Mattress, Carpet, and Paint Stewardship Laws Ran the EPR Experiment First. Packaging Is Copying the Wrong Parts.

Which products actually have stewardship laws, and where?

Start with the map, because most operators I talk to have it backwards. Product stewardship laws in the United States don't cover "durable goods" or "bulky items" as a category. They cover specific products a legislature decided were expensive to landfill and easy to trace back to a manufacturer. Mattresses, carpet, and paint are the three with the longest records, and they behave nothing like the packaging EPR everyone's been writing about in 2026.

Here's the shape of it before we get into who pays:

Three products, three stewardship organizations, one visible fee on the receipt. Put numbers on it and the spread is wide: California's mattress fee climbs to $18/unit on April 1, 2026, per the Mattress Recycling Council, while a paint can carries anywhere from $0.30 to $2.75 depending on size. That pattern of one product, one steward, one flat fee is the whole story, and it's why these programs hold where others stall. California's carpet stewardship program under AB 2398 was the first product-specific EPR law in the country when it passed in 2010. The mattress recycling EPR programs came next, and the PaintCare model has spread the widest, which tells you something about how tractable a paint can is compared to a queen boxspring.

Who actually pays, and who gets fined if the fee isn't collected?

The consumer pays. The fee sits on the receipt as a separate line item, which is deliberate: legislators wanted it visible, not buried in the shelf price. So far so simple. The part that trips people up is the enforcement side, because "producer responsibility" splits into three jobs that land on three different parties.

The producers fund and run the stewardship organization. The stewardship organization reports to the state and contracts out the actual collection and processing. But the legal duty to collect and remit the fee at the register lands on the seller, and that includes an out-of-state online retailer shipping a mattress into California. Sell that mattress without adding the line item and you're the one exposed, not the manufacturer whose name is on the statute. In my compliance reviews the recurring surprise, every time, is a retailer who assumed the brand "handles EPR" and finds out the collection obligation was theirs.

This is the same split I walked through for whether a 2026 battery EPR obligation lands on you or the producer: the entity named in the law often isn't the entity that gets the notice of violation. Read the definition of "producer" and the definition of "retailer" in each state's statute before you assume which one you are. They're not always what the marketing says.

Do these programs actually recycle the material, or just collect a fee?

Depends which number you trust, and this program publishes more than one. So which should you actually believe? California's carpet program hit a 38.5% recycling rate in 2024 against a statutory goal of 34%, its fifth straight year of improvement, per Resource Recycling's write-up of the CARE data. Genuinely good. Collection climbed to roughly 83 million pounds. So the program collects more and recycles more every year.

And yet CalRecycle failed CARE's 2022, 2023, and 2024 annual reports in its published compliance determinations. Three consecutive years. That looks like a contradiction until you remember what the compliance object is in an EPR program: it's the plan and the report, not the tonnage. You can beat the recycling goal and still file a report the regulator rejects, because the rules bite on process and documentation, not on how many pounds you moved. Most ESG-compliant waste projects fold the same way the moment an auditor asks them to document a number they've been reporting for years, and the carpet program is the cleanest public proof of it I know.

Actually, be careful with that rate. It's the share of carpet sold that came back as recycled output. CARE also reports a recycling efficiency figure above 90%, but that one is the share of what they collected that got recycled. Different denominator, and the space between the two numbers is the collection problem nobody has solved. The economics only close for one fiber anyway: nylon 6, which Aquafil depolymerizes back to caprolactam under its ECONYL process. You have to identify the polymer before you can route it, so a serious carpet line runs near-infrared sorting, Tomra or similar, to split nylon 6 from the polyester and polypropylene that mostly head to lower-value fill or waste-to-energy (a fine outcome, just not the one the statute is counting). That gap is why AB 863 in 2024 bolted a carpet-to-carpet recycled-content mandate onto the program, 5% by 2028, to build a pull market instead of just subsidizing collection.

Mattresses recycle better on paper because a mattress is mostly steel, foam, wood, and fiber, and the steel pays. Roughly three-quarters of the unit is recoverable by weight. The processors who actually pull those fractions apart, running the shredders and magnetic separators behind these programs, are the ones building genuine zero-waste-to-landfill solutions for bulky streams. But none of these "recycling rates" is landfill diversion the way a municipal recycling coordinator means it. A mattress that gets shredded for its steel and sends the foam to a boiler counts very differently from one remanufactured, and the programs don't all draw that line the same way. Read the denominator before you quote the rate.

Why do mattress, carpet, and paint get EPR when most products don't?

Three features these products share, and packaging shares none of them. First, each is one identifiable product, not tens of thousands of SKUs. Second, each had a dominant trade association ready to become the monopoly steward: the mattress industry stood up the Mattress Recycling Council, the carpet industry stood up CARE, the coatings industry stood up PaintCare. Third, each is bulky or hazardous enough that landfilling it is a visible, unpopular waste, so the political case writes itself. Bulky waste producer responsibility is an easy sell precisely because the product is hard to hide in a dumpster.

So why not bolt the same fee onto a soda bottle and move on? Because packaging has the opposite profile on all three counts. Millions of SKUs across every material. No single trade association that can credibly speak for a soda bottle and a shipping mailer at once. And it's light, compactable, and already inside the curbside stream, so the political urgency is softer. The mistake I see in the 2026 packaging laws is treating the mattress-carpet-paint model as a template. A flat, visible, one-size fee works when there's one product and one steward. Packaging fees have to be eco-modulated across thousands of materials, and there's no ISPA equivalent to run it. The mature programs are a bad blueprint for packaging exactly because they're so clean, and the ones held up as proof that EPR works are the least like the packaging law they're cited to justify.

Which states are adding product stewardship laws in 2026, and how should a facility position for it?

Paint keeps spreading: Maryland's PaintCare program launched on April 1, 2026, the newest state to switch it on, and more than 88 million gallons of architectural paint have moved through the PaintCare system to date. Mattresses are quieter. New York, Washington, and Massachusetts have all floated mattress bills, but none had passed as of early 2026, so the count stays at four states. Carpet remains a California experiment other states are watching rather than copying, which given the compliance record is a defensible instinct. State product stewardship in 2026 is still, overwhelmingly, a paint-and-packaging story with mattress and carpet as the older cousins.

If you're a processor or an investor, here's the position: the money was never in the fee. It's in being the contractor the stewardship organization hires. MRC, CARE, and PaintCare don't own trucks or run shredding lines; they contract throughput out. So the offtake question for a bulky-waste recovery line isn't "what's the gate fee," it's "will a steward organization sign a multi-year processing contract at my throughput, and can my downstream document where the material went." That second clause is what turns a recovery line into one of the ESG-compliant projects a lender will actually fund. If you're scoping that math against one of these programs, the contract terms with the steward decide the model, and it's worth getting a read on the offtake before anyone signs off on capex.

I'll admit where this kind of read has bitten me. In 2021 I told a client their RCRA Subtitle D path was clear because the state permit history looked clean; in 2022 EPA Region 4 reopened it on a derived-from-rule interpretation I hadn't seen coming, and they lost the better part of a year. Different program, same lesson: a clean surface signal and a compliant record are separate objects, and the regulator grades the record. Fifteen years in, the carpet program recycles more every year and still can't file a report the state will accept. That's not a paradox. It's what producer responsibility looks like once the press releases stop, and it's the part the 2026 packaging laws forgot to copy.

Sources & Notes

• The California mattress fee moving to $18 a unit on April 1, 2026, and the four-state footprint, come straight from the Mattress Recycling Council's fee-adjustment notice.
• For the carpet numbers I leaned on Resource Recycling's report on the 2024 CARE results: the 38.5% rate, the 34% goal, and the collection tonnage.
• The noncompliance finding is not my characterization; it's on CalRecycle's carpet program results page, which posts the annual-report compliance determinations.
• Paint fees by container size and the state list are published on the PaintCare fee page; the Oregon mattress assessment of $22.50 is documented by Oregon DEQ.
• For who's moving next, the Product Stewardship Institute's 2026 legislative tracker is the cleanest running list of bills by product category.