There is no binding standard, so each brand sizes against its own fit model

The argument

The first explanation is that a clothing size is not a measurement at all but a label, and nothing obliges two companies to attach the same label to the same garment. A brand develops its range from a fit model, a single body whose measurements define the middle of the size run, and then grades upward and downward from there by increments it chooses itself. Change the fit model or change the grading increments and every size in the range moves, which is why a size that fits in one shop can be tight or loose in the next without anyone having done anything unusual. The explanation extends to who actually makes the clothes. A retailer's range is typically cut by several suppliers, often in different countries, from different cloth and to tolerances agreed separately. Woven and knitted fabric behave differently, stretch alters effective fit, and a pattern adjusted at the factory to save material comes out slightly different again. Proponents of this explanation argue that the result is variation not only between brands but within a single brand, between its ranges and even between batches of the same garment. On this account the erratic quality of sizing is an accumulation of ordinary inconsistency rather than a strategy, and that also indicates how it could be fixed: a published standard tied to stated body measurements, applied by retailers to their own grading, with those measurements printed on the label so the number means something a shopper can check. Until a standard binds, proponents argue, the number on a label records a decision taken inside one company and cannot be compared with the same number anywhere else.

Premises

Counter-arguments

The competing explanation accepts all of this and says it is incomplete. Inconsistency alone would produce scatter in both directions, with labels running small about as often as they run large, whereas the drift shoppers complain about has a direction: the body measurements attached to a given number have tended to grow. Critics of the inconsistency account argue that a one-way drift is evidence of a shared incentive rather than of independent error, because every retailer faces the same reason to flatter the number on the label. On that view the lack of a standard is what makes the drift possible, but the drift itself is a choice, and treating it as mere sloppiness misdescribes the mechanism and would leave any published standard to be quietly undercut in the same way.

Rejecting the premises

[Rejecting P1] A size label is not arbitrary in practice: retailers publish size charts in body measurements and buyers rely on them, so the label is tied to stated dimensions even without a legal standard. [Rejecting P2] Grading from a fit model explains why brands differ from one another, but not why the differences have persisted in the same direction, which is the pattern that needs explaining. [Rejecting P3] Supplier and cloth variation is a tolerance problem measured in small fractions of a size, too small to account for a garment that fits a shopper two sizes away from its label. [Rejecting C] At most it follows that inconsistency contributes to the spread of sizing, not that it is the reason sizes are erratic.