Payment for ecosystem services, explained without the jargon

What follows is an argument about payment for ecosystem services, and about where the received version of it stops being true.

Paying for outcomes rather than crops

Traditional agricultural support paid for production or for area farmed. Ecosystem service payments instead pay for clean water, stored carbon, flood attenuation or habitat.

The shift changes what land is economically rational to do, which is the entire point. The catch is that an outcome has to be measurable at a cost proportionate to the payment, which is why carbon and water quality lead the field while soil health and biodiversity, both slower and more expensive to measure, lag behind them.

Who actually pays

Water companies fund catchment management because it is cheaper than treating the water downstream. Governments fund public goods through agri-environment schemes.

On the ground, private buyers fund carbon and biodiversity units, with quality and credibility varying enormously. That private demand rests largely on voluntary corporate commitments, which can be revised or quietly dropped, so a landowner signing a thirty-year agreement is taking a view on a market that has only existed for about a decade.

Additionality is the hard part

A payment should fund something that would not otherwise have happened, or it is simply a transfer. Proving that counterfactual is genuinely difficult and is where most criticism of these schemes lands.

In practice, weak additionality is the main reason some carbon offsetting has been discredited. The perverse version of the same problem is that paying for improvement rewards whoever starts from the worst position, which is why credible standards set the baseline from historic records rather than from the state of the land on the day somebody applies.

Permanence and term length

Carbon stored in trees or peat only counts if it stays there, which requires long-term commitments. Landowners reasonably resist agreements that bind successors for decades without adequate compensation.

Short contracts, meanwhile, do not justify the capital cost of changing a land use. The mechanisms for holding it are covenants that run with the land and buffer pools in which a share of every project's credits is withheld against reversals elsewhere, and the size of that buffer tells you how much confidence a standard has in its own permanence.

What makes schemes work

Clear measurement, long terms, reliable payment and advisory support are the features common to successful ones. Schemes that pay for actions rather than outcomes are simpler to administer and less precisely targeted.

Locally, both models are in use, and the trade-off between them is the live policy argument. Aggregation is the other practical requirement, because a single small holding cannot generate a volume worth transacting, so brokers and farmer clusters have appeared to bundle them — and how much of the payment survives that bundling is a question worth asking early.

How much of that applies depends on the terrain you are actually in.

Two ways the accounting goes wrong

Leakage is production moving rather than stopping: take a field out of cultivation here and, with demand unchanged, the crop is grown somewhere with weaker environmental rules. Stacking is selling several services off the same hectare — carbon to one buyer, a biodiversity unit to another, flood attenuation to a third — which is defensible only if every buyer knows what they have actually bought. Standards and registries address both by requiring disclosure and by applying a discount for leakage, but those discounts are modelled estimates rather than measurements.

The two questions that expose most of this in any scheme prospectus are what happened to the displaced production and whether the same hectare appears in another register.

The takeaway

Ask who is buying, for how long, and what would have happened anyway.

Go slowly enough to notice, and most of this stops being advice and starts being obvious.