Supply Chain Action Points
Read this first — the conclusion, and the moves to make:
- By 31 July, pull every customs entry from the past 18 months and record the authority cited for the duty on each one; a blank authority field counts as unresolved and goes to the broker in writing.
- Within five working days, obtain the execution text of the Section 301 action and the Section 122 termination notice, and log which trigger each uses, entry, arrival or export.
- Re-date 100 percent of open orders against that trigger by 31 July, so no shipment crosses the 24 July seam on an assumption the text does not support.
- Register the three duty cohorts as separate ERP lines with separate GL accounts, so no refund, protest or indemnity question is answered from an aggregate figure.
- Collect a written origin statement per line at the 10-digit level from every supplier by 15 August; a certificate that does not cite the origin rule is unanswered.
A July 21 Xinhua analysis says the U.S. is swapping legal bases to keep tariffs in place: the Supreme Court struck down IEEPA 'reciprocal' tariffs, so Washington used Section 122's 150-day 10% global tariff (expiring July 24) and now pivots to Section 301 'forced-labor' duties of 10%-12.5% on 60 economies - covering 99.4% of U.S. imports.
The Analysis
Strictly speaking, the headline that these tariffs expire but never leave describes a rate, not a rule. On 21 July the account from Xinhua sets out a sequence: the Supreme Court struck down the IEEPA reciprocal tariffs, Washington fell back on Section 122 for a 150-day global tariff of 10 percent that runs out on 24 July, and it now pivots to Section 301 forced-labour duties of 10 to 12.5 percent on 60 economies, covering 99.4 percent of US imports.
What matters to anyone filing entries is not the continuity of the number. It is the seam. A duty at the same rate resting on a different statute is a different legal object: a different finding behind it, a different review path, and a different answer to what happens to money already paid. The rate held. The cause of action did not.
I will not put a motive on any of this. What I can do is mark where the boundaries sit.
Begin with the predicates, because the three instruments are not interchangeable and the swap works only because each has its own trigger. IEEPA rested on an emergency determination, and that determination is what failed at the Supreme Court. Section 122 is a balance-of-payments authority whose life is written into the statute as 150 days; the clock is not a policy choice and cannot be extended by announcing that it will be. Section 301 requires a determination following an investigation. Nothing carries over except the intent, and intent is not a legal basis.
Coverage of 99.4 percent of import value is a coverage ratio, not a hit rate, and the questions that decide money sit at its edge. A duty aimed at 60 economies does not automatically reach goods entering under a preferential programme; whether a preference survives, stacks with, or is displaced by an additional duty is a question for the execution text, and it will not be uniform across programmes. Goods whose origin is not among the 60 are outside by construction, which puts the weight on the origin rule rather than the routing: a shipment consolidated in a third country does not acquire that country's origin, and a certificate saying otherwise is a document, not a fact. The announced figure is also expressed as an additional duty, so the rate at the line is that figure plus the column one rate. Two importers in the same country, both reading 12.5 percent, can face materially different totals.
Then the seam, which is where I would spend the next two days. The 122 tariff runs out on 24 July and the 301 duty takes effect at midnight on 24 July, so there is no gap day on the calendar. There is still a seam in the paperwork. Whether a shipment falls on one side depends on the trigger the text uses, and the three candidates give three answers: an entry date, an arrival date, an export date. A container loaded on 22 July, arriving on 26 July and entered on 27 July sits in a different cohort under each reading. Do not assume the entry date governs because that is the field your system already holds.
Who feels this first is not a question about size. An importer with one origin and a high duty-to-value ratio feels it at the entry line on the first filing after 24 July; one holding a valid preferential claim may feel nothing until the text is read. A US manufacturer buying imported inputs feels it through the supplier's price a contract cycle later, and an exporter selling FOB feels all of it indirectly, because the buyer's landed cost moves and the buyer reopens price. What predicts exposure is the ratio of duty to net margin, not revenue: at 10 to 12.5 percent, a 4 percent net margin has nothing to absorb.
Run a small number, and take the assumptions as mine rather than as anything in the source. Assume USD 5 million a year of imports from one of the 60 economies, spread evenly, one origin, no preferential claim. The 150-day window covers about 41 percent of the year, so roughly USD 2.06 million passed under the 10 percent Section 122 tariff, about USD 205,000 of duty. The remaining USD 2.94 million falls under the 301 duty; at the top of the band, 12.5 percent, that is about USD 368,000. The year totals roughly USD 573,000, against nothing under the pre-ruling baseline. At 10 percent for the full year the same volume would have paid about USD 500,000, so the band alone is worth about USD 73,000. Those are my assumptions and my arithmetic.
Here is the part the coverage does not reach. The swap did not create one duty; it created cohorts, and the cohorts answer differently the only question that survives after the money is paid, which is whether it can be recovered. Entries assessed under the authority the Supreme Court struck down form one cohort, with whatever refund or protest route that authority leaves open. Entries inside the 150-day window form a second, on a statute that was valid but whose clock has run. Entries after midnight on 24 July form a third, on a fresh action. None of the three answers is derivable from the rate. Most operations keep exactly one duty field in the ERP, and that field erases the only attribute that matters. If you cannot produce, entry by entry, which authority the duty was assessed under, you cannot join a refund claim, cannot support a protest, and cannot answer an indemnity question six months later. The record of why the duty was charged has become the asset.
Had nothing replaced the Section 122 tariff when its clock ran out, the 10 percent would have lapsed and a genuine zero-duty window would have opened, short but real, against which contracts could have been re-priced. The swap closed it without closing the mechanism that produced it, which is what to model now: not the rate, but the next seam date.
The expensive mistakes are not about arithmetic. Treating 99.4 percent as proof that everyone is hit identically is one. Treating the country of export as the country of origin is another, and it survives longest inside a company, because the transport documents all agree with each other and none of them is the origin rule.
Where this reading stops is worth being precise about. If the execution text keys the duty to export date rather than entry date, the in-transit cohort moves to the other side and this seam becomes far cheaper than it looks today. If it carves out goods entering under a preferential programme, a Canada- or Mexico-sourced line is a different calculation. If a court stays the 301 action, the 24 July cohort becomes a refund question of its own. Get the execution text, log every entry by the authority cited on it, and keep the cohorts apart before anyone asks you to.
↑ Back to the key points