Offset or illusion? A new study puts numbers on what biodiversity net gain actually delivers

When the United Kingdom made biodiversity net gain a legal requirement for most new developments in February 2024, the policy arrived with two convictions baked in: that growth did not have to come at the expense of nature, and that a damaged hectare somewhere else could balance a hectare flattened somewhere here. A peer-reviewed paper published on 17 July 2026 puts hard numbers on the second conviction, and the picture it draws is conditional.

The paper, summarised that afternoon by Phys.org, identifies the ecological conditions under which biodiversity offsetting actually delivers what its proponents promise, and the conditions under which it becomes an elaborate accounting fiction. The distinction matters because offsetting has shifted from voluntary corporate gesture to a load-bearing element of planning law in several jurisdictions, including England, France, and a growing list of US states.

What the study actually measured

The research team set out to do something the offsetting industry has largely avoided: test whether the promised gains materialise on the ground, under what conditions, and on what timescale. The headline finding is that offsetting can work, but only when three conditions are simultaneously met. The receiving habitat is ecologically equivalent to the one destroyed. Restoration starts early enough that the lag does not exceed the planned development's lifetime. And the offset site is protected in perpetuity, not for the thirty-year window common in current schemes.

Without those three conditions stacked, the offsets under-deliver or fail outright. The study treats the lag problem as the most consequential. A forest replaced in twenty years is not equivalent to a forest flattened today; the species that depended on the original habitat do not wait. The arithmetic of "one hectare lost, one hectare gained" only holds when the clock starts in the developer's favour.

The counter-narrative developers prefer

Industry groups have spent the better part of a decade arguing that any offset is better than none, and that strict equivalence requirements raise costs without raising conservation outcomes. The position has a grain of truth: a restored wetland is not nothing, and demanding perfect equivalence can block projects whose net impact is genuinely modest. The counter-position, also articulated in planning literature, holds that without equivalence the whole framework becomes a permission slip. Once one habitat can substitute for another on price, every developer has an incentive to push for the cheapest substitute and call it compensation.

The study does not resolve that tension so much as it quantifies it. Where equivalence is preserved, offsetting produces measurable gains. Where it is not, the gains are statistical noise. The framing question for policymakers is therefore not whether to allow offsetting, but how strictly to police the equivalence threshold. So far the political answer in most jurisdictions has been loose.

Why the legal architecture lags the science

Biodiversity net gain policies were drafted when the evidence base was thinner and the political pressure to keep building was stronger. England's framework, for instance, sets a 30-year protection horizon for most offsets and permits substitution across broad habitat categories. Both choices made the law easier to administer and cheaper for developers, and both choices run directly against what the July 2026 paper identifies as success conditions. The gap is structural: planning law optimises for transaction speed and political durability, while ecology optimises for time horizons measured in centuries.

The study's contribution is to give regulators a defensible reference point. A government that wants to tighten equivalence rules, lengthen protection horizons, or restrict substitution categories now has peer-reviewed evidence rather than NGO assertion. That shifts the burden of proof in a way that should worry developers and reassure conservation groups, though neither side will read the paper as a clean win.

Stakes for the next planning cycle

The practical question is whether any jurisdiction moves to close the gap between its biodiversity law and the conditions the paper identifies. England is the closest test case, with a planning regime already under review and a treasury always alert to construction-sector friction. France operates a similar system with stricter equivalence requirements on paper and a weaker enforcement record in practice. Several US states run offset programmes with even weaker protection horizons and even broader substitution allowances.

The structural pattern is familiar: a policy innovation designed under one set of scientific assumptions, locked in by political economy, and then slowly forced to reckon with evidence that arrived after the law was written. The study does not settle the offsetting debate. It does, however, give both sides something they have not had until now: a common empirical baseline against which the next round of rulemaking can be measured.

How Monexus framed this: the wire reports on biodiversity offsetting tend to either celebrate the policy as innovative or critique it as greenwashing. The more useful framing is the one this study enables, which is that offsetting is a conditional instrument whose success depends on design choices that most jurisdictions have not yet made.