Supply Chain Action Points
Read this first — the conclusion, and the moves to make:
- By 21 July 2026, hold the signed instrument and its exemption annex on file with instrument number, publication reference, effective date and product scope recorded entry by entry, and zero lines in the register missing an effective date; owner is legal or trade compliance, not the broker.
- Within five working days of publication, map the previous twelve months of entry data, classification, origin and entered value, against the annex, and report non-exempt value as a share of total value for each 12.5% origin, covering 100% of entered value and not only the top twenty lines.
- By 31 July, issue a one-page applicability memo that puts every premise in writing: which origin rule is assumed, which reading of the annex is assumed and which date governs the entry; any premise the text does not settle must be listed as open, to a target of zero unstated assumptions.
- Run a 24 July seam check on every shipment with an entry date on or after that day but a price fixed before it, recording which regime was priced against which was entered, and target zero discrepancies without a documented price-adjustment clause.
- By 14 August, have outside counsel answer in writing, one question per product family, whether a pending classification or exclusion issue would change the annex result, at one written answer per family rather than per shipment.
Dorsey & Whitney analysis says the Trump administration will finalize Section 301 tariffs of 10% or 12.5% covering more than 60 economies, effectively replacing the 10% global Section 122 surcharge that expires 24 July. Rates of 12.5% apply to China, Vietnam, Thailand and others; 10% to the EU, UK, Taiwan, Mexico, Canada and more, with extensive product exemptions.
The Analysis
On 24 July the 10% global surcharge under Section 122 expires, and on the same day a Section 301 action on forced labor takes its place: duties of 10% or 12.5% across more than 60 economies, with 12.5% for China, Vietnam, Thailand and others, 10% for the European Union, the United Kingdom, Taiwan, Mexico, Canada and others, and what the analysis calls extensive product exemptions. That is the outline. What follows is about the parts of it the outline does not settle.
Strictly speaking, this is not a rate change. A global surcharge and a country-specific action are different instruments, and they fail in different places. The surcharge is a flat percentage with an expiry date built into it and no product list attached. The action is written against named economies, is assessed line by entry line, and carries an annex of goods outside its scope. When one replaces the other, the headline number may move by two and a half points, but the operative change is that your duty stops being a function of where the goods were made and starts being a function of what your entry says.
So the question worth asking first is not how much the rate went up. It is which of your lines the annex reaches, which of your entries the effective date catches, and whether either answer changes if the annex is amended between publication and entry. Those questions have different answers for different importers, and for some of them the new action costs less than the old surcharge. The arithmetic is below, with the assumptions stated.
Take the instrument apart, because there are at least three documents inside the announcement. There is a 10% surcharge expiring on 24 July. There is a Section 301 action taking effect the same day, with two rates and a list of more than 60 economies. There is an exemption schedule described as extensive and not enumerated. Three documents do not necessarily publish together or carry the same effective date, and until you hold the signed text and the annex you are working from a description of a document. A description is not a basis for a declaration. That sounds like procedural fussiness; it decides how much of the rest of this analysis survives contact with the border.
Origin is the first of these premises, and it is the one most often assumed rather than read. The rates are written against economies, which invites the reading that a shipment is taxed according to where it was made. The annex will state which origin rule it applies, and origin rules are where actions of this kind are contested rather than administered. A product assembled in a 10% economy from components made in a 12.5% economy can fall on either side of the line depending on the rule the text adopts, and the same question returns at the level of the individual line rather than the shipment as a whole. Do not infer the rule from the rate table. The rate table gives you the consequence, not the test.
The exemption annex is the other premise, and it is the part that coverage compresses into a phrase. An exemption is not a discount. It is a limitation of scope, which means the goods it names are not subject to the action at all, and goods outside it are subject to the full rate with no phase-in and no relief at the margin. Two importers buying the same product family from the same 12.5% economy can therefore face different effective rates, because one holds a line the annex names and the other does not. That is what the headline number cannot carry. The rate multiplies a base that is shrinking, and the size of the base is a fact about your invoice lines, not about your sourcing map.
Which brings me to the number that does not appear in the coverage I have read. A 12.5% rate on a base reduced by exemptions is not comparable with a 10% rate on an unreduced base, and the comparison can run in either direction. It runs the wrong way for the importer who reads a higher headline rate as a higher bill, reprices the customer, and discovers afterwards that most of its lines were outside the action. I am not arguing that the new action is cheaper. I am saying the sign of the change is an output of your line mix, and nobody outside your entry data can compute it for you. That is why the first deliverable is a mapping and not a negotiation.
Where this lands is uneven, and uneven for reasons that have nothing to do with size. An importer with a narrow range concentrated in exempt lines will feel almost nothing, may see a reduction, and will spend a quarter being told by its own finance team that the tariff story was overdone. An importer with a broad consumer range, no in-house trade counsel and a broker who reports at month end will find out four to six weeks later, on a duty statement it cannot reconcile to a purchase order. Exporters selling into the United States on delivered terms sit on the other side of the same fact: in that structure they are the party clearing the goods, so the annex question is theirs and not the buyer's, and no buyer volunteers to share it. Exporters selling free on board carry none of the duty and should expect the conversation to arrive as a price request rather than a legal question.
The timing does not follow the sailing schedule. Duty is assessed when the entry is made, so the seam on 24 July runs through entries rather than bookings, and a container that left Asia in the third week of July can be entered on either side depending on when the entry is filed and what it covers. The exposed contracts are the ones priced in June on a 10% assumption and delivered in August: the price was fixed under one instrument, the entry is assessed under another, and whether that gap belongs to you depends on a clause most purchase orders do not contain. If I had a legal budget for this, it would go there rather than into a study of the rate table.
Take a book of USD 4 million of annual entered value from a 12.5% economy. Treat everything in this paragraph as my assumption rather than as anything in the source: that the 10% surcharge was charged on the whole value, that the Section 301 action replaces it instead of stacking on top of it, that 60% of the value sits on lines the annex does not exempt, and that no other duty changes. Under the old surcharge the bill is USD 400,000. Under the new action it is 12.5% of USD 2.4 million, which is USD 300,000. The headline rate rose by a quarter and the bill fell by a quarter. Push non-exempt value to 90% of the book and the same importer pays USD 450,000. The distance between those outcomes is USD 150,000, and the only variable that moved was a fact about invoice lines.
The error I expect to see is the imported assumption that an exemption is a category. It is a list, with descriptions and often with dates, and the descriptions are drawn at a specificity that will not match a commercial product name. The second error is treating an exemption granted under a different action as evidence about this one. Neither is visible from a rate table, and both are expensive once the entry has been liquidated. Everything above holds on a condition: that the action takes effect as described, on the date described, with the annex as published. If the effective date slips, or the annex narrows between publication and entry, or the two instruments overlap for a period instead of changing over on one day, the conclusion changes, and the importer who built a model on the announcement rather than on the text is the one who finds out at the border.
↑ Back to the key points