Certification describes what a cooperative is allowed to sell, not what it manages to sell. Those are different things, and the gap between them is where most of the argument about whether certification works actually lives.
A certified cooperative typically sells only a minority of its coffee on certified terms. The rest goes onto the ordinary commodity market at the ordinary price, because certified supply substantially exceeds certified demand. So a farmer can be fully certified, pay the full cost of being certified, and receive the certified price on a fraction of what they grow.
Why the gap exists
Certification is supply-led. A cooperative decides to certify, pays the fees, meets the standards, and is then eligible to sell at certified prices. Nothing about that process creates a buyer.
Demand comes from the other end entirely: roasters and retailers choosing to buy certified, in the volumes they choose, at the times they choose. When the certified volume available exceeds what buyers want, the surplus does not sit in a warehouse. It is sold as conventional coffee, at the conventional price, and the certification simply does not apply to it.
Fieldwork in Nicaragua found cooperatives routinely selling a large share of their crop on conventional markets for exactly this reason. The certification was real. The buyer was not there.
The number, and the sentence after it
Fairtrade publishes that cooperatives sell on average around 35% of their coffee on Fairtrade terms. That figure circulates widely, usually as reassurance: a third is not everything, but it is substantial.
The next sentence in the same document is the one that matters. It explains that this proportion has risen in recent years mainly because cooperatives with a low share of Fairtrade sales have been voluntarily dropping certification.
Read that again, because it inverts the meaning of the number entirely.
The average improved because the cooperatives for whom the economics did not work left the system. The figure is not measuring certification getting better at finding buyers. It is measuring who remains after the ones it failed walked away. That is a survivorship statistic, and quoting it as evidence of success gets the causation exactly backwards.
There is corroboration that is difficult to argue with. Fairtrade's Latin American producer network operates a subsidy fund, and its eligibility criteria explicitly include organisations that were de-certified within the past year for failure to pay. A scheme does not build a fund for that scenario unless the scenario is common enough to plan for.
Why exit is stronger evidence than any study
Most of the argument about certification is conducted through impact studies: surveys of certified households, comparisons against non-certified neighbours, income effects measured over a season or two. Those studies matter, and they disagree with each other a great deal.
Producers leaving is a different kind of evidence. It is a costly decision made by the people with the most complete information about their own economics, and it points in one direction. A cooperative that drops certification is not registering an opinion in a survey. It is giving up the price floor, the premium and the market access, having concluded that what it was paying to keep them was more than they were worth.
That is a revealed preference, and it comes from the producers rather than from anyone selling them a scheme or studying it.
What the label will not do for you
The certification mark on a bag will not tell you what price this particular coffee fetched, will not tell you what share of that cooperative's crop found a certified buyer this season, and will not tell you whether the growers who could not make the economics work are still members. None of that is knowable from the front of a bag, and no amount of reading the label harder will produce it.
It does not mean certification is a fraud. The mechanism is real: there is a genuine minimum price, a genuine premium paid to the cooperative, and genuine standards on labour and environment. For a cooperative with reliable certified buyers, it works roughly as described.
It does not mean you should stop buying certified coffee. The alternative for most shoppers is coffee with no traceability and no standards at all, which is worse on every dimension.
And it does not mean the 35% figure is false. It is accurate. It is simply measuring something other than what people quote it to mean.
What it means is narrower and more useful: the label tells you a cooperative was eligible, not that the coffee in your hand was sold at the certified price, and not that the scheme is working for the farmers who are no longer in it.
The question to ask instead
If you want to know whether more money reached the grower, certification is the wrong thing to look at. Two better signals:
Does the roaster publish what it actually paid? Not a policy statement, a price. A roaster that publishes the price per pound, the farm or cooperative name, the harvest year and the lot size has exposed itself to contradiction by the cooperative, by the exporter, and by any journalist who cares to check. That is a falsifiable claim, which puts it in a different category from a logo.
And does the report cover everything they buy, or three hero lots? This is the question that separates disclosure from marketing, and almost nobody asks it. A transparency report covering 100% of green purchases means something. One showcasing the three lots that make the best story means very little, and both are called transparency reports.
Why we can say this
We do not sell coffee. We have no bean to defend, no subscription to protect and no sourcing story of our own riding on the answer.
That matters here more than it does anywhere else on this site, because almost everyone writing about certification is also selling the coffee it certifies. We are not, which is worth stating plainly rather than leaving you to work out.
It also cuts the other way, and we should say so: we sell objects whose own supply chains we are still in the process of documenting properly. Holding coffee to a standard we have not yet met on ceramics would be an easy thing to do quietly, so we would rather name it. Our own provenance work is published separately, including the parts that currently read "unknown."