Impact

"Impact begins where delivery ends and the evidence must explain not only what changed, but why the claim belongs to the intervention."

Impact is one of the most overclaimed words in sustainability. It is attached to anything that can be counted: farmers trained, seedlings distributed, hectares enrolled, audits completed or suppliers mapped. These figures may show effort and delivery. They do not, by themselves, show that people, ecosystems or systems changed in a significant way.

The OECD Development Assistance Committee defines impact as the extent to which an intervention has generated, or is expected to generate, significant positive or negative, intended or unintended, higher-level effects. Its guidance describes impact as the large 'so what?' question: whether the intervention created change that genuinely mattered.

The definition is demanding because it moves beyond the programme's immediate products and even beyond its early outcomes. It asks about significance, distribution, duration and consequence.

The difference becomes clearer when the results chain is kept intact. Seedlings distributed are an output. Trees surviving and growing may be an outcome. Improved ecosystem function, reduced exposure to climate risk or more secure livelihoods may represent impact, but only when the evidence supports the connection. A positive indicator is not automatically an impact, and an impact is not automatically positive.

A payments-for-ecosystem-services trial in western Uganda shows why the counterfactual matters. Between 2011 and 2013, forest-owning households in 60 of 121 villages were offered payments to conserve forest, while households in the remaining villages were not. Satellite imagery showed that tree cover declined by 4. 2 per cent in participating villages and 9. 1 per cent in comparison villages.

Researchers found no evidence that tree cutting shifted to nearby land.

The programme did not merely count contracts or hectares; it asked what happened relative to what would probably have happened without it. Not every sustainability programme can or should be evaluated through a randomised trial. The discipline of the counterfactual still applies. What other forces could explain the observed change?

Did rainfall improve, commodity prices rise, enforcement tighten or another programme operate in the same communities? A credible claim may rely on experimental, quasi-experimental or theory-based evidence, but it should confront alternative explanations rather than treating sequence as causation. Contribution is often a more honest standard than sole attribution. Complex changes rarely belong to one organisation.

A company may contribute to lower deforestation through purchasing requirements, supplier support and monitoring while public enforcement, community action, market conditions and land tenure also shape the result. The claim should be proportionate to the organisation's influence and to the strength of the evidence. 'Contributed to' is not weaker language when it is more accurate.

A second question is equally important: who benefited, and who carried the cost? A programme can improve an aggregate indicator while transferring risk elsewhere or excluding those least able to participate. A practice change may reduce emissions in a buyer's disclosure while producers finance the working capital, absorb the yield risk and provide additional data without compensation.

The environmental result may be real, but the account is incomplete if the distribution of benefits and burdens is ignored. This is also where evaluation language meets corporate reporting. Under the European Sustainability Reporting Standards, undertakings in scope assess actual and potential positive and negative impacts on people and the environment across their own operations and value chains.

Actual positive impacts are considered through scale and scope; negative impacts through severity. ESRS does not turn every corporate claim into an impact evaluation or require a randomised counterfactual, but it makes activity counts an increasingly inadequate substitute for evidence about material effects. The same movement is visible in consumer law.

From 27 September 2026, the EU rules introduced by Directive (EU) 2024/825 apply, strengthening protection against greenwashing. The rules address unsupported generic environmental claims and claims about an entire product or business where the evidence concerns only one aspect. The proposed Green Claims Directive is separate and should not be confused with these adopted rules.

The direction is clear: language once challenged mainly as a matter of rigour is increasingly tested as a matter of substantiation. Impact therefore includes the inconvenient parts of the story. It includes negative and unintended effects, benefits that fade after funding ends, gains enjoyed by one group while costs fall on another, and changes that would have occurred anyway.

The purpose is not to make every claim impossible. It is to make claims commensurate with evidence.

For practitioners, the discipline is simple to state and difficult to apply: report activities as activities, outputs as outputs and outcomes as outcomes. Reserve impact for significant higher-level change supported by a credible account of contribution, distribution, duration and uncertainty. The strongest impact statement is not the boldest one.

It is the one that remains defensible when someone asks what would have happened anyway and who paid for the change.

Practical application

Create an impact-claim register before drafting the report. For each proposed claim, specify the affected people, ecosystem or system; the baseline; the expected counterfactual; the pathway of contribution; the timeframe; the distribution of benefits and costs; possible unintended effects; the evidence source; and the level of confidence.

Separate observed change from interpretation and identify what would cause the claim to be revised. Use mixed evidence where appropriate. Quantitative data can show magnitude and distribution; qualitative evidence can explain mechanism, lived experience and unexpected effects. Revisit the claim after incentives or project support end.

A result that disappears with the programme may still be an outcome, but its durability and higher-level significance should not be overstated.

Why it matters

Impact claims influence investment, procurement, regulation and public trust. When delivery is presented as impact, resources can continue flowing to activities that do not solve the problem. Clearer impact language improves accountability by directing attention towards significant change, credible contribution and the people or ecosystems carrying the consequences.

Common misconception

Impact is often treated as any positive result observed after an intervention. Timing does not establish causation, scale does not establish distribution and a favourable indicator does not reveal unintended costs. An impact claim should explain what changed, compared with what, for whom, for how long and why the intervention can credibly claim a contribution.

Connections

Outcome, the next chapter, describes the nearer-term changes through which impact may become possible.

Materiality determines which impacts are significant enough to shape governance and disclosure. Stakeholder engagement provides evidence about consequences that central reporting systems may not see. Later chapters on attribution, contribution, additionality and counterfactual examine the evidentiary tools behind credible impact claims.

A question worth asking

Which of your organisation's impact claims would remain credible after someone asked what would have happened anyway, who benefited, who carried the cost and how long the change lasted?

Selected references

OECD. 2023. Glossary of Key Terms in Evaluation OECD. 2021. Applying Evaluation Criteria and Results-Based Management for Sustainable Thoughtfully. Development, Second Edition. Jayachandran, S. et al. 2017. Cash for Carbon: A Mayne, J. 2012. Contribution Analysis: Coming of Randomized Trial of Payments for Ecosystem Age? Evaluation 18(3): 270-280. Services to Reduce Deforestation. Science 357(6348): 267-273. EFRAG.

2024. ESRS Implementation Guidance 1: European Parliament and Council. 2024. Directive Materiality Assessment. (EU) 2024/825 on Empowering Consumers for the Green Transition.