Even if PackUK corrects a possibly miscalculated EPR fibre composite fee, this would only fix the smaller of the scheme’s two problems. Sameer Kulkarni of Ecofy argues the real issue is a 5% plastic threshold.
Under the year one base fees for Extended Producer Responsibility (EPR) confirmed by PackUK in June 2025, a fibre pack with a plastic liner is charged £461 a tonne. The plastic pack it may substitute for is charged £423. If you strip the liner out, the same item is paper and card at £196.
A scheme built to make producers carry the cost of managing their packaging, and to reward recyclability, charges more per tonne for lined fibre than for plastic itself.
I should say at the outset that what follows runs against my own interest. I work for a moulded fibre manufacturer whose uncoated products sit below the 5% threshold and are charged at the paper and card rate. The cliff edge I am about to criticise currently works in our favour.
In June, Circular Online reported that those fibre-based composite fees may have been significantly overcharged.
Research by Beyondly, commissioned by ACE UK, the Alliance for Beverage Cartons and the Environment, argued the £461 figure had been ‘incorrectly calculated’ and put the true cost between £34 and £92 a tonne lower.
PackUK had not publicly responded at the time of writing. When the allegation was put to Defra in the Commons on 1 July, the minister acknowledged the sector and did not address it.
That argument may be right, and it matters who is making it: the trade body for the material being repriced. But my point is different: if PackUK accepts every figure and reprices the category tomorrow, the thing that makes this fee behave strangely will still be there.
What a recalculation reaches, and what it does not
Apply the Beyondly correction and the inversion mostly disappears. It survives only between £34 and about £38 of reduction; past that, fibre composite falls below plastic. On their own numbers, most of the range they argue for would fix it.
So take the inversion away, and the structural problem is untouched.
The rule is a threshold, with RAM 2027 being explicit: ‘You must assess paper or board packaging with plastic content greater than 5% by mass as FBC materials.’
RPS 351 already allowed producers in England to apply that line early, and it has been statutory since the 2025 amendment regulations came into force on 1 January 2026.
The burden of evidence sits with the producer: paper or board status depends on showing the plastic layers are ‘not more than 5% of the packaging material by mass’. Below the line is not automatic. You have to prove it.
One specification decision moves a pack between two bands, and at year one fees the gap between them is £265 a tonne.
The scheme does band elsewhere. RAM 2027 rates paper or board amber above 10% non-paper content and red above 15%, and applies the same test to fibre composite.
However, those triggers sit inside a material band and change the multiplier applied to it. Only the 5% line moves a pack from one fee table to another. That is the difference between being modulated and being reclassified, and it is the whole argument.
Recalculating £461 changes how big the cliff is. It does not turn the cliff into a slope.
Not every category boundary is this one
A reasonable objection: all material bands are thresholds. Glass against aluminium, steel against wood. But those edges fall between materials that are genuinely different.
Plastic content within a fibre pack is a continuum. A pack can sit at 2% or 30%, and the property the scheme wants reduced is priced as a binary. That is where a threshold distorts, and an ordinary material boundary does not.
The strongest counter is that 5% proxies something binary: whether the pack repulps. If so, the evidence for the line should be published, because the behaviour it produces is not binary at all.
The second objection is stronger: base fees are blunt by design, and modulation is the instrument meant to deliver the behavioural signal. Year two fees are RAG-modulated, so the gradient is arguably arriving.
Except that RAG is red, amber and green. Three bands. That is not a gradient; it is three more thresholds, and the material classification is settled before modulation applies.
A pack is assigned to fibre-based composite or to paper and board first, and only then does its rating move it within that band.
A producer cutting from 30 to 20% may gain something within the band, but movement inside a band is bounded by it: the whole green to red spread for fibre composite is £155, against the £315 for crossing the line.
Modulation softens the cliff; it cannot flatten it. The green discount is applied as an equal percentage across every material category, so the proportional distance between the bands survives at every rating.
What a threshold rewards
Take a producer at 6% plastic by mass. Moving to 4% reclassifies the pack and is worth £265 a tonne.
Take a producer at 30%. Cutting to 20% is a substantial redesign and worth a fraction of that. The pack stays in the same band, so the only saving is the mass removed, still charged at £461.
On a thousand tonnes, that is roughly £46,000, against roughly £265,000 for the producer who crosses the line. A third of the plastic gone, for about a sixth of the reward.
The incentive is strongest exactly where the environmental gain is smallest.
However, there is a second effect. Once a threshold exists, the rational response is not to minimise plastic but to design to just under 5% and stop. The number stops being an environmental target and becomes a compliance boundary.
So what does a producer in front of that cliff do? There are three responses, and only one is the behaviour the scheme wants.
The first is to reformulate below the line, worth the full £265 a tonne but available only to those already near it. The second is to absorb the fee, which is rational for anyone too far above 5% to reach it, and which removes no plastic at all.
The third is to leave the category, either for an unlined fibre format where the application allows it, or back to plastic, which at £423 is cheaper than lined fibre and is the outcome the scheme was built to prevent.
Only the first response reduces plastic. The second does nothing. The third can make things worse.
Year two widens it
The illustrative year two disposal fees, published by Defra in December 2025, do not narrow any of this. Fibre-based composite is 475, 525 and 630 for green, amber and red. Plastic is 415, 455 and 545. Paper and board is 190, 210 and 250.
The inversion widens from £38 at year one to £70 at amber. The band gap runs from £285 at green to £315 at amber to £380 at red. Those figures are illustrative.
The government’s own note says they ‘are intended to help with planning but are likely to change significantly as producers submit more data and compliance is monitored by regulators’, and confirmed year two base fees had not been published when this was written, though PackUK has said it does not intend to issue a further illustrative set. But the direction is not ambiguous.
The trade-off, stated
A charge proportional to plastic content would remove all of this. Every kilogram removed would be worth the same, and nobody would be designing to a boundary.
It would also be harder to administer. A threshold needs one declaration a year; a gradient needs measured, verified content across every line. That is why thresholds exist, and it is not a trivial objection.
But the scheme already bands within materials, so the question is not whether banding is possible, only where the bands sit and what they attach to.
I have not found where the trade-off between administrability and behavioural distortion is set out in the published consultation material. It may be there, and I may have missed it.
If not, it is worth stating plainly: the scheme has accepted a significant distortion in exchange for a simpler system, and that is defensible only if it is made openly.
ACE UK has asked PackUK to review the figures, and it should. But if that review ends with a corrected number and an unchanged structure, the scheme will have fixed the smaller of its two problems.