Material Passports

A material passport is a structured record bound to a physical object that states what it is made of, where the material came from, how it is fastened, and what may be done with it when it stops being useful. Established Three distinct artefacts travel under the name, and conflating them is the commonest error in the field. The regulatory product passport is mandated by statute for a product class, and only two exist in law: the European digital product passport under the Ecodesign for Sustainable Products Regulation, and the battery passport under the 2023 Batteries Regulation. The building material passport is a voluntary platform-hosted inventory of a structure’s components. The chain-of-custody certificate is older than both and is rarely called a passport at all: forestry, minerals and food have run attribute attestation for decades.

Frontier This brief owns the information layer and nothing else. Industrial Ecology owns the metabolism and the material-flow accounting that produces the circularity denominator; Circular Infrastructure Systems owns the built environment and the re-certification wall; Circular Economies owns decoupling and producer responsibility. Each touches the passport question and none owns it. The unoccupied joint question is whether making a material legible changes what happens to it — an empirical question about an information instrument, which the policy literature has not gone looking for.

Frontier The answer this brief lands is that a passport is an attestation instrument whose verification stops where its physical binding stops, and that the chain breaks at the shredder, upstream of every sorting decision the passport was built to improve. It can tell a customs officer that a recycled-content figure was declared, a repairer which torque to apply, a demolition contractor that a beam exists. It cannot tell a recycler what is in a shredded fraction, because the carrier died in the machine that destroyed the shape. Where an attribute is cheap to measure directly the passport is redundant; where it is expensive to measure the passport is unverifiable. Between those conditions lies a real but narrow band, and nearly all the claimed benefit is asserted to sit inside it without measurement.

Established A note on sourcing. This brief was commissioned in September 2026 from the Institute’s research base. Reading-list entries without links are cited from the bibliographic record rather than re-fetched, and claims are dated no later than early 2026 unless carried by a linked source.

Established Two passport mandates exist in law and no others do. Regulation (EU) 2024/1781 entered into force in July 2024 and created the digital product passport as a horizontal device whose substance is delegated: the Regulation says a passport shall exist and what it must be capable of, and each product group acquires real requirements only when a delegated act arrives. Regulation (EU) 2023/1542 is the one instrument with a hard date and a defined content list — a battery passport from 18 February 2027 for electric-vehicle, light-means-of-transport and industrial batteries above 2 kWh, attached to quantified recycled-content minima, recycling-efficiency floors and element-level recovery floors. Frontier Everything else called a passport in 2026 — textiles, steel, aluminium, construction products, tyres — is a working-plan priority or a pilot, not an obligation, and the delegated acts and calculation methodologies that make even the battery figures operable have slipped repeatedly. Dating this regime from the Regulation rather than from its delegated acts dates it from the wrong document.

Established The most useful document in the field is not about passports at all: ISO 22095 defines five chain-of-custody models, and the difference between them decides what any passport can honestly claim. Identity preserved keeps a consignment separate and traceable. Segregated keeps certified material apart from uncertified but allows mixing among certified sources. Controlled blending permits a stated proportion of uncertified material. Mass balance reconciles certified input against certified output in aggregate while permitting physical mixing, so the molecule delivered may have no relation to the certified molecule. Book and claim severs the physical link and trades the attribute as a certificate. Established Only the first two support the sentence “this object contains that material”. Most industrial recycled-content claims in polymers and chemicals run on mass balance, so the number on a passport is often a statement about a company’s books rather than about the object in your hand — and no passport format requires that distinction to reach the reader.

Established The one large mandated traceability regime with a measured outcome record is Section 1502 of the Dodd-Frank Act, and the record is not encouraging. It required issuers to determine whether tin, tantalum, tungsten or gold originated in the Democratic Republic of the Congo or adjoining states. The Government Accountability Office reviewed filings annually and found a large majority of filers — consistently around two-thirds, varying by year — could not determine country of origin despite conducting mandated due diligence. Frontier Evaluations of the ground-level effect disagree sharply: one PLOS ONE study reported increased violence in affected territories, another increased infant mortality attributed to a de facto embargo, and defenders argue the comparison period conflates the statute with a separate Congolese mining suspension. The determination rate is not disputed: a mandate to know where material came from produced, for most filers, a documented inability to know.

Established The closest economic analogue with a clean econometric answer is extended producer responsibility, and it is close to a null result. Across twenty-five EU countries from 1998 to 2015, using the fees producer responsibility organisations actually charged for four packaging materials, the finding was “very little (though statistically significant) packaging reduction” and no systematic substitution between materials. Frontier That instrument is stronger than a passport because it moves money rather than information. If a priced obligation moved design that little, the prior for an unpriced information obligation should be lower, not higher.

Established The denominator defeats the instrument before it starts. The sociometabolic baseline reports, for 2005, roughly 4 gigatonnes per year of waste materials recycled against about 62 Gt/yr processed and 41 Gt/yr of outputs, with 44 per cent of processed materials used for energy and therefore unavailable for recycling, and stocks growing at 17 Gt/yr in net additions. A passport cannot make combusted material recoverable or make a road built last year reach end of life. Established Nor is the stock static: the UN International Resource Panel’s 2024 accounting has extraction tripling over five decades and rising 60 per cent by 2060 on current trends, and the OECD’s independent projection across 61 materials has primary materials use doubling by 2060.

Established In the one sector where the passport arrives first, the near-term feedstock is the stream a passport adds least to. The International Energy Agency reports recovered volumes above 40 per cent of available feedstock for nickel and cobalt and 20 per cent for lithium in 2023, announced 2030 recycling capacity at roughly seven times available feedstock, manufacturing scrap at about two-thirds of 2030 feedstock, and end-of-life vehicle batteries not exceeding 90 per cent of feedstock until around 2050. Frontier Manufacturing scrap never leaves the factory that made it and its composition is known by construction, so across the decade in which the battery passport takes effect its marginal informational contribution is near zero on two-thirds of the tonnage.

Frontier There is no published study demonstrating that a material passport raised a recovery rate. The research behind this brief located no randomised, quasi-experimental or matched-cohort evaluation linking passport availability to recovered mass, recovered value or sorting yield, in any jurisdiction, for any product class. Established Absence of a located study is weaker than a null result and is stated here as absence. But forestry chain of custody dates to the 1990s, and what the literature supplies instead is compliance throughput: passports issued, fields populated, platforms onboarded. A field measuring its own throughput rather than its effect has not tried to falsify itself.

Frontier The organising question is whether legibility changes behaviour, and it is assumed rather than tested. The theory of change has five links: a record is created; a downstream actor reads it; reading changes a sorting, repair or procurement decision; the decision raises recovered mass or value; the gain exceeds the cost of maintaining the record. Frontier Link one is demonstrated by pilots. Link two has weak evidence, since recyclers report not reading what is already printed on products. Links three, four and five have no published measurement this research could find.

Frontier The confidentiality tension is structural, not transitional. Full composition — additive packages, alloy micro-chemistry, catalyst loadings — is the thing a supplier sells. The regulatory answer is tiered access: public fields, fields for authorities and customs, fields for repairers and recyclers. Whether tiering holds is untested, and the failure is asymmetric: a leak is permanent, so a rational supplier populates each tier with the least informative compliant answer. That tiering resolves the trade-secret problem appears in policy documents as assertion, with no enforcement evidence attached.

Frontier Whether a passport is a circularity instrument or a market-access instrument is genuinely contested, and the second reading explains the politics better. On that reading it is a customs artefact, letting a border authority condition entry on an attribute the way the carbon border adjustment mechanism conditions entry on embedded emissions. It predicts the observed sequencing — batteries and metals first, both import-exposed strategic sectors — better than a recovery-maximising reading, which would have started with construction and demolition waste, the largest arising stream by mass and the one with no mandate at all.

Frontier Whether the identifier survives the process that destroys the object is the physical question underneath everything else. Tracer-based sorting — fluorescent, chemical or magnetic markers compounded into the material, so identity travels in the material rather than on a label — is the only route by which a passport survives comminution. Pilots exist in packaging polymers; no jurisdiction mandates a tracer, no tracer set has an interoperable registry, and a foreign marker in a recyclate is a contamination question the buyer prices.

Established The binding constraint is identifier survival, and it is a materials problem dressed as a data problem. A two-dimensional barcode survives shelf life and fails a shredder. A radio-frequency tag survives longer and fails a furnace. A cast-in datamatrix on a steel section survives demolition and fails the first cut to length. Every process that recovers material rather than components destroys the carrier, and the moment the passport would be most valuable — sorting a mixed stream — is downstream of that destruction.

Established Second, the manufacturer frequently does not know the answer. A converter buying compounded resin, a fabricator buying a coil, an assembler buying a module: each knows its supplier and not its supplier’s supplier. The conflict-minerals determination record is the measured demonstration — a legal duty to determine origin produced, for most filers, a documented inability, because the information does not exist in a form that propagates upward through four tiers of trade.

Frontier Third, dynamic attributes need hardware the regime does not mandate. A battery passport’s state-of-health field decides second-life value and is a measurement, not a record: it requires onboard sensing, a management system willing to expose it, and a reference protocol for what the number means. Absent a standard protocol, a state-of-health figure is comparable only within one manufacturer’s fleet, which is precisely the comparison a second-life market does not need.

Frontier Fourth, registry permanence is unfunded. No mandate located here says who holds the record when the issuer is liquidated, who pays for resolution across a fifty-year building life, or what happens when an identifier method is deprecated. The measured warning is link rot in legal citation: roughly half the URLs cited in United States Supreme Court opinions no longer resolved to the cited material, in institutions with permanent budgets and strong incentives.

Frontier The result this brief most needs and cannot find is the displacement elasticity: how much primary production a marginal tonne of secondary supply displaces. The Institute’s built-environment brief records the same gap and states that no estimate was obtainable. Without it, every claim that a passport reduces extraction is a claim about an unmeasured quantity, because recovered material can add to total supply rather than substitute for primary supply — the circular-economy rebound mechanism.

Frontier Second, a re-certification pathway for recovered components. Circular Infrastructure Systems reports that no such document could be obtained for any component class in any jurisdiction. That is what makes component-level reuse — the only mode in which a passport is genuinely load-bearing, because it is the only mode where the object survives intact — legally unavailable for structural applications. A passport documenting a beam nobody may specify is a catalogue of scrap.

Frontier Third, sorting performance on real streams rather than spiked laboratory feeds. The passport’s value is the residual after direct sensing takes what it can: X-ray fluorescence and laser-induced breakdown spectroscopy identify alloys at line speed, near-infrared sorts polymer families and fails on carbon black. A published curve of accuracy and throughput against contamination, per material family, would fix the width of the passport’s useful band, and no such curve appears in the open literature.

Frontier Fourth, a price response. Recyclate has traded at a premium to virgin resin in some periods and a discount in others, driven by oil prices rather than by attributes. If attested provenance has no detectable price effect once quality is controlled for, the passport’s market-pull theory fails on its own terms.

Frontier The decisive test is already scheduled and nobody has designed it: the battery passport becomes obligatory on 18 February 2027, which creates a dated before-and-after on a defined product class whose recovery statistics are already collected. It is the one natural experiment in this field with a hard date, a bounded scope, a legal trigger and an outcome variable reported under the same Regulation. Frontier A pre-registered design would fix baselines per Member State and per chemistry before the trigger, take recovered mass per tonne collected and recovered value per tonne as primary endpoints, and use non-EU recyclers processing comparable chemistries as the comparison arm. The window for clean baselines closes in 2026; after the trigger the counterfactual must be reconstructed rather than observed.

Frontier The second experiment is cheap, physical, and would settle the bottleneck argument in months: run marked and unmarked identical waste streams through a commercial shredder line and count surviving readable identifiers by fraction. Speculative The expected result is near-total loss for printed and adhesive carriers, partial survival for cast-in marks on heavy sections, and material-borne tracers as the only carrier with a meaningful survival fraction. If survival is near-total loss across all carriers, the recycling rationale collapses to pre-shredder disassembly, and the case becomes a labour-cost argument rather than an information one.

Frontier The third is an audit: sample declared fields and assay them destructively. Recycled content, element concentrations and hazardous-substance declarations are all testable. The informative split is not overall accuracy but accuracy by verification cost — if cheap-to-test fields are accurate and expensive-to-test fields are not, the regime is producing exactly the behaviour a self-declaration regime predicts and the environmental fields are decoration.

Frontier The fourth is a demolition trial a single city could run. Match building pairs by age, structure type and jurisdiction; give contractors a full component inventory for one of each pair; measure diverted mass, resale revenue and contractor time. No such trial has been published. Its likely finding is that inventory changes the composition of diversion — more fittings, fixtures and services, not more structure — because structure is where re-certification binds.

Established The carrier stack is settled and uninteresting; the identity stack is not. Two-dimensional barcodes carrying resolvable web identifiers, passive radio-frequency tags and near-field tags all work, and the retail migration to two-dimensional codes at point of sale gives the barcode route an installed reader base no competitor has. What is unsettled is what the identifier resolves to and who guarantees resolution: a vendor platform, a sector registry, a decentralised identifier method, or a persistent web address with an institutional custodian.

Frontier Verifiable credentials are the right shape for the trust layer and inherit a known set of failure modes. The format lets an issuer sign a claim and a verifier check it without contacting the issuer, which is what an offline customs check or a scrapyard without connectivity needs. Established The audit literature on the nearest deployed analogue — the content-provenance specification used for media authenticity — found optional revocation checking, timestamps not bound to the signed content, validators disagreeing on identical inputs, certificate expiry colliding with legal retention windows, and self-reported conformance. Every one transfers unchanged, because they are properties of the trust machinery rather than of the payload; Content Authenticity Infrastructure owns that analysis and is the engineering prior here.

Frontier Semantic interoperability is the expensive engineering and the unglamorous one. A passport is machine-useful only if “recycled content” means the same thing in two records: same chain-of-custody model, same system boundary, same allocation rule for multi-output processes. Product circularity data-sheet standardisation exists and is a genuine step, but it standardises the sheet, not the definitions under the numbers on it. Without those, aggregation across passports yields a number with a decimal point and no referent.

Frontier Material-borne identity is the only path that survives the process. The requirements are severe: detectable at line speed, stable through thermal processing, unique enough to encode a class, cheap enough for bulk material, and acceptable to a buyer who now has a foreign additive in their feedstock. Claims that tracers make the identifier problem solved in principle skip the last two requirements, which are the commercial ones.

Established The nearest neighbour is media provenance, and the resemblance is structural rather than thematic. Content Authenticity Infrastructure covers a system that binds a signed assertion to an artefact, faces an adversary who can strip the binding, and finds the weakest points in revocation, timestamping and self-reported conformance. Frontier A material passport is that architecture applied to an object that can be melted — harder in one respect, since ordinary industry destroys the binding without an attacker, and easier in another, since a physical assay can sometimes adjudicate what no cryptographic check can.

Frontier Upstream, the passport meets the midstream and the border. Critical Mineral Midstream owns refining concentration, which is why origin attestation is politically valuable: where one country holds most of a processing step, an attribute that pins origin is a trade instrument before it is an environmental one. Future Trade Systems owns the border reading of the same machinery, and Autonomous Supply Chains owns the case where the reader is software, which is where machine-readable fields stop being decorative.

Established Downstream, the passport is one input to the three circularity briefs and changes none of their conclusions. Industrial Ecology supplies the corrected denominator, Circular Economies the decoupling evidence and the producer-responsibility near-null, Circular Infrastructure Systems the re-certification wall. This brief’s contribution to that set is the observation that the information instrument is not the binding constraint in any of them.

Established The European Union is building this alone, which is a fact about the instrument and not only about geography. No comparable mandate exists in the United States, China, Japan or India as of early 2026; what exists elsewhere is voluntary chain-of-custody certification and sectoral labelling. Frontier A single-jurisdiction transparency mandate on globally traded goods has two futures: it becomes a de facto world standard because exporters build to the strictest market, or it becomes a border dialect satisfied at customs and ignored internally. Both have precedents in European chemicals and data-protection law, and they are distinguishable only by measuring whether non-EU domestic products carry the attribute too.

Established Standardisation is the rate-limiting institution, not legislation. A standardisation request to the European standards bodies and a dedicated technical committee is where the content is actually written, and standards committees run on consensus among the firms being regulated. That is how product standards have always been made, and it predicts the observed pattern: fast agreement on syntax, carriers and identifiers; slow, contested progress on the semantics that decide what a number means.

Frontier Enforcement is the gap, and it has a precedent attached. The EU Deforestation Regulation is the closest recent instrument demanding granular supply-chain data from every operator, and its application date has been postponed more than once under implementation load and trade objection. Speculative The realistic base case for passports is the same: dates hold in principle, delegated acts arrive late, and first enforcement lands on large, visible, EU-domiciled firms whose data was already good. An attestation regime without an adversarial testing function converges on the accuracy of its cheapest-to-audit field.

Established The strongest ethical warning here is a measured one: mandated traceability can harm the people at the origin end of the chain. The conflict-minerals record includes published findings of increased violence and increased infant mortality in affected mining communities after the due-diligence mandate, attributed to buyers withdrawing rather than verifying. Frontier The causal attribution is contested, partly over whether a contemporaneous mining suspension confounds the comparison. The mechanism’s plausibility is not contested: when verification is expensive and non-compliance is legally dangerous, the cheapest compliant act is to stop buying from places that are hard to verify.

Frontier The same mechanism threatens the informal recycling sector, which performs a large share of the world’s actual recovery. A regime conditioning market access on documented chain of custody makes undocumented recovery unsaleable into the regulated market. If formal recovery does not expand to absorb the displaced volume, net recovered tonnage falls while the reported recovery rate rises, because the reported rate counts only documented flows.

Frontier The greenwashing exposure runs through the accounting model rather than through lying. A truthful mass-balance claim and a truthful identity-preserved claim can be printed in the same field with the same words, and a consumer cannot tell them apart. Unless the chain-of-custody model is a mandatory, prominent, machine-readable field, the passport legitimates claims it was built to discipline. And because data-governance cost is roughly fixed per product line, the honest version of the policy states its consolidation effect rather than calling the cost negligible.

Speculative The long-run prize is an inventory of the anthropogenic stock: a civilisation that knows, at component resolution, what it has already built. Material stock is currently estimated in aggregate tonnes by material-flow accounting; full passport coverage would in principle replace estimation with enumeration. Stocks grow at roughly 17 Gt/yr in net additions, and that stock is the only ore body whose grade rises as extraction elsewhere falls. An enumerated stock would turn urban mining from prospecting into logistics.

Handwave The step from enumeration to recovery is always assumed and never shown. Knowing where the copper is does not make retrieving it cheaper than mining it; retrieval cost is demolition labour, transport and separation, none of which an inventory reduces. The one place the inference is sound is component reuse, where the object is retrieved intact and its value is its shape — exactly the path blocked by the missing re-certification pathway.

Speculative The pessimistic scenario deserves stating plainly. A full-coverage regime costing tens of billions of euros across a decade, achieving high field-population rates and moving recovered tonnage by a statistically undetectable amount, would be a consequential failure — not because the records were false but because information was never the binding constraint. On current evidence that scenario cannot be ruled out, and the instrument’s advocates have not proposed the measurement that would rule it out.

These horizons track what can be verified about a passport regime, not what can be announced about one.

• 10 yr: Frontier Battery passports are in force and at least one Ecodesign delegated act has bitten; the first before-and-after evidence on battery recovery exists; semantic interoperability remains partial, with recycled-content fields comparable within sectors and not across them. Expect a first audit of declared against assayed fields, and expect it to find a verification-cost gradient.
• 25 yr: Speculative Either passports have become a global market-access norm because exporters built to the strictest market, or they have settled into a border dialect with no domestic effect outside the European Union; the discriminating observation is whether non-EU domestic products carry the attribute. Component reuse remains structurally blocked unless a re-certification pathway is written, an institutional act with no current sponsor.
• 50 yr: Speculative The first passports issued for long-lived built assets reach the age at which registry permanence is tested for real, and the test is whether an identifier issued in the 2020s still resolves. Material-borne tracers are either a mandated part of the stack for at least one polymer family or abandoned as a contamination cost.
• 100 / 250+ yr: Handwave A complete queryable inventory of the anthropogenic stock at component resolution, used to plan extraction, is coherent but rests on unbroken institutional continuity, format migration and funding across a period longer than most states have kept any continuous digital record.

• Depends on the material-flow accounting and corrected circularity denominator owned by Industrial Ecology, which sets the ceiling on what any numerator instrument can reach; the re-certification wall documented in Circular Infrastructure Systems, which decides whether component reuse is legally available at all; and the producer-responsibility and rebound evidence in Circular Economies, which supplies the prior for how far an information instrument moves behaviour.
• Requires (not on this map) a published estimate of how much primary production a marginal tonne of secondary supply displaces, without which no extraction claim can be made; a written re-certification route for recovered structural components in some jurisdiction, without which reuse is scrap; an adversarial testing function able to assay declared composition, without which every field is a self-declaration; a carrier that survives shredding and smelting, without which the record dies before the sorting decision it was meant to inform; a demonstrated price response to attested provenance in a secondary-material market, without which the market-pull theory has no mechanism; and a statutory custodian obligation for registries outliving the issuing firm, without which fifty-year assets carry ten-year records.
• Enables component-level reuse markets where re-certification exists, second-life battery trade conditioned on comparable state-of-health measurement, border instruments conditioning market access on an embodied attribute, liability that follows a component across owners, and a stock inventory precise enough to convert urban mining from prospecting into logistics.
• Adjacent to Content Authenticity Infrastructure, which owns the trust machinery and its measured failure modes; Carbon Removal Verification, which owns attestation of unobservable attributes; Critical Mineral Midstream and Future Trade Systems, which own the trade reading of the same machinery; and Autonomous Supply Chains, where the reader is software.

Handwave “The digital product passport will make the European economy circular.” The claim appears in consultancy and trade material and does not survive the arithmetic. Circularity is bounded by combustion and stock growth — 44 per cent of processed materials go to energy and 17 Gt/yr is added to stock — and neither term is informational. A passport can raise the share of recoverable material actually recovered; it cannot enlarge the recoverable set.

Handwave “Blockchain solves chain of custody.” A distributed ledger makes a record tamper-evident after it is written. Every failure identified here is upstream: whether the declared fact was true when written, whether the identifier is still attached to the object, and whether anyone tests. The audit of the deployed media-provenance specification found weaknesses in revocation, timestamp binding, validator disagreement and self-reported conformance — none of which a ledger addresses.

Frontier “Recycled content on a label means recycled molecules in the product.” Under mass balance and book and claim, both recognised chain-of-custody models and both widely used in chemicals and polymers, it does not. The five models are formally defined and differ precisely in how much physical linkage survives. Until the model is a mandatory front-of-record field, the passport transmits the ambiguity rather than resolving it.

Frontier “It is just a QR code — the cost is trivial.” The carrier costs almost nothing; the recurring cost is supplier-data collection across tiers that do not have the data, legal review of disclosure, vocabulary mapping, and keeping the record true as bills of materials change. The conflict-minerals determination record is the measured cost of the part that is not the code.

Frontier “The 2027 date is fixed.” The obligation sits in a regulation and the operative content sits in delegated acts, methodologies and standards. Delegated acts under the Batteries Regulation have already slipped, and the EU Deforestation Regulation has been postponed more than once. A regime arriving with its methodology unfinished produces populated fields of uncertain meaning, which is worse than arriving late.

Speculative “Nothing here is worth doing.” That is not this brief’s finding, and the steelman is real: repair access, spare-part identification, hazardous-substance declaration and second-life battery trade are all cases where an identity-preserved record does work no assay can do cheaply. The finding is narrower and harder — the benefit concentrates in component-level reuse and repair, the rhetoric is written about material-level recycling, and nobody has measured the part that is claimed loudest.