Supply Chain Action Points
Read this first — the conclusion, and the moves to make:
- Obtain the final determination and its annexes, and open a line-level register recording, for every affected tariff line, the economy column, the effective date and the exclusion status, by 31 July 2026.
- Confirm country of origin rather than country of export for every line in the register, and put the factual basis for any substantial-transformation claim in writing, by 7 August 2026.
- Mark the lines that sit inside a Section 232 measure and therefore outside the stacking rule, and state their share of annual entered value as a percentage, by 14 August 2026.
- Check each exclusion line against your own entries and establish whether any request or extension channel exists for lines not covered, recording the answer and its date, by 21 August 2026.
- Rewrite every contract clause that prices off the Section 122 duty so that it names the operative instrument, and have the revised wording signed before the next quarterly price review, by 30 September 2026.
The US Trade Representative's final Section 301 action took effect on 24 July 2026, imposing additional duties of 10% to 12.5% on 60 economies that together account for 99% of US goods imports, replacing the 10% global Section 122 tariff that expired the same day after its 150-day statutory limit. Thirty-eight economies including China, Vietnam, Brazil, Australia and Saudi Arabia face 12.5% on top of MFN rates, while 17 including Canada, Mexico, India, Indonesia, Malaysia and the UK face 10%; Japan, South Korea and Switzerland receive a net 12.5%. The duties stack on MFN rates but not on Section 232 measures, and USTR added several hundred new exclusion lines.
The Analysis
The final Section 301 determination took effect on 24 July 2026, the same day the 10% Section 122 duty expired at the end of its 150-day statutory limit. The replacement covers 60 economies that together account for 99% of US goods imports. Thirty-eight of them, China, Vietnam, Brazil, Australia and Saudi Arabia among them, carry 12.5% on top of MFN rates. Seventeen, including Canada, Mexico, India, Indonesia, Malaysia and the UK, carry 10%. Japan, South Korea and Switzerland are recorded at a net 12.5%.
Strictly speaking, the rate is not the news. The substitution of legal authority is. Section 122 ran on a statutory clock and had to lapse on a fixed day. Section 301 rests on a determination and does not expire on a calendar. A duty that had to end has been replaced by one that does not, and importers now operate inside that difference.
The question worth answering this week is narrower than whether tariffs rose. For a meaningful share of entries the rate did not move at all. What moved is which column of the schedule your goods sit in, and what evidence you would have if someone asked you to defend that placement.
Origin is the gate that decides everything downstream, and origin is not the country of export. The schedule is written against economies, so a shipment consolidated in a third country still takes the rate of the economy where the goods were wholly obtained or substantially transformed. Where a supplier has moved finishing operations, whether that move amounts to substantial transformation is a legal question with an evidentiary answer, and it is the most common place where a company's own reading of its exposure parts company with the customs reading of the same facts.
The interaction rule with Section 232 is where the arithmetic stops being intuitive. Additional duties stack on MFN rates but not on Section 232 measures. Read plainly, a line inside a 232 measure carries no 301 increment at all. For a book weighted toward 232 products, replacing a flat 10% duty with a duty that does not reach those lines is not an increase. It is a reduction. That sentence will look wrong to anyone who took the headline at face value, and it follows from the interaction rule exactly as stated.
Exclusions change the rate for the excluded line only, and several hundred new lines were added. This is where aggregate figures and line-level figures separate. It is also where time matters, because a line excluded today can be narrowed later, and a line not excluded today can be opened to requests later. A register built once and never revisited will be wrong inside a quarter.
Evidence is a gate of its own. A rate that turns on origin, on product scope and on exclusion status requires proof of each, and the standard of proof is not set by the schedule. It is set by whatever verification follows, which usually arrives as a request for records months after the entry was liquidated.
A single-origin importer in the 38-economy column, with no 232 products and no exclusion coverage, meets the full 12.5% at the next entry and has nothing to interpret. A multi-origin importer spread across Asian and North American sources sees a blended movement and may not notice for a quarter. An importer whose book is mostly 232-covered may see a decline and spend a month convinced the invoice is wrong. Size cuts a second way, because an origin file and an exclusion register are fixed costs: a company entering twenty lines a year builds neither and absorbs the headline rate, while a larger one absorbs the schedule.
The seventeen economies at 10% are the group most often misread. Their number is the number the expiring Section 122 duty carried. Nothing about the cash cost changed on 24 July. What changed is the authority under which the money is collected, and that is not bookkeeping, because the authority determines the review route, the basis on which a refund could ever be claimed, and whether an exclusion is available at all. An importer in that group who tells the board tariffs went up has described the aggregate correctly and their own position incorrectly.
Japan, South Korea and Switzerland are recorded at a net 12.5%, and the word net is doing real work. A net figure implies something has been netted out. Whether that something is a preferential rate, an existing measure or an offsetting adjustment is exactly what the execution text has to say, and until you have read the annex for your own lines the only defensible position is that you do not know which construction applies. That sounds like hedging and is not. Twelve and a half points on top of MFN and a net twelve and a half are the difference between a moderate increase and a large one.
What the coverage statistic does not tell you is where the money lands. Ninety-nine per cent is the share of US goods imports accounted for by the 60 economies, which is a statement about the reach of the instrument, not about incidence on any one importer. Incidence is set by the gates above. A company can sit entirely inside the 60 and still find a third of its value untouched, and a company in the 10% group can find its legal position changed more than its cash position. The part the coverage missed is that this instrument's cost is now mostly administrative. An origin file, an exclusion register and a 232 mapping are the price of being able to state a rate at all, and that price falls hardest on the importer with the fewest entries and no trade counsel.
Run the arithmetic on a book that is only partly exposed. Assume an annual entered value of USD 40 million, all of it from one economy in the 12.5% column, an average MFN rate of 3%, and 30% of that value inside a Section 232 measure. Those three figures are my assumptions for illustration and are not in the source. Under the expired 122 duty, 10% of 40 million is 4.0 million. Under the 301 action, the 232 portion falls outside the stacking rule, so the increment applies to 28 million at 12.5%, which is 3.5 million, and the 12 million of 232 goods carry no 301 duty. The comparison is 4.0 million against 3.5 million: the replacement is cheaper by half a million, and the direction of the change flipped because of a carve-out rather than a rate. Set the 232 share at zero and the same book pays 5.0 million, an increase of one million. Both results come from one schedule, and neither is an error.
If the Section 122 authority had been extendable, the trade would be paying a flat 10% with no origin analysis, no exclusion list and no 232 interaction. That did not happen for a statutory reason rather than a political one: a 150-day limit is a limit. Everything now consuming compliance time exists because the replacement instrument is differentiated where the old one was flat, and differentiation is what creates the questions a flat duty never asked.
So where does the obvious conclusion fail? It holds only for a single-origin book in the 38-economy column, with no 232 exposure and no exclusion coverage, and that is a narrower set of importers than the headline implies. It fails for the 17 economies, where the rate is unchanged and the exposure is procedural. It fails in the other direction if a supplier's substantial-transformation claim does not survive scrutiny, because then the rate is higher than modelled rather than lower. It fails if an exclusion narrows, and it fails if a 232 measure is extended to cover a product group that was outside it. Read the schedule as a rate card and all four failures look like surprises. Read it as a determination with annexes, an interaction rule and an exclusion process, each on its own date, and they look like the ordinary movement of an instrument that is still being administered.
↑ Back to the key points