Temporary U.S. Section 122 tariffs set to lapse July 24, fueling import front-loading

The Analysis

The temporary 10% Section 122 tariff on global imports lapses at midnight on 24 July. It was imposed after the Supreme Court struck down the earlier reciprocal tariffs, and it is now being replaced by something that has not been published. Importers are pulling cargo forward ahead of that deadline and ahead of a possible new forced-labour Section 301 round expected in early August, and US ports are posting record volumes as a result.

Strictly speaking, this is two events and the trade press is treating it as one story about rates. Section 122 is a temporary authority with a clock written into it, which is the reason it has an expiry date at all. A Section 301 round is a different statute, a different process, a different scope and its own effective date. Between the two sits a gap in the legal basis, and the money is in the gap rather than in either rate.

The first question is not whether the duty goes away. It is which date fixes which rate for which entry, because those three things are not the same date and only one of them is in the headline.

A lapse is not a repeal and it is not a refund. The duty attaches when the entry is made. Entries lodged before midnight on 24 July are at 10%; entries lodged after are at whatever applies on that date, which may be zero, may be a new instrument, may be a transitional measure with its own conditions. The operative variable is therefore the entry date, not the arrival date and not the sailing date, and the gap between those three is where importers lose money they had already budgeted.

The boundary is worth drawing before anyone reorganises a supply chain. Goods entered before 24 July are not reached by the lapse, and the duty already paid on them is not restored by it. Goods that were never subject to the Section 122 surcharge are unaffected in either direction. Goods whose value is mostly services or intangibles are outside a tariff on merchandise, whatever the invoice calls them. Goods moving under a regime where duty is suspended or deferred present a different question entirely, and the date that matters for them is when the suspension ends, not when Section 122 ends. An importer with no US exposure is a spectator, unless a supplier redirects volume into its market, in which case it becomes a competitor's problem rather than a customs one.

The causal chain is short and it runs through capacity. A known lapse date pulls volume forward. Pulled-forward volume compresses into the last ten days of July. Terminal, drayage and broker capacity do not expand for the occasion. Those who cannot get a slot do not get the old rate, which is the part of front-loading that is never in the plan and always in the outcome. Then an August instrument lands with its own scope, and inventory bought at 10% meets a rate that may be zero or may be higher than what was paid.

The distinction the coverage has flattened is that front-loading is not a saving, it is a bet with two live outcomes. If the duty lapses to zero and nothing succeeds it, the importer paid 10% on goods that would have been duty-free a fortnight later, which is a loss and not a deferral. If a successor arrives in August at a comparable rate, the front-load was correct and the only cost was carrying. Both are plausible on the facts available, so the real decision is not whether to pull forward but what it costs to be wrong in each direction. Most mid-size importers are answering that with a feeling, and a feeling is not a hedge.

Layered by capacity rather than by size, three positions emerge. Large importers with bonded or zone-type arrangements have a third option, holding goods without entry, which converts a tariff decision into a storage decision. Mid-size importers without that option face a binary and usually choose the more expensive branch, because the cheaper one requires paperwork they do not have. Small importers face a throughput problem rather than a price problem, since the last ten days of a tariff window are exactly when everyone else is also trying to lodge. Exporters on the other side of the border feel the same event as a pull-forward in orders followed by a hole in August, and the hole is the harder part to explain to a factory.

Timing has three separate cliffs and only one is published. The legal cliff is midnight on 24 July. The commercial cliff is earlier, at the point where a vessel's arrival can no longer be guaranteed before the deadline, and that is a schedule question that a broker answers and a lawyer cannot. The accounting cliff is later, at liquidation, when the rate is finally applied to the entry and the cash leaves. If a successor is published in early August with its own effective date, a fourth cliff appears, and the interval between the third and the fourth is the only window in which an importer knows both prices and can still choose.

The point the coverage misses is a measurement error with consequences. Record port volumes are being read as evidence of record entries at the old rate, and they are not the same series. A container that arrives on 23 July and is entered on 25 July pays the new rate while appearing in July throughput statistics. Port volume measures cargo moving; entry volume measures duty attaching. Anyone using the first to forecast the second is forecasting a number that will not reconcile. The second uncovered point is structural: a temporary authority expiring is a statutory event, so there is no discretion to extend it quietly. Any successor has to be a new instrument with a new justification, and a forced-labour instrument is justified differently from a balance-of-payments one, which means its scope will be narrower and its targeting sharper. Importers modelling the successor as ten per cent again are modelling the wrong variable. The rate may well match; the coverage almost certainly will not.

A small account, and the assumptions are mine. Take a six million dollar annual import book, evenly spread, all of it at 10% under Section 122. One month of goods is 500,000 dollars and one month of duty is 50,000. Pulling two months forward means a million dollars of goods entered early and 100,000 dollars of duty paid about six weeks sooner; at an 8% cost of capital that carry is a little over 900 dollars. Against that, if the duty lapses to zero, the same decision costs 100,000 dollars of duty on goods that would have been free. The asymmetry is the whole argument: the saving is trivial, the loss is not. Front-loading pays only if the probability of a comparable successor is high, and that probability is a judgement rather than a calculation, but the arithmetic around it is neither.

Run the counterfactual. Without a sunset there would be no spike, no record volumes and no gap, and also no reason for anyone to read a successor's scope clause until it bit them. The sunset is doing the work of a stress test, and most supply chains are failing it in the same place, which is the week before the entry rather than the month before the sailing.

The trap is treating 24 July as a shipping deadline. It is an entry deadline, and entry capacity in the final week of a tariff window is the least elastic thing in the chain. The second trap is the accounting treatment: duty paid is cash out, but whether it becomes cost of goods or a recoverable depends on facts that will not be known until the successor is published.

All of this holds if the entry-date rule is applied as written and no retroactive relief is granted. If a court orders refunds on the lapsed measure, or if a successor carries a retroactive effective date, the front-loading calculation changes sign and the importers who moved fastest are the ones with the largest claim. Read the effective-date clause before the rate, and read the scope clause before either.

↑ Back to the key points

— By Dr. Ingrid Voss