it. Across all three, the same pattern matters more than the exact figure, three and a half percent included: a committed minority that persists can draw in a much larger population once it passes a critical point.
A count of participants also understates opposition. Timur Kuran described why a government can appear stable until shortly before it falls: where people fear the state, many conceal their actual views, so observable support exceeds the genuine level. Once enough people reveal those views, the concealment ends quickly, and a population that appeared compliant proves to have opposed the government for years, which is part of why such collapses are difficult to predict.
Concealed opinion is one element. The outcome also depends on the people who carry out the government’s orders. The factor most strongly associated with success in Chenoweth and Stephan’s data is defection. A government’s orders are executed by specific actors: soldiers, police, civil servants, judges, broadcasters, and the firms on which the economy depends. Security-force defections appeared in more than half of the successful campaigns and made a campaign forty-six times more likely to succeed than a comparable campaign without them.
Defection is also why Chenoweth emphasizes breadth over size. Security forces are drawn from the society they police, and they are less willing to repress a movement that resembles their own communities than to repress a narrow or isolated group. That is a reminder of why we will never win if we only preach to the choir. Every successful movement in history that benefited the public required bringing your message to the out-groups and building public consensus.
If you can’t convince your neighbor then you’ll never convince your governor.
Chenoweth attributes part of the decline since 2010 to adaptation by governments, which now train and screen their forces against defection, and some of which deploy units with no local ties to the people they confront.
Tactics differ in their effect. A demonstration assembles people in one place, where authorities can observe and contain them. A strike or a boycott disperses participants and withholds the labor or revenue an institution depends on. A strike that halts rail traffic, or a boycott that reduces a firm’s revenue, imposes a sustained cost, and the target concedes when that cost exceeds the cost of the change being demanded.
Information or rumor can spread through a single contact, but Damon Centola and Michael Macy found that costly behavior spreads differently: a person undertakes a risky and demanding action, such as joining a high-stakes movement, only after several of their existing contacts have done so. Repeated confirmation from close contacts, rather than a single signal from a distant one, is what leads a person to act. A large movement built from weak and dispersed ties can fail to move its own members to accept significant risk. To succeed, we need to build relationships that have depth and meaning through shared values, knowledge, and action. Those relationships need to extend and expand, that’s the work of community building because ultimately it makes a lot of sense that humans will do extraordinary things when they recognize themselves as part of a shared identity and value structure.
When we find our values, our courage lives in that same place. When we share those values with others, we influence the world to be a better place for everyone.
This is as good a place as any to make it clear that our greatest tool for breaking through the oligarch controlled algorithm and media is the same tool that created the U.S. 250 years ago, that abolished slavery, that built organized labor, women’s suffrage, the civil rights movement, and the same tool that has toppled deeply entrenched regimes. That tool is talking to people in real life. Even better if you can hand them something educational and they can see you practicing what you preach. It’s simple. Not easy, but simple.
It is absolutely necessary to recognize that successful campaigns applied these elements over years rather than in a single event, and each illustrates a different one. Let’s explore some of those campaigns.
Poland’s Solidarity movement lasted nearly a decade. It began in 1980 as a strike in the Gdansk shipyards, grew into an independent trade union of roughly ten million members, sustained an underground press and a network of institutions outside state control, survived the imposition of martial law, and prevailed in the 1989 elections that preceded the end of communist rule across the region. Its decisive feature was duration combined with organized economic power.
In Serbia, Otpor directed its preparation at the security forces. In the two years before October 2000, its organizers trained thousands of people in decentralized local cells and established contact with the police well before the final confrontation, so that when crowds converged on the capital, the police declined to attack. Slobodan Milosevic conceded within a day. Its distinguishing feature was preparation aimed at the people who would carry out the orders rather than at the public alone.
The campaign against apartheid in South Africa applied pressure from outside the country. Alongside internal strikes and boycotts, an international movement of divestment and sanctions, joined by dozens of American states and cities, contributed to the capital flight that weakened the government’s finances in the late 1980s. This outside pressure supplied economic cost, including the withdrawal of external support.
The Montgomery bus boycott of 1955 and 1956 sustained noncooperation well beyond brief inconvenience. Black residents stayed off the city buses for more than a year, and the lost revenue, together with a federal court ruling, ended segregated seating. The boycott lasted because churches and local associations gave participants enough trust in one another to share its cost.
Chile’s opposition turned the plebiscite Pinochet had scheduled for himself against him. Seventeen rival parties agreed to campaign jointly, registered millions of new voters, fielded their own poll watchers, and built an independent vote count to detect fraud. The No vote prevailed with about fifty-six percent. The campaign required agreement among rivals and a means of contesting a result the government controlled.
The campaigns that failed were often large, and each lacked one of these elements.
Egypt in 2011 removed Hosni Mubarak in eighteen days, and the uprising went no further. It did not last long enough or reach deeply enough to affect the army and the other institutions the government depended on. The army withdrew its support from Mubarak to protect its own position rather than to support the protesters, and it later assumed full control. The crowd had removed the ruler while the institutions he relied on remained in place.
Hong Kong’s protests in 2019 were large and continued for months, but the authority that could decide the outcome was in Beijing, and the police remained loyal. In 2020 Beijing imposed a national security law, and the protests ended.
Occupy Wall Street spread to more than nine hundred cities during a single day of action in 2011 and drew sustained attention to economic inequality. It operated, by choice, without named leaders or specific demands, and depended on physical encampments. When police cleared the camps after about two months, the movement ended with them, because the camps were its entire structure.
Iran is a case still unresolved. The protests after the death of Mahsa Amini in 2022 reached more than ninety cities across nearly every province, the security forces did not defect, and after more than five hundred deaths and thousands of arrests the government continued to function. That wave was one in a recurring cycle. A far larger uprising began in December 2025 amid economic collapse, the largest since the 1979 revolution, spread to every province, and again the forces did not divide. Israeli and American strikes on the country, and the threat of more, let the government present the movement as a foreign operation and gave the security forces reason to act as national defenders. The state responded with mass killing, and the protests were largely suppressed by early 2026, though unrest has continued. Each wave has been larger than the last, and each has ended with the security forces intact, so whether they remain loyal will decide an outcome that is still open.
The same pattern appears across the successes and the failures. A trained and connected minority must be in place before turnout matters, because it recruits the rest and creates sustained and increasing pressure on the regime.
The effective tactics withhold cooperation and impose a sustained cost, and they are directed at the full set of actors a power depends on: its officials and enforcers, its corporations and investors, its media and influential figures. Public education and consensus building enlarge the committed minority over time and lead those actors to withdraw their support.
A campaign that does the work and knows that the process is far bigger than a single election will strengthen its own position over time, because an entrenched power must prevail in every confrontation while a challenger need prevail only once. That formula makes success inevitable.
The article you just read is some of the research that formed The Existentialist Republic’s Effective Activism system: Educate, Activate, Recruit, Repeat. You can find the booklet as a free PDF download or a physical booklet in the sentence right below this one. You have our express permission to print, photocopy, and share any of our training materials. That’s why we make it all available in free versions. The ER library is at the bottom of this article.
Grab Them By The E.A.R.R.: How to Get Politicians to Do What You Want — physical copy / free download
The Existentialist Republic needs around ten subscribers per article to keep all the activist journalism moving. Don’t let this be the reason you miss rent or skip a meal. For everyone else, you can be one of the ten today and be the reason the articles, books, legislation, and training keep coming for everyone.
Buy The ER Team a Cup of Coffee
You can get a free PDF of the book in the BMAC shop for $0.00 by clicking on this sentence.
The Existentialist Republic Library-
The American Reformation: How states can use the U.S. Constitution to break authoritarian capture — physical copy / free download
Conservatism: America’s Personality Disorder — physical copy / free download
Intro to Soft Secession — physical copy / free download
Oppositional Federalism and You — physical copy / free download
Toppling Tyrants: A Field Guide to Removing Fascism in America — physical copy / free download
Grab Them By The E.A.R.R.: How to Get Politicians to Do What You Want — physical copy / free download
Being Dangerous: Go From Activist to Operative
More Free downloads:
Soft Secession: 100 Policies That Pass
All Four Completed Model Legislation Bills
The Opposition Guide to Tax Warfare
Six-Panel Soft Secession Brochure
Thankyou for your persistence, research, insight and work on behalf of the disappointed, the fearful and the angry.
Did you say an 11 minute read? Speed reading maybe. BUT, very good article and you supported your hypothesis very well IMO. Thanks for your "cathedral" brain work. I also see so many alcoves and places to go and see in my "cathedral" brain. Tieing all the variables together is the hard work. You seem to do it well.