In 2007, after a two-year fight, Ethiopia won legal ownership of three of the most valuable words in coffee: Sidamo, Harar and Yirgacheffe. Then it set the licence fee at zero. Any company willing to sign the agreement could use the names without paying a cent of royalty. That looks like a country winning a war and declining the reparations, and the reason is uncomfortable in a way that explains more about the coffee trade than most price charts do. Ethiopian officials feared that charging for the names would push buyers toward other origins, so the price of keeping the customers was giving the victory away. A rights-holder that cannot afford to charge for its rights does not have market power. It has paper. This article is the story of how the paper was won, why the zero was rational, and what nearly two decades of evidence since then says about whether owning a name changes what a farmer gets paid.
Why three words were worth fighting over
The economists Benoit Daviron and Stefano Ponte made an argument in The Coffee Paradox (2005) that has held up badly for farmers and extremely well as a description of the trade: an increasing share of coffee's value is attached to symbolic attributes, meaning reputation, origin story and brand, and to in-person service, meaning the cafe. Both are produced in the consuming country. A grower in the Gedeo Zone cannot export the story a roaster in Portland tells about her coffee, and the story is where the margin lives.
Ethiopia's trademark campaign, which began in 2005, was a direct attempt to claw a piece of that symbolic layer back to origin. The government's claim, as Gavin Fridell recounts it in Coffee (Polity, 2014), was straightforward: Ethiopia is one of the birthplaces of coffee, the names Sidamo, Harar and Yirgacheffe refer to Ethiopian places and peoples, and the country and its farmers should hold the intellectual property in them. Roasters were printing those words on bags and menus precisely because the words commanded a premium. Ethiopia wanted to own the words.
It chose trademarks rather than certification marks, and the distinction matters. A certification mark, which is how Jamaica protects Blue Mountain, pairing the name with a single controlled export channel, certifies that the product meets a standard and comes from a place. A trademark is a brand: it can be licensed, managed and marketed the way a company manages a logo. Ethiopia was not trying to police geography. It was trying to become the brand owner of its own reputation, which is a different and more commercial ambition. Disputes over what a coffee name is allowed to mean are older than either approach: a 1912 court case had to decide what "Mocha" could legally describe, and the question has never fully closed.
The fight, 2005 to 2007
The application did not glide through. Fridell's account, drawing on contemporary reporting and an Overseas Development Institute study, records that the US National Coffee Association and other industry lobbyists pressed the US Patent and Trademark Office to reject or delay the registrations. Starbucks initially rejected the idea and said so publicly through a media campaign. Some industry critics suggested at the time that Starbucks had also pressed the NCA behind closed doors to take its hard line, given how much the company relied on exactly the kind of high-grade beans Ethiopia grows. Starbucks denied that claim, and the denial belongs in the record as much as the allegation does. What is not in dispute is the public sequence: the industry's most visible specialty buyer opposed the initiative openly.
Oxfam International then ran a consumer campaign in support of Ethiopia and persuaded more than 96,000 people to contact Starbucks by email, fax, phone, postcard and in-store visit. In 2007 Starbucks signed an agreement to use and promote Ethiopian coffee brands under the initiative's terms. Ethiopia registered its trademarks in the United States that year, and in roughly 30 other countries in the years that followed.
Two years of lobbying, a public pressure campaign and a negotiated settlement, all to establish something that sounds almost tautological: that Ethiopia owns the names of Ethiopian places. The scale of the resistance is itself the best evidence of what the names were worth.
The zero at the centre of the deal
Here is the part that most retellings skip. The trademarks are owned by the Ethiopian Intellectual Property Office, which manages them in consultation with a stakeholder committee of Ethiopian exporters, cooperatives and government agencies. Companies must be licensed to use the names. And the licence is royalty-free. Licensees pay nothing. What they agree to instead is to market and promote Ethiopian coffee brands in various ways.
Fridell states the reason plainly: Ethiopian officials feared that a royalty would chase companies away from buying Ethiopian beans at all. Look at the position they were negotiating from and the fear was well founded. In the figures Fridell reports from the early 2010s, coffee accounted for over 31 percent of Ethiopia's export earnings while the country supplied under 3 percent of world exports. A roaster can substitute origins. It can move a blend toward Peru or Kenya or a different Ethiopian grade and most customers will never notice. Ethiopia cannot substitute customers at anything like the same speed, because coffee is a large share of everything it sells to the world. When one side of a table can walk away and the other cannot, the price settles where the side that can walk away wants it. In this case the price of a licence settled at zero.
That is what negotiating from weakness looks like even when you win. The trademark fight established the principle that an origin country can own its names. The royalty decision demonstrated, in the same breath, that owning a name and being able to charge for it are separate things, and the second one depends on market power the ownership certificate does not confer.
What the evidence since then shows
Did it work? The most quoted answer comes from Light Years IP, the non-profit that assisted Ethiopia with the trademarking, which claimed that Ethiopian coffee farmers' incomes doubled, with an additional $100 million arriving in 2007 to 2008. Fridell's caution about that number is worth repeating because it is a lesson in reading every impact claim you will ever meet: 2007 to 2008 was also a year of sharply rising international coffee prices, and it is not clear how much of the gain came from the trademarks and how much would have arrived anyway on the price alone. A benefit measured in the same year as a market-wide windfall, by the organisation that ran the project, is not evidence of the project's effect. It is a number waiting for a control group.
The harder test is whether Ethiopian prices detached, even slightly, from the commodity cycle. They did not. Fridell noted in 2014 that Ethiopian beans continued to follow international prices much as other origins' beans did, and the pattern has held right through the recent price regime. ICO market data, current as of the June 2026 market report, put the indicator price relevant to Ethiopian coffee at an average of 343.08 US cents per pound across February to May 2025, near the peak of the historic spike, and 265.89 cents across the same months of 2026, a fall of 29 percent, tracking the general market down. Owning the brand did not exempt anyone from the C-price.
What Ethiopia does have, two decades on, is scale and a growing claim on its own crop. USDA's Foreign Agricultural Service, in its circular of July 2026, forecasts a record Ethiopian harvest of 12.1 million bags for 2026/27, with exports at 7.1 million bags, restrained not by supply but by domestic consumption climbing toward 5 million bags. Ethiopians increasingly drink Ethiopian coffee, which is its own kind of value capture and owes nothing to a trademark office in Washington. In 2008 the government also created the Ethiopian Commodity Exchange, Africa's first commodity exchange, as the mandatory channel for most of the country's coffee. How much extra value these institutions have created, and how much of it reaches small farmers, remains, in Fridell's careful phrasing, unclear.
What a trademark will not do
It is worth being precise about the limits, because the Yirgacheffe on a shelf tempts a shopper into a chain of assumptions the trademark does not support. The registration will not set a price for anyone's cherry. It will not tell you what a licensee paid for the beans in the bag, to whom, or in what year. It does not verify quality, which is a separate grading question, and it does nothing at all for the picker, who sits below the farm gate in a part of the chain no consuming-country paperwork reaches. The licence obliges a company to promote Ethiopian coffee. It does not oblige the company to pay more for it, and at a royalty of zero, the direct revenue to Ethiopia from the marks themselves is, by design, nothing.
None of that makes the campaign pointless. The principle it established has been borrowed and argued over ever since, and there is a real, if unmeasurable, value in the fact that three Ethiopian names are now legally Ethiopia's in 30 or so jurisdictions. But the gap between the principle and a farmer's income is the entire subject, and pretending the gap closed in 2007 is the kind of feel-good compression this story usually gets.
So read the word Yirgacheffe on a bag for what it is: a place name, a flavour reputation earned over generations, and a brand owned by the Ethiopian state and lent to the roaster for free. It tells you where, roughly, and it tells you the roaster signed a licence. A single origin statement carries less information than most people assume, and a name on a bag is a signal that needs corroborating before it becomes a reason to pay more. The corroboration that actually touches the farmer is a published price for the lot with a date on it. When a roaster gives you that, the beautiful word on the label finally has a number under it, which is the thing Ethiopia fought two years to make possible and could not, on its own, guarantee.