Leakage (offset projects)

“A problem has not been solved when it has merely crossed the line on the map or the organisation chart. ”

Leakage is best known in climate policy and carbon accounting. A project reduces emissions in one place, but activity shifts and increases emissions elsewhere. The reported boundary improves while the wider system changes less, not at all or in the opposite direction.

The Greenhouse Gas Protocol describes leakage as a secondary effect in which changes in the availability or quantity of a product or service lead to greenhouse gas changes elsewhere. The IPCC uses the concept in relation to trade, land use, production and policy spillovers. A forest-conservation project can restrict clearing inside its boundary while agricultural expansion moves to neighbouring land.

A jurisdiction can reduce domestic industrial emissions while importing more carbon-intensive goods. A company can lower reported operational emissions by outsourcing production to a supplier.

In each case, the location of pressure changes. Leakage is not limited to carbon. A buyer can eliminate child labour findings by excluding smallholders and shifting sales to less visible markets. A wage policy can apply to direct employees while work moves to contractors. Water use can fall at one facility while water-intensive processing is relocated upstream.

Boundary improvement can coexist with system deterioration. The causal link matters. Harm elsewhere is not leakage merely because it occurs at the same time. The intervention should create incentives or constraints that contribute to displacement. Market prices, land demand, labour substitution and supply-chain reconfiguration are common pathways. Direct leakage may be observable.

Farmers denied access to one area clear another. Market leakage can be diffuse: reduced supply raises price and encourages production elsewhere.

Estimates become uncertain as the system expands, but ignoring the effect because it is difficult to measure is not neutral. Design can reduce leakage. Landscape and jurisdictional approaches widen boundaries. Alternative livelihood support reduces pressure to relocate harmful activity. Supply-chain engagement can prevent outsourcing to lower-control contractors.

Policy coordination limits movement across jurisdictions. Accounting rules should specify relevant sources, methods and deductions. Conservative estimates may be appropriate when evidence is weak and crediting claims are high stakes. Monitoring should extend beyond the immediate intervention zone where displacement is plausible.

Positive spillovers also occur. Technology can diffuse, norms can change and neighbouring farms can adopt better practice. These benefits should be evidenced separately rather than used casually to offset leakage. Symmetry in accounting matters. Leakage can reveal a deeper design flaw: the intervention treated a location rather than a driver.

If deforestation is caused by commodity demand and land inequality, protecting one parcel without addressing demand or access may only redirect pressure. System diagnosis is part of prevention. Organisations sometimes resist wider boundaries because they cannot control all effects. Control is not the standard for relevance.

Due diligence and credible claims require attention to effects caused or contributed to through incentives and relationships, even where direct authority is limited. Leakage takes several forms.

Activity-shifting leakage occurs when the same producer or company moves pressure elsewhere. Market leakage occurs when reduced supply in one place raises prices and encourages expansion in another. Ecological leakage occurs when an intervention changes water, fire, wildlife or pollution pathways beyond the site.

Social leakage appears when cost, hazardous work or exclusion is transferred to less visible workers and suppliers. Responses therefore need an appropriate scale. Project boundaries may be sufficient for direct implementation but not for judging net effect. Jurisdictional monitoring, demand-side measures, supplier-wide rules and policy coordination can reduce displacement that parcel-level projects miss.

Not every wider change can be measured exactly, but plausible leakage pathways should be identified, monitored and reflected conservatively in claims.

The discipline is to follow pressure beyond the reported success. Where did production, labour, emissions or cost move? Who absorbed it? Leakage tests whether the intervention changed the system or only its visible surface.

Practical application

Map displacement pathways before implementation, including land, labour, markets, contractors and value-chain stages. Define monitoring zones and indicators beyond the project boundary and use conservative accounting where leakage is material but uncertain. Address drivers through coordinated policy, landscape approaches, supplier engagement or demand-side action.

Report residual leakage and avoid claims based solely on boundary performance.

Map plausible displacement pathways before implementation: other land, suppliers, workers, products, markets and time periods. Assign indicators and geographic boundaries to the highest-risk pathways and use conservative deductions where effects cannot be measured directly. Claims should report net change after material leakage, not only the improvement inside the project fence.

Review leakage over a period long enough to detect delayed displacement. Pressure may move only after contracts, enforcement or local prices have had time to adjust.

Why it matters

Leakage can convert local success into system-wide illusion. It overstates credited outcomes, transfers harm to less visible groups and rewards interventions that optimise reporting boundaries.

Common misconception

Leakage is often treated as any negative effect outside a project. It is specifically an effect caused by displacement or market response linked to the intervention.

Connections

Boundary choices in Baseline, Metric and Target affect whether leakage is visible. Additionality and Counterfactual determine claimed change. Rebound Effect describes increased use after efficiency gains, while Landscape Approach can reduce spatial displacement.

A question worth asking

Which pressure did your intervention remove, and where would that pressure most plausibly reappear if its underlying driver remained unchanged?

Selected references

Greenhouse Gas Protocol. 2005. The GHG Protocol for Project Accounting. Greenhouse Gas Protocol. 2026. Land Sector and Removals Standard and Guidance. Intergovernmental Panel on Climate Change. 2022. AR6 WGIII Annex I: Glossary.

Meyfroidt, P. et al. 2020. Focus on Leakage and Spillovers: Informing Land-Use Governance in a Tele-coupled World. Environmental Research Letters 15(9). Atmadja, S. and Verchot, L. 2012. A Review of the State of Research, Policies and Strategies in Addressing Leakage from Reducing Emissions from Deforestation and Forest Degradation.