The Sneaker Economy Has a Structural Recycling Problem, Finds Fashion for Good Study

The non-rewearable fraction is where footwear circularity either begins or ends, yet it remains the least understood and least acted-upon part of the waste stream. Half of all post-consumer footwear collected through European textile systems is classified as non-rewearable, a share that already exceeds the equivalent figure for apparel, where 45% falls into the same category. That gap matters, because non-rewearability is not a fixed characteristic of the incoming stream. It is, in significant part, a constructed outcome, shaped by disposal behaviour, sorting thresholds, and the absence of any systematic intervention between the moment a shoe is discarded and the moment it is graded.

The damage data makes this legible. Across the 1,200-shoe sample, 72% of shoes displayed some form of damage on the upper, but the dominant forms were soiling, at 50.5%, and discolouration, at 27%. Structural failure, tears or holes, accounted for only 16%. Sole damage was less prevalent still, appearing in just 26% of the sample, with wear from use the most common presentation at 15%. The picture that emerges is not one of products exhausted by use. It is one of products disqualified by surface condition, by aesthetic deterioration that cleaning or refurbishment could plausibly address, and by the absence of infrastructure capable of making that intervention at scale.

The 24% figure, shoes carrying no visible damage at all yet classified as non-rewearable nonetheless, makes the problem legible in a single number. These are shoes removed from the reuse stream not by structural failure but by system failure: disposed of without their pair, soiled in ways that crossed market thresholds, or simply absorbed into a waste category that, once applied, closes off every pathway except export, incineration, or landfill. Non-rewearable and contaminated footwear is predominantly exported to Pakistan and India, or directed to incineration and landfill. It does not enter circular recovery pathways.

The archetype breakdown adds a further layer. Lifestyle shoes dominated the sample at 37.5%, followed by sandals at 25.2%, slippers and slides at 14.9%, and performance shoes at 11.2%. Within lifestyle shoes, trainers accounted for 81% of the sub-archetype, the category that constitutes the most common second-hand footwear and the one most heavily represented in European collection systems. These are also the shoes most likely to carry soiling and discolouration as their primary damage type, and most likely to arrive as single, unpaired items. The dominance of this archetype in the non-rewearable stream is not incidental. It reflects a consumption pattern, high-volume, short lifespan, casual disposal, that the current infrastructure has no means of interrupting before the reuse window closes.

Footwear arrives at sorting facilities as a categorically different object from apparel, where 93% of the non-rewearable fraction is mono-layered and the material complexity that compounds end-of-use decisions is largely absent. Multi-component, permanently bonded, and carrying damage profiles that existing infrastructure was not designed to triage, it presents a sorting challenge of a fundamentally different kind. Sorting remains market-driven rather than material-driven, oriented toward the reuse grades that generate revenue, with no economic rationale to assess whether a soiled or unpaired shoe could be redirected before it is written off.

The non-rewearable designation is applied at the point of sorting, but the conditions generating it are set much earlier, at the moment a consumer discards a single shoe, or drops a soiled pair into a collection bin without cleaning it. Until the intervention point moves upstream, the fraction classified as non-rewearable will continue to be larger than the material condition of the shoes warrants.

Even among non-rewearable shoes that have exhausted their reuse potential, the path to recycling is blocked not by the chemistry of what is inside them but by the structural logic of how those materials were assembled and pigmented. The barrier is construction complexity and pigmentation choices made at the design stage, optimised for performance and aesthetics, with no consideration of what happens when the shoe reaches a sorting facility.

The near-infrared (NIR) scanning data is unambiguous on this point. Of sole materials in the sample, 37% could not be identified by the near-infrared technology the sector relies on for automated material classification. The cause is specific and traceable: carbon black pigments, used in the majority of black soles as both a performance additive and a colourant, absorb near-infrared light rather than reflecting it. A sensor that depends on spectral reflection cannot read a material that produces none. The consequence is that 97% of scanned black soles returned an "unknown" result, not occasionally, not in edge cases, but as the systematic output of a pigmentation decision embedded in standard design practice. Black soles accounted for 24.6% of the sample, and the cause resides in the pigment, not the scanner.

The adhesive system operates in a different register. Glue-based bonding dominated assembly across 51.9% of the sample, with stitching the second most common method at only 19%. The presence of adhesive does not merely complicate disassembly. It actively interferes with NIR material identification, since the adhesive layer obstructs detection of the material beneath it. Even where a material could theoretically be identified, the glue introduces a contamination risk that persists through shredding and into the recycled stream. Nearly all performance shoes were assembled using glue, as were the majority of lifestyle shoes. The exceptions, slippers and slides where stitching dominates, are precisely the archetypes with the simplest construction and the fewest layers, a pattern that holds across every archetype where complexity is highest.

The sole composition data extends the picture further. Only 10.5% of analysed shoes shared the same material in both the midsole and outsole, meaning that in nearly nine out of ten shoes, the sole alone presents a multi-polymer separation challenge before any recycling process begins. Among the categorised sole materials, SBS accounted for 16%, EVA for 15%, and rubber and SBR combined for 14%, each requiring different processing conditions, each incompatible with the others if mixed into a single stream. Upper materials presented an equivalent challenge: 52% of uppers consisted of blended compositions, with polyester/cotton, acrylic/wool, and polyester/elastane the dominant combinations. Both mechanical and chemical recycling require high levels of material purity. Blended inputs, by definition, cannot.

The PICVISA automated sorting trials put these design consequences into yield data. Performance shoes lost 73% of their material as light fraction during shredding and density separation. EVA recovery dropped from 33% under manual scanning conditions to just 6% in automated sorting, as the polymer was misclassified into lighter fractions after shredding altered its physical state. TPU fell from 14% to 5% by the same mechanism. Rubber was the exception, emerging as the most reliably recoverable polymer across all test batches, consistent in its behaviour precisely because its properties do not shift unpredictably through the shredding process. The trials also confirmed what the composition data implied: when input is heterogeneous and unsorted, recovery rates collapse. The unmarked batch, the largest by volume, had the lowest recovery rate of all five test conditions. Homogeneity of input is the precondition for viable output.

The disruptor data completes the picture. External disruptors, logos, patches, trims, metal hardware, were present on 89.9% of shoes in the sample. Internal disruptors, detectable only by X-ray, appeared in 21.8%. These are not incidental features. They are the accumulated consequence of design decisions that prioritised branding, aesthetics, and performance enhancement without any accounting for what their presence means at end-of-use. Each disruptor increases preprocessing cost, reduces recyclate purity, and in some cases poses direct risks to recycling machinery. The recycling readiness gap was designed in, not left by accident.

One pilot demonstrates technical viability. The conditions it required to function expose how far the industry remains from a circular infrastructure capable of operating at the volumes the non-rewearable fraction generates. THE 8 IMPACT's end-to-end sneaker recycling study, processing over 2,000 pairs through a fully validated process, confirms that closed-loop material recovery from post-consumer footwear is technically achievable. What it also confirms is that none of those conditions currently exist at system scale in Europe.

The process THE 8 IMPACT deployed was not a simplified proof of concept. It was a precisely sequenced chain of dependencies. Sneakers were categorised into three types, complex, standard running, and vulcanised, and each category was processed separately to maintain material consistency. Pre-sorting was conducted by local French social enterprises, achieving approximately 98% accuracy, with statistical verification reducing sorting errors from roughly 7% to 3%. Materials were mechanically separated, then micronised from granulates of approximately 4mm to particles of approximately 500 µm, enabling optimal interlocking with virgin materials while minimising energy use. IR spectroscopy was employed throughout to verify material identity. Pre-industrial batches containing up to 40% recycled rubber and EVA met brand Bills of Properties and passed toxicological requirements under REACH compliance. Rubber emerged as the most reliably recoverable polymer, consistent with the PICVISA trial findings. EVA recovery was viable but dependent on controlled batch conditions that the broader non-rewearable stream cannot currently guarantee.

A further structural condition governed the economics. Extended Producer Responsibility (EPR) frameworks, specifically France's Refashion-administered Textile, Household Goods and Footwear scheme, helped offset costs for smaller batch runs. Without that cost support, the process economics do not close. France's regulatory architecture, currently the most developed EPR scheme in Europe for footwear alongside textiles, is a structural element of the business case. Replicating the pilot outside that regulatory environment means replicating the environment first.

The PICVISA data is explicit on this point. When input is heterogeneous and pre-sorting is absent, the system collapses into material loss rather than material recovery. The unmarked batch, the largest by volume, the closest approximation to what arrives at a typical sorting facility, produced the lowest recovery rate across all test conditions. The gap between the THE 8 IMPACT outcomes and the PICVISA unmarked batch outcomes separates two input conditions, not two technologies: one controlled, category-consistent, socially sorted, and EPR-subsidised; the other representative of what the non-rewearable stream actually looks like when it arrives without preprocessing. The distance between those two conditions is the distance the industry has yet to travel.

Category-based pre-sorting is the foundational condition for consistent material quality, yet no economic rationale currently exists for sorting facilities to perform it—their business model is oriented toward the reuse market, with no regulatory signal distinguishing footwear from apparel or non-rewearable from rewearable for recycling purposes. Brand offtake commitments remain the exception rather than the embedded expectation. Reverse logistics at category-consistent scale does not exist, and the data transparency required to verify material composition through the chain is absent for the overwhelming majority of shoes currently entering the non-rewearable stream.

Proof-of-concept recycling exists. The THE 8 IMPACT study has demonstrated that the material value is real, that recycled rubber and EVA can meet brand performance specifications, and that the process can achieve price parity through volume and margin control. What does not exist is the preprocessing layer, the offtake architecture, the regulatory framework, and the data infrastructure that would allow that proof of concept to operate at the volumes the 1.2 million tonnes of footwear discarded annually across the EU demands. The sector is at the threshold of the conditions that would make scale possible, not scale itself.

The Reckoning the Industry Needs

The footwear sector has the evidence it needs. It is waiting for the industry to accept that the shoe as currently designed was built for performance and disposability in equal measure. The forthcoming Phase 2 circular design guidelines and Phase 3 pre-processing and recycling validations are where that acceptance will first be tested—or quietly avoided.