Participation Thresholds And Defection

Two things predict outcomes better than anything else researchers have measured. Neither is intensity, and neither is how right the cause was.

The First Is How Many, And Who

Size matters, and there is a famous number attached to it that you should learn along with its caveats.

Looking at the historical cases, no campaign that got sustained active participation from about three and a half percent of the population failed to achieve its goal. That figure has traveled widely and is usually stated much too confidently, so here is the rest of it.

It is an observation about a small set of past cases, not a law. At least one campaign has passed the threshold and lost. Most successful campaigns never reached it at all, which means it is nowhere near necessary. And three and a half percent of a large country is an enormous number of people, sustained over time, not assembled for one afternoon.

Chenoweth calls it a rule of thumb. Treat it as a rough marker of scale, not a target and not a promise.

Who participates matters as much as how many. A campaign of students is a campaign of students. A campaign that includes students, pensioners, shopkeepers, bus drivers, clergy, and off duty police is a different object, and the reason is Lesson 7.2. Broad participation means every pillar contains somebody with a relative in the movement. That is the transmission line. A pillar does not withdraw because it read an argument. It withdraws because the people inside it start hearing the same thing at dinner.

The Second Is Defection

The clearest signal in the whole literature is the moment people who are supposed to enforce an order stop enforcing it.

Soldiers who will not fire. Police who stand aside, or who take off their helmets, or who go home. Civil servants who stop implementing. Business owners who stop paying. State broadcasters who report the crowd honestly.

When that starts and spreads, outcomes follow. When it does not, campaigns can be very large, very sustained, very brave, and still lose. This is the finding that separates the cases, and it is why so much of Lesson 7.2 was about who is holding the roof.

The Soldier Has The Farmer's Problem

Now put yourself inside the pillar, because this is where two modules connect.

A soldier in the line has a decision. If he refuses alone, he is arrested, and the man beside him fires anyway, and nothing changes except his life. Refusing is expensive and useless. If most of his unit refuses at the same moment, he is safe and the order is dead.

You have seen this exact structure before. It is the farmer standing over his well in Lesson 4.1. Going first is worse than useless. The action is only rational if enough others are doing it, and nobody can see whether they will.

Which tells you what the missing ingredient is. Not information. He knows what he has been ordered to do. Not moral awareness. He is aware.

What he lacks is knowledge of what everybody else is about to do, and that is a problem only visible numbers can solve.

This is what a crowd of a hundred thousand actually produces. Not persuasion. Evidence. It tells every person in every pillar something they could not otherwise learn, which is how many other people have already moved. It is the reason movements care about turnout in a way that looks obsessive from outside, and the reason regimes work so hard to make crowds invisible, uncountable, or discredited before anyone sees them.

Why Violence Cuts The Other Way

This also explains the finding in 7.1 from underneath.

Firing on unarmed people is costly for the person who fires. It splits units, because some will not do it. It is hard to justify at home. Officers know that if the government falls, they will be asked about that day.

Firing on people who are shooting is not costly in any of those ways. It is easy to justify, it unifies the security forces, and it converts a political problem into a military one, which is the arena where the government has every advantage. A movement that turns violent has usually just solved its opponent's hardest problem.

There is a related pattern worth knowing. Repression against a disciplined unarmed movement sometimes backfires, producing more participants and more defection than it prevents. Sometimes. It is not automatic and it should not be counted on. Backfire requires witnesses who matter, and a government that controls what is seen, or that faces no consequences from anyone who sees it, can repress a movement into silence and frequently has.

One More Honest Note

Large crowds may partly be a symptom rather than a cause.

When people conclude that a government is finished, it becomes safer to appear in the street, so more of them do. In that case the size of the crowd is measuring the collapse rather than producing it, and the truth is probably that it runs in both directions at once. Researchers have not fully separated these, and you should hold the finding with that in mind.