A week ago, Guinea had an elected president.
Today, it is governed by soldiers.
On 5 September, gunfire erupted around the presidential palace in Conakry. Hours later, Colonel Mamady Doumbouya appeared on television surrounded by members of Guinea’s special forces and announced that President Alpha Condé had been removed from power. The constitution was dissolved, the government was dismissed and Guinea’s borders were temporarily closed.
Condé was detained.
Doumbouya justified the coup using language that has appeared repeatedly throughout political history: corruption, poverty, institutional failure and the need to save the country from a government that had stopped serving its people. Quoting former Ghanaian leader Jerry Rawlings, he argued that if people are crushed by their elites, it becomes legitimate for the army to intervene.
There is an obvious democratic response to this. Soldiers do not get to decide that they dislike a government and therefore replace it. Guinea held an election less than a year ago. Whatever its problems, political leaders are supposed to be removed through political institutions, not rifles.
Except there is a complication.
Alpha Condé had already manipulated those institutions to remain in power.
That does not make the coup democratic. But it does make the story much more interesting than democracy being suddenly destroyed on 5 September.
Guinea’s democratic crisis began before the soldiers arrived.
Condé’s own political history makes that particularly striking. For decades, he was an opposition figure challenging authoritarian government. He was imprisoned under previous regimes. Then, in 2010, after years of dictatorship and military rule, he became Guinea’s first democratically elected president.
That should have been an extraordinary democratic achievement. A country accustomed to leaders obtaining power through authoritarian rule had transferred authority through an election.
Condé won another term in 2015. Under the existing constitutional framework, however, he faced a two-term limit. Instead of preparing to leave office, he promoted a new constitution in 2020. The government argued that Guinea needed constitutional reform. Opponents argued that the real purpose was much simpler: resetting the presidential term count so Condé could run again.
He did.
The process produced months of protests and violence. Opposition groups boycotted the constitutional referendum. Condé then contested the October 2020 presidential election and was declared the winner with roughly 59% of the vote. His principal opponent, Cellou Dalein Diallo, rejected the result, and further unrest followed.
So when soldiers detained Condé last Sunday, they did not remove the same democratic reformer who entered office in 2010. They removed a president whose willingness to alter constitutional arrangements in order to remain in power had already damaged the democratic legitimacy he once represented.
This creates an uncomfortable distinction that I think matters far beyond Guinea: elected government and democratic government are not necessarily the same thing.
Elections answer one question: how did somebody obtain power? Democracy also asks what they do with it afterwards. A leader can win a legitimate election and subsequently weaken courts, manipulate electoral rules, restrict opponents or alter constitutional limits. If we define democracy only by whether elections occur, an elected government can gradually dismantle democracy while continuing to claim democratic legitimacy.
But recognising that problem creates an extremely dangerous temptation. If the president has already violated democratic principles, why shouldn’t the military remove him?
Because then somebody has to decide when democracy has deteriorated enough to justify suspending democracy.
And if the answer is the army, democracy has acquired a much larger problem.
Doumbouya’s argument is attractive precisely because some of his criticisms are credible. Guinea is extraordinarily rich in natural resources, particularly bauxite, yet much of its population remains poor. Corruption and weak institutions have been persistent problems. Condé’s third-term project created genuine anger. There are Guineans celebrating his removal rather than mourning it.
It would therefore be too easy to describe everyone celebrating in Conakry as anti-democratic. Many may believe the coup removed a president who had already betrayed the democratic rules under which he was originally elected.
But public support cannot transform a military seizure of power into a democratic transfer.
That is the same distinction I encountered in Tunisia in July: popular does not automatically mean democratic.
Democracy is partly designed for moments when people desperately want an immediate political outcome. Constitutions deliberately make some things difficult because political enthusiasm can be temporary while institutional changes last much longer.
If a sufficiently popular military coup becomes legitimate simply because the government it removes is unpopular, every future general gains the same argument.
The military does not need to say, “We want power.”
It says, “The politicians failed, so we reluctantly had to take responsibility.”
That sentence has introduced an enormous number of dictatorships.
The problem is not necessarily that Doumbouya is lying about his intentions. He may genuinely believe intervention was necessary. The problem is that political institutions cannot depend entirely on the personal virtue of the person who has just abolished them.
A benevolent soldier with unlimited power and an ambitious soldier with unlimited power possess exactly the same institutional safeguards:
none.
This week, the Economic Community of West African States, ECOWAS, responded by suspending Guinea and demanding a return to constitutional order. That reaction matters because Guinea’s coup is not occurring in isolation. Mali experienced military coups in August 2020 and again this May. West African governments therefore have a strong incentive to prevent military intervention from becoming politically normal.
Coups can spread through demonstration effects. Not because one colonel watches another coup and simply copies it, but because successful coups change expectations. If officers elsewhere observe soldiers overthrow a government, face limited consequences and eventually receive international acceptance, the perceived cost of attempting the same thing falls.
ECOWAS therefore has to think beyond Guinea. Punishing the coup is partly about changing the calculations of soldiers who have not yet attempted one.
But regional pressure creates another problem. What exactly should ECOWAS defend: the previous constitutional order, or constitutional government as a principle?
Restoring Condé personally would be difficult to present as an uncomplicated victory for democracy when his third term was itself the product of a deeply contested constitutional change. Yet allowing the junta simply to remain in power risks signalling that soldiers can resolve constitutional disputes by removing civilian governments.
The better objective is therefore probably not restoring the political situation of 4 September exactly as it existed. It is creating a credible route towards a new civilian political order.
That means the transition matters more than the coup leaders’ speeches.
Doumbouya has begun meeting political parties, civil-society groups and business representatives and has promised consultation over Guinea’s future. Those steps may be positive. They are not sufficient evidence that the military intends to leave power.
Military governments almost always describe themselves as temporary.
The useful questions are more concrete. Will there be a clear transition timetable? Who writes the rules for the next election? Can Doumbouya himself run? Will opposition parties operate freely? Will journalists be able to criticise the junta? Will independent institutions exist before elections occur? Most importantly, when civilians are eventually elected, will the military actually return to its barracks?
That last question determines whether Guinea has experienced a temporary rupture or the beginning of another political system.
There is also an economic dimension that makes Guinea particularly important internationally. The country possesses the world’s largest known reserves of bauxite, the principal ore used to produce aluminium, and is one of its largest exporters. China is a major purchaser.
That means political instability in a country many people outside West Africa rarely discuss can affect global industrial supply chains.
It is another reminder of how misleading political maps can be. Guinea may have a relatively small economy compared with the United States, China or European states, but control over strategically important resources creates influence that GDP alone does not capture.
Natural-resource wealth also creates a more difficult domestic question. Guinea possesses extraordinary mineral resources, yet resource abundance has not automatically produced broad prosperity.
Economists sometimes describe versions of this problem as the resource curse. Valuable natural resources can generate government revenue and foreign investment, but they can also encourage corruption, rent-seeking and political competition over control of the state. When governments can obtain substantial revenue from extracting resources rather than taxing citizens, the relationship between rulers and citizens can change as well.
Resources do not doom countries to bad government. Botswana is an obvious African counterexample to any deterministic version of that argument. Institutions determine how resource wealth is managed.
That brings Guinea back to the same problem.
The country does not principally lack valuable assets. It lacks political institutions trusted enough to determine how power and resources should be distributed without every major dispute threatening the system itself.
Condé’s constitutional manipulation weakened those institutions.
The coup has weakened them further.
This is why I don’t think the correct choice is between defending Condé and defending Doumbouya.
Both positions personalise a fundamentally institutional problem.
Guinea needs rules strong enough that it does not matter whether the president wants a third term or whether a colonel believes he could govern better. The president should be unable to rewrite constraints merely because leaving office is inconvenient. The military should be unable to abolish civilian government merely because politicians have violated those constraints.
Democracy requires both.
Civilian leaders must obey constitutional limits.
Armed forces must remain subordinate to civilian politics.
If either principle disappears, elections become much less meaningful.
There is an irony here that appears repeatedly in politics. Leaders weaken institutions because institutions prevent them from doing what they want. In doing so, they also weaken the institutions that protect them.
Condé wanted to remain president beyond the original two-term limit. Changing the constitutional order helped him accomplish that.
But when soldiers eventually arrived at his palace, the legitimacy of the constitutional order he needed everyone else to defend had already been damaged partly by his own actions.
That does not justify what happened to him.
It demonstrates why institutional constraints protect leaders as well as restrain them.
A president who accepts a term limit loses power.
A political system in which presidents accept term limits gains something much more valuable: the expectation that losing power is normal rather than existential.
That expectation makes peaceful transfers possible.
And once politicians destroy it, the question of who governs increasingly stops being answered by rules and starts being answered by whoever possesses enough power to impose an answer.
Guinea’s soldiers say they intervened because the political system had failed.
Perhaps parts of it had.
But abolishing the constitution is not evidence that constitutional government has been repaired.
The real test will come when the military is asked to do something much harder than seizing power:
give it back.
