US 12.5% Section 301 Duty Reshapes Import Flows as Second Tariff Probe Awaits Ruling

The Analysis

On 24 July a 12.5% Section 301 duty took effect on Chinese-origin goods, and the load-bearing word in that sentence is permanent. This is meant to replace rather than bridge: it succeeds the IEEPA tariffs the Supreme Court struck down in February, and it succeeds the 10% stop-gap that was running out. Coverage reaches roughly sixty trading partners. Goods on the Section 232 steel and aluminium lists are exempt.

Read the coverage line twice, because sixty partners is not a China story. Whatever the headline says, the test this instrument applies is origin, and origin is a question about the goods rather than about the importer's address or the carrier's rotation.

A second investigation, into excess capacity, is still open with no ruling. Until it produces one, any plan that assumes a single rate for a single product is assuming a text that does not yet exist.

Strictly, what changed in February was not the policy but the legal basis for it. The IEEPA tariffs were struck down, which left a gap that a 10% stop-gap filled on borrowed time. A Section 301 action rests on a different statutory footing and has survived the kind of challenge that ended the earlier measures. That is the durable part of this, and it matters more than the rate. Do not read permanent as final, though. In trade measures, permanent describes an instrument's intended lifespan, not its resistance to amendment. A statutory action can be modified, its scope can be widened by a subsequent determination, and it can be met with a fresh investigation running on its own clock. What you have is a stable legal basis with an unstable perimeter.

The applicable premise is narrower than the announcement suggests, and it has two parts: origin and entry. The duty follows the origin of the goods, which is why the sixty-partner coverage matters more than the China framing. It also follows the entry, not the invoice date, the booking date or the date the purchase order was confirmed. A container that left Asia in the third week of July and is entered in August is inside the duty; a container entered on 23 July is outside it. Front-loading was the rational response to that, and it is also finished.

The forecast fall of 4.5% in August import volumes, to about 2.22 million TEU, is the visible half of the story. The invisible half is the queue of cargo that made the date, and the shippers who missed it by four days and are now arguing over who carries the difference. Whether goods already in transit on 24 July were given a transition window is exactly the sort of detail that lives in the operative text and not in a summary. Until somebody reads that clause, the answer for that queue is unknown.

Here is the part the coverage will skate past. The Section 232 carve-out is not a safe harbour; it is a handover. A steel or aluminium article that sits on the 232 lists is outside the 12.5% and inside a different instrument, with a different rate, a different origin rule and a different procedure for changing it. So the question an importer should be asking is not whether the goods are exempt. It is which instrument governs the line, and whether a change in the bill of materials or in the sourcing country moves that line from one schedule to the other.

The second question almost never gets asked, because exemption lists are treated as whitelists: you check the box, you file the page, you stop thinking about it. A single product family can sit on both sides of that boundary at once, with one variant under 232 and its near-identical sibling under 301, differing by alloy content or by a coating. The rate difference is real, and it attaches to a specification rather than to a category.

Who feels this first is mostly a function of paperwork rather than of size. Large importers with in-house customs capacity, bonded space or foreign-trade-zone space can sequence entries and split declarations; their advantage in the first month is procedural, and it does not last. Mid-sized importers who priced a season on 10% are carrying a 2.5-point gap that has to land somewhere. Where it lands depends on the Incoterm and on a clause most contracts get wrong. Under DDP the seller carries the duty; under FOB or DAP the importer of record does.

A change-of-law clause drafted in an era of VAT adjustments is often silent on a duty that is neither a VAT change nor a named schedule, which means the gap resolves by negotiation or by whoever holds the weaker position. Exporters selling FOB feel this as order volume rather than cost, which is a slower and less visible kind of pain. Steel and aluminium producers sitting outside the 301 line feel nothing from this instrument, which is not the same as feeling nothing.

Three clocks are running and they do not agree. The 12.5% is live now, so its effect shows up in the August volume number and then in September cash flow. The contract effect shows up at the next renewal, when the change-of-law clause is either fixed or quietly rolled over. The second investigation has not started its real clock; the sequence from the record to a preliminary determination to a final one runs in months, and each stage is a separate chance for the covered perimeter to widen. Those are three dates landing on three budgets, and the middle one gets missed because nobody owns a renewal that is eight months away.

Take a declared value of 200,000 US dollars per container. That figure is my assumption for the arithmetic, not a number from the reporting. At 12.5% the duty is 25,000. The 10% bridge it replaces would have been 20,000. The delta is 5,000 per container, or 2.5 points. At forty containers a month that is 200,000 a month and 2.4 million across a year, on one product line, before the second investigation has said anything at all. The number is not the point; its size relative to a gross margin is. A 2.5-point delta is roughly the whole of a distributor's margin on the same goods, which is why the argument about who bears it gets loud so quickly.

Two errors recur. Classifying by product family rather than by tariff line is the first, and it is how a variant ends up on the wrong side of a carve-out and is discovered at entry. Treating the rate on the entry summary as settled is the second. It is settled for that entry; it is not settled for the product, because the second investigation can name the same line and stack a separate liability on top of a rate everyone has already absorbed. The first error costs a week. The second costs a year of margin that was never reserved.

So where does this reading stop working. It holds if the operative text is what the summary implies: one 12.5% ad valorem line, a carve-out defined by the 232 lists, and liability triggered at entry. It fails in three places, and each is checkable inside a week. If the text contains a transition provision for goods in transit on 24 July, the queue I described sits outside the duty and the arithmetic above overstates it. If the carve-out is drawn by tariff line rather than by chapter, some steel and aluminium goods fall inside both instruments and the handover argument collapses for them. And if the second investigation lands with an overlapping scope and its own liability date, then the rate everyone has budgeted is a floor rather than a rate. None of those three is a prediction. They are the clauses to read before anyone commits a number to a forecast.

↑ Back to the key points

— By Dr. Ingrid Voss