Supply Chain Action Points
Read this first — the conclusion, and the moves to make:
- By 26 June 2026, pull the entry file for your ten largest entries since the 10% rate took effect and score each one for completeness of declaration basis and supporting documents, because provability is fixed at entry.
- By 10 July 2026, confirm in writing with your broker the retention period and the exact fields held for entries declared on your behalf, and list the fields a refund claim would need that are missing.
- By 24 July 2026, separate entries where a first-sale or preference claim is genuinely supported by documents from entries where the claim would rest on reconstruction, and stop counting the second group as recoverable.
- By 14 August 2026, reconcile the dual tracking, live 10% against the proposed 301 schedule, at line level for your top 50 lines, so a change in the applicable rate does not force a fresh exercise.
- By 30 September 2026, brief finance on a range rather than a point estimate for recoverable duty, with the range driven by record completeness, so the contingency is not booked at face value.
A federal appeals court ruled that the U.S. government may continue collecting the 10% global tariffs while litigation challenging their underlying authority proceeds, extending a period of policy uncertainty for importers. The decision preserves a key revenue stream but keeps the Section 122 framework vulnerable to a future Supreme Court review, similar to the IEEPA ruling earlier in 2026. Importers and brokers are maintaining dual compliance systems, tracking both the existing 10% global rate and the proposed Section 301 replacement structure. Customs brokers report increased requests for transactional tariff engineering, including first-sale valuation and FTA preference qualification reviews.
The Analysis
A federal appeals court has ruled that the U.S. government may keep collecting the 10% global tariffs while the litigation challenging their underlying authority runs its course. The decision preserves a revenue stream and keeps the Section 122 framework alive, but it leaves that framework exposed to a future Supreme Court review, in much the same way the IEEPA ruling earlier in 2026 did. Importers and brokers are running dual compliance arrangements, tracking the live 10% rate alongside the proposed Section 301 replacement, and brokers report a rise in requests for transactional tariff engineering, first-sale valuation and free trade agreement qualification reviews.
Read the holding for what it is. This is a decision about collection pending appeal, not a decision about authority. The money is being taken now and the legal question is unanswered. Almost every practical consequence follows from that split, and almost every importer I have spoken with has collapsed the two into one.
The distinction is not academic. A ruling that permits collection while the case proceeds says nothing about whether the underlying power exists. It says the court declined to stop the machinery in the interim. That is a procedural posture, and it is the posture most likely to be misread as a merits win, because the visible result is that nothing changed at the border. Nothing changing is precisely what a transitional ruling produces. When the merits arrive the answer can go either way, and the IEEPA ruling from earlier in 2026 is the reason to take that possibility seriously rather than as a lawyer's reflex.
If the challenge ultimately succeeds, the question turns into refunds, and refunds rest on a different body of law from the one being litigated. A right to refund in principle is not the same as a provable claim in fact. What makes a claim provable is the record created at the time of entry: what was declared, on what basis, with which supporting documents, and whether the entry has been liquidated. That record is fixed at the moment of entry and does not improve afterwards. This is the exposure nobody is pricing. Litigation risk sits with the government. Evidentiary risk sits with the importer, and it accrues one entry at a time, starting now.
Dual compliance is usually described as running two systems. It is more accurate to describe it as one entry with two sets of numbers behind it. The entry itself can declare only one rate, so the second set lives outside the declaration, in the reconciliation file, and its only purpose is to remain usable later. That changes what the work is. It is not a software problem and it is not a staffing problem. It is a records problem, and records problems have a property that systems problems do not: they cannot be solved retrospectively. A file that was thin on the day of entry stays thin, no matter how good the analysis is two years later.
First-sale valuation and FTA qualification both change the base to which a rate applies rather than the rate itself, which is why they attract attention when the rate is contested. They also share a dependency. Both rest on facts that exist before entry and documents that must be assembled at entry. A first-sale claim needs the transaction chain to be genuine and documented; a preference claim needs origin evidence from the producer. Neither can be manufactured after the fact without producing exactly the kind of record that fails on examination. The requests brokers are reporting are therefore best understood as documentation projects with a hard deadline, and the deadline is the entry date, not the ruling date.
Exposure divides unevenly. A large importer with an in-house trade compliance function already holds entry-level records as a matter of routine, and for them the appellate ruling changes the size of a contingency rather than the nature of the work. A mid-sized importer that outsources declarations to a broker holds whatever the broker retained, which may be enough to file and not enough to claim. The smallest importers hold almost nothing, and they are the ones for whom 10% is a margin rather than a line item. Direction matters too. Under DDP terms the importer of record carries the legal exposure while the cost has already been passed to the buyer, so the party holding the claim and the party holding the incentive are not always the same entity.
The clock that matters is not the appellate calendar. It is the entry calendar. Every entry filed while the rate is contested becomes a data point that will either support or undermine a later claim, and the volume of those entries grows every week the litigation runs. A final ruling in eighteen months would find an importer holding eighteen months of records, some usable and some not. The proportion is decided now, not then. Waiting for the ruling before organising the file is the one strategy that guarantees the file is unusable, because the file is a by-product of the entry rather than a task you can schedule afterwards.
Here is the arithmetic, on assumptions I am setting rather than borrowing. Take an importer with USD 12 million of annual entries at the 10% rate. That is USD 1.2 million a year collected, USD 100,000 a month. If the challenge succeeds in part, the refundable amount is bounded by that 10%, but the amount actually recoverable is bounded by the quality of the entry records, and those are two different numbers. A file that supports seventy per cent of entries and not the rest turns a USD 1.2 million question into roughly USD 840,000 of claim plus USD 360,000 of argument. The figures are mine, the assumptions are mine, and the seventy per cent is illustrative rather than measured.
Where this stops holding is easy to state. If the Supreme Court declines to hear the case, or hears it and upholds the authority, the refund question disappears and the documentation work loses its litigation purpose. It does not lose its other purpose, which is that the same records support first-sale and preference claims that stand on their own footing. The preparation is robust to the legal outcome, and that is the only reason it is worth doing before the outcome is known.
↑ Back to the key points