A Balkanised Africa: The Arithmetic

(Your inaction is financing their freedom.)

I. The joke is on us

When a European says "So in Africa…", we are quick to interject and correct them, "Africa is not a country!" we say, with the small satisfaction of the well-schooled. He apologises. He moves on. He buys our cobalt at ten dollars a kilogram and sells it back to us at two hundred, embedded in a phone.

We won the sentence. He won the century.

I want to be careful here, because I am about to spend many thousands of words arguing something that sounds like the opposite of that correction. Africa is not a country. That is true, and it is trivially true, and it is the least interesting true thing anyone says about this continent. Europe is not a country either. Europe has a parliament, a court whose rulings override national law, a common external tariff, a currency shared by twenty states and a passport that lets a Portuguese woman take a job in Helsinki on a Tuesday. Nobody thinks Europe is a country. Everybody notices that Europe acts like one when money is on the table.

So when we deliver our little correction, what exactly are we defending? The fifty-five flags? The forty-odd currencies? The visa I need to visit a neighbour whose grandmother shares a language with mine?

There is a version of this conversation that begins with a map and ends with a sigh. Fifty-five states. Nearly one and a half billion people. A landmass that could swallow China, India, the United States and most of Europe with room left over. And roughly two per cent of global trade.

But the map is the symptom, not the disease. The real question is not why doesn't Africa unite. It is who benefited from Africa not uniting, and who is still collecting.

Because fragmentation is not weather. It did not happen to us. It was manufactured, at a specific address, on a specific date, by people who wrote down what they were doing — and then, sixty years later, ratified by people who looked exactly like us, using words we taught our children to be proud of.

This is an essay about that second group.

II. The word that gave it away

I had a small, embarrassing revelation recently, and I am going to lead with it because embarrassment is the only credential I trust.

Boundary. Border. We use them interchangeably. Both mean to limit. A boundary is a thing that stops you. A border is the edge past which you may not go without permission, paperwork, and a fee.

Now set those beside the words we use for the other thing. Union. United. Community. Words that mean to join, to bring in, to make access easier.

Nobody hid this. It is sitting in the dictionary. It is sitting in the name of every institution on both sides of the argument. And I — a woman who would describe herself, with the usual modesty, as learned — only stumbled onto it in my thirties, in the middle of a paragraph, like someone tripping over a suitcase in her own hallway.

Which taught me the operating principle of this entire essay: the most effective place to hide an intention is casually in the open. Not in a vault. In a tariff schedule. In a footnote. In a word so ordinary you stopped seeing it in primary school.

Hold that thought. We will come back to it with receipts.

III. The manufacture: Berlin, 1884

In November 1884, European powers convened in Berlin under Otto von Bismarck to set the rules for partitioning Africa. Germany, France, Britain, Belgium, Portugal, the Netherlands and Spain attended. No African leaders were present. Not one. Not as observers. Not as witnesses. Our grandfathers were the subject of a meeting they were not told was happening.

The conference did not itself draw most of the lines — it established the framework that governed their creation over the following three decades. The critical provision was effective occupation: to claim a territory, a European power had to physically hold it. That produced a scramble, and the scramble produced lines.

The geographer Ieuan Griffiths put it plainly: Africa's boundaries were drawn by Europeans, for Europeans, with scant regard for Africa, let alone for Africans.

Sit with the human arithmetic. The Ewe were split between Ghana, Togo and Benin. The Somali were divided among five colonial entities. The Luo were split between Kenya, Uganda, Tanzania and Sudan. The Afar were divided among Ethiopia, Eritrea and Djibouti.

Julius Nyerere named the consequence without decoration: Africans inherited artificial nations carved out at Berlin, then spent independence struggling to build those artificial units into stable societies — with the continent, in his words, in danger of becoming the most Balkanised in the world.

Two structural facts follow, and both matter enormously for everything that comes after.

First, the borders did not only divide — they also combined. Peoples with no shared political history, sometimes with histories of open conflict, were enclosed inside a single administrative unit and informed that they were now a nation. What emerged at independence was not a nation-state. It was a colonial administrative container with a flag stapled to it.

Second, the economic geography was built to extract, not to connect. Rail ran from mines to ports, never city to city. Roads served the coast, not the interior. To this day it is easier to move goods from Nairobi to London than from Nairobi to Lagos — because the infrastructure was never designed for Africans to trade with each other. It was designed for Africa to trade with Europe.

That is the inheritance. Not merely bad lines on a map. An entire physical and financial architecture pointed outward, like a room where every door opens into the corridor and none into the next room.

And then we took that architecture, kept it exactly as it was, and called it freedom.

IV. But wait — did they really divide us?

Here is where I have to interrupt myself, because if I do not, someone smarter and less sympathetic will do it for me, and they will be right.

We tell the division story as though Berlin cut through a unified body. It did not. Before 1884 there was no African polity. There was Ashanti and Buganda and Ethiopia and the Sokoto Caliphate and a thousand others, and several of them made war on each other with enthusiasm. We spoke — we still speak — more than two thousand languages. The Zulu and the Kikuyu were not one people waiting for a passport. Ethiopia was never colonised at all and has spent the intervening century looking nothing like a candidate for surrendered sovereignty.

So no, they did not carve a nation into fifty-five pieces. There was no nation.

But that is not the argument, and pretending it is, is the cheapest move in this entire debate.

Europe was not one people either. The Germans and the French spent a thousand years killing each other and got very good at it. Nobody argues that the European Union rests on a shared Frankish soul. It rests on a calculation: that coal and steel administered jointly in 1951 made a fourth Franco-German war mechanically difficult, and that a common market was worth more to each of them than a national tariff.

Unity is not a claim about ancestry. It is a claim about arithmetic.

Africa is not owed a union because we were once one. Africa needs a union because fifty-five separate negotiating counterparties facing four coordinated buyers is a losing configuration in any bargaining problem ever studied, and it stays losing no matter how many languages the fifty-five speak at home.

Keep the romance if you like it. I am arguing the balance sheet. The balance sheet is worse, and it does not require you to believe anything sentimental.

V. The argument we already had — and lost

Here is the part of the story most people skip, and it is the most important part.

Africa had this argument. Properly. At the highest level. At the moment of maximum possibility. And the federalists lost.

By the early 1960s two camps had formed.

The Casablanca Group — Ghana, Algeria, Egypt, Guinea, Libya, Mali, Morocco — argued for political integration first, on the theory that economic integration without political union would simply be captured by the former colonial powers.

The Monrovia Group — led by Liberia's William Tubman and Côte d'Ivoire's Félix Houphouët-Boigny — wanted a loose confederal arrangement and were prepared to maintain strong economic ties with the very powers they had just left.

Kwame Nkrumah led the Casablanca position and was not subtle. In I Speak of Freedom (1961) he warned that allowing the fragmentation to continue might well prove disastrous for all of them, noting that of the twenty-eight African states then existing, no fewer than nine had populations under three million. His whole political theology compressed into a line: Seek ye first the political kingdom, and all things shall be added unto you.

He wanted a continental government. An African monetary authority. An African High Command — a single continental defence force. One market area.

He lost the vote before it was taken. At Monrovia in May 1961, Houphouët-Boigny and Tubman carried 22 of the 27 independent African states — many of them Francophone countries that had opted to remain inside the Paris-backed currency zone.

Read that parenthetical again, because it is the whole essay in miniature.

The states voting against political union were, in large part, states whose currency was already administered from Paris.

The monetary architecture of dependency was voting on whether to dismantle dependency. It voted no.

That is not a metaphor. That is a procedural fact. Ask a room full of tenants to vote on abolishing landlords, but only give ballots to the ones whose rent is being quietly paid by the landlord, and you do not need to rig anything. You just count.

The 1963 Addis Ababa Charter founding the Organisation of African Unity was the compromise, and it bore the clear hallmark of the Monrovia position: conciliatory rather than militant, conservative rather than revolutionary — and, most consequentially, it compromised on the one issue that would have marked a genuine rupture from the colonial past: the inherited colonial boundaries.

Emperor Haile Selassie, hosting, tried to steer between the camps, naming Nkrumah's political union as the eventual destination while embracing the gradualism of the majority. Nkrumah's warning at Addis was that delaying continental government would produce fragmentation and strengthen neo-colonialism. Sixty-three years on, that reads less like rhetoric and more like a forecast filed early.

And Nyerere identified the trap that would define the next half-century: the OAU Charter spoke for Africans under colonialism and racial domination — but once countries won freedom, the charter stood for the protection of heads of state.

That sentence deserves rereading.

The institution built to liberate Africans became the institution that protected the people governing them. Sovereignty — the sacred principle, the word we put in the anthems — became, in practice, a shield against accountability.

And here is the bitterest irony, the one that should stop every pan-Africanist mid-sentence. Legal scholarship on this period argues that Nkrumah — ideologically the most committed pan-Africanist of the lot — played a pivotal if unintentional role in entrenching colonial-era borders, because in accepting the OAU framework he accepted inherited boundaries as the basis of the new order.

The prophet of unity signed the partition. Not from treachery. From the belief that he could fix it later.

There is no later. There has never been a later. Sixty-three years of later.

VI. Were they stupid, or were they bought?

Twenty-two of twenty-seven. Every time I return to that number I want to explain it away with the kind explanation: they were afraid. Comfort versus risk. Human nature. They had just come out of one arrangement and were not eager to enter another before the flags had finished rising.

Perhaps they could not see their own value. Perhaps four decades of being told your continent was a burden makes it hard to believe you are a bargaining position. Perhaps that is simply what ignorance of one's own capacity looks like from the inside.

I no longer find that explanation adequate, and I do not think it is generous. I think it is condescending.

Because look at who voted, and look at what they kept.

They were not stupid. They were correct — about their own interests. A head of state in 1961 with a Paris-guaranteed currency, a French garrison, a marketing board, and personal control over an export licence was being asked to trade all of that for a seat on a continental council where he would be one voice among thirty. In exchange, his people would get a larger market. He would get a smaller kingdom.

He did the sum. He voted accordingly. He was right.

And here is the thing I need you to hold onto for the next ten thousand words, because everything else in this essay hangs on it:

The interests of an African state and the interests of the man running it are not the same thing, and the entire architecture of fragmentation survives in the gap between them.

Was it by design? Partly, yes — the currency zone was designed, the marketing boards were designed, and we will get to Togo in a moment, where a president was killed two days before a signature. But conspiracy is the lazy half of the answer. The durable half is that you do not need to buy a man who already wants what you are selling.

External powers did not have to prevent African unity in 1963. They only had to make sure the people voting on it had more to lose than to gain. And then wait.

They are still waiting. They are very comfortable.

VII. What the Americans actually did — and which eleven states said no

Before you judge the direction this essay is heading, a question: did you know that the United States of America was not assembled voluntarily?

The US did not become a single market by agreement. It became one through a constitutional convention in 1787 that replaced a failed confederation — and then, when eleven states attempted to leave, through a civil war that killed roughly 600,000 people and settled by violence the question of whether membership was optional.

The Commerce Clause, and two centuries of federal court enforcement, is what actually created the American single market. Not goodwill. Enforcement. A state cannot raise a tariff against its neighbour because a federal court will strike it down and a federal government will make the ruling stick.

The lesson for Africa is uncomfortable: the American single market was not built by asking states nicely to lower barriers. It was built by removing states' ability to raise them.

Now — the sharpest question in all of this, and I promised myself I would answer it. Which eleven states, and what were they actually afraid of?

Seven left first, before Lincoln was even inaugurated: South Carolina, Mississippi, Florida, Alabama, Georgia, Louisiana, Texas. Four followed after Fort Sumter: Virginia, Arkansas, Tennessee, North Carolina.

Here is what matters, and it demolishes the assumption I arrived with.

They were not the poor states afraid of being leached by the union. They were the rich ones. Cotton was roughly sixty per cent of US exports in 1860. By many estimates the enslaved population constituted the single largest asset class in the United States — worth more than the country's railroads and factories combined.

They did not secede because they were weak and feared subsidising the strong. They seceded because they were wealthy, and their wealth depended entirely on a property regime that a continental authority might one day vote to abolish. Their fear was not of leaching. Their fear was of enforcement.

Do you see it?

So run the mirror. When African states resist a supranational court, a common external tariff, a continental monetary authority, enforceable rules on elections — what is the fear? Is it that Malawi will drain Nigeria? Is it that a poor member will leach a rich one?

No. The fear is enforcement, and the thing being protected is not the state. It is the arrangement.

And that answers the second half of my own question — who risks being leached, the country or the chaperons?

The chaperons. Always the chaperons. A common external tariff does not cost Uganda. It costs the man who controls who gets the import waiver. A continental court does not cost Cameroon. It costs the person for whom the national court is presently a convenience. Free movement does not cost Kenya. It costs whoever is selling the permit.

Unity is not expensive for Africa. Unity is expensive for approximately four hundred people, and they are the four hundred people who would have to sign it.

VIII. So should the enlightened do it by the gun?

I promised to put my own worst thought on the table, so here it is.

If a people cannot see the advantage of unity, should those who can see it impose it? I am not a promoter of war. But where discussion fails, a strong arm has historically prevailed. The Americans did not debate their way to a single market. They shelled their way to it and then wrote the court decisions afterwards.

I want to make the case honestly before I dismantle it, because a provocation you refuse to examine is just a mood.

The case: sixty-three years of summits have produced an agreement most members have not operationalised and an air transport market signatories decline to honour. Every incremental path is designed to be non-threatening, and it is non-threatening precisely because it changes nothing. Meanwhile a child in Kolwezi washes cobalt by hand for a dollar a day so that a battery can be assembled in Shenzhen. At some point patience becomes complicity.

Now the dismantling.

First: Africa has already run this experiment, more than two hundred times. Since 1960 this continent has had well over two hundred coups and attempted coups. Force is not the road not taken. Force is the road we have been on. And the men with guns have not, as a class, delivered federation. They have delivered Mobutu. They have delivered Blaise

Compaoré, who shot Thomas Sankara and then governed for twenty-seven years. The gun does not select for Sankara. It selects for whoever shoots Sankara. That is not a moral objection. That is an engineering objection, and it is fatal.

Second: read the American case properly. Six hundred thousand dead settled that membership was not voluntary. It did not settle that the federation was just. A century of segregation followed. If your model requires a civil war and then a hundred years of unfinished business, you have not found a shortcut. You have found a longer road with a mass grave at the entrance.

Third, and this is the one that should stop us: the claim "the people cannot see it, so we must act for them" is precisely the sentence Berlin used. It is the civilising mission with the pronouns swapped. When Afrobarometer surveys tens of thousands of people across thirty-four African countries, the finding is not that citizens crave strongmen who stay. It is the opposite: in principle, leaders who remain beyond two terms are not what people want. The same research notes that pressure on long-serving presidents comes not primarily from foreign capitals but from African citizens who watch these men frustrate development and govern for a small elite.

So the premise is false. It is not the people who cannot see it. It is the people who cannot see it who happen to be holding the pen.

Which turns the whole provocation inside out. The problem was never a population too dim to want a continent. The problem is a governing class with a rational, bankable, entirely comprehensible interest in there not being one.

You do not need a gun for that. You need the one thing this continent has systematically been prevented from having: enforcement that a head of state cannot switch off.

And — one uncomfortable footnote, which I will return to later — the only governments currently attempting a genuine monetary rupture are juntas. Mali, Burkina Faso, Niger. Soldiers, not elected governments, put exit on the table. That should trouble anyone who wants both sovereignty and democracy, and it should trouble me most of all, because it is the closest thing to evidence my provocation has.

I am going to sit with that discomfort rather than resolve it. Anyone who resolves it quickly is selling something.

IX. The arithmetic of fragmentation

Enough philosophy. Numbers.

Africa's GDP was roughly $2.8 trillion in 2024 — under 3% of world GDP and about 2% of global trade, from a continent of nearly 1.5 billion people.

But the external number is the distraction. The structural problem is internal.

Intra-African trade accounts for only about 16% of Africa's total trade. Over half the continent's imports and exports are tied to just five economies, every one of them outside Africa.

Compare the regions we insist on being measured against. Intra-regional trade runs at roughly 67% in Europe. 61% in Asia. 47% in the Americas.

Africa trades with everyone except itself. That is the fragmentation, expressed as a number.

There has been real movement — intra-African trade rose 12.4% to about $220 billion in 2024, rebounding from a contraction the year before. Hold the applause. $220 billion is roughly the annual revenue of one large multinational corporation. That is not a continental market. That is a rounding error with a communiqué attached.

The aviation picture, or: why you have flown through Dubai to reach your neighbour Aviation is where fragmentation becomes physically legible to anyone who has tried to book Lagos–Nairobi.

Intra-African fares run about 60% higher than comparable European routes. The mechanics are not mysterious. Passenger taxes, charges and fees for regional departures in Africa average $68, against $32 in Europe and $34 in the Middle East. Passenger Service Charges within Africa frequently exceed $50–$100 per international departure — and in many cases the taxes and fees exceed the base fare itself.

The structural cause is that many African states still operate restrictive bilateral agreements limiting flight frequencies, which suffocates the competition that produces European low-cost pricing. A fragmented air market functions as a chain of isolated islands rather than a network.

The African Union's answer is the Single African Air Transport Market, launched in 2018. Progress is real and partial: 38 states have signed, and the share of intra-African routes operating under fifth-freedom traffic rights rose from 14.5% in 2018 to 23% by 2026. But the persistent complaint from airlines is that some countries that have signed SAATM still refuse the traffic rights needed to actually fly.

That gap — between signature and compliance — is the entire African integration story in one sentence. Write it on the wall of every summit venue.

Visa liberalisation has genuinely accelerated: Rwanda and The Gambia have achieved full openness, with Kenya, Ghana, Namibia, Zambia, Zimbabwe and Malawi among the most improved. But as Nairobi immigration lawyer George Mucee observes, visa-free policy needs matching integration elsewhere — Kenya removed its visa regime and then introduced a $30 electronic travel authorisation, since waived for 52 African passports.

Open the border. Keep the $600 airfare. You have opened nothing. You have built a door in a wall and left the wall.

AfCFTA at five: the market we have agreed to and refuse to use And now the flagship, as of this month.

Of the fifty-four signatories, fifty have deposited their instruments of ratification as of September 2026, with Somalia approved but not yet complete, and Benin, Libya, South Sudan and Sudan still to approve at all. Eritrea has not signed. Magnificent. A continental market on paper, near-universally endorsed.

Now the operational reality, which is where the bodies are buried. By July 2025, forty-nine provisional schedules of tariff concessions had been submitted and forty-eight adopted — yet only seventeen states had filed complete schedules; twenty-four had gazetted notices confirming they were actually ready to trade under AfCFTA, a figure raised to twenty-five in early 2026. A shipment moving between two operational countries can claim preference today. A shipment moving between an operational and a non-operational country cannot — even if both have ratified.

And the number that should be read aloud at the next AU Assembly: by mid-2025, participating nations had issued 8,561 certificates of origin — the documents that actually let goods cross a border under preferential terms.

Eight thousand five hundred and sixty-one. For 1.4 billion people. Over several years. A single mid-sized port in Europe processes more paperwork before lunch.

There is genuine movement, and I will not pretend otherwise: the Guided Trade Initiative has grown from eight founding countries to around thirty-nine by mid-2025, and more than forty countries now issue certificates of origin; in February 2026 heads of state approved harmonised rules of origin for vehicles and components, requiring 40% African content to qualify as "Made in Africa." Dedicated air cargo corridors under the initiative have cut freight costs by 50% to 75% into East and Southern Africa.

Fine. Good. Real. And still: five years after trading commenced, roughly a quarter of the continent can actually use the free trade area it signed.

We did not fail to build the thing. We built it, framed the signing photograph, and then declined to switch it on.

X. Where the money actually goes

This is the heart of it. This is where the abstraction becomes a wage.

Africa holds an estimated 30% of the world's critical minerals — roughly 75% of global cobalt production, 62% of manganese, plus major lithium, copper and rare earth deposits. Revenues from just four minerals — copper, nickel, cobalt and lithium — are projected at $16 trillion over the next twenty-five years.

And African countries currently collect only about 40% of the potential revenue. Less than 5% of critical minerals are refined on the continent.

Follow one kilogram

Cobalt:

Raw ore: ~$10/kg
Refined: ~$60/kg
Battery cathode: over $200/kg

A tonne of Zimbabwean lithium:

As spodumene concentrate: ~$900–1,500
As battery-grade lithium carbonate: $12,000–20,000
Embedded in a finished EV battery: over $150,000 equivalent

UNCTAD documents roughly a hundred-fold increase in value occurring almost entirely outside the continent.

The DRC produces around 70–80% of the world's cobalt and earns roughly $1 billion from it — while the lithium-ion battery market that depends on that cobalt exceeds $60 billion annually. Mining captures about 5% of total value chain revenue. Cell manufacturing and pack assembly capture 40%.

China processes 65% of the world's lithium, 74% of its cobalt and 90% of its rare earths. Chinese state-owned enterprises and policy banks control roughly 80% of the DRC's total cobalt output.

And it is not only the minerals

The obvious question is whether this pattern runs across sectors — minerals, coffee, everything. It does, and the agricultural numbers are, if anything, more humiliating, because nobody can claim a coffee roastery requires a nuclear-grade power grid.

Coffee. In 2024 Germany exported more than 473,000 tonnes of coffee valued at €6 billion. Germany earned about $5.9 billion from coffee exports in 2023; all of Africa earned roughly $3.6 billion in 2022. Germany does not grow coffee. Ethiopia — the literal birthplace of the plant, the place the word coffee comes from — is outbid in its own inheritance by a country whose climate would kill the seedling.

The per-bag figures make the mechanism explicit. Between 2000 and 2010, Ethiopia, Uganda and Côte d'Ivoire received an average of $138, $71 and $68 per bag of coffee exported respectively, while Switzerland — Europe's most profitable re-exporter — earned over $700 per bag. In 2009 the International Coffee Organisation found coffee added $31 billion to the economies of the nine leading importing nations — twice the total export earnings of every coffee-producing nation on earth.

That is not a market outcome. That is a toll booth.

Tea. Kenya. Kenyan tea averaged USD 2.28 per kilogramme across the first twenty-four auction sales of 2026; roughly 186.2 million kilos were offered between January and June, earning about KES 55 billion. Now do the arithmetic nobody does at the auction house. A kilogram of tea yields somewhere north of four hundred cups. Four hundred cups, sold in London or Cairo or Karachi at anything from thirty cents to three dollars, against $2.28 paid at Mombasa for the whole kilo.

It is said that Kenyan farmers see less than one per cent of what their labour and land are ultimately worth, and that this approaches ownership rather than employment. Measured against retail value in the consuming market, that is not rhetorical excess. It is roughly the right order of magnitude.

Cocoa. We will come to cocoa in a moment, because cocoa contains the single most important experiment in modern African economic history, and almost nobody talks about it.

Gold — and here is the one the mainstream has not picked up, the one I circled for a long time without being able to name it.

The question is this: why do the Chinese, or anyone else, mine our gold and export it raw? Why not build the refinery here?

Part of the answer is the ordinary one: power, capital, scale. But the larger part is a gate almost nobody outside the trade knows exists. A gold bar is not simply gold. To trade at full value in the global market it must come from a refinery on the London Bullion Market Association's Good Delivery List. That accreditation is administered in London, requires years of throughput history, audited provenance, and minimum annual volumes. You can build a technically excellent refinery in Accra or Kampala tomorrow. If your bar is not Good Delivery, the major buyers will discount it or decline it, and your refinery becomes an expensive shed.

So: is there a memorandum between the outside powers? A secret agreement to keep processing offshore?

No. And that is so much worse.

XI. There is no memorandum. There is a tariff schedule.

Here is where my earlier revelation pays its debt.

The best way to hide an intention is to lay it casually in the open.

The mechanism that keeps value-addition out of Africa is not a conspiracy. It is published policy, sitting in plain sight in the tariff books of the world's richest markets, and it has a boring technical name that guarantees nobody reads about it at dinner: tariff escalation.

It works like this. A wealthy market charges you approximately nothing to import your raw beans, your raw ore, your unprocessed leaf. Then it charges you more if you roast, grind, refine, or package that same product before shipping. The duty rises with every step of processing you perform at home.

Read that as a sentence addressed to an African government, because that is exactly what it is:

Send us the raw material and you may enter free. Add value first, and you will be charged for the privilege.

Nobody had to meet in a smoky room. You simply make raw extraction the profitable option and processing the penalised one, publish the schedule, and let a thousand rational firms make a thousand rational decisions that all point the same direction. The Chinese firm that mines and exports rather than refining is not conspiring. It is reading the price signal you wrote.

And tariff escalation is only the front door. Behind it: rules of origin that disqualify your goods; sanitary and phytosanitary standards written for producers with laboratories you do not have; sustainability and traceability regulations whose compliance costs are trivial for a multinational and existential for a cooperative; accreditation regimes like LBMA Good Delivery; shipping and marine insurance priced by firms in two cities; and commodity exchanges in London and New York that set your price before your farmer wakes up.

None of it is secret. All of it is published. That is the design. A locked door announces itself. A staircase that only goes down does not.

So when someone tells you Africa fails to industrialise because of corruption and poor governance — and I will spend a whole section arguing that corruption is real and catastrophic — ask them also to open the tariff schedule of the market they are lecturing you on behalf of, and read the line for roasted coffee next to the line for green.

Both things are true. Only one of them gets said out loud.

XII. The one time two African states tried it — and what happened

Now the story that should be taught in every economics department on this continent, and is taught in almost none.

In 2018, Côte d'Ivoire and Ghana — who together grow roughly sixty per cent of the world's cocoa — did the thing. They stopped negotiating separately. They founded a joint initiative and agreed to add a premium of US$400 per tonne on every ton exported: the Living Income Differential. They promised farmers a minimum of USD 1,820 per tonne.

The press dubbed it "COPEC" — a snide reference to OPEC. Snide, yes. Also accurate. Two African states had just discovered they were a cartel.

Understand what was at stake. Cocoa farmers in Ghana make roughly a dollar a day; in Côte d'Ivoire, around $0.78. Against a global chocolate industry worth over a hundred billion dollars. The LID was not radical. It was $400 a tonne. It was a rounding error on a Mars bar.

And it was beaten. Comprehensively, and within months.

Only a month after launch, authorities in both countries disclosed that confectionery multinationals were refusing to pay the agreed amount. Companies found their way around it: they paid the $400 on top of the futures price, then significantly reduced the "country differentials" normally paid according to origin — driving the country differential negative for both Côte d'Ivoire and Ghana, so that the targeted farmgate price was never reached. One major chocolate manufacturer bought large volumes through the futures market — a venue used for hedging and speculation rather than actual bean procurement — allowing it to buy less directly from the two countries and avoid the premium entirely. Others negotiated with the governments to bring the floor price down.

The outcome: by April 2021 Côte d'Ivoire had cut its guaranteed mid-season producer price by 25% and set the following season below the target; Ghana held nominal prices with subsidies while inflation ate the real value. In Ghana, producers were still waiting to be paid the differential.

Now. Why did it fail?

The convenient answer is corporate ruthlessness, and there was plenty. But the structural answer is the one that matters for everything else in this essay:

Two is not enough.

Two countries controlling sixty per cent of supply can be routed around — by a futures exchange, by a rival origin, by a discount on a different line of the contract, by simply waiting for one of the two to blink under fiscal pressure. And one of them will blink, because cocoa taxes fund the budget and the budget funds the salaries and the salaries fund the peace.

But scale it. Fifty-five states holding roughly 70% of global cobalt, the majority of the world's platinum group metals, a controlling share of manganese, and — through a continental agreement rather than two exhausted ministries — sixty per cent of the world's cocoa, cannot be routed around. There is nowhere to route to. That is what a bloc is. That is the entire product.

The LID is not a story about how African coordination fails. It is the strongest available proof that African coordination works — and that the only reason it was defeated is that there were two of us instead of fifty-five.

We ran the experiment. We got a clean result. We filed it and went back to signing memoranda.

And it is already happening at gunpoint elsewhere

Lest anyone claim leverage is theoretical: since 2023, at least fourteen African countries have restricted exports of raw or semi-processed minerals. Zimbabwe banned raw mineral exports — and Zhejiang Huayou Cobalt promptly installed a $400-million lithium sulfate refining plant in-country rather than lose access.

In February 2025 the DRC imposed a four-month cobalt export ban — the first major market intervention in its mineral history — and by October 2025 moved to a quota system capping exports at 96,600 tonnes for 2026 and 2027, roughly half the 2024 level, with analysts projecting it would raise the country's cobalt export value by around a quarter.

Read that last clause twice. Export less. Earn more. That is what happens the moment a seller remembers it is a seller.

Farai Maguwu of Zimbabwe's Center for Natural Resource Governance puts the stake plainly: a country exporting raw lithium is enriching China at its own expense, when it should be building its own mine-to-market ecosystem.

Now imagine that exercised by fifty-five states at once, under one harmonised code.

That is the whole argument. Everything else is commentary.

Why fragmentation is the mechanism, not the backdrop

Here is the connection most analyses miss, and it is worth being precise.

Refining is capital-intensive, energy-intensive and scale-dependent.The logic writes itself: lithium mined in Zimbabwe, refined in Zambia using DRC copper and cobalt, assembled into cells in Kenya — all moving tariff-free under harmonised rules of origin. An internal market that insulates the value chain from external shocks. (Allow for me to be as they would say delulu!)

Fragmentation forecloses it. Fifty-five mining cod It needs reliable baseload power, deep infrastructure, technical skills — and, critically, a market large enough to justify the plant.

No single African country is large enough on its own to justify the full battery value chain. A coordinated one absolutely is.

Fifty-five tax regimes, fifty-five sets of customs paperwork, and rules of origin still incomplete in the sectors that matter most. The regional value chain cannot form. So the ore leaves.

And the strategic asymmetry is total: Africa negotiates one country at a time, against buyers who negotiate as coordinated blocs — China's state-linked enterprises, the EU's Critical Raw Materials Act, US strategic minerals policy. A single African state facing a Chinese SOE has essentially no leverage. A continental bloc holding 70% of global cobalt has enormous leverage.

Unity would not end great-power appetite for African minerals. It would convert fifty-five asymmetric bilateral deals into one symmetric bloc-to-bloc negotiation.

That is the entire difference. And it is enormous.

XIII. The currency: poor but wealthy, rich but poor

Now the oldest wound, and the one I return to most often.

Look at France. A nation whose economic weight rests, in part, on the backs of West African states who are, in tandem, rich but poor. Countries sending their reserves abroad, to back another country's balance sheet, and then borrowing — so that they can access their own money— buffoonery (should I state it plainly).

Let us get the mechanics exactly right, because the mechanics are more damning than the slogan.

The CFA franc, created in 1945, remains the currency of 14 countries and roughly 210 million people. Pegged to the euro at a fixed 655.957 to €1. Printed in France. Convertibility guaranteed by the French Treasury. Until recent reforms, member states were required to deposit 50% of their foreign exchange reserves with that Treasury.

It was imposed on France's colonies in 1945. Eighty-one years ago. Sixty-six years after independence.

A November 2024 Sciences Po survey found 95% of West Africans want to leave the system. Ninety-five. There is no policy question in the world on which ninety-five per cent of an affected population agrees, and this one has survived six decades of that agreement.

The Senegalese musician Tièmoko Koné's formulation is the one that circulates: whoever controls a country's currency controls its economy.

And now the part where I argue against myself again, because this essay is worthless if it only flatters its own thesis.

The defenders are not frivolous. The CFA zone has delivered genuinely low inflation and exchange-rate stability in a neighbourhood where Ghana's cedi and Nigeria's naira have suffered severe instability. Fixed convertibility means an Ivorian importer knows next year's price. A Nigerian importer does not. That is not nothing. For a trader, that is close to everything.

But notice the trap — and notice that it is precisely a fragmentation trap.

You cannot quietly leave a shared monetary union the way you cancel a subscription. Exit requires either a coordinated departure by the entire bloc, or a costly unilateral rupture — and a single country that leaves forfeits the common market in the act of leaving.

The system is built so that individual exit is irrational and collective exit requires the coordination that dependency prevents.

That is not an accident of design. That is the design.

Which brings me to the men who tried anyway.

XIV. The list

Whilst drafting this piece, I set out to find and name them. Them — the men who tried anyway, and let me be scrupulous about what is documented and what is inference, because this is exactly the territory where a righteous argument gets destroyed by one overreach.

Sylvanus Olympio. Togo. 13 January 1963.

Olympio was a trained economist and the first president of Togo. In early 1963 he was contemplating leaving the franc zone and creating a Togolese currency — backed by the Deutsche Mark. By balancing between powers, Togo risked offering a model of emancipation to every other former French colony. Jacques Foccart, de Gaulle's man for African affairs, later remarked that Olympio had not been among their friends — noting, by contrast, how much more cordial his relations were with Nicolas Grunitzky, the man installed after the coup (hopefully the irony isn't lost on you!).

He was assassinated on 13 January 1963 — the first military coup in France's and Britain's African colonies — reportedly days before he was due to sign Togo's withdrawal from the CFA zone.

He is said to have declared that he would rather be poor but free than rich but enslaved.

The documented facts: the currency plan, the timing, Foccart's preference for the successor. The claim that Paris or Washington ordered it circulates widely and is not established. I will not assert it. I do not need to. The pattern is loud enough without a signature on it: a man moved to leave the currency, and he did not live to sign.

Patrice Lumumba. Congo. January 1961. Murdered weeks after his country's independence, with Belgian officers present at the execution and a documented American desire to see him gone. Belgium has since formally apologised for its role. He was thirty-five years old, and he had governed for eighty-one days.

His offence was not incompetence. It was that he said out loud, at the independence ceremony, with the King of the Belgians sitting in front of him, what everybody in the room already knew about what had been done to the Congo — and then proposed to run the country's minerals for the Congolese. Dead.

Thomas Sankara. Burkina Faso. 15 October 1987. Four years in power. He renamed the country "Land of Upright People." He sold the government's Mercedes fleet and made the Renault 5 the ministerial car. He vaccinated two and a half million children in a fortnight. He asked, at the OAU in 1987, why African states should honour debts contracted by colonialism, and predicted he would not be there to speak again if the others did not join him.

He was killed three months later — and this is the detail that should never be softened — by Blaise Compaoré. His closest comrade. His brother in the revolution. A fellow African. Who then governed Burkina Faso for twenty-seven years, and was convicted in absentia of complicity in the murder in 2022.

Put those two beside each other and the pattern clarifies into something more uncomfortable than a story about foreigners.

Lumumba was killed with foreign hands in the room. Sankara was killed by a man who had shared his cell, his risk and his ideology. Same motive, two different instruments — and the second instrument is the one that has done most of the work since 1963, because it is cheaper, it is deniable, and it requires no visa.

That is the efficiency of the arrangement. Once the incentive structure is built, it no longer needs Belgians. It grows its own.

And the others, because the list is long and the length is the argument: Félix-Roland Moumié, Cameroonian independence leader, poisoned with thallium in Geneva in 1960, a French secret service agent implicated. Ruben Um Nyobè, killed 1958. Eduardo Mondlane, Mozambique, killed by parcel bomb, 1969. Amílcar Cabral, Guinea-Bissau, one of the finest political minds the continent has produced, shot in 1973. Mehdi Ben Barka, Morocco, abducted in Paris in 1965 and never seen again. Samora Machel, killed in a 1986 plane crash near the South African border that has never been satisfactorily explained.

And Muammar Gaddafi. 2011. Because he deserves a section of his own.

Start with what is not in dispute, because the documented part is stronger than the legend and almost nobody bothers with it.

On 2 February 2009, in Addis Ababa, Gaddafi was handed the chairman's gavel of the fifty-three-nation African Union by Tanzania's Jakaya Kikwete. In his inaugural address he told the assembled heads of state that he would go on insisting their sovereign countries work towards a United States of Africa. He proposed, explicitly, a single African military force, a single currency, and a single passport allowing Africans to move freely across the continent. He conceded in the same speech that they were nowhere near agreement, that they remained independent states, and that responding to the call was their decision to make.

Strip the gold brocade and the sunglasses off that and read the list again.

A single continental defence force. An African monetary authority. One market for people.

That is Nkrumah's 1961 proposal, item for item, delivered to the same institution, in the same city, forty-eight years later. The Casablanca programme had been dead for half a century, and here it was again, being read out by a man most of the room could not stand.

And watch what happened to it — because this is the most instructive scene in modern African diplomacy and it is played, invariably, for comedy.

He talked for hours. It ran past midnight. Museveni eventually walked over and whispered that it was past everyone's bedtime. Gaddafi left remarking that silence gives consent — that unless they openly opposed him at the next summit, the United States of Africa would simply come into being. It did not. The AU works by consensus, and nearly all fifty-three heads of state regarded the proposal as a terrible idea. A week before his death, South Africa's Jacob Zuma expressed relief at the regime's fall, complaining that Gaddafi had intimidated African leaders to gain influence and suggesting the AU would work better without him and his repeated demands for a unitary government (the humour).

Now — one of the writers reporting that midnight scene noted, in passing, the cynics' explanation: that the leaders in the room opposed it because in an Africa with a single army, a single currency and a single passport, their own jobs would vanish. He raised it in order to dismiss it, on the grounds that a thousand years of collective political experience was sitting in that room and they were not all cynics.

I have spent this entire essay building the case that on this specific point the cynics were simply correct, and that the thousand years of experience is precisely how they knew. Experienced men are not worse at identifying their own interests. They are better at it.

Now the gold, and here I am going to be scrupulous, because this is where good arguments are destroyed.

The currency Gaddafi wanted was not to be another paper promise. The proposal was a gold-backed dinar — oil and other exports sold for a currency backed by physical metal rather than the dollar, open to African and Muslim-majority states, and pitched to Francophone West and Central Africa as a direct alternative to the CFA franc. By March 2011, days after the uprising began, the Financial Times and NPR were reporting that Libya's central bank held roughly 143 tonnes of gold — worth about six and a half billion dollars at the time, among the largest reserves on the continent.

What the email speculated. On 2 April 2011, Sidney Blumenthal — a longtime Clinton confidant, then working for the Clinton Foundation and not employed by the State Department — sent Hillary Clinton a memo titled, in essence, France's client and Qaddafi's gold. It has since been released publicly.

Its claims, as reported, were these. That Gaddafi's government held 143 tonnes of gold and a comparable quantity of silver, sourced to individuals with access to advisers around Saif al-Islam Gaddafi. That the metal had been accumulated before the rebellion and was intended to establish a pan-African currency based on the Libyan gold dinar. That the plan was explicitly designed to give Francophone African countries an alternative to the CFA franc. That French intelligence discovered the plan shortly after the rebellion began. And that this discovery was among the factors influencing Sarkozy's decision to commit France to the attack — alongside, the memo listed, securing Libyan oil, preserving French regional influence, improving Sarkozy's domestic standing, asserting French military power, and limiting Gaddafi's reach in what Paris considers Francophone Africa.

The email is real. It is in the public record. The claims inside it are not established (again — the humour).

Africa Check found the assertion difficult to verify precisely because the original sourcing is not publicly available — it is a memo citing anonymous individuals, not an intelligence product. And Patrick Haimzadeh, a Libya specialist who accepts that Gaddafi had genuine pan-African currency designs, argues the timeline does not hold: in his assessment Sarkozy had decided to intervene by 21 February, well before the gold plan surfaced.

So here is my position, and I will not be moved off it in either direction.

Do not say NATO bombed Libya to save the CFA franc. You cannot prove it. The moment you assert it, every serious reader is entitled to discard the twenty thousand documented words around it — the tariff schedules, the Living Income Differential, the eight thousand certificates of origin — all of which are true and none of which requires a conspiracy.

And do not pretend nothing is there. Because what is established needs no embellishment:

A man stood before the African Union, as its chairman, and proposed a single currency, a single army and a single passport. He proposed backing that currency with metal rather than trust. He named the CFA franc as the thing it would replace. He held 143 tonnes of gold. The other heads of state waited him out and went to bed. Within two years he was dead in a drainage ditch, his country was dismembered, a US Secretary of State joked about it on camera, and a South African president expressed relief that the AU could now function without him.

And his own people — this is the part I keep returning to, and the part I have no comfortable answer for — were persuaded at extraordinary speed that he was the principal enemy. Some of that was earned; he was a dictator for forty-two years and the list of his victims is long and real. But a continent that cannot hold two thoughts at once — this man was a tyrant and this man was killed while proposing the only continental monetary architecture anyone has tabled since 1963 — is a continent that will keep being surprised.

You do not have to believe in a plot. You have to notice a pattern. Olympio, 1963, days from signing Togo out of the franc zone. Sankara, 1987, four years after asking why the franc. Gaddafi, 2011, two years after proposing to replace it.

Three attempts at monetary sovereignty. Three graves. And a currency still printed in France.

Correlation is not causation. Agreed. But when the same variable shows up in every fatal case, an honest analyst does not delete the column. She writes down her uncertainty and keeps it in the model.

A word on the American president with Kenyan roots

He belongs in this argument, and dodging him would be cowardice.

Libya's destruction occurred under Barack Obama. Whatever any of us hoped a son of Kogelo might mean for this continent, the office is not a person — it is an apparatus with interests that predate every occupant and outlast him. Obama himself has described the failure to plan for post-Gaddafi Libya as the worst mistake of his presidency. I do not think that makes him a pawn, and I do not think he was ever ours to claim.

I think it proves the most important point in this essay: the system does not require bad men. It only requires that good men occupy positions whose incentives are fixed.

Which is exactly what I have been arguing about our own presidents. It would be dishonest to apply that standard to Washington and exempt Kampala.

XV. The counterfactual: what if Côte d'Ivoire said no?

I questioned the inverse reality. Not the slogan — the sequence. What actually happens if a CFA country walks? Let me take the biggest: Côte d'Ivoire. Largest economy in the zone, largest cocoa producer on earth, the very state whose president led the vote against Nkrumah in 1961. Poetic, and analytically useful.

I am going to give you the honest version, which includes the part nobody posting about this online ever includes.

Years 0–2: the punishment phase

The peg breaks. The new currency — call it whatever you like — depreciates hard, probably 30–50% against the euro, because markets price a new central bank's credibility at zero until proven otherwise. Imported fuel, wheat, medicines and machinery become sharply more expensive within weeks. Inflation, historically low in the zone, spikes into the teens or worse. Urban Abidjan feels it first and loudest, and urban populations are the ones that bring down governments.

Euro-denominated debt becomes more expensive to service in local currency terms overnight. Spreads widen. Some external financing dries up. Capital flight begins the moment exit becomes credible — probably before it is announced.

And critically: exiting the monetary union means exiting or renegotiating the WAEMU common market. The neighbour becomes a foreign market. That is the trap doing its work in real time.

There will be a moment — somewhere in month eight — when every newspaper in Paris and every technocrat in Abidjan says we told you so. That moment is the design feature. It exists to be survivable-looking from outside and unsurvivable-looking from inside.

Years 2–5: the pivot, if the state holds

Here is what the depreciation also does, which nobody says out loud.

It makes Ivorian processing internationally competitive for the first time. A weaker currency means Ivorian labour, land and power are cheap in dollar terms. Grinding cocoa in San Pédro becomes dramatically more attractive against grinding it in Amsterdam. The country already has grinding capacity. It has never had the exchange-rate advantage to scale it against European incumbents.

Monetary policy comes home. The central bank can now set rates for Ivorian conditions rather than importing the European Central Bank's stance — a stance calibrated for German manufacturing and French services, not for a cocoa-and-cashew economy with 40% of its population under fifteen. Every recession in the zone since the euro's creation has been fought with someone else's rate.

The reserves stay. That is not a small line item; it is a sovereign wealth position that can capitalise a development bank.

Years 5–15: the fork

If the state holds — meaning fiscal discipline survives the political temptation of a printing press — you get a mid-income economy with domestic monetary policy, a growing processing sector, and, for the first time, the ability to negotiate cocoa terms without an exchange-rate straitjacket. Think Vietnam's trajectory, not Ghana's.

If the state does not hold, the printing press solves this year's salary problem and creates next decade's inflation catastrophe. This is the actual risk, and it is not theoretical. It is why the defenders of the CFA are not fools.

The political outcome, which is the real one

The political consequences dwarf the economic ones.

A successful Ivorian exit does not stay Ivorian. It is a demonstration. It is Olympio's plan, sixty years late, finally executed by a state too large to quietly remove. Senegal, Cameroon, Burkina — the calculus changes for all of them the moment one large economy survives the crossing.

Which is precisely why the crossing has never been permitted to be attempted by anyone large enough to survive it.

The uncomfortable coda

Since 2024, the juntas of Mali, Burkina Faso and Niger — as the Alliance of Sahel States — have made exit explicit policy, expelling French troops, leaving ECOWAS, and in December 2025 creating a confederal investment bank to prepare a currency provisionally called the sira.

Note who finally attempted it. Not elected governments. Soldiers.

Genuine exit became real policy only at the outer edge of the normal political process. It took men with guns to put on the table a thing ninety-five per cent of the population wants.

That is the most frightening sentence in this essay, and I have not found a way to make it less frightening. If constitutional politics in Francophone West Africa cannot deliver what 95% of people want, and only juntas will attempt it, then the democratic argument and the sovereignty argument have been placed in opposition — and they were placed there deliberately, by a design that made collective exit impossible and unilateral exit suicidal.

Someone chose that. It did not grow on a tree.

Someone chose that. It did not grow on a tree.

XVI. Would owning Congo liberate us all?

Let's invoke the delusion once more — imagine if Africa owned the DRC — would that alone free the continent?

I want to answer the arithmetic first and the ownership second, because the arithmetic is where I have to be disciplined and the ownership is where I have to stop being polite.

The arithmetic. The DRC's untapped mineral resources are routinely valued at $24 trillion — repeated by the US government, the Atlantic Council, the World Bank's orbit and every article ever written on the subject. It is also an extrapolation from known reserves that lacks robust empirical foundation; valuations swing with market conditions and extraction capacity, and conservative institutional assessments are far more hedged. And in-ground value is not money. Spread across fifty years and a hundred and ten million people, even the optimistic figure is transformative rather than continentally redemptive.

So: no, the DRC alone does not liberate 1.5 billion people. I will not inflate a number to win an argument I can win without it.

Now ownership — and here I withdraw the word I used earlier, because it was a lawyer's word and this is not a legal question.

I wrote that the DRC already owns the DRC. That is true in the sense that a UN seat is true, and false in every sense that pays a wage.

Ask the only question that matters: where does the $24 trillion circulate?

Not in Kinshasa. Not in Kolwezi. It leaves as ore. The country produces 70–80% of the world's cobalt and earns roughly a billion dollars from it, against a lithium-ion battery market exceeding sixty billion a year. Mining captures about 5% of the value chain; cell manufacture and pack assembly capture 40%. Less than 5% of Africa's critical minerals are refined on the continent at all. Chinese state enterprises and policy banks control roughly 80% of the DRC's cobalt output; China processes 74% of the world's cobalt and 90% of its rare earths.

So what, precisely, does Congolese "ownership" consist of? The right to hold the asset while somebody else sets its price, moves it, refines it, and books the margin. That is not ownership. That is custody. A man who holds the deed to a house he may not enter, rent out or sell at his own price does not own a house. He is being made to guard one.

And the guarding is expensive, which is the second half of the machinery and the part I shall name correctly.

The war is not an interruption of the extraction. It is a component of it.

I am not claiming every militia was manufactured in a foreign office. I am claiming something more durable: that a state permanently at war cannot regulate, cannot tax properly, cannot enforce a contract, cannot insist on a refinery, and cannot say no. Instability is not a bug in the extraction model. It is the operating condition under which the extraction model achieves its best prices.

Look at how it is currently financed and you can see the loop closing. M23's control of the coltan-rich Rubaya mine yields roughly $300,000 a month, with at least 150 tonnes of tantalum ore smuggled to Rwanda per UN experts. In gold-rich Ituri, militias earned at least $140 million in mineral revenues in 2024. Rwanda's mineral exports rose from $772 million in 2022 to $1.1 billion in 2023 — a figure analysts note exceeds what its own production could plausibly generate — with UN experts finding Rwandan minerals mixed with those from M23-held territory. Kigali denies direct involvement. Kinshasa's finance minister put the annual smuggling loss at close to a billion dollars. M23's 2025 capture of Goma and Bukavu displaced 700,000 people; over 7,000 were killed in the Goma assault alone.

Note carefully what this is. This is not primarily white men taking African minerals. This is an African state — one of the best-governed on the continent by conventional metrics — accused by UN experts of laundering the minerals of a neighbour through its own export figures. Fragmentation does not only invite external predation. It manufactures internal predation. There would be no smuggling economy in the Kivus if there were no line to smuggle across.

And then diplomacy arrives — which is the guise

The parallel is exact, and worth drawing slowly. In the nineteenth century the Bible went ahead and the flag followed: one instrument prepared the ground, the other took the title, and the sequence was always presented as service. The instruments have been retired. The sequence has not.

Today the war goes ahead and the agreement follows. February 2025: the European Parliament voted to suspend the EU's raw materials memorandum with Rwanda over the military presence and traceability failures — while the 2024 Global Gateway agreement facilitating mineral trade was not formally revoked. March 2025: President Tshisekedi reportedly proposed to President Trump military assistance to end M23 in exchange for US access to cobalt, tantalum and lithium, framed as advancing America's technological edge. December 2025: the Washington Agreements for Peace and Prosperity — security guarantees and troop withdrawals traded for direct access to Congolese mineral wealth, with a deep-sea port and land for US military bases attached (again—I hope the irony isn't lost on you!).

Read it as a sentence and it is unmistakable. Peace was sold to the Congolese, and they paid for it in cobalt.

Not brokered by the African Union. Not resolved by any continental security architecture — because there isn't one, because Nkrumah's African High Command was voted down at Monrovia in 1961 and never revived. A bilateral transaction between a fragmented state and a superpower, conducted over seven thousand bodies in Goma, and announced as a peace initiative.

The word "aid" does similar work. When a country's security, its mineral licensing, its health budget and its humanitarian relief are all supplied by external actors flying a multilateral flag, the argument about whether this counts as colonisation becomes semantic. A state that cannot fund its own army, its own courts or its own clinics has already transferred sovereignty; the only remaining question is the identity of the recipient, and the answer is written on the side of the Land Cruiser (Hint—dreamy blue eyes!—Again—I hope the sarcasm isn't lost on you!).

Blum, and why I will not lean on him

William Blum's America's Deadliest Export: Democracy argues that America's stated commitment to democracy promotion is cover for economic and strategic interest — that the secret of US foreign policy is that there is no secret, the objective being dominance by available means. He documents the National Endowment for Democracy, founded 1983, as Washington's foremost non-military regime-change instrument.

Blum is a polemicist. The book is advocacy, not dispassionate scholarship — reviewers call it a sustained tirade, and it should be read as argument. Serious scholars of democracy promotion — Michael McFaul, Christopher Coyne — contest both his framing and his conclusions, and there is substantial literature arguing that democracy promotion has mixed motives and mixed results rather than uniformly cynical ones.

But you do not need Blum's strongest claims to make the structural point, and the structural point survives every objection to him:

A fragmented continent presents fifty-five separate negotiating counterparties, each individually weak, each with leaders whose political survival can be influenced, and each holding resources external powers need. Whether or not any given intervention is cynically motivated, the opportunity structure for external leverage exists because of fragmentation, not despite it.

That is the sentence. It requires no villain. It survives good faith. It would be true if every foreign minister in the world were a saint.

So let me restate the point I was too careful to make the first time.

The constraint was never who holds the title. It is who holds the refinery, who sets the price, who writes the tariff schedule, who accredits the bar — and who is permitted to be at peace. Congo is not poor because it does not own its minerals. Congo is being kept in a condition in which ownership cannot be exercised.

Which is why a continental answer is not sentiment. One buyer's market and fifty-five sellers is a price. One seller's bloc and four buyers is a negotiation. The Congolese cannot fix this alone, and that is not a failure of Congolese character. It is the arithmetic of being made to stand alone.

XVII. How a banana republic gets its name

I reached for this analogy early, and having looked properly at where the phrase comes from, I am not sure I have ever encountered a cleaner parable for what is being done to us.

The term was coined by the writer O. Henry in 1904, describing Honduras. But the machinery it names belongs to the United Fruit Company, and the full machinery looked like this: a foreign corporation acquires the land, then the railways, then the ports, then the telegraph. It sets wages. It selects the crop. It influences who governs. The country's economy becomes an extension of one company's logistics chain — and the people who live there become, functionally, tenants on ground their grandfathers walked freely.

Then Guatemala elected Jacobo Árbenz, and Guatemala tried to stop it.

Two per cent of the population held seventy-two per cent of the arable land, and only twelve per cent of private land was being used. United Fruit owned forty-two per cent of Guatemalan land and cultivated fifteen per cent of what it held. Árbenz's Decree 900 redistributed uncultivated land, compensating owners in bonds maturing over twenty-five years, valued according to the claims those owners had made on their own previous tax returns.

Stop there. That is the most elegant policy instrument I have ever read about.

We will pay you exactly what you told the tax office it was worth.

United Fruit had been serially undervaluing its holdings to avoid tax. Árbenz offered $1.2 million based on the company's own declared figures; United Fruit and the State Department demanded $16 million.

Secretary of State John Foster Dulles was a former lawyer for United Fruit. His brother Allen, director of the CIA, had sat on the company's board. In August 1953 Eisenhower authorised Operation PBSUCCESS — a $2.7 million covert programme to overthrow Árbenz, which armed, funded and trained a force of 480 men under Carlos Castillo Armas.

By 1954, 570,000 hectares had been redistributed to a hundred thousand families. The coup deposed Árbenz and set off decades of civil war — thirty-six years of it. Roughly two hundred thousand dead. Over a fruit company's tax declaration.

Now bring it home

Put plainly: if you are not careful, you become a squatter in your own land. The people became labour for the company. The land was locked into growing one thing. The community got nothing and was told it was development.

Ask yourself what distinguishes that from a Zambian copper belt where the ore leaves as concentrate, or a Congolese cobalt corridor where eighty per cent of output is controlled by foreign state enterprises, or an Ivorian countryside locked into a single bean whose price is set on an exchange in London, or a Kenyan hillside growing a leaf that earns $2.28 a kilo and reaches a shelf in Hamburg at forty times that.

The differences are real: we hold the title deeds. We have flags and anthems and seats at the UN.

And the similarities are load-bearing: one crop, one buyer, terms set elsewhere, value captured elsewhere, and a political class whose position depends on the arrangement continuing.

That is not a banana republic because the fruit is different. It is a banana republic because the fruit is the only thing that is different.

XVIII. What single rule actually costs — and the ten-times question

Now the question I have been circling, and the one everything else has been leading to: would our presidents not be richer if they stopped the bleeding?

The intuition is seductive. If you liberated the value chain, national income rises tenfold. Surely a president's share of a tenfold-larger pie beats his share of the current one? Surely corruption is just short-sightedness? Surely we could persuade them?

I spent a long time wanting this to be true. It is not, and understanding why it is not is the most important paragraph in this essay.

A rent is not a percentage of GDP. A rent is a toll on a chokepoint you personally control.

The income of an entrenched political class does not come from the size of the economy. It comes from discretion over a flow: who gets the mining licence, whose container clears the port, which firm wins the road contract, which importer receives the waiver, which bank handles the sovereign account. That income scales with opacity and discretion, not with national output.

Now apply every item on the federation list:

A supranational court whose rulings override national courts
A continental parliament with binding legislative authority
A common external tariff, removing national trade policy
A monetary authority, removing national monetary policy
A continental defence command, removing national military autonomy
Free movement, removing control over who enters and stays
Enforceable standards on elections and governance

Every single one of those is a direct attack on discretion. Harmonised rules of origin mean the customs officer cannot decide. A common tariff means the minister cannot grant the waiver. A continental court means the national court is no longer a convenience. Free movement means the permit is worthless.

Unity would make the country ten times richer and make the men currently running it far poorer — not in the long run, not eventually, but on the day the treaty enters into force.

So no. They would not accrue ten times more. They would accrue a fraction, transparently, at a salary. And we should stop constructing arguments that assume our rulers are confused. They are not confused. They have done the sum, exactly as their predecessors did in Monrovia in 1961, and they keep arriving at the same answer because it keeps being the correct answer for them.

The incumbency landscape

Cameroon's Paul Biya has held power over forty years. Equatorial Guinea's Teodoro Obiang over forty years. Uganda's Yoweri Museveni since 1986 — nearly four decades — having overseen the removal of presidential term limits in 2005 and age limits in 2017. Rwanda's Paul Kagame, in power since 2000, oversaw a constitutional amendment permitting him to remain potentially until 2034 (Design or coincidence?—No offence intended).

And the detail that ought to be carved above the door of the AU headquarters: shortly after taking power in 1986, Museveni wrote that the problem of Africa in general, and Uganda in particular, was not the people but leaders who wanted to overstay in power (Need I repeat myself? irony — humour).

He then stayed thirty-nine years (Kettle? Pot? Black!).

I do not offer that as a cheap shot, and I want to be precise about why. It is not evidence that Museveni is a hypocrite. It is evidence that the container shapes the contents. He arrived believing what he wrote. The incentive structure of the fragmented African state converted a leader of genuine reformist conviction into a defender of the sovereignty that protects his position. That is the mechanism, in one biography.

And the AU is institutionally incapable of constraining it, for exactly the reason Nyerere named in 1963. The African Charter on Democracy, Elections and Governance exists. It has been widely ratified. Enforcement against a sitting head of state is essentially nonexistent. We wrote the rule and declined to build the bailiff.

The counter-argument, given its fair hearing

There is a case on the other side, and dismissing it would make me exactly the kind of writer I do not want to be.

Scholarship examining Mugabe, Museveni and Kagame argues that indigenous African leadership systems often prioritised lineage and lifelong appointment, and that Eurocentric models emphasising fixed terms and elections were imposed by colonial administrations with little regard for existing structures — producing a durable tension between traditional practice and modern statecraft. Long-tenure leaders argue that stability and development justify continuity, and Rwanda's governance rankings are genuinely strong — eleventh of fifty-four on the Ibrahim Index, against Uganda's twenty-second, both above countries with cleaner term-limit records.

So: the cultural-legitimacy argument is real. It is also extraordinarily convenient for the people making it, and the polling does not support it. Afrobarometer's surveys across thirty-four countries and more than forty-eight thousand interviews indicate that, in principle, leaders who stay beyond two terms are not what people want.

Which means: whatever our ancestors did with chieftaincy, the living have been asked, and the living have answered.

The hypothesis

So here is the proposition at the centre of everything I have written.

The primary obstacle to African unity is not Western interference.

Western interference is real, documented and consequential — Berlin, the franc zone, the tariff schedules, Guatemala, the Washington Agreements. I have spent twenty thousand words establishing it. But it operates through a domestic incentive structure that would resist unification even in the West's complete absence.

The people who would have to surrender power to create a United States of Africa are precisely the people who have spent their entire careers ensuring they never have to surrender power to anyone.

Nkrumah's federalism failed in 1963 because heads of state would not surrender sovereignty. It would fail today for the identical reason — with the added complication that today's leaders command patronage networks, security services and resource revenues that the class of 1963 could not have imagined.

External powers do not need to prevent African unity. They need only wait, and negotiate bilaterally in the meantime.

They are extremely good at waiting. They have had five hundred years of practice.

XIX. Wakanda, and the thing we already have

Someone always brings up Wakanda, and I have brought it up myself, so let me deal with it and then bury it.

The fantasy is that we possessed a miracle metal, kept it, learned to build with it, and traded it on our own terms. And the irony cuts, because of course we do have the vibranium. We have cobalt, coltan, platinum, manganese, lithium, uranium, gold, and the world's second-largest rainforest sitting on a river system that could power half a continent.

But notice what the fantasy actually requires, and notice that it is not the metal.

Wakanda is not rich because it has vibranium. Wakanda is rich because it has a single sovereign authority over the entire vibranium deposit, an enforced external border, its own defence force, and no foreign concession-holders. Strip away the spaceships and Wakanda is, precisely and exactly, Nkrumah's 1961 proposal with better costumes. Continental government. Single market. African High Command. It is not a fantasy about minerals. It is a fantasy about institutions, and we rejected the institutions in a conference hall in Monrovia while the metal was already in the ground.

That is why I will not spend more time on it. The dream is not that we might one day find our vibranium. We are standing on it, and we have signed fifty-five separate contracts for it, and the dream is that we might one day stop.

XX. Then what? Four roads, in descending order of comfort

If the maximalist version is blocked — and I have just spent an entire section arguing that it is blocked by design — what remains?

Path One: Functional integration without political union

Make AfCFTA real. Enforce SAATM. Harmonise rules of origin so regional mineral value chains can form. Build the corridors. Leave sovereignty formally untouched while making the economics of separation progressively more expensive.

This is the current strategy. Its ceiling is limited — the World Bank projects AfCFTA implementation could cut intra-African trade costs by 16%, with full implementation raising Africa's GDP by around 10% by 2043 — but its floor is safe, because it never requires a president to surrender anything he cares about.

The critique writes itself: it is precisely because it threatens nothing that it changes little. Signature without compliance is the pattern. Eight thousand five hundred certificates of origin is the pattern. SAATM is the pattern.

Path Two: Regional blocs first, federation later

Let the regional economic communities lead. The EAC and SADC already have real economic and political cooperation and could liberalise deeply within their regions before anyone attempts fifty-four at once. This is the EU's actual historical model: start small, among states of comparable interest, deepen, then widen.

The critique: ECOWAS is fraying, not deepening. Burkina Faso, Mali and Niger withdrew in 2024. Regional blocs are not obviously more stable than the continental project, and the EU analogy flatters us — six states of comparable wealth is not fifty-four with five colonial languages, forty currencies and debt profiles ranging from investment-grade to default.

Path Three: Federation under external pressure

The historical pattern for genuine political union is that it follows catastrophe or acute existential threat. America federated properly only after a civil war. Europe integrated from the ashes of 1945 — not from romantic pan-European sentiment, of which there was none, but from exhaustion and terror.

Africa's plausible forcing function is the resource competition now accelerating. If the continent watches its lithium, cobalt and rare earths extracted bilaterally at five per cent of value while conflict escalates in the Kivus, the case for coordinated negotiation becomes harder for even entrenched incumbents to dismiss — not on moral grounds, but on revenue grounds.

The cynical version of the optimistic case: African leaders will federate when fragmentation costs them personally more than unity does. Not before.

Path Four: the one nobody writes about, which is you

Every path above is addressed to heads of state. Which is why every path above has failed for sixty-three years. We keep writing petitions to the beneficiaries of the problem.

So let me propose the fourth road, and it is the only one that does not require asking permission from the people who have already answered.

The gap that matters is not the signature gap. It is the enforcement gap — and enforcement gaps close under observation.

Concretely:

Publish the compliance league table. Not ratifications — ratifications are theatre. Publish, quarterly and by name, which states have gazetted their tariff schedules, which have filed complete Category A/B/C offers, which have issued certificates of origin and how many, and which SAATM signatories have denied traffic rights to which carriers. Twenty-five countries are operational. Name the other twenty-nine. Every week.

Make the non-tariff barrier reportable and public. A trader blocked at Malaba or Beitbridge should be able to file, publicly, in an hour. Aggregate it. Rank the borders.

Demand a beneficiation floor in every new mineral contract. Zimbabwe's ban produced a $400 million refinery. The instrument works. It is not a theory.

Ask your MP a single question and do not accept an answer about aspiration: Has our country gazetted its AfCFTA schedules? Yes or no? If no, why not, and when?

Treat the COPEC lesson as policy, not history. Two producers cannot hold a floor price. A continental commodity authority for cocoa, cobalt, copper, coffee and platinum can. Ask why the AU has never attempted it. The answer will tell you who it is for.

Because here is the thing about the sovereignty shield: it works against courts, against treaties, against other heads of state. It has no defence at all against being watched by your own citizens, in numbers, continuously, with specific questions and public scorecards.

That is not a revolution with a gun. That is a revolution with a spreadsheet. And the spreadsheet is the only one of the two that has never once put a Compaoré in power.

XXI. The map, and the size of us

I want to end where my curiosity paused, on something that sounds small and is not.

Even the drawing of us is shrunk. The Mercator projection, created in 1569 for navigation, preserves shape and destroys size — inflating everything near the poles and compressing everything near the equator. On Mercator, Greenland and Africa appear roughly the same size. In reality, fourteen Greenlands fit inside Africa. Africa's land area exceeds 30 million square kilometres.

Every classroom on this continent hangs a map that makes us small. We have been looking at a diminished version of ourselves for four hundred years, in geography lessons, on news broadcasts, in the corner of every weather report, and we have called it a map.

As I wrote this I questioned whether our education systems are built to keep us inside an illusion. I will not indulge the full conspiracy — most of your teachers are not agents of anything; they are overworked people using the textbooks they were given. But a curriculum does not need an author to have an effect. If a child spends twelve years looking at a projection that shrinks her continent, learning a history that begins at contact, and taking an examination that never once asks her what raw cobalt costs against refined cobalt, she will emerge fluent, credentialed, and structurally unable to see the thing.

That is not a plot. That is inertia, which is more durable than any plot, because nobody has to maintain it.

And then, this month, something happened.

The African Union backed the Correct the Map campaign, and the United Nations passed a resolution — 164 votes in favour — encouraging wider use of equal-area projections such as Equal Earth in classrooms, media and international institutions, so that the world map finally reflects Africa's true scale. Carlos Lopes, one of the sharpest analytical voices on African economic strategy, called it an emotional day.

And I am glad. I am not being sardonic. Representation is real, and a child who sees the true size of her home carries something she would not otherwise have.

But I want us to hold both facts in one hand, because holding only one is how sixty-three years disappear.

In September 2026, one hundred and sixty-four nations voted to acknowledge that Africa is bigger than they had been drawing it.

And in September 2026, a Kenyan still cannot fly to Lagos without paying sixty per cent more than the equivalent European route, in taxes that frequently exceed the fare itself. Twenty-nine African states have not gazetted the schedules that would let their traders use the free trade area they ratified. Eight thousand five hundred and sixty-one certificates of origin have been issued for a market of 1.4 billion people. The cobalt is leaving at ten dollars a kilo. The tea is leaving at $2.28. The bean is leaving raw, and Germany — which cannot grow a single one — out-earns the entire continent that gave the world coffee.

We won the map. We have not yet won the kilogram.

XXII. The honest conclusion

The romantic case for African unity is Nkrumah's: that we are one people, artificially divided, and that reunification is a restoration.

The realist case is colder and, I think, more useful.

Fragmentation is not primarily a wound to be healed. It is a business model to be dismantled.

It benefits external powers who negotiate bilaterally and price accordingly. It benefits domestic elites whose power depends on the sovereign container and the discretion it confers. It benefits intermediary states running smuggling economies that exist only because borders are simultaneously arbitrary and enforced. It benefits the roaster in Hamburg, the refiner in Zhejiang, the grinder in Amsterdam, the assayer in London, and roughly four hundred men on this continent with residences in three countries.

Every one of those beneficiaries has a more immediate incentive to preserve the arrangement than any diffuse population has to change it.

That is why sixty-three years of pan-African rhetoric has produced a free trade agreement most members have not operationalised and an air transport market its own signatories decline to honour.

Nkrumah said: we unite or we perish. The continent found a third option he did not anticipate. Neither unite nor perish, but persist — fragmented, individually negotiating, collectively underpriced, and always, always about to integrate next year.

So let me close by addressing the two audiences this was written for, because they require different sentences.

To the presidents

You are not confused. I have stopped pretending you are, and I would ask you to stop pretending too. You have run the numbers, exactly as Houphouët-Boigny ran them in Monrovia in 1961, and you have concluded that a smaller kingdom fully held is worth more than a larger one shared.

You are right about your own position and wrong about its duration. The arrangement that protects you is not stable. It is merely slow. Ask Compaoré, who held it twenty-seven years and now hears his conviction read in absentia. Ask the men who thought the CFA was permanent, and then watched three of their neighbours walk out and start printing. Ask any head of state on this continent under the age of sixty what the median citizen's age is, and whether that citizen — who is nineteen, unemployed, on a phone containing Congolese cobalt she will never be paid for — is going to spend her forties asking politely.

Sovereignty, the thing you are protecting, was handed to you in 1963 as a shield. It was never a sword. It has never once been used to get a better price for anything your people grow or dig.

Sign the schedules. Open the skies. Harmonise the codes. Not because it is noble. Because the alternative is to be the last generation of men who could have done it, and everyone will know your names.

And to everyone else

We have been told the obstacle is Europe, and it is partly Europe. We have been told the obstacle is China, and it is partly China. These are comfortable enemies, because they are far away and nothing is required of us but anger.

The nearer obstacle is a class of men who look like us, speak our languages, were educated in our schools, and have arrived — rationally, repeatedly, over three generations — at the conclusion that our fragmentation is worth more to them than our unity.

That is harder to say at a funeral or on a podcast. It is also actionable, which the other version is not.

They vote against unity in every forum they enter, and they will keep doing it until the price of voting that way exceeds the price of voting the other way. You are the only mechanism on earth that can raise that price. Not a court they can ignore. Not a charter they can ratify and shelve. Not a foreign power that profits from their weakness.

You. Watching. Counting. Naming. Refusing to accept a communiqué in place of a customs gazette.

Sixty-three years ago, twenty-two of twenty-seven men decided that their comfort outweighed our arithmetic, and we have paid for that meeting every single day since — in a cup of tea that earns a Kenyan grandmother half a cent, in a kilogram of cobalt that leaves Kolwezi at ten dollars and returns at two hundred, in a flight to Lagos that costs more in tax than in fare.

The question for this generation is not whether Nkrumah was right. He obviously was.

The question is whether anything short of catastrophe can make his argument politically survivable for the people who would have to act on it.

And the only honest answer I have — the one that keeps me writing at this hour — is that catastrophe and attention are the only two forces that have ever moved a ruling class, and exactly one of them is available to us before it arrives.

Choose the spreadsheet. Choose it loudly. Choose it now.

Because the borders were drawn to limit us. The word has been sitting there, in plain sight, since 1884.

Everything they intended was written down. We just never read it aloud.

Notes on sources and further reading

Primary texts

Kwame Nkrumah, Africa Must Unite (1963); Neo-Colonialism: The Last Stage of Imperialism (1965); I Speak of Freedom (1961)
William Blum, America's Deadliest Export: Democracy (2013) — read as polemic, not neutral history
Colin Legum's documentation of the Casablanca/Monrovia debates

Living voices worth following

Achille Mbembe (Wits) — On the Postcolony (2001); the concept of necropolitics
Carlos Lopes (former UN ECA executive secretary; Oxford, UCT) — the sharpest analytical voice on African economic strategy, and a central figure in Correct the Map
Ngozi Okonjo-Iweala (WTO Director-General) — the trade-institutional perspective
Jason Stearns (Simon Fraser) — the authoritative researcher on eastern DRC conflict dynamics
Farai Maguwu (Center for Natural Resource Governance, Zimbabwe) — resource beneficiation

Institutional sources

UNCTAD Economic Development in Africa Report; UN ECA regional integration assessments
tralac (Trade Law Centre) for AfCFTA legal texts and ratification tracking
AFRAA taxes and charges studies for aviation data
Afrobarometer for public opinion on term limits and governance
UN Group of Experts reports on the DRC

Counter-reading, for evenhandedness

Michael McFaul, Advancing Democracy Abroad — the case for democracy promotion
Christopher Coyne, After War: The Political Economy of Exporting Democracy — skeptical but non-conspiratorial

Web sources for the added material

The Conversation — why the cocoa Living Income Differential hasn't worked
IIED — civil society perspectives on the LID
Oxfam Novib — the LID, futures markets and price
Development Research — the rough start for living income cocoa prices
Ghana Business News — Germany out-earns African coffee producers
The Elephant — Europe's re-export boom and Africa's earnings
The Africa Report — who killed Sylvanus Olympio
The Citizen — Olympio and the vision of a united Africa
Al Jazeera — the African Union names Gaddafi as head (2 February 2009)
Gwynne Dyer — Gaddafi, King of Africa (the midnight scene)
Africa Check — the Gaddafi gold-dinar claim examined
Vice — the Blumenthal email
Nairametrics — AfCFTA at five
Rio Times — AfCFTA operational status 2026
Habari Network — AfCFTA implementation progress
The Africa Logistics — AfCFTA in 2026
US ITA — DRC market overview
Fact-check on the $24 trillion figure
Africana Voice — DRC cobalt quotas
Elisa Exporters — Mombasa auction prices 2026
NPR — the UN backs Equal Earth
Fortune/AP — AU endorses Correct the Map
Zinn Education Project — Árbenz and United Fruit
History in Charts — United Fruit in Guatemala
Penguin History — the CIA coup that ended Guatemala's democracy