How large is the free-riding problem, and how is it addressed?¶

Free-riding — obligated producers who do not register, do not report, or under-report — is packaging EPR's integrity problem: every tonne that escapes the obligation shifts cost onto compliant producers and corrodes the fairness on which collective systems depend. It is invoked constantly in policy debate, by producers complaining of unfair burdens and by critics alleging leaky systems. The remarkable fact underneath the rhetoric is that its size is unmeasured everywhere: no jurisdiction publishes an estimate of its unregistered population or its under-reported tonnage, and the best available enforcement records suggest the capacity to find out is largely absent. This article defines the problem's three distinct forms; assembles the fragmentary evidence on scale, led by the Ontario Auditor General's findings and the German historical record; reviews the enforcement and design instruments that actually work; and explains why the question's unfalsifiability is itself the finding that matters.

1. Three problems under one name¶

Free-riding conflates three behaviours with different mechanics and different remedies.

Non-registration — the obligated producer that never enters the system at all. Its packaging is collected and managed; compliant producers pay for it. This is the classic form, concentrated among smaller, foreign and online sellers whom registries struggle to see.

Under-reporting — the registered producer that declares less than it places on the market. Fees are levied per reported tonne, so under-reporting is a price discount awarded to dishonesty, paid by honest members of the same pool (the verification article treats the audit machinery this demands).

Inter-system leakage — specific to competitive markets: a compliance scheme that under-declares its clients' volumes shrinks its share of shared-system costs, shifting them onto rival schemes. Germany's 2000s experience — the self-collection and interim-solution loopholes through which large volumes escaped the dual systems — is the canonical case, and it contaminates the field's best natural experiment: some unknown portion of Germany's celebrated 63% post-competition price decline reflects obligation leakage rather than efficiency (the German article).

A fourth, historical variant deserves a line because it shows the problem is not unique to fee-based schemes: the United Kingdom's tradable-certificate system (packaging recovery notes) — the one Western compliance-note market — carries a documented fraud history in evidence generation, free-riding's mirror image: instead of escaping the obligation, parties manufactured proof of meeting it. Every compliance architecture has a characteristic leak; the leak's location follows the design.

The three fee-scheme forms share one property: each is invisible in the system's own data, because the system's data is what the free-rider avoids or falsifies.

2. What is actually known about scale¶

The honest inventory is short.

Ontario supplies the best-documented enforcement record, and it measures capacity rather than scale. The provincial Auditor General's December 2025 audit found the regulator had identified 1,804 potentially unregistered producers across its programs since 2020 — and placed 71% of those cases on hold; 92% of open cases were more than a year old, with 541 unresolved after five years; only 4% of producers with outstanding reports received compliance orders, and an average of 262 days elapsed from identifying non-compliance to issuing an order (the capacity article). Against this stands one number that frames the packaging question without answering it: 1,918 Blue Box producers registered as of April 2025. Whether the obligated population is 2,200 or 8,000 — whether free-riding is a rounding error or a third of the fee base — is precisely what no registry can say about itself, and the Blue Box program was excluded from the audit's scope.

Germany supplies the historical demonstration that leakage can be system-threatening. Through the 2000s, the exploitation of self-collection and interim-solution provisions eroded dual-system volumes to the point of endangering the financing model — the crisis that produced the 2019 Packaging Act, the central register, and the clearing-house apparatus. Germany is the one jurisdiction where free-riding's scale was, for a period, visible — because it grew large enough to destabilise the system that could not measure it.

Nowhere is there a published misreporting rate. No packaging jurisdiction publishes an estimate of under-reported tonnage; this library located no enforcement action anywhere specifically for false supply reporting (Ontario's substantial record concerns failures to register or report at all). France's published practice of auditing at least 15% of producer declarations annually, and Italy's thousand-plus targeted controls a year, are the only public audit-intensity benchmarks — and neither publishes the correction rates the audits find.

Bounding the plausible range is possible in principle and undone in practice. The obligated population under a producer cascade is, roughly, every brand holder, importer and large retailer selling packaged goods into the jurisdiction — a set that business registries and tax rolls could approximate. No regulator has published such a reconciliation: registered producers against an independently derived obligated population. Germany's public register makes the exercise crowd-sourceable — any competitor can check any seller — which is part of why German registration counts run to the hundreds of thousands while Ontario's Blue Box registry holds under two thousand entries above its revenue threshold. The two numbers describe differently drawn obligations and cannot be compared directly; what can be said is that one jurisdiction has built the instrument that would measure its perimeter and the others have not.

The consequence deserves flat statement: claims that free-riding is large, and claims that it is small, are currently both unfalsifiable. Producers invoking free-riders to argue fees are unfairly loaded, and scheme defenders minimising the problem, are arguing about an unmeasured quantity — and the measurement gap favours whoever benefits from the status quo in any given argument.

The equity stakes explain why the question matters beyond system finance. Free-riding is a transfer from compliant to non-compliant competitors: the honest producer's fee covers the free-rider's tonnage, which means compliance is priced at a competitive disadvantage exactly proportional to enforcement weakness. This is the argument compliant industry makes for enforcement spending — and it is sound — but it also means producer complaints about free-riding function politically as arguments for lower fees rather than for better enforcement, since the unmeasured free-rider conveniently explains any fee level as unfairly loaded. An enforcement regime that published its perimeter estimate would deprive both uses of the ambiguity.

3. What actually works: the enforcement record¶

The instruments divide into detection, deterrence and design, and the record ranks them.

Detection by registry. The precondition for everything else is a register that defines the obligated population. Germany's LUCID register is the maximal version — universal registration with no de minimis threshold ("no de minimis threshold or exemption for companies with low packaging volumes," per the register authority), publicly searchable, so that "unregistered" and "non-compliant" are synonyms. The public searchability matters: it enables competitors, customers and platforms to police the perimeter at zero regulatory cost. Registries with exemption cliffs are structurally weaker detectors, because establishing that an unregistered firm is obligated first requires establishing its size — adding an investigative step to exactly the caseload regulators shelve (the small-producer article).

Deterrence by enforcement. The record here is thin on both sides of the Atlantic. Ontario's is the best documented and shows what under-resourced deterrence looks like — though the same regulator's penalty orders against battery producers (totalling $2.78 million for missed targets) demonstrate that consequences, when actually imposed, register: the orders remain North America's most significant EPR enforcement action and established that individual producer liability can be real. France's inspectorates found no sanction ever issued for missed targets, with a €30,000 maximum that is negligible against scheme budgets. The general condition — paper powers exceeding exercised powers — is the free-rider's structural ally.

Design that recruits third parties. The strongest recent instrument attacks the hardest population — small, foreign and online sellers — through the platforms that host them. Germany's marketplace obligation (since July 2022) prohibits marketplaces from enabling sales by sellers who cannot show LUCID registration and a system-participation agreement; Ontario makes the marketplace facilitator itself the obligated producer for marketplace sellers; the EU's packaging regulation carries platform verification duties into directly applicable law (the definitions article). This converts the enforcement problem's worst segment into a checkbox in platform onboarding — the regulatory cost approaches zero, and the platform's compliance incentive does the work.

Mutual audit in competitive markets. For inter-system leakage, Germany's post-2019 answer recruits the market's own incentives: the clearing house determines market shares in agreement with the competition authority, and audit guidelines require auditors to verify competing systems' volumes — each participant having a direct financial interest in its rivals' honesty (the clearing-house discussion).

4. The internal free-rider: under-reporting as the quiet form¶

Non-registration dominates the rhetoric; under-reporting plausibly dominates the money, and it is the form the current assurance architecture is least equipped to catch.

The arithmetic of temptation is straightforward: a producer under-declaring 10% of tonnage buys a 10% fee discount, at detection probabilities that — on the published record — approach zero in most of North America. Ontario wrote a serious producer-audit procedure and waived it for the foundational 2025 and 2026 submission years; the US programs launched with no attestation or third-party audit regime at all; the UK builds validation into payment processing, with materiality thresholds designed to catch large distortions (the verification article). Because under-reporting redistributes cost within the producer pool, its natural constituency of victims is honest producers — who cannot see it, because aggregate audit results are published nowhere. A scheme that let compliant members see even sampled correction rates would create the internal demand for verification that no external constituency currently supplies.

5. Designing against free-riding: the checklist¶

The record supports six measures, ordered roughly by evidence of effect:

Universal registration with public searchability — the German model; exemption from payment need not mean exemption from existence in the register.
Marketplace obligations — verification duties or facilitator liability, closing the population regulators cannot reach.
Producer cascades that always land on a reachable party — the definitional layer as enforcement design.
Published audit coverage and correction rates — France's 15% is the benchmark precisely because it is public; a sampled misreporting estimate is the single most useful unpublished number in the field.
Sanctions proportionate to scheme budgets, actually used — the Ontario battery penalties as proof of concept; the French €30,000 ceiling as the counter-example.
Mutual-audit architecture wherever multiple schemes share costs — turning rivalry into compliance pressure.

6. Where the argument stands¶

Free-riding occupies a peculiar position in EPR debate: universally invoked, nowhere measured, and structurally convenient in its immeasurability — producers cite it to contest fee fairness, schemes minimise it to defend system integrity, and no party with the data has an interest in producing the estimate that would settle the matter. The German history shows the problem can grow large enough to threaten a mature system; the Ontario record shows detection and enforcement capacity in a new system running years behind the obligation; the design record shows the effective instruments are known, cheap, and — outside Germany — mostly unadopted.

The synthesis this library can defend: free-riding's true scale is unknown; the capacity to discover it is demonstrably absent in the jurisdictions examined; and the honest default is that a system which does not measure its perimeter should be presumed leaky in proportion to the weakness of its registry, its audit intensity, and its marketplace coverage. The problem, unusually for this field, has a complete and inexpensive solution set — universal registries, platform duties, published audit statistics — whose slow adoption says less about difficulty than about the absence of any constituency, inside or outside the schemes, whose interests are served by finding out how much is missing.

References¶

Auditor General of Ontario (December 2025), value-for-money audit of the Resource Productivity and Recovery Authority — the 1,804/71%/92%/262-day enforcement findings and the Blue Box scope exclusion; RPRA registration data (1,918 Blue Box producers, April 2025); RPRA penalty orders against battery producers ($2.78M) and their 2026 resolution.
German record: the self-collection and interim-solution loophole era as documented in the VerpackG legislative history and ZSVR materials; Rasek & Smuda (2018), De Economist 166(1) — the undecomposed confound; ZSVR guidance on universal registration ("no de minimis threshold") and the marketplace verification obligation (in force July 2022).
IGF/IGEDD/CGE (June 2024) — no sanction ever issued for missed targets; Cour des comptes (2020) — the €30,000 sanction ceiling.
Laubinger, F. & Brown, A. (2021), OECD Environment Working Paper No. 184 — the French 15% audit-intensity and Italian control figures.
RPRA Blue Box Verification and Audit Procedure (2021) and the November 2024 waiver; Holland & Knight analysis of the June 2026 US consolidated reporting round — the absence of attestation regimes; PackUK validation protocols and materiality thresholds.
O. Reg. 391/21 producer hierarchy (marketplace facilitator liability), as restated by RPRA; Regulation (EU) 2025/40 platform provisions.

Verification note: the central quantitative claims are absence findings — no published misreporting rate, no located enforcement action for false supply reporting, no measured unregistered population — established by targeted search and stated as such. See Sources and method.