Validation

“A validated forecast is a defensible proposition, not a promise that the future will comply. ”

Validation is the future-facing counterpart to verification. It examines whether a claim, model or plan is plausible for an intended future use. Because the event has not yet occurred, the validator cannot confirm truth in the same way as a historical result. The task is to test assumptions, methods, evidence and logic.

ISO/IEC 17029:2019 defines the general framework for validation and verification bodies. Its distinction is temporal: validation confirms the plausibility of claims regarding intended future use or projected outcomes, while verification confirms the truthfulness of claims about events or results already obtained.

Carbon projects offer a practical example. Under mechanisms such as the Clean Development Mechanism, a project design was validated before registration. The designated operational entity assessed the proposed methodology, baseline, monitoring plan, additionality and stakeholder requirements. Later verification examined monitored emission reductions that the project claimed to have achieved.

The sequence shows why the terms cannot be exchanged. Validation can conclude that a method and forecast are reasonable under stated assumptions. Weather, behaviour, policy, implementation or market conditions may later differ. The validated project can underperform without the original validation having been dishonest, and it can succeed for reasons not predicted.

Assumptions are therefore the centre of the work. A livelihood programme may assume that training leads to adoption, adoption raises yield, prices remain stable and additional income stays with the intended household members. A climate model may assume a baseline, technology performance and operating life. Validation should make these dependencies visible and test whether evidence supports them.

Counterfactuals deserve particular scrutiny. A projected impact often compares the proposed future with a baseline describing what would happen without the intervention. If the baseline is too pessimistic, the projected benefit is exaggerated. Validators should examine data, alternatives, policy trends and incentives rather than accept the project developer's preferred scenario.

Optimism bias is not always deliberate. Sponsors tend to underestimate cost and time while overestimating uptake and benefit. Bent Flyvbjerg's work on major projects has documented persistent forecast error and strategic misrepresentation. Independent validation introduces challenge, but independence must be real and the programme must allow a negative conclusion.

Sensitivity analysis helps users understand which assumptions matter most. A project may remain viable across a range of prices and yields, or collapse if one adoption rate changes slightly. Reporting only the central forecast creates false precision. Scenarios, ranges and trigger points make the future claim more useful.

Validation has a date and scope. Methods, law, technology and context change. A design validated before a major policy shift, conflict or climate event may no longer be plausible. Programmes should define when material change requires revalidation rather than allowing the original statement to travel indefinitely.

The validator's conclusion should not become a guarantee. Words such as approved, proven or assured may imply that the projected outcome will occur. The appropriate statement concerns plausibility for the intended use under defined assumptions. Responsibility for implementation remains with the organisation.

Validation can improve theories of change, targets, transition plans, environmental claims and investment decisions. Its greatest value is not the stamp at the end. It is the disciplined exposure of assumptions before money, rights and reputations become committed to them.

The discipline is to treat the future as conditional. A credible validation asks which evidence supports the model, what alternatives were considered, what would falsify the assumptions and how performance will later be verified. A plan should not be called validated without making its conditions visible.

Practical application

State the intended use, criteria, assumptions, baseline, methods, scenarios, data and uncertainty. Use independent subject expertise and require sensitivity analysis for variables that drive the conclusion. Document alternatives and reasons for selecting the proposed approach.

Set triggers for revalidation when scope, law, technology, geography or underlying assumptions change. Link the validated design to a monitoring plan and future verification. Communicate the conclusion as plausibility under conditions, not a guarantee of delivery.

Why it matters

Sustainability strategies depend on claims about what interventions, investments and transition plans are expected to achieve. Validation provides structured challenge before implementation, when assumptions can still be changed and avoidable failure is less costly.

Common misconception

Validation is often treated as approval that a projected outcome will occur. It confirms that a future-oriented claim is plausible for its intended use under stated assumptions. Performance must still be monitored and later verified.

Connections

Theory of Change makes causal assumptions visible. Impact asks what higher-level change occurred, while verification tests historical results. ISO/IEC 17029 provides the common conformity-assessment frame for both validation and verification.

A question worth asking

Which single assumption would most weaken your projected sustainability outcome if it proved false, and was that assumption genuinely challenged during validation?

Selected references

ISO/IEC 17029:2019. Conformity Assessment - General Principles and Requirements for Validation and Verification Bodies. ISO 14064-3:2019. Greenhouse Gases - Part 3: Specification with Guidance for the Verification and Validation of Greenhouse Gas Statements. UNFCCC. Clean Development Mechanism Validation and Verification Standard for Project Activities. Flyvbjerg, B. 2009.

Survival of the Unfittest: Why the Worst Infrastructure Gets Built - and What We Can Do about It. Oxford Review of Economic Policy 25(3): 344-367. OECD. 2021. Applying Evaluation Criteria Thoughtfully.