A land manager is told there is money in nature now. A buyer will pay for the biodiversity units on the back field, or the carbon in the replanted gill, or the flood water held back from a town downstream. The question that arrives with the cheque sounds practical and is anything but: who will pay, and why? It is now the organising question of UK nature-recovery policy — the shift, as BioCap’s 2026 report Private Finance for Nature Recovery puts it, from “build it and they will come” to “who’s going to pay for it?” — and it carries a buried premise. To finance an action is to assume it would not otherwise happen. The moment you accept payment for something, you concede that, unpaid, you would not do it.
That premise is true less often than the market assumes, and where it is false the money does real damage. The same payment might be buying three quite different things. The first is a practice that already pays for itself: healthier soil, fewer inputs, a sounder rotation. These are right and also profitable, so financing them subsidises a decision the manager should reach unaided. The second is a genuine transition with a cost hump — a felling, a wetland, a decade before the work earns its keep — where temporary capital clears a real barrier. The third is a public good with no private return at all, which can rest only on ongoing payment or on duty. Private finance is built for the middle case, and blended de-risking capital serves it well. It is routinely mis-sold for the other two, dressing up the profitable as heroic and unable to hold the unprofitable for long.
The holding is the deeper problem. Ecosystems need indefinite continuity; the instruments do not have it. Biodiversity Net Gain secures thirty years. Amazon Web Services’ wetlands in the Kennet catchment were driven, on the report’s own account, by AWS’s target of being “water positive” by 2030 — a date past which the reputational driver simply expires. So every financed recovery has a cliff-edge written into its contract, a moment when revenue and pressure both lapse and the habitat is left to fend for itself in a landscape reorganised around the assumption of payment. This is the failure mode set out in what-gets-removed-does-not-come-back: a system optimised to a metric loses the slack that let it persist without one.
Then there is where the money goes. The advice is candid: start with business exposure, not just biodiversity need. Route capital to where a high-value asset is at risk, because that is where a willing payer sits. As efficiency this is unarguable. As conservation it is a quiet surrender, because the most degraded land is exactly the land with no exposed corporate asset to defend it. The market allocates not to need but to proximity to money, which is how proxy-capture works in every domain it touches. And the orthodoxy admits its own strain: that buyers “have rarely materialised at the expected scale,” that average BNG unit prices have “steadily declined as supply has expanded,” that voluntary carbon “will rarely generate sufficient returns” — before recommending the market be built anyway.
Underneath all of this is a finding the conversation prefers to forget. When Gneezy and Rustichini studied a fine for collecting children late from a nursery, late collections rose: the fine became a price, and an obligation became a transaction one could choose to buy. Pricing an intrinsically motivated behaviour can reduce it, because it moves the question from “is this right?” to “is this worth it?” — and the second question, once asked, does not switch off when the payment stops. A market for stewardship risks teaching farmers and firms that nature is worth protecting only when someone pays. That is an expensive lesson to un-teach, and it points the same way as from-optimised-to-resilient: the things that make a system endure are the things a price erodes.
None of this argues against finance. It argues for a rule. Stop paying for what already pays; name it as good practice and expect it. Pay once, as bridging capital, for transitions with a real and temporary cost — the Meon Springs wetland clearing a nutrient-neutrality hump — and design the exit before the entry. Pay forever, through public mechanism rather than corporate whim, for the public goods that have no private return and never will, as the Wyre’s flood schemes ultimately need. Stop pretending a thirty-year contract can stand in for forever. The free lunch was never free; it was paid for by a duty we are now busy converting into a price. The work is to know which of the three you are looking at before the cheque arrives, and to refuse the cases where the money buys less than it costs.
Further reading
the-legibility-trade — why making nature legible to a market changes what survives in it.