What counts as adequate substantiation?

The question every claim rule delegates

Every rule in this theme eventually reduces to the same delegated question. "Misleading" claims are those the evidence cannot carry; "generic" claims are those no evidence could specify; per se rules are legislated shortcuts past evidence assessment. What counts as adequate substantiation is therefore not one topic among the claims rules but their common denominator — and it is the least harmonised element of the entire field. The regimes agree on the slogan (claims must be substantiated) and diverge on everything operational: what kind of evidence, generated when, by whom, disclosed to whom, and judged against what meaning of the claim. This article assembles the standards as the primary sources state them, then maps the four axes on which they genuinely differ.

The standards, as written

The United States: "competent and reliable scientific evidence". The FTC's Green Guides state the most-quoted formulation in the field: substantiation for environmental claims ordinarily requires evidence consisting of "tests, analyses, research, or studies that have been conducted and evaluated in an objective manner by qualified persons and are generally accepted in the profession to yield accurate and reliable results", and "sufficient in quality and quantity based on standards generally accepted in the relevant scientific fields", assessed in light of the entire body of relevant evidence (16 CFR § 260.2). Three demands are packed into the formulation: methodological orthodoxy (generally accepted methods), professional competence (qualified persons, objective evaluation), and evidential proportionality (quality and quantity matched to the claim against the whole record). The standard is claim-relative by construction — the evidence must fit "all express and implied claims" conveyed, which imports the consumer-understanding question into the substantiation question.

The United Kingdom: substantiation as a code principle. The CMA's Green Claims Code makes substantiation its sixth principle — "Claims must be substantiated" — glossed by the requirement that businesses hold "robust, credible, relevant and up to date evidence" (CMA, 2021). The formulation is less technical than the FTC's and broader in one respect: "up to date" writes evidential currency into the standard itself, a point the American formulation reaches only through the whole-record assessment. Under the DMCC regime, the same principles are enforced through the misleading-actions and omissions framework, with the CMA's guidance treating environmentally concerned consumers as potentially "more susceptible to misleading environmental claims" (CMA, 2025) — a vulnerability framing that effectively raises the evidential bar for green claims relative to ordinary marketing.

Canada: "adequate and proper", before the claim. Canada's Competition Act states the standard as a statutory element with a reverse onus: product environmental-benefit claims must be "based on an adequate and proper test, the proof of which lies on the person making the representation" (Competition Act, s. 74.01(1)(b.1)); business-level environmental claims must rest on "adequate and proper substantiation" (s. 74.01(1)(b.2), as amended in 2026 — the Canada article traces the changes). The Competition Bureau's guidelines add the two operational glosses that give the standard its teeth: the testing "must be conducted before the claim is made", and courts read "adequate and proper" as "fit, apt, suitable or as required by the circumstances", judged against "the general impression that the representation conveys to consumers" (Competition Bureau, 2025). The Competition Tribunal upheld the constitutionality of the prior-testing requirement in the Imperial Brush litigation, decided in 2008 — the jurisprudential anchor for the proposition that post hoc validation cannot cure an untested claim (Competition Bureau, 2008).

The Netherlands: substantiation as guidance architecture. The ACM's guidelines distribute the substantiation duty across its rules of thumb — "Substantiate your sustainability claims with facts, and keep them up-to-date" standing second among five (ACM, 2023) — with the distinctive feature that the guidance governs ambitions as well as assertions: future claims must be "concrete and verifiable", importing the substantiation logic into the promissory register that the comparative-and-future article examines.

The standards layer: ISO 14021. Beneath the regulatory formulations sits the standard the regimes historically borrowed from. ISO 14021 requires that self-declared claims "shall be substantiated and verified" (ISO, 1999, cl. 5.7), places responsibility squarely on the claimant — "[t]he claimant shall be responsible for evaluation and provision of data necessary for the verification" (cl. 6.1) — and states the prior-evidence rule in standards language: "Prior to making the claim, evaluation measures shall be implemented to achieve reliable and reproducible results necessary to verify the claim" (cl. 6.2.1). The Canadian statutory standard is, in substance, this clause with penalties attached.

The EU: substantiation by prohibition — and the road not taken. The EU's current architecture reaches substantiation obliquely: the blacklist prohibits claim forms that cannot be substantiated (generic claims absent demonstrated excellence), the misleading-actions test catches unsupported specifics, and PPWR Article 14 routes packaging-claim substantiation into technical documentation policed by market surveillance (the PPWR-claims article). The direct answer — specified substantiation requirements plus ex ante verification by accredited bodies — is the blocked Green Claims Directive's contribution, suspended with the proposal (the Green Claims Directive article). The EU thus presently operates the field's strictest claim-form rules atop its least specified evidence standard: a combination whose practical meaning national enforcers like the ACM have filled with guidance.

Axis one: what evidence

The regimes' first divergence is the evidence type the claim demands. Performance claims — degradability, compostability, recyclability-by-design — resolve to testing: protocol-based, laboratory-generated, the territory of the standards this library's third theme documents; the FTC's formulation and Canada's "test" language were written for exactly this register. History claims — recycled content, sourcing — resolve to records and custody: audit trails, supplier documentation, chain-of-custody certification, where "testing" is category-inapt and the substantiation question becomes documentary integrity. System claims — recyclability-in-practice, with its access thresholds — resolve to infrastructure statistics: population-access measurement, programme surveys, the contested numerator-and-denominator questions this library treats elsewhere. Lifecycle claims — footprints, comparative environmental superiority — resolve to modelling: LCA under declared methodology, where the substantiation dispute is methodology choice itself, and where France's carbon-neutrality decree shows the regulatory response — prescribe the method (lifecycle reporting per the standard), mandate its publication, and let the claim stand or fall on the disclosed model (Légifrance, 2022).

The axis explains a persistent practitioner error: holding the wrong genre of evidence. A laboratory certificate cannot substantiate an access-threshold recyclability claim; a supplier declaration cannot substantiate a degradation claim. Adequacy is genre-matching before it is quality.

Axis two: evidence when

The temporal rule divides the regimes more sharply than any other. Canada's statute and ISO's clause 6.2.1 state a hard prior-evidence rule: substantiation must exist before the claim is made, and the reverse onus makes its absence dispositive. The American position is functionally similar — the FTC treats substantiation as a duty existing at the time of the claim — but reaches it through enforcement doctrine rather than statutory element. The UK's "up to date" gloss adds the mirror-image temporal duty: evidence must not only precede the claim but survive with it, decaying claims requiring re-substantiation as facts change. The practical synthesis this field's compliance literature has converged on — an evidence file predating the claim and reviewed on a cycle — is not stated in any single instrument; it is the intersection of all of them.

Axis three: evidence to whom

Disclosure separates the regimes into three tiers. The baseline tier holds evidence for the regulator on demand — the American, British and Canadian position. The middle tier mandates public disclosure for defined claims: France's carbon-neutrality reports, the EU future-claims rule's requirement that third-party verification findings be "made available to consumers" (Directive (EU) 2024/825, art. 6(2)(d) insertion). The far tier — ex ante submission to an accredited verifier before any communication — is the Green Claims Directive's suspended design. The tiers price claims differently: a hold-on-demand regime makes weak substantiation a latent risk; a publication regime makes it a visible one; a verification regime makes it a barrier to entry. The direction of travel across the decade's instruments is outward through the tiers — and packaging's PPWR documentation requirement, filed rather than merely held, sits already in the middle distance.

Axis four: evidence against what meaning

The deepest axis is interpretive: adequacy is measured against the claim's meaning, and every regime assigns the meaning question to the consumer's understanding — the "general impression" in Canada's formulation, the express-and-implied-claims scope in the FTC's, the average consumer of the European tests. The consequence, developed throughout this theme, is that substantiation adequacy is not a property of the evidence alone but of the evidence-meaning pair: the same file that adequately substantiates "30% recycled content (mass balance attributed)" under-substantiates "30% recycled" read molecularly; the same access data that carries a qualified recyclable claim fails the unqualified one. The largest substantiation failures in the enforcement record — the bamboo cases, the fashion undertakings, the airline commitments — were not evidence-free claims but evidence-meaning mismatches: real files, attached to representations whose conveyed meaning outran them.

That is also, finally, why "adequate substantiation" resists the checklist form practitioners request. The regimes can specify the evidence genres, the temporal rule, the disclosure tier; they cannot pre-specify the meaning consumers will take, and so adequacy remains a judgment made claim by claim, market by market — the irreducible case-by-case core this theme's per se rules exist to spare regulators from litigating, one vague word at a time.

How substantiation fails: a taxonomy from the enforcement record

The standards' abstractions become concrete when the enforcement record is read as a catalogue of substantiation failure modes, because the cases sort into a small and repeating set.

The absent file. The simplest failure — no evidence at all — dominates the sweep findings: the Commission's 2021 screening found that "in 59% of cases the trader had not provided easily accessible evidence to support its claim" (the prevalence article examines the figures' limits). Pure absence is the failure mode the reverse-onus regimes annihilate: under Canada's provisions, an unevidenced claim loses by construction, with nothing for the enforcer to prove.

The wrong-genre file. The evidence exists but answers a different question — the laboratory certificate offered for an infrastructure claim, the supplier declaration offered for a performance claim. This is the failure mode axis one predicts, and the compostability and recyclability case law this library documents elsewhere is largely its jurisprudence: capability evidence deployed to carry fate claims.

The stale file. The evidence was adequate when generated and the world moved — collection programmes changed, reprocessing markets collapsed, the science turned. The UK's "up to date" requirement names this failure directly; the recyclability record supplies its scale, with access and outcome statistics shifting under labels printed years earlier.

The overrun file. The most consequential mode, and the subtlest: genuine evidence, outrun by the claim's conveyed meaning. The fashion undertakings' vague ranges, the bamboo cases' processing-erased virtues, the airline commitments' offset promises — each involved real underlying facts (recycled fibres existed, bamboo was grown, offsets were purchased) whose evidentiary reach the marketing exceeded. Axis four's evidence-meaning pair is the diagnostic: the file substantiated a claim; the package made a different one.

The undisclosed-model file. Evidence adequate under one methodology, communicated without the methodology — the mass-balance percentages and access-based recyclability claims whose accounting basis the label omits. The regimes' emerging answer is disclosure conversion: France's published lifecycle reports, the EU future-claims rule's public verification findings, scope statements under PPWR Article 14 — moving the model from the file into the claim.

The taxonomy's practical value is triage. A reader assessing any environmental claim can locate its likeliest failure mode from the claim type alone: system claims fail stale or wrong-genre; content claims fail undisclosed-model; aspiration claims fail overrun; and the vague claims the blacklists now capture failed, historically, absent. The substantiation standards read as abstractions; the failure record reads as a checklist — and the checklist, unlike the standards, is the same in every jurisdiction. That convergence is this article's quiet finding: regimes that cannot agree on what adequate substantiation is agree almost perfectly on how its absence looks, which is why the failure taxonomy, not the standards text, is the transferable knowledge — the part of this article a reader in any jurisdiction can apply unmodified tomorrow, to any claim this library documents, or indeed to any package that presently carries one.

References

ACM (Autoriteit Consument & Markt) (2023) Guidelines regarding sustainability claims, 13 June. Available at: Open source (Accessed: 18 August 2026).

CMA (Competition and Markets Authority) (2021) Environmental claims on goods and services (Green Claims Code guidance), 20 September. Available at: Open source (Accessed: 18 August 2026).

CMA (Competition and Markets Authority) (2025) Unfair commercial practices (CMA207), 4 April (as updated 18 November 2025). Available at: Open source (Accessed: 18 August 2026).

Competition Act, R.S.C. 1985, c. C-34, s. 74.01 (as amended). Available at: Open source (Accessed: 18 August 2026).

Competition Bureau (2008) Competition Bureau wins court order against chimney products company (Imperial Brush; Tribunal decision announced 11 February 2008). Available at: Open source (Accessed: 18 August 2026).

Competition Bureau (2025) Environmental claims and the Competition Act (final guidelines), 5 June. Available at: Open source (Accessed: 18 August 2026).

Directive (EU) 2024/825 of the European Parliament and of the Council of 28 February 2024, OJ L, 2024/825, 6.3.2024. Available at: Open source (Accessed: 18 August 2026).

ISO (International Organization for Standardization) (1999) ISO 14021:1999 Environmental labels and declarations — Self-declared environmental claims. Geneva: ISO. Text as adopted in IS/ISO 14021:1999, available at: Open source (Accessed: 18 August 2026).

Légifrance (2022) Décret n° 2022-539 du 13 avril 2022 relatif à la compensation carbone et aux allégations de neutralité carbone dans la publicité. Available at: Open source (Accessed: 18 August 2026).

16 CFR § 260.2 (FTC Green Guides, substantiation). Available at: Open source (Accessed: 18 August 2026).

Note on sources and verification

The § 260.2 formulation is quoted in the phrase-level form retrieved from the eCFR; readers should consult the section for full sentence context. The CMA evidence gloss ("robust, credible, relevant and up to date") and the CMA207 vulnerability passage are as retrieved from the cited guidance. Canadian statutory language is verbatim from the consolidated Competition Act; the Bureau's interpretive glosses are from its 2025 guidelines page, and the Imperial Brush characterisation is from the Bureau's contemporaneous release. ISO 14021 clause quotations are from the archived 1999 adoption with the standing edition caveat. The French decree's requirements are summarised from the Légifrance text without verbatim article quotation. The four-axis framework, the three disclosure tiers, and the evidence-meaning-pair analysis are this library's synthesis. The characterisation of enforcement-record failures as evidence-meaning mismatches is an interpretive claim about the cited cases, not a statement from any authority.

Last verified: 18 August 2026.