A Record Is Not Enough: The record, and the arithmetic behind it

A Record Is Not Enough

Ethiopia's coffee windfall, and what reached the farm gate

By Wondwossen Mezlekia

Part 1 of a five-part series. Figures current through May 2026.

I started writing about Ethiopian coffee when the Commodity Exchange was still an argument, before it became a building, a trading floor, and finally a system that touched nearly every coffee farmer in the country. I have kept at it, on and off, for the better part of twenty years.

This government came to office promising reform, transparency, prosperity, and a different economic future. Nearly a decade later, it has a record coffee year to point to.

The number is real. What it proves is another matter.

I am not returning to settle that question with a slogan. I am returning to count the record: what produced it, who gained from it, what it cost, and what reached the farm gate.

The question is not whether Ethiopia exported more coffee, or whether coffee earned more foreign currency. It did. The question is whether the grower, after the reforms and the promises, kept more of the value. That is the question behind this series. Not the headline at the port, but the account at the farm gate.

On sources, since they do not always agree, and I will not hide that. I use the Ethiopian Coffee and Tea Authority for the official account; the USDA Foreign Agricultural Service and the International Coffee Organization for figures that are not the government's to shape; the Intercontinental Exchange for market prices; the IMF and the Ethiopian Statistical Service for the wider economy; and the Institute for Security Studies for the parts of the trade officials would rather treat as somebody else's problem. Where they diverge, I will say so.

Five posts. The first four run now, before the coffee year closes on July 7. The fifth, on what the farmer keeps, follows once the Authority publishes the final account and it can be measured against independent trade data.

Part 1: The record, and the arithmetic behind it

The Coffee and Tea Authority says Ethiopia earned 2.65 billion dollars from coffee in the fiscal year that closed in July 2025, on 468,967 tons. Its director general called the result historic. It was historic. But a historic figure deserves more than applause. It deserves an accounting.

Start with the fact that the country cannot quite agree with itself on the number. The USDA, working on a marketing-year calendar and a different trade series, puts the same broad period nearer 2.89 billion dollars on 7.43 million sixty-kilogram bags. The ICO, which has tracked these things since the 1960s, tells the story through price. Three serious sources, three different totals, built on different clocks and different definitions of what counts as an export. That is not proof anyone is cooking the books. It is proof that Ethiopia still does not publish one reconciled public account of what it shipped, when, and in what form. When a government announces a record to the dollar and the most-watched outside estimate lands a quarter of a billion away, the footnotes are not a side argument. They are the story.

Set the bookkeeping aside and ask what moved the money. Most of it was not made in Ethiopia. The world ran short of coffee, with the ICO counting market deficits in three of the four coffee years through 2024/25, close to eighteen million bags in all, and its composite indicator reached the highest monthly average on record, 354.32 cents a pound, in February 2025. Ethiopia sold into that shortage. A higher price for the same bean is welcome. It is not, by itself, proof that the coffee economy was rebuilt.

The global market did much of the lifting: ICO Composite Indicator Price, January 2024 to May 2026

The peak did not hold. By May 2026 the same index had slipped to 256.05 cents, roughly 28 percent below the February 2025 high, as the market began pricing in a record Brazilian crop. Still a strong price by any historical measure, but a record built on an exceptional and already fading price cycle is a windfall, not a foundation.

Then the birr. In July 2024, under its IMF program, the government moved to a more flexible exchange rate, and the official rate slid from about 57 to the dollar at the end of June 2024 to about 136 a year later. The record is denominated in dollars, and a weaker birr did not conjure 2.65 billion of them. The coffee still had to be sold, and buyers still had to pay. But the float changed who gained and who paid once those dollars came home. It swelled the birr value of export earnings and sharpened the incentive to export, and it raised the price of fuel, fertilizer, transport, food, and medicine, everything a farming household buys in birr. In plain terms it was a transfer, from everyone who holds birr to everyone who earns dollars. The exporter earns dollars. The farmer earns birr. The household pays in birr. That is where the arithmetic becomes political.

The currency moved sharply: official birr/US-dollar exchange rate, June 2024 to June 2025

Handle the inflation honestly, in both directions. It came down hard from the worst years, which the government has earned the right to claim. But across 2021 to 2025, consumer prices still rose by something close to two hundred percent in total, so the cost of living nearly tripled in five years, and lately the relief has reversed. The Statistical Service put year-on-year inflation at 13.4 percent in May 2026, up from 9.4 percent as recently as March and 11.7 in April, with food alone running near 15. So when an official tells me cherry prices tripled this season, I do not hear a tripled income. I hear a price running hard to stay in place. The household told to celebrate a record is the same household that paid triple for cooking oil while doing it.

Everything returns to the figure the government does not publish. The Authority reports revenue at the port. It does not publish a consistent series, region by region and year by year, showing what the grower received per kilogram of cherry and what that money could still buy. Until it does, I will read the record as a fact about buyers in Jeddah, Hamburg, and Shanghai, and not yet as a fact about a household in Gedeo, Jimma, Guji, or Sidama. The people holding the farm-gate data carry the burden of proof, and so far they have answered it with a press release.

The missing series is the point: export revenue and inflation are reported, farm-gate income is not

A note on timing. Ethiopia's fiscal year closes on July 7, and the final revenue figure is not yet in. The questions in this series do not wait on that press release. The world price has already begun to fall, the cost of moving coffee has already risen, and the farmer's share remains unreported. When the final number lands, read it against those facts, not instead of them.

Source note: USDA FAS, Ethiopia: Coffee Annual (May 2026) | ICO, Coffee Market Report, February 2025 and May 2026 | IMF, Ethiopia Fourth Review (2026) | Ethiopian Statistical Service, May 2026 Inflation Release

 

Next in the series, Part 2: Who is in charge of coffee?