Recycling Cannot Repair What the Production Phase Has Already Locked In

The cotton-versus-polyester comparison rests on a category error. It assumes that two fibres competing for the same environmental brief can be ranked on a single axis, when full lifecycle scoring shows they carry different signatures rather than lesser ones. The Concordia team ran each fibre through the ReCiPe 2016 midpoint method across eighteen impact categories, and what came back showed divergence rather than victory or defeat. Polyester came out worse on some measures; cotton came out worse on others. The single-axis framing has been the wrong framing throughout.

The comparative values make the divergence concrete. Polyester T-shirts produce 9.62 kg CO₂ eq per functional unit; cotton, 6.1 kg CO₂ eq. Fossil resource scarcity follows the same pattern, at 2.74 kg oil eq for polyester against 1.49 kg oil eq for cotton, reflecting the petrochemical base of synthetic fibre. Cotton reverses the balance on the categories tied to cultivation. Water and land invert it. Its water consumption reaches 2.48 cubic metres per functional unit, roughly seventeen times polyester's 0.15, and its land use is more than three times higher. Human carcinogenic toxicity for polyester, at 37.87 kg 1,4-DCB eq against cotton's 0.83, remains sensitive to how completely pesticide inventories for cotton are modelled.

Across the eighteen midpoint categories the ReCiPe method covers, polyester carried the higher relative contribution in most, with values typically ranging between 50 and 90 per cent of the combined impact. Cotton scores worse on only two. Cotton overtook chiefly in land use and water consumption, where the agricultural stage dictates the totals. The imbalance is what makes the single-axis comparison so misleading; the fibre that scores worse on most counts is also the one against which those counts have been repeatedly narrowed to carbon alone.

Polyester's higher greenhouse gas footprint reflects the energy-intensive petrochemical processing that carries it from crude oil to fibre; cotton's lower carbon number is bought at the cost of agricultural water demand and land occupation that fall on different ecosystems and different communities. The two fibres do not compete in the same environmental register; they draw down different resources. Ranking them on carbon alone compares the wrong quantities. Any judgement that treats one as the greener option resolves the sum by declining to look at the columns that would complicate it.

The category error would matter less if the two fibres were being drawn into different production systems. They are not. Both feed the same operating logic of high-frequency manufacture and rapid replacement, which compounds each fibre's respective burden in proportion to how much of it the system moves through in a season. A fibre with a lower per-garment footprint scaled across ten short-lived collections generates more total impact than one with a heavier footprint held to a longer wear life. The relevant question is where in the lifecycle those units accumulate their damage.

Broken down by lifecycle stage, the carbon footprint reveals a distribution the fibre debate has kept out of view. For both polyester and cotton, emissions concentrate in the same two zones: the manufacturing processes that turn raw material into fabric, and the transport network that moves the finished garment across continents. The share carried by each zone differs by fibre, but the pair-versus-rest hierarchy does not. Production and transport dominate the profile; use and disposal contribute marginally. The internal balance between the two zones inverts between fibres, with polyester production heavier than its transport and cotton transport heavier than its production.

Transport alone tells much of the story. Air freight accounts for roughly 60.7 per cent of a cotton T-shirt's total emissions and 31.5 per cent of a polyester one. In cotton's case, a single logistical choice outweighs every stage of manufacture combined. For polyester, the manufacturing footprint is heavier and more concentrated: yarn production contributes 27.85 per cent and dyeing 26.63 per cent, together carrying more than half the total. Cotton's yarn production contributes 18.41 per cent, with the remaining production stages contributing modestly. Washing and drying during the use phase contribute 5 to 6 per cent for cotton and under 3 per cent for polyester.

When individual parameters were varied by ±10 and ±20 per cent, air-transport distance came through as the single most influential variable in the entire lifecycle model, outweighing every production input tested. A polyester T-shirt air-freighted from China to Australia carries roughly 120 times the emissions of one shipped by sea over the same route. What looks in the aggregate CO₂-eq value like a fixed cost per garment is in fact a function of the logistical choice made at the point of dispatch.

The way emissions were spread traced the shape of the supply chain rather than the properties of the fibre. Polyester's burden concentrated in the energy-heavy upstream operations that convert petrochemicals into knitted fabric. Cotton's burden concentrated in the long-distance freight that carries it from cultivation regions to points of retail sale on other continents. In neither case did the material itself carry the emissions; the industrial architecture around it did, and that architecture is common to both. A globalised, time-compressed system produces the same emissions geography whichever fibre is fed into it.

The structural consequence is direct. Interventions aimed at fibre substitution operate on the least sensitive variable in the model. The two largest sources of emissions in the lifecycle, air freight and production intensity, remain unaffected by whether the T-shirt in question began as crude oil or cotton lint. Any strategy that leaves those two sources intact is negotiating with the margins of the problem while its centre continues to expand.

Once the emissions geography was established, the question of what can be done followed a different logic. The Concordia team tested five scenarios against the baseline, moving the model's own variables one at a time: transport mode, waste disposal pathway, and garment lifespan. The exercise produced a clear ranking of interventions, and that ranking inverts several of the priorities that industry-facing sustainability communications from major fast-fashion retailers have emphasised for years.

Two interventions delivered by orders of magnitude what the others delivered at the margins. Switching air freight to sea freight cut polyester T-shirt emissions by 30.42 per cent and cotton T-shirt emissions by 62.57 per cent, with cotton benefiting more because its transport share is larger to begin with. Doubling garment lifespan cut lifecycle emissions by roughly half across both fibres, at 47.75 per cent for polyester and 52.03 per cent for cotton; halving lifespan reversed the calculation, raising polyester's footprint by 95.51 per cent and cotton's by 51.16 per cent. The lifespan variable, in other words, is symmetrical: what extension gives back, compression takes away, and takes it away faster.

The disposal scenarios collapsed by comparison. Shifting polyester from landfill to incineration or to recycling changed emissions by +5.82 per cent and +0.69 per cent respectively, a variation small enough to fall within the model's uncertainty range. Cotton responded more, with 13 to 15 per cent reductions across the two alternatives, because its material behaviour in landfill differs. Even those percentages represent a modest recovery against a lifecycle already heavily loaded upstream. The recycling calculation itself was a simplified single-cycle pathway that did not capture multi-cycle recovery, quality degradation, or virgin-material substitution effects, so its outputs should be read as indicative rather than definitive.

What holds the ranking in place is structural. Emissions embedded in the production phase cannot be recovered at the end of life, which is why recycling, the intervention most visible in retailer sustainability disclosures, was the weakest option identified. Transport optimisation operates on a variable that carries more than half the cotton T-shirt's total burden and nearly a third of polyester's. Lifespan extension operates on the multiplier that determines how many times the whole cycle has to run. Both interventions engage the stages where the emissions actually accumulate, which is why their leverage dwarfs anything available at disposal.

The constraint identified in the five scenarios operates commercially rather than technically. Air freight exists because the model requires rapid turnover; short garment life exists because the model depends on frequent replacement. The two interventions that carry the greatest environmental return are precisely the two that the business logic of fast fashion is structurally designed to resist. Slowing the freight and lengthening the wear cuts against the mechanism by which the industry generates its margins. What comes through as decisive is a set of decisions the sector has consistently treated as commercial trade-offs rather than environmental ones.

The Commercial Terms of Reduction

Cleaner materials and expanded recycling operate at the margins of a system whose emissions are governed by production frequency, transport mode and wear life. Meaningful reduction requires the interventions identified as decisive: slower supply chains, fewer collections, garments built to outlast a season. The reframing changes what the industry is being asked to defend. Whether a business logic organised around speed can accept the terms remains the open question. Speed and durability rarely cohabit.