Leakage in Nature-Based Carbon Projects, Explained Simply

Imagine a project that protects a stretch of coastal wetland from development and sells carbon credits on the resulting benefit. Six months later, the developer builds the same marina two bays down the coast, destroying the same amount of carbon-rich habitat. The Earth's atmosphere sees no benefit at all — emissions simply moved from inside the project boundary to outside it. The credits, however, were still sold and claimed as real.

That displacement is leakage, and it is one of the most important — and most misunderstood — integrity concepts in carbon markets. A project can have perfect measurements, genuine additionality, and honest permanence accounting, and still deliver little to the atmosphere if its carbon-saving activity pushes the emitting activity somewhere unmonitored. Understanding leakage is understanding why where a project stops mattering as much as what happens around it.

What leakage is, in one sentence

Leakage is the increase in emissions (or decrease in removals) outside a project's boundary that happens because of the project — the atmospheric carbon ledger's way of insisting that moving a problem is not the same as solving it.

The three kinds of leakage

Leakage comes in three recognisable species, and good accounting treats each differently.

Activity shifting

This is the intuitive kind: a specific actor moves their activity from inside the boundary to just outside it. The fisher excluded from a newly protected bay takes the same boat to the neighbouring reef. The developer denied one mangrove parcel purchases the adjacent one. The activity — and its emissions — is conserved, not reduced, and the project's climate benefit is correspondingly smaller than its accounting claims.

Market leakage

This is subtler and often larger. When a project removes product from a market — timber not harvested, fish not landed, land not developed — prices shift, and other producers expand supply to fill the gap. A forest carbon project that takes a timber concession out of production does not reduce global wood demand; it reshuffles who cuts which forest. Market leakage operates through price signals, crosses borders effortlessly, and cannot be fixed by watching the boundary — only by estimating how supply responds.

Ecological leakage

Least formalised but very real in coastal systems: the project changes physical or biological conditions beyond its boundary. Restoring a tidal wetland may alter sediment flows or erosion patterns downstream; protecting one reef displaces fishing effort that was itself controlling a species, letting urchin populations boom on the next reef. The climate system doesn't know the difference between a policy effect and a physical effect — the carbon moved is the carbon moved.

Why leakage is so hard to see

Leakage's cruelty is that it happens where monitoring is weakest: outside the project, outside the report, often outside the country. Measuring it requires answering a layered counterfactual — what would the displaced actors have done instead, and what emissions follow? — across actors who were never surveyed and may actively avoid attention. A project's own monitoring fence, however rigorous, cannot answer that question from inside.

There are honest responses. Some are statistical: registries use leakage deductions, multiplying project credits by a discount factor based on the assessed risk of displacement (high-market-integrated commodities get bigger deductions than place-bound activities). Some are structural: the strongest is to widen the accounting boundary until displacement has nowhere to hide.

Jurisdictional approaches: shrinking the loophole

Jurisdictional carbon accounting — crediting at the scale of a whole state, country or fisheries region rather than a single parcel — deals with leakage by making it theoretically impossible: emissions within the jurisdiction fall or they don't, and a tonne displaced internally still shows up in the ledger. Where national or jurisdictional programs "nest" individual projects inside them, the project's claims reconcile with the jurisdiction's total, and internal leakage stops mattering quite so much. External leakage — displacement across jurisdictional borders — remains, but it is visible at a scale that trade and production data can actually track.

Our explainer on the full life of a blue carbon credit covers how leakage assessment slots into project design and verification at every stage.

Leakage where the sea meets the ledger

Blue carbon makes leakage vivid because coastal economies are so dense with displaced users. A mangrove protected for carbon displaces fuelwood cutters and pond aquaculture; a seagrass meadow protected from trawling displaces the trawl fleet; a restored reef displaces the fishers who worked it. The kelp-specific anatomy — how restoration shifts fishing and aquaculture pressure from stone to stone along a coast, and why good projects design against it — is detailed in our companion piece on leakage in kelp carbon.

The principles generalise beyond kelp:

• Displaced effort concentrates. Removing pressure from a patch of ocean doesn't remove the pressure — it intensifies it everywhere nearby, often on the most vulnerable remaining habitat.
• Livelihood alternatives are leakage controls, not charity. A project that funds displaced fishers into restoration work, aquaculture or monitoring is not being generous; it is doing carbon accounting. The alternative is that the same effort reappears, uncounted, somewhere unmeasured.
• Natural leakage still counts. Marine systems also leak biologically — a protected reef exports larvae, drifts nutrients, changes food webs down-current. Most of these flows are benefits ("leakage" running in the climate's favour), which is why serious accounting attempts to estimate both directions, not assume them away.

What good project design does about it

A credible leakage response is written into the project from the feasibility stage, not retrofitted to answer a verifier:

1. Map the displaced users. Document who uses the site and what they will do without it — before the project starts, not after the credits are claimed.
2. Design absorption. Provide credible alternatives (gear shifts, adjacent livelihoods, employment) sized to the displaced effort, not to a brochure.
3. Apply the deduction honestly. Take the methodology's leakage percentage where it applies, and argue for more where realism requires it — not less where marketing prefers it.
4. Monitor outside the boundary. Sample the neighbouring coast, at least as an indicator. Leakage you never look for never appears — which is not the same as never happening.
5. Prefer jurisdictional nesting where it exists, so internal leakage stays visible in the official ledger.

The atmosphere doesn't care which side of a boundary a tonne sits on. Every integrity mechanism in carbon markets, at bottom, is an elaboration of that sentence — and leakage is the one that polices the map, not the territory.

Key takeaways

• Leakage is emissions displaced outside a project's boundary by the project itself — a real benefit inside the line paired with a hidden cost outside it.
• It flows through three channels: activity shifting (actors move), market leakage (prices respond), and ecological leakage (the system itself reacts). All three reduce the project's true atmospheric benefit.
• Standard treatments combine leakage deductions with, increasingly, jurisdictional accounting that abolishes internal displacement as a category.
• In blue carbon, displaced fishing and development effort is the dominant channel — which is why livelihood alternatives are carbon controls, and why monitoring must extend past the boundary.
• Any project claiming zero leakage without a jurisdictional boundary and positive findings outside its fence deserves a harder question: where did you look?

FAQ

Is leakage the same as impermanence?

No. Impermanence (or reversal) means credited carbon returns to the atmosphere from inside the project later — our article on reversal risk covers that. Leakage means emissions rise outside the project because of it. Both erode the claim, but through different mechanisms and with different fixes.

Do all carbon methodologies account for leakage?

Most credible modern methodologies either require a leakage assessment and deduction or exclude project types with untracked leakage risk entirely. Older and weaker methodologies didn't always — one reason vintage and methodology choice matters when evaluating credits.

Can leakage be positive?

Yes — "positive leakage" (sometimes called spillover) happens when a project reduces emissions beyond its boundary: protection that lets larvae seed recovery elsewhere, or a working blueprint that neighbouring communities copy. Good projects try to claim it; none can rely on it to offset negative leakage without matching evidence.

Why don't projects just measure leakage directly?

Sometimes they can and should — displaced effort surveys and neighbouring-site monitoring are real things. But market leakage through price signals, and displacement across distant borders, can't be measured at any single site; it has to be estimated from production and trade data, which is why deductions exist.

How big are typical leakage deductions?

They vary widely by project type and market integration — from negligible for tightly place-bound activities to large where displaced commodities are highly traded. Any specific percentage quoted without a methodology context should be treated as out of context: what matters is that the deduction be derived, not decided.