Should producers pay for the recycling system, or run it?¶
When producers become responsible for packaging waste, a jurisdiction must decide whether they merely pay for a system someone else operates, or take over the operations themselves — and what share of the system's costs their payments must cover. This article reviews the evidence on both questions together, because they are the two variables that the best independent comparative research identifies as actually predicting scheme performance. It examines the operational-versus-financial choice and the hybrids between them; the "full net cost" standard and what hides inside it; who keeps the revenue from selling collected materials; and the early Canadian evidence that operational transfer concentrates cost pressure on producers in ways that can destabilise the policy itself.
1. The choice, stated concretely¶
The financial-only form leaves service delivery where it was — municipal collection contracts, municipal or regional processing — and routes producer money to those incumbents. The United Kingdom is the largest example: producers pay modelled costs to local authorities, which continue operating their services.
The operational form transfers the contracts themselves. The producer organisation procures collection, selects sorting facilities, markets the recovered commodities, and bears performance risk. British Columbia has operated this way since 2014; Quebec since its modernised system went live on 1 January 2025; Ontario since its transition completed on 1 January 2026.
Between the poles sit hybrids. Quebec's pre-2025 arrangement reimbursed municipalities close to fully while they continued operating. Oregon's model has producers funding enumerated system elements while ratepayers continue paying for collection itself. Maine reimburses municipalities at median benchmark rates. Each hybrid is a different answer to the same underlying question: how do you create cost discipline without moving the contracts?
The reason the choice matters is that contracts carry incentives. A municipality reimbursed for its actual costs has weak reasons to compress them — the classic moral hazard of reimbursement. An organisation that holds the contracts internalises every efficiency gain, and every service failure. A cost-reimbursed operator optimises against the payer. The allocation of the contracts is the allocation of the incentive to economise.
2. The evidence: the two variables that predict performance¶
The most analytically useful recent finding in this field comes from a 2025 comparative study of EU packaging systems by adelphi, a Berlin-based independent environmental research institute: operational responsibility and full-cost internalisation predict scheme performance better than either producer-organisation market structure or eco-modulation intensity — better, that is, than the two variables on which the policy debate spends most of its energy.
The proposed mechanism is the contract logic above. A body that owns the system's costs end to end optimises across collection, sorting and marketing jointly; a payer-only body optimises nothing; a cost-reimbursed operator optimises against the payer.
Three cautions should travel with the finding. It is cross-sectional: systems with operational responsibility differ from systems without it in age, wealth, infrastructure and landfill policy, so the study identifies an association within a structured comparison, not a causal effect. It is unreplicated: one study, however carefully constructed, is thin ground for a strong design recommendation. And the performance metrics are non-comparable across jurisdictions — recycling rates are calculated on different bases with different denominators, a problem examined below through the British Columbia dispute. What the finding does most usefully is reorder the agenda: if it is even directionally right, the first-order questions are who holds the contracts and whether the money covers the whole system.
3. The Canadian evidence: what operational transfer actually does¶
Canada has run all three forms sequentially, which makes it the best observational site — and the early record cuts against triumphalism.
British Columbia is the mature case. Since 2014 its producer organisation, Recycle BC, has held full financial and operational responsibility: contracting municipalities as paid collectors at its own rates, taking commodity price risk, and paying fixed contractor rates regardless of market swings. Its reported performance leads North America — though the reported rates are contested. A 2020 review disputed the denominators (non-obligated supply excluded, single-family housing bias, depot reliance for flexible plastics), and a published rebuttal defended them. That exchange between informed parties is the standing exhibit for a general proposition: performance comparisons across responsibility models rest on non-comparable measurement, and cross-jurisdictional claims carry an error margin nobody can quantify.
Quebec transferred operations on 1 January 2025 after a decade of near-full municipal compensation — and its first operational year ran materially over cost expectations, prompting staggered-payment arrangements to ease the impact on producers.
Ontario completed its geographic transition on 1 January 2026 — but by then its producers had already lobbied the provincial government successfully for a 2025 regulatory amendment deferring target increases and dropping planned service expansions (multi-residential buildings, schools, long-term care facilities and public spaces) on cost grounds.
The pattern deserves precise statement: operational responsibility concentrates cost pressure on producers visibly — which disciplines the system, or destabilises the policy, depending on the political response. The same concentration of cost that creates the efficiency incentive creates the constituency, and the standing, to demand weaker obligations when costs exceed projections.
4. What share of costs? The "full net cost" standard and its fine print¶
The companion question is coverage: what fraction of the system's costs producer payments must meet. Four genuinely different machines hide behind the percentages jurisdictions quote.
Full net cost. Producers pay the whole net cost of the defined system — gross cost minus commodity revenue. British Columbia, Quebec, Ontario and Colorado implement it, and it is the EU baseline under Article 8a(4)(a) of the Waste Framework Directive, which requires coverage of the "necessary costs" of separate collection, transport and treatment, net of revenues. The routinely forgotten qualification: the directive's derogations permit member states to drop to 80% for schemes established after 2018 and 50% for pre-2018 schemes serving national-only targets. "The EU requires 100%" is wrong as stated.
Modelled-efficient-cost payment. The UK pays local authorities 100% — but of modelled "efficient costs," grouped by housing type, rurality and deprivation, rather than of their actuals. The design deliberately avoids reimbursement's moral hazard by substituting a regulator's estimate for an owner's incentive. The price is that the model becomes the battleground: the local authority sector has publicly disputed the calculation's transparency and representativeness (the transparency article).
Median-cost reimbursement. Maine computes median per-tonne rates within tiers of similar municipalities and reimburses at the median — municipalities beating it profit, laggards absorb the difference. It is reimbursement with embedded yardstick competition, untested in operation at the time of writing; its predictable weakness is that tier construction becomes the contested step.
Escalators and enumerated elements. Minnesota steps producers from 50% to 90% of covered costs between 2029 and 2031; Washington ramps reimbursement to 90% by 2032. Oregon is the outlier and the most misdescribed: its statute contains no percentage at all. Producers fund specified elements — long-haul transport, service expansions identified by a needs assessment, contamination programs, and a processor commodity-risk fee buffering sorting-facility economics — while residential collection stays on ratepayers. The circulating "~28%" figure for Oregon is a modelled estimate of the resulting share, not a legislated number.
Does the level matter? The adelphi finding says yes — full-cost internalisation is one of its two predictive variables — and the mechanism is that coverage percentages are incentive allocations in disguise: a 50% scheme leaves half the marginal cost of a badly designed system on parties who did not design the packaging and cannot change it. The finding is consistent with the historical pattern: a 2014 study for the European Commission by BIO Intelligence Service and Deloitte found coverage ranging from roughly 10% to 100% across member states, with the strongest performers clustered at the top, and every second-generation statute since has moved upward.
5. The boundary: where the real money moves¶
The percentage is the headline; the boundary decides the amount. Every scheme must resolve which costs are inside the producer-funded system, and each of the following is a live fight somewhere:
| Cost item | Contested because |
|---|---|
| Litter and public-space collection | Open-ended liability; producers resist |
| Bins and containers | Capital rather than operating cost — Ontario's 2026 bin-responsibility confusion arose precisely here |
| Consumer education | Genuinely part of the system; also a marketing budget |
| Regulator cost recovery | Producer-paid but not program cost; routinely conflated in fee reporting |
| Residual disposal | Covered material that ends up in the garbage stream |
| Multi-residential and away-from-home service | Higher cost per tonne; the first thing dropped under cost pressure (as Ontario did in 2025) |
The United Kingdom supplies the sharpest documented boundary fight, and it is a drafting detail with large consequences: in the Year 2 fee formula, the costs of packaging "commonly disposed of in public bins or littered" are excluded from the cost numerator while that packaging's tonnage remains in the fee denominator — producers of bin-prone formats pay on tonnage whose management the payment does not fund. The general lesson: the boundary is where the distributional decisions are made, and it is settled in technical documents that attract far less scrutiny than headline percentages.
A related political observation: coverage levels are set at adoption and renegotiated under cost pressure — almost always downward in scope rather than in headline percentage. Ontario kept 100% coverage in 2025 and narrowed what was covered. Because scope changes are less visible than percentage changes, this is the path of least resistance, and it is where a jurisdiction serious about durable coverage should focus statutory protection.
6. Commodity revenue and price risk¶
Netting costs against material revenue sounds mechanical; it allocates one of the system's largest risks. Four live arrangements illustrate the choices: in the UK, revenue is netted off local-authority payments, so producers hold the price risk and authorities are indifferent to it. In Ontario, material title transfers to the producers' system, which owns and markets the commodities and holds both upside and downside (inter-organisation revenue-sharing terms are unpublished). In British Columbia, the producer organisation takes commodity risk explicitly, paying contracted collectors fixed rates regardless of market swings — the producer system as price insurer for municipalities. In Oregon, a processor commodity-risk fee, set through rulemaking, buffers sorting-facility exposure directly — socialising price risk across producers by design.
The principle underneath: whoever holds commodity risk needs either the scale to absorb it or an explicit smoothing mechanism. Reimbursement models that leave the risk with fragmented municipalities reproduce exactly the fiscal fragility that China's 2018 import restrictions exposed in North American recycling.
7. How to choose¶
Choose operational responsibility if the aim is to locate the efficiency incentive with the party that also controls design and procurement — accepting that producers will then have both the means and the standing to argue for weaker obligations when costs rise, and that the organisation must be capable of procurement at scale, contract management across hundreds of counterparties, commodity marketing, capital planning, and resident-facing service delivery. The last is the one that generates political trouble: operational transfer moves the contracts but not the public's mental model of who is responsible (the transition article).
Choose financial-only if service continuity and democratic control of collection matter more than end-to-end cost optimisation — and then invest heavily in the payment benchmark, because it becomes the entire battleground. There is no benchmark-free option: actual-cost reimbursement embeds moral hazard, modelled cost requires a model everyone trusts, and median-cost reimbursement makes tier construction the fight.
In either case, note what the evidence does not support: the claim that these transfers deliver savings to households. Municipal budget relief is real and documented — Ontario's transition moved the full operating cost of the residential Blue Box system off municipal budgets as of January 2026, and the sums flowing through the new systems are large (roughly £1.46 billion in the UK's first fee year; Can$457 million in Quebec's first modernised year; US$167.9 million in Oregon's first program year). But savings reaching households have not been demonstrated anywhere — a distributional question examined in Theme 4.
Finally, for the reader assessing any jurisdiction's coverage claim, six questions extract what the headline number conceals:
- Percentage of what? Full net cost, gross cost, or enumerated elements only?
- Measured against actual, modelled, or benchmark costs?
- What is inside the boundary — litter, bins, education, regulator costs, residual disposal, away-from-home collection?
- Who keeps the material revenue, and is it netted before or after the percentage is applied?
- What is excluded by sector — multi-residential, institutional, commercial?
- Is the coverage level in statute or in regulation? Ontario's 2025 experience shows the difference: regulation-based scope was narrowed by the government of the day under producer cost pressure, with no legislative process required.
References¶
- adelphi (2025). Efficiency and Performance of Packaging EPR Systems in the EU. The operational-responsibility and cost-internalisation finding; cross-sectional and unreplicated, as discussed in Section 2.
- Directive (EU) 2018/851, Art. 8a(4)(a) — necessary costs, net of revenues, and the 80%/50% derogations.
- BIO Intelligence Service & Deloitte (2014). Development of Guidance on Extended Producer Responsibility. Report for DG Environment, European Commission — the 10–100% historical coverage range.
- Recycle BC stewardship plans (2019; 2022 consultation draft); the 2020 external review of Recycle BC's reported performance and the published rebuttal (2019) — the measurement dispute discussed in Section 3.
- Québec MELCCFP deployment documentation; Éco Entreprises Québec first-operational-year reporting (2025).
- Ontario O. Reg. 391/21 and the 2025 amendments (ERO 025-0009) — the target deferral and scope narrowing.
- GOV.UK, local authority payment calculation methodology and Year 2 illustrative fee documentation — the numerator/denominator asymmetry; LARAC public criticism (July 2025; advocacy source).
- Maine DEP municipal reimbursement methodology (38 MRS §2146); Oregon Recycling Modernization Act and DEQ program documentation — no statutory percentage; Minnesota (2024) and Washington SB 5284 escalators.
Verification note: the Oregon "~28%" figure is identified as a modelled estimate rather than a statutory number, and the EU derogations are stated alongside the full-net-cost baseline — both corrections of common misstatements in the policy literature. See Sources and method.