The Anatomy of a Coup

At one o’clock in the morning on 24 February 1981, a Spanish king put on his old military uniform, sat down in front of a television camera, and saved a democracy with a broadcast. Hours earlier, a lieutenant-colonel of the Civil Guard had marched into the Spanish parliament with a pistol and two hundred men, firing into the ceiling while the entire elected government dived under their desks. Tanks rolled through the streets of Valencia. For a few hours, Spain’s young democracy hung on nothing at all.

And then King Juan Carlos went on air, in uniform, and ordered the armed forces — his armed forces — to stand down and defend the constitution. By dawn it was over. The colonel surrendered. The tanks went home.

Notice what actually decided it. Not a battle. A broadcast. The men with the guns had already taken the parliament; what they could not take was the story about who was legitimately in charge — and the moment the king’s version of that story reached every barracks in the country, the coup was finished. This is the thing almost nobody understands about how power changes hands overnight: a coup is not a fight. It is a performance. And it is won or lost far less in the streets than in the minds of a few thousand people deciding, in real time, which way to jump.

A Coup Is Coordination, Not Conquest

For all its drama, a coup is a surprisingly small event. The plotters almost never have the strength to fight the whole state — so they don’t try. What they’re really running is a coordination game.

Put yourself in the boots of a single officer on the morning of a coup. You wake to an announcement that the government has fallen and a new authority is in charge. You have no idea whether it’s true. What you do next depends almost entirely on what you think everyone else will do. If you believe the coup has already won, resisting is both suicidal and pointless, so you comply — and your compliance becomes one more reason for the next person to comply. If you believe it will fail, resisting is safe, even heroic. The plotters’ whole job is to tip that calculation, across thousands of people, in the first few hours: to manufacture the belief that it’s already over, so that it becomes over.

The political scientist Naunihal Singh calls this “making a fact.” You don’t defeat your opponents; you convince them they’ve already lost. Which is why the first thing coup plotters reach for, again and again, is not the barracks or the treasury. It’s the radio station.

It is the same machinery The Longest Con described empires using across whole centuries — manufacturing the belief that their rule is natural, total, inevitable — only compressed into a single violent night.

The Playbook

Strip away the local details and coups tend to follow the same sequence, in roughly the same order. Analysts have watched it repeat across continents and decades.

Every step serves one master goal: to make the new reality look inevitable before anyone can test whether it’s real. Step four is the heart of it. The announcement isn’t a report on the coup’s success — it largely is the coup’s success. A calm voice on the national frequency saying order has been restored does more work than a tank, because it tells every waverer that the decision has already been made and they are simply late to it.

And step five is where most coups are actually settled. The plotters rarely need the whole army to rise with them; they need it not to move against them. An army that stays in barracks “to await orders” has, in effect, already chosen the coup — from the outside, neutrality and victory look identical. This is why the loyalty of a handful of key commanders matters more than the raw size of any force. A coup is a whispered negotiation with people who have guns, conducted in the hours before anyone will say out loud which side they’re on.

Why Coups Fail

If manufacturing inevitability is how coups win, then anything that punctures the illusion is how they lose. Strip away the specifics and almost every failure comes down to two questions.

The deadlier of the two is the rival broadcast — a competing voice with enough legitimacy to give the undecided both a reason and permission to refuse. That is exactly what the Spanish king was in 1981. It is what Charles de Gaulle was in 1961, when four generals seized Algiers and de Gaulle went on French radio in uniform; conscripts pressed transistor radios to their ears in the desert, heard their president forbid them to follow the rebels, and simply declined to march. The generals held the city and lost the country in four days, because they never held the story. And it is what Boris Yeltsin became in 1991, climbing onto a tank outside the Russian parliament while the men who had seized Moscow gave a televised press conference so nervous that one plotter’s hands visibly shook. The tanks were real. The authority behind them was not — and everyone watching could see it.

Line the three up and the pattern is almost mechanical: each seized the buildings and lost the broadcast.

What a Coup Reveals

There’s a deeper reason all of this works the way it does, and it’s the thread running through this whole series. A coup is the most naked demonstration you will ever see of what political power actually is. Not the palace, not the treasury, not even the guns — but the shared belief about who has the right to give orders. The Longest Con watched empires build that belief over centuries; How Empires Die watched it corrode over decades. A coup is the same belief seized in a night, at gunpoint, by people betting they can rewrite it faster than anyone can stop them.

Two honest cautions before you treat this as a universal key. First, the machinery is morally blank: the same playbook has toppled genuine tyrants and strangled fragile democracies, and the anatomy alone tells you nothing about which. Second, the “coup as pure perception” model can be pushed too far. Raw force still counts — sometimes the side with more guns simply wins — and so do the unglamorous structural facts. Coups cluster in poorer countries with weak institutions and a history of previous coups, a grim pattern political scientists call the “coup trap.” A country’s vulnerability is built long before the night in question. The choreography decides who wins on the night; the conditions decide whether there is a coup to begin with.

The Spell and How It Breaks

Still, the choreography is where the lesson lives, and it’s a strangely hopeful one. Because a coup runs on manufactured belief, it is only ever as strong as that belief — which means it can be undone by anyone who breaks the spell before it sets. A king with a television. A president with a radio. A crowd that fills the square instead of staying home. In each case the counter-move was not a bigger army. It was a better claim on the same story, delivered before the plotters’ version could harden into fact.

Which returns us to the quiet truth underneath all three of these pieces. A coup is simply someone trying to rewrite who’s in charge before the rest of us wake up and decide whether to believe them. Sometimes it works. It works exactly as often as we let it.