Can you find out what a coffee farmer was paid?

How to read the ratings A row carries a rating only when a dated document is on file for it. A row that has been researched and then withheld, because the evidence chain does not close, is shown as held rather than as unchecked: those are different findings and the map keeps them apart. An origin with no rating at all has not done badly: it means no row on it has a dated document on file yet. A benchmark price is what growers were offered; the farm-gate row asks what they actually received. The Grower price layer covers both of those rows at once, and a country can stand at a different point on each. Where they differ the country is drawn at the furthest-along of the two, in this order: a published rating, then held, then reviewed without a determination. A cell marked reviewed has been searched and closed for this cycle without a rating: the evidence found, the decisive blocker and the condition that would reopen it are all recorded. The 2026 research cycle is closed and the evidence record was frozen on 11 September 2026: all 306 cells across the 51 origins were reviewed to a defined release status. Held cells, reviewed cells and the one unresolved rating dispute are outcomes of the method, not signs of an unfinished build. Comparative results drawn from the 102 producer-price cells are separately under independent validation. That is a finding, not unfinished work, and it is deliberately not converted into a rating merely because the search is over. Colombia is the current example: its benchmark row carries an unresolved rating dispute, where the evidence supports competing states and neither has been adopted, while its farm-gate row was reviewed without a determination. Disputed outranks reviewed, so it is drawn as disputed.