Does EPR increase recycling rates?¶
The claim that extended producer responsibility raises recycling rates is the instrument's central performance promise, repeated in every legislative debate and program announcement. The state of the evidence behind it is more unsettling than either side of those debates acknowledges: no difference-in-differences or synthetic-control study of packaging EPR adoption exists anywhere in the peer-reviewed literature — a targeted search for this library found none — and the evidence that does exist consists of cross-sectional comparisons with deep confounds, panel studies of policy bundles that cannot isolate EPR, before/after program records whose measurement bases changed mid-series, and mature systems whose rates have plateaued or declined. This article assembles that record in full: what the studies actually show, what the program histories actually show, the attribution problem that contaminates every causal claim, and the precise statement of what can and cannot currently be said.
1. Why the question is hard¶
EPR never arrives alone. In Europe, packaging EPR spread in the same two decades as the Landfill Directive, landfill taxes, mandatory separate collection, deposit systems and recycling targets; in Canada, provincial EPR overlaid decades of municipal programs; in the United States, the new state programs land on mature kerbside infrastructure. Any recycling-rate change following EPR adoption is therefore consistent with several causes, and attributing it to EPR requires a counterfactual — what rates would have done otherwise — that only quasi-experimental designs can construct.
There is also a conceptual ambiguity the debate rarely surfaces: what would it even mean for EPR to "cause" recycling? The candidate mechanisms differ sharply. EPR might raise rates by funding more collection (the financing channel — well documented as a transfer, per Theme 4); by fee incentives changing producer or consumer behaviour (the price channel — theoretically weak at observed magnitudes, per Themes 1 and 5); or by institutional consolidation improving system management (the governance channel — plausible, unmeasured). A finding that "EPR jurisdictions recycle more" cannot distinguish these, and the policy implications differ: the financing channel could be replicated by any adequate funding source; only the others are specific to producer responsibility.
2. What the peer-reviewed literature actually contains¶
The absence finding comes first, because it frames everything else. Searches for quasi-experimental evaluations of packaging EPR adoption — difference-in-differences across adopting and non-adopting jurisdictions, synthetic controls, regression discontinuity around program starts — returned none. Any claim of the form "EPR causes an X-point increase in recycling" is, on the identified record, unsupported by peer-reviewed causal evidence. The closest candidate — a 2026 statistical assessment of the EU packaging directive's 2004 amendment on plastic recycling, in Resources, Conservation and Recycling — could not be accessed for this library and is flagged for retrieval rather than characterised.
What the literature does contain falls into four families.
Panel studies of policy bundles. The econometric work on European waste outcomes — the reference example being Massimiliano Mazzanti and Roberto Zoboli's panel analysis of EU landfill diversion — finds policy mattered: EU waste policy was "a strong driver" of diversion from landfill. But the identified driver is the disposal-side pressure of the Landfill Directive era, and the studies neither isolate packaging EPR's contribution nor find upstream prevention effects. The European recycling achievement, on this evidence, is a bundle achievement in which EPR's specific share is unidentified.
The fee-incentive test. One study directly tests whether EPR's price mechanism moves recycling rates: a 2016 fixed-effects panel analysis by Calvin Lakhan in the journal Recycling, covering 23 Ontario Blue Box materials over roughly a decade with instrumental-variables methods. Its finding: packaging fee rates have no effect on material recycling rates — a one-cent fee change moves rates by an economically null 0.0001% — while rates track material revenue and management cost. Whatever raised Ontario's rates, it was not the fee signal — convergent with the design-change literature of Theme 5, from the recycling-rate side.
Ex-ante and framework analyses. A 2025 assessment in Resources, Conservation and Recycling by Jan Pruess and Rachael Garrett, comparing plastic-packaging EPR in Belgium, France and Germany, concludes that effectiveness "depends heavily on broader regulatory, socioeconomic, and value chain contexts — not just the policy itself" — a formalisation of the attribution problem rather than a solution to it.
Cross-sectional benchmarking. The most-cited quantitative claim on the pro-EPR side is The Recycling Partnership's 2023 seven-jurisdiction study: programs in British Columbia, Belgium, Spain, South Korea and the Netherlands achieving over 75% collection or recycling of targeted materials, with modelling that US EPR could raise residential rates "by as much as 48 percentage points." The figures are real; the design cannot carry the causal weight put on it. The comparison jurisdictions differ from the US in deposit coverage, density, landfill pricing, collection maturity and measurement method (the measurement article); the study is benchmarking by an industry-funded pro-EPR organisation; and no counterfactual is constructed. It shows that high-performing systems have EPR — not that EPR made them high-performing.
3. The before/after records¶
The program histories are the second-best evidence available, and each carries a defect that the field rarely quotes alongside the numbers.
British Columbia is the flagship, and its record is genuinely strong on its own terms: recovery rates of 77–80% in the program's first years (2014–15), roughly 85% by 2020–21, 79.6% in 2023 and 83.3% in 2024 (92% for paper; 45% for plastics). The program's own first annual report is instructive about the inferential limits: it recorded a recovery-rate decline from 80% (2014) to 77% (2015) that reflected growth in steward-reported supply — the denominator — rather than any collection deterioration, while the program simultaneously missed its own plan target of maintaining 208,700 tonnes of collection (186,509 tonnes were collected). From its first full year, in other words, the flagship program's headline rate was moving for denominator reasons its own reporting had to explain away. Two structural defects then prevent any causal reading of the series. There is no pre-2014 baseline on the same measurement basis — the program's own first reports state no prior-system recovery rate, and the denominator (steward-reported supplied tonnes) came into existence with the program. And the metric is a collection measure against an obligated-supply denominator, not reprocessed output against generation (the full anatomy).
Ontario provides the longest consistent series anywhere, and it is the plateau exhibit: Blue Box recovery rose from roughly 53% (2003) to a peak of 68% in 2010, then declined for a decade to 57.3% by 2019 — below the 60% statutory target — under the shared-cost EPR model, with industry analysts projecting only ~58% by 2030 under the pre-transition structure. Whether full producer responsibility (from 2023–26) bends that curve is exactly the question the transition's data disruption has made hard to answer: no official post-transition recovery rate had been published at the time of writing, with 2024 collection data still pending (the evaluation article).
Oregon, the first US test, has reported only a first six months: 144,771 tons collected, 20 depots opened, 24.7% measured contamination — infrastructure and tonnage, no rate, and the state's pre-program recovery rate (39.4% of generation in 2022, on a generation basis) is not comparable to anything the program will report about covered products.
The mature-system plateau is the most under-reported pattern in the performance record. Eurostat's EU-27 packaging recycling series is essentially flat for over a decade — 65.2% in 2012, 67.6% in 2016, a methodology-driven dip to 64.0% in 2020–21, 67.5% in 2023 — and the European Environment Agency states plainly that progress "has recently slowed down," with packaging rates having "barely changed" since 2010. France is the sharpest single counterexample to any simple EPR-performance link: three decades of continuous packaging EPR under Eco-Emballages/Citeo, and a Eurostat plastic packaging rate of 25.7% (2023) — among the worst in the EU-27.
4. The improvement that did happen — and what drove it¶
Against the plateaus, the record contains one large, verified, recent improvement, and its anatomy matters more than its size: German dual-system plastics mechanical recycling rose from 42.1% (2018) to 68.9% (2023) — a 27-point gain in five years, in the world's oldest packaging EPR market, where EPR itself had been constant throughout.
What changed in 2019 was not producer responsibility but its enforcement architecture: the Packaging Act imposed escalating statutory recycling quotas on the dual systems (rising to 63% for plastics from 2022, on a strict mechanical-recycling basis), created the central register that closed the free-riding channels, and gave the register authority annual published quota verification with enforcement consequences. The dual systems responded by contracting for sorting and reprocessing capacity because the law left them no alternative — and the same 2023 verification shows the mechanism's limits, with glass missing its quota by 6.9 points and beverage cartons by 8.7.
The German episode is the cleanest available demonstration of a proposition that reframes the causal question: where recycling rates moved in a mature EPR system, the mover was binding quotas with enforcement, operating through the EPR structure — not the producer-responsibility financing itself. EPR was the delivery vehicle; the target regime was the engine. That reading is consistent with everything else in this section: Ontario declined for a decade under an EPR whose targets were unenforced; France's plastics stagnated under an EPR whose sanctions regime the Cour des comptes called negligible (Theme 2); and the e-waste EPR literature — the analogue stream with the longest evaluation history — reaches the matching conclusion that institutional and enforcement context, not the EPR form, explains outcome variation.
5. What the institutional reviews conclude¶
The European Court of Auditors' 2025 municipal-waste report — with EPR now mandatory for all EU packaging — found 23 member states at risk of missing at least one 2025 recycling target; for the 55% municipal target, only six member states had met it or come within five points. Its EPR-specific findings are about implementation shortfall: fees not covering costs in Poland and Portugal, schemes not fully operational in Greece, and secondary-material markets collapsing beneath the collection systems. The EEA's early-warning assessment identified only nine member states likely to meet both 2025 targets, and nineteen struggling to reach the 50% plastic-packaging target. The pattern across both: the presence of EPR does not predict target attainment; infrastructure, enforcement and market conditions do — consistent with the adelphi finding (Theme 2) that operational responsibility and cost internalisation, not scheme existence or structure, track performance.
6. What can and cannot be said¶
Assembling the record, with confidence levels:
| Proposition | Confidence | Basis |
|---|---|---|
| EPR reliably finances collection systems at scale | High | The documented transfers (Theme 4); universal program records |
| Jurisdictions with mature EPR report high collection/recovery rates | High | BC, Belgium, Germany, Netherlands program data — with measurement caveats |
| EPR caused those high rates | Unsupported | No quasi-experimental study anywhere; deep bundle confounds |
| The fee mechanism raises recycling rates | Contradicted | The one direct test (Ontario panel) finds a null |
| Mature-system rates have plateaued | High | Eurostat 11-year series; EEA; the Ontario decline |
| Money buys collection; collection is not recycling | High | The yield-loss and measurement evidence (next article) |
The defensible synthesis runs as follows. EPR demonstrably builds and funds collection infrastructure — that is the financing achievement documented throughout this library, and where no such infrastructure existed, funding it will raise collection, almost definitionally. What the evidence does not support is the stronger claim in circulation: that EPR adoption as such raises recycling rates relative to an adequately funded alternative, or that the producer-responsibility form contributes anything beyond the money. The European plateau, the French plastics record, the Ontario decline under shared-cost EPR, and the null result on fee incentives together suggest that EPR's recycling effect is the effect of its budget — and that once collection is funded and mature, the binding constraints move elsewhere: to sorting yields, end markets, and the measurement practices that decide what gets counted at all. Those constraints are the subject of the rest of this theme.
References¶
- Mazzanti, M. & Zoboli, R. — panel analyses of EU waste generation, landfill diversion and policy effectiveness (working-paper and journal versions) — the policy-bundle finding.
- Lakhan, C. (2016). "Do Eco-Fees Encourage Design for the Environment? The Relationship between Environmental Handling Fees and Recycling Rates for Printed Paper and Packaging." Recycling 1(1), 136–146 — the fee-rate null result (fixed-effects panel with IV, Ontario).
- Pruess, J.T. & Garrett, R.D. (2025). Ex-ante policy impact analysis of plastic-packaging EPR in Belgium, France and Germany. Resources, Conservation and Recycling — the context-dependence conclusion.
- The Recycling Partnership (2023). Increasing Recycling Rates with EPR Policy — the seven-jurisdiction benchmarking and modelled 48-point claim; industry-funded pro-EPR organisation, identified as such.
- MMBC/Recycle BC annual reports (2015 through 2024) — the BC recovery series and the absence of a pre-2014 baseline; trade-press summaries of the 2024 report (83.3%).
- RPRA, Datacall Report 2019 (the 57.3% recovery rate) and resource-recovery data publications (2023–26); trade and industry analyses of the Ontario 2003–2019 series (peak 68% in 2010) and 2030 projections.
- Circular Action Alliance, first Oregon annual report (2026), as summarised in trade coverage; Oregon DEQ 2022 Material Recovery Report (the 39.4% baseline, non-comparable basis).
- Eurostat, packaging waste statistics (series 2012–2023); European Environment Agency, early-warning assessment (2023); European Court of Auditors, Special Report 23/2025.
- Citeo, published French recycling figures (2023–25); Eurostat country data — the French plastics counterexample.
- A 2026 statistical assessment of the EU packaging directive amendment (Resources, Conservation and Recycling) was identified but could not be accessed; it is flagged for retrieval rather than characterised.
Verification note: the central claim of this article — that no quasi-experimental evaluation of packaging EPR adoption exists — is an absence finding from targeted searches conducted in August 2026, stated as such; the one inaccessible candidate study is disclosed. See Sources and method.