Country of origin
One word carries two questions here, and they are answered by different authorities for different purposes. A customs determination is made under customs law about duty and trade measures. A line on a product page is a statement to the person reading it, and the first answer does not settle the second.
On this page
Two questions sharing one word
Customs rules of origin and origin statements made to shoppers are different tests, applied by different authorities, for different purposes. Merging them is the single error this page exists to prevent, and it is easy to make because both produce a country name and both get typed into the same box.
A customs determination answers a customs question. It exists so that duty, quota and trade measures can be applied to a consignment, and it is made by a customs authority under customs law. For garments there is one rule inside that body of law worth knowing, because it settles how far up a supply chain anybody has to look, and the section after next states it. Making the declaration is still your customs agent's work, and the answer they give you is an answer to their question.
A line on a product page is something else. It is a statement made to the person reading it, and what matters about it is whether that person is being told something accurate. A customs determination is not evidence that they are, because it was made under a different test for a different reason. How the estate treats a question it has not settled at the source is set out on the evidence page.
Cut in one country, sewn in another
Take an invented jacket with an ordinary supply chain. The yarn is spun in one country. The fabric is woven in a second. The panels are cut in a third and shipped to a fourth, where the garment is assembled and finished. The trims come from a fifth, and the whole thing is packed in the fourth again before it ships.
Nothing about that is unusual and nothing about it is a problem until somebody needs one country name. Three parties then need one, and they need it for three different reasons.
| The question | Who answers it | What the answer is for |
|---|---|---|
| What origin goes on the customs paperwork | A customs authority, on a declaration made by the importer or its agent. | Applying duty, quota and trade measures to a consignment. It is a determination about goods crossing a border. |
| What a shopper is told | The brand, in the words it publishes. | Telling a customer something true about the product. It is a representation, and the brand carries it. |
| What goes in the passport field | Nobody yet, in law. No adopted act for textiles defines this field or its test. | An attribute in a record. Until an act defines it, what fills it is a decision the business makes and should be able to explain. |
The three answers can coincide. They frequently do, and when they do nobody notices there were three questions. The failure mode is quiet: a value produced for the first row is copied into the second, arrives on a product page under a brand name, and is now a claim made to a customer that nobody has checked against the customer's question.
A determination made for duty is an answer about a consignment. A line on a product page is an answer about a product.
The customs rule for garments: origin is conferred at making up
Under the non-preferential rules of origin, a garment takes the origin of the country in which it was completely made up. Making up means everything that follows cutting, knitting or crocheting to shape, so spinning, weaving and dyeing are not origin-conferring operations for these goods.
Applied to the invented jacket above, that gives one answer and removes the argument. The fourth country, where the panels were assembled and the garment was finished, is where the making up was completed. The spinning in the first country and the weaving in the second do not enter the test at all, and a mill certificate from either of them does not make the customs answer better supported than it already was.
One boundary travels with that rule and it is the reason this page exists. It is a customs origin statement, made under customs law for duty and trade measures, and it never answers an origin claim made to a shopper. A garment can take a customs origin from the country where it was made up while the question of whether a sentence written for a customer is accurate remains completely open, because that question is asked under different rules by a different authority. Compliance with the customs test is not a defence to the consumer one, and the customs test is not evidence about it either.
We read the guidance for that one rule. We did not read it for the harder cases it also covers, so a chain that splits the making up itself across two countries is a question for a customs authority rather than one this page answers.
The practical consequence is a subtraction. Evidence about this field has to reach the making-up site and the records held there, and it does not have to reach the loom, so a sourcing programme buying visibility two steps further upstream is buying it for some other reason. Which attributes need that depth, and which do not, is set out in how far up the chain you actually need to go.
A judgment on labelling compliance, and why nothing here rests on it
There is a judgment on the relationship between the sectoral labelling regime for textiles and the general regime on unfair commercial practices. It is the authority a reader is most likely to be shown when they ask whether doing everything the label rules require settles what may be said about a product elsewhere.
This estate has not obtained its text from the court's own database. The route we tried was blocked and no second route was completed before this build closed, so the judgment is named here and relied on in neither direction. Nothing on this page turns on it.
Two accounts of it circulate and they point opposite ways. In one, the two regimes run side by side: the general instrument does not preclude penalising conduct that both infringes the sectoral information article and is a misleading practice, while a practice complying with all the requirements of that article may not, in principle, be prohibited under the general instrument. The other account inverts that into the flat proposition that labelling compliance is no defence at all.
We have read neither the judgment nor a reproduction of it we would trust, so this page adopts neither account. Saying so is more useful than staying silent, because the inverted version is the expensive one. A business acting on it either over-corrects at real cost or advises itself into a position the text may not support. If somebody cites the judgment to you, ask for the paragraph number and read it before anything is decided on the strength of it.
What the passport framework does at the border
The framework we did read is careful about this in a way that is worth borrowing. It sets up a registry that stores, for goods entering free circulation, the identifier and the commodity code. It then has customs verify those against the registry when the interconnection is running.
And it says, in terms, that releasing the goods is not to be deemed proof of compliance. The check is a check. It is not a finding about the product, and the framework refuses to let it be read as one. A great deal of confusion in this field would go away if that sentence were applied to origin as well.
One further distinction is worth keeping, because the two values usually live side by side. A commodity code is a classification. It answers what a thing is, by placing it in a nomenclature, and its last digits in textiles turn on what the garment is and what it is made of. It does not answer where the thing came from. Two garments with the same code can be attributed to different countries and two garments from the same factory can carry different codes.
We do not look this up and we do not fill it in
This is a standing position rather than a limitation, and it is worth stating with its reason attached. Country of origin is never web searched, never inferred from anything else in the record and never auto populated. Where the merchant has not told us, the field stays empty with a reason on it. The making up rule narrows the question usefully to one point in the chain. It does not answer it: the rule tells you which fact to go and get rather than what the fact is.
Four reasons, and the first one carries the rest.
- It is not a property of the object. Composition can be tested in a laboratory and a weight can be put on scales. Origin cannot be read off the garment at all, so a system that produces a value is not reading, it is inventing a fact about somebody's supply chain.
- It moves without anything visible moving with it. The same style, in the same colour, with the same barcode, can be made in a different country in the next production run. A value that was right last season is then silently wrong, and nothing on the product says so.
- A search returns adjacent facts that look like the answer. Where a brand is registered, where its studio is, where a similar product is made by somebody else: each of those is retrievable, none of them is the origin of this garment, and all of them read convincingly in a field.
- The answer sits with whoever placed the order. It lives in purchase records rather than in product data, and it is the one attribute on this list where a guess means asserting something about a third party's factory.
The one thing an empty field is not
An empty origin field is not a defect in the catalogue. It is the correct output when nobody has stated the value, and it is more useful than a plausible country, because it can be routed to the person who can settle it. A filled field that nobody checked cannot be, and it looks finished to everybody who sees it afterwards.
Where the value actually lives in a shop
One platform is worth naming, because a good deal of this estate's work runs through it. In Shopify, country of origin does not sit where merchants expect. It is held on the inventory item rather than on the product, alongside the commodity code, which is a sensible place for it if you think of both as customs fields and a confusing one if you think of origin as something a shopper reads.
Three consequences follow and all of them are practical. The value is set per stock item, so two variants of one style can legitimately differ, which is the correct behaviour and rarely the behaviour anybody has configured. It is not edited on the screen where the rest of the product copy is edited, so it goes stale quietly. And it sits outside the product export most merchants use to audit themselves, so a business looking at its own spreadsheet cannot see whether the field is populated at all.
The practical order of work is short. Find out whether the value is set, on the object it is actually set on. Record where it came from, which is a purchase order rather than a guess. Decide, once, whether what you publish to a shopper is the same string as the one held for customs, and write down why. Then treat a change of factory as a change to the record, because that is the moment the field goes wrong and there is no other signal that it has.
Which fields a shop can already answer, and where each of them is held, is covered in the practice family, which starts at the platform guide.
Which channels ask for origin, and what that tells you
Commercially, origin behaves like a customs field rather than a sustainability field, and the plainest evidence for that is where it gets asked for. Across four sales channels we looked at, an origin attribute turns up in four different states.
| The channel | State of an origin attribute | What the documentation we read says |
|---|---|---|
| A large retailer's supplier data specification | Required | The reason given is customs reporting. It is asked for as a trade field and not as an environmental one. |
| A second retailer of the same kind | Optional | The field is offered and a supplier may leave it empty without the record being rejected. |
| A third channel's product schema | Present, requiredness not stated | The attribute exists in the schema we read. That documentation does not say whether it has to be filled, so this page does not say it has to be. |
| One major shopping surface | Not found by us | We did not find an origin attribute in the product data it accepts. That is an absence found by us in what we read rather than a statement that no such attribute exists. |
Read down the last column and the pattern is consistent. Where origin is required, the reason given is customs. Where it is not required, it is offered and frequently ignored. No channel in those four asks for it as an environmental attribute and none of them asks for a passport at all, so nothing in this table should be read as a channel requiring passport data.
That has a practical consequence for anybody planning this work. The pressure to fill this field arrives from the part of the business that ships goods rather than the part that writes product copy, and it arrives as a customs question with a customs answer attached. Copying that answer onto a product page is the failure this page opened with, and it happens most easily when the value turns up already populated in a channel feed and nobody remembers who put it there.
A field's presence in a channel's schema is also not the same thing as an attribute a passport is expected to carry. What each attribute is, and how origin sits among them, is on the attributes we track. Where a shop's own fields stop lining up with what a regulated format asks for, under names that look identical in both systems, is a separate problem and it has its own page in the practice family.
You might want to read next
Sources
-
Arts. 13, 15CELEX 02024R1781-20240628In forceRelevant provisions reviewed
Read at the consolidated text for the registry and the customs articles. Two things on this page rest on it: that the registry stores a commodity code for goods entering free circulation, and that release for free circulation is expressly not to be deemed proof of compliance. Nothing in the articles we read sets a test for origin.
-
In forceRelevant provisions reviewed
The classification the commodity code comes from. It is here to hold one distinction apart: a commodity code says what a thing is, in a classification, and it does not say where the thing came from. The two are routinely stored in the same place and read as one field.
-
Art. 5(1), Annex ICELEX 02011R1007-20180215In forceRelevant provisions reviewed
The instrument that governs what a textile label must say about fibres. It is named here for what it does not do: nothing we read in it resolves the country a garment is attributed to, which is why origin does not travel with the composition.
-
Platform documentationRelevant provisions reviewed
The platform documentation for where product data is held. One claim on this page rests on it, that a country of origin and a commodity code sit on the inventory item rather than on the product, which is why the two values behave differently from everything a merchant edits on a product page.
-
Commission guidance on non-preferential rules of origin for textiles and garmentsIn forceRelevant provisions reviewed
Commission guidance on the non-preferential rules, read for one rule and used here for that rule alone. For garments, origin is conferred by complete making up, meaning everything following cutting or knitting or crocheting to shape, and spinning, weaving and dyeing do not confer it. It is a customs origin statement. It never answers an origin claim made to a shopper, and that boundary is stated in the same section as the rule rather than left to the reader. This build did not establish the guidance's own address, so the source is named without a link, and the harder cases the guidance also covers were not read.
-
Marketplace and platform seller policy documentationPlatform documentationOfficial source confirmed, detailed review pending
Seller and supplier policy documentation for a set of sales channels. It carries one thing on this page: the state of an origin attribute across the four channels we looked at. The read behind those rows is uneven, part of it reached through integrator documentation rather than a platform's own policy page, and none of it was re-verified for this build. Nothing in it supports any statement that a channel requires passport data, and this page makes none.
Help someone else make sense of product passports.