A forest remains standing after a conservation project begins. The fact is observable, but the project's climate contribution is not identical to the carbon stored in those trees. Some or all of the forest might have remained standing without the project. To claim an additional benefit, the project must compare what happened with a credible account of what would otherwise have happened. The difficulty is structural: one landscape supplies the observed outcome, while the relevant alternative remains counterfactual. A satellite can help describe the forest that exists; it cannot photograph the absent history against which the project's effect is assessed.
Additionality names this requirement that the credited benefit exceed the outcome expected without the intervention. It is conceptually straightforward and empirically demanding. If a forest faced imminent clearance, protection may change its trajectory substantially. If the land was already secure, the same visible forest may represent little additional change. The project boundary and starting date can influence the comparison. Choosing a particularly pessimistic baseline makes ordinary outcomes look exceptional, while an unrealistically optimistic baseline can understate a worthwhile intervention. The central question is therefore not merely whether monitoring is accurate but whether the comparison being monitored is defensible.
A hypothetical developer might estimate that a tract would lose half its trees within ten years, then claim credit for avoiding that loss. Even perfect observation of the actual tract cannot validate the assumed rate of future clearance by itself. Evidence about comparable areas, legal protections, economic pressures, and planned activities becomes relevant. Each source has limitations. Nearby areas may differ in access; stated plans may be strategic; historical trends may not persist. The baseline is a model, and its uncertainty should remain visible. Certification cannot transform a disputed projection into an observed fact simply by assigning it an official category.
Leakage creates a further problem. If protecting one area shifts logging or agricultural expansion into another, local success overstates the wider benefit. The displacement may occur through identifiable activities or through market responses that are harder to trace. This does not mean every project experiences complete displacement. It means the relevant accounting boundary must extend far enough to examine plausible transfers. A conservation map can display a protected zone with impressive clarity while leaving the consequences beyond its edge unmeasured. The environment does not necessarily respect the administrative boundary that makes the project convenient to report.
Permanence concerns time. Carbon retained in vegetation can later be released through fire, disease, clearance, or other disturbances. A benefit lasting several decades is not equivalent in every respect to one lasting much longer. Accounting arrangements may use reserves, replacement obligations, or conservative estimates to address reversal risk. Their credibility depends on the resources and institutions available when the reversal occurs, not only on a promise made at issuance. A guarantee extending beyond the lifetime of its guarantor requires explanation. The duration of the environmental claim and the duration of the institution supporting it must be considered together.
These issues become especially consequential when a credited benefit is used to compensate for emissions elsewhere. Funding conservation can be valuable without establishing that a separate emission has been neutralised. The compensatory claim requires a stronger equivalence: the credited quantity must correspond to a real, additional, appropriately durable effect, with relevant displacement addressed. This distinction is often blurred because the project itself may have evident benefits for biodiversity or livelihoods. Those benefits deserve recognition on their own terms. They do not automatically prove the precise carbon equivalence required by an offsetting statement.
The language of fungibility intensifies the challenge. Markets work more easily when units can be treated as interchangeable, yet ecological interventions differ in location, timing, risk, and social consequences. Standardisation can make comparison possible, but it also decides which differences will be ignored. A common unit is an accounting achievement rather than proof that every underlying activity is identical. The question is whether the excluded differences are immaterial to the purpose of exchange. If they alter the reliability of the climate claim, treating the units as equivalent simplifies the market by weakening the environmental meaning it was supposed to preserve.
There is a distributional dimension to baseline construction as well. Communities that have historically protected forests may appear to offer fewer additional reductions than actors threatening to destroy them. A poorly designed incentive can therefore reward the credibility of prospective damage more than the continuity of stewardship. This is not an unavoidable consequence of conservation finance, but it is a risk inherent in paying solely for improvements against projected harmful behaviour. The mechanism must consider how today's baseline rules affect tomorrow's incentives. An accounting system can shape the conduct from which its future comparisons will be constructed.
Uncertainty does not imply that all conservation claims are worthless. That conclusion would confuse imperfect measurement with the absence of any defensible evidence. Some projects can provide strong support for their effect, while others cannot. The appropriate response is differentiation: conservative baselines, transparent methods, independent scrutiny, and claims proportionate to the evidence. A range may be more honest than a single confident figure. Where the compensatory claim is weak, finance might still be justified as a contribution to conservation rather than as a licence to declare an unrelated emission cancelled. The wording changes because the evidential burden changes.
Buyers also have incentives that deserve inspection. A cheap unit may be attractive precisely because its baseline or durability assumptions are generous. Competition on price alone can favour projects whose claims are easiest to inflate rather than whose benefits are most reliable. Better information helps, but the purchaser must actually value integrity over the appearance of compliance. If the principal objective is a convenient public statement, additional disclosure may be absorbed as paperwork. Market design must therefore align rewards with the environmental purpose instead of assuming that a transaction labelled green necessarily does so.
Monitoring technology can reduce some uncertainties without dissolving the conceptual structure. More frequent observations can identify changes sooner, improve estimates of vegetation, and reveal some forms of displacement. They cannot independently settle whether a threatened activity would have occurred, whether future institutions will replace reversed benefits, or which communities possess legitimate claims over a landscape. Technical measurement and institutional judgement are complementary. Treating one as a substitute for the other encourages either unjustified technological confidence or an equally unjustified dismissal of useful evidence because it is not omniscient.
The forest's visible persistence is thus only the beginning of a serious climate account. The account must explain an invisible alternative, a spatial boundary, a time horizon, and the relationship between a local intervention and a wider claim. Each element can be scrutinised without denying the value of keeping forests standing. Indeed, scrutiny protects that value from being converted into an exaggerated certificate. Conservation deserves financing because of what it achieves. The more ambitious assertion that it cancels something elsewhere must earn its authority through an equivalence that survives examination, not merely through the reassuring presence of trees.
Source notes
[9] IPCC, Land Use, Land-Use Change and Forestry (2000), project baselines and additionality. Background: the without-project comparison. Open primary source
[10] IPCC, AR6 Working Group III Technical Summary (2022). Background: carbon-removal storage timescales and reversal risks. Open primary source