Why cacao yield figures are so unreliable

Compare a research-station figure for a well-managed block of improved planting material with a national average derived from dividing a country's exports by its estimated planted area, and the two can differ several-fold.

Both numbers are real. Neither is wrong. They are answers to different questions, and almost every argument about cacao productivity consists of quoting one at the other.

A yield figure has to make several decisions before it exists, and the sources rarely record which ones they made.

Planted area or bearing area — a farm with a third of its trees too young to bear looks unproductive on the first basis and normal on the second. Whose farms — a trial block, a demonstration farm, a survey of smallholders, or a national statistic derived from export tonnage. What age — cacao yield varies enormously across a tree's life, and a district mid-way through replanting reads low for reasons that are a sign of investment rather than neglect. Which season — cacao is strongly seasonal and many regions have two crops of different sizes, so a figure covering the wrong window is not comparable. Dry or wet beans, and at what moisture — a several-fold difference that is purely definitional.

And underneath all of it: how the planted area was estimated at all. In several major producing countries this is genuinely uncertain, which means the denominator of the most-quoted yield figures in the world is an estimate.

Low quoted yields are the standard justification for intensification programmes: if the average farm produces far less than a trial plot, the gap looks like an opportunity and the recommended response is improved material, fertiliser and reduced shade.

That argument is sometimes right. The reason to be careful is that the yield gap is doing a lot of rhetorical work for a figure with this much definitional slack, and closing it has costs that the yield number does not price: shade reduction changes disease pressure and pollinator habitat, fertiliser requires cash a household may not have, and the returns are collected years after the outlay.

So an intensification case built on a yield gap is a case that has to be argued on the ground it is being applied to. It is not settled by the arithmetic.

The directions are well established even where the magnitudes are not, and directions are what transfer between environments.

Yield rises then falls across a tree's life. Disease pressure reduces it, sometimes drastically, and black pod is the largest single cause of loss worldwide. Shade reduces light available for production and buffers stress, and which effect dominates depends on the site. Good planting material outperforms poor planting material on the same ground. Replanting costs production before it adds any.

Each of those is a mechanism a reader can apply. None of them requires a number, and this catalogue states them and stops — the same discipline the pests section applies to loss percentages, and for the same reason: a single figure would be repeated everywhere without the conditions that produced it.