Tighter CBP Cargo Holds Push Demurrage Bills as High as $45,000 per Shipment

The Analysis

The date to hold onto is 3 June 2026, not 1 August. On 3 June an executive order directed Customs and Border Protection to intensify its scrutiny of importers with overseas headquarters, and the week to 1 August is when that direction began appearing on invoices. Importers and forwarders reported hold and storage charges of 8,000 to 12,000 dollars in some port markets, delays running from several days to several weeks, and one case at 45,000 dollars after an entire bill of lading was detained as a single unit.

The charge is not the interesting part. A physical examination is a lawful exercise of a border function, and nobody in this discussion argues otherwise. The charge is a private claim under a private contract, and it falls on a party who did nothing except hold title to the cargo at the wrong moment. Two legal systems are touching each other here, a public one that authorises the hold and a private one that bills for it, and the seam between them is where the money leaks.

Precisely stated, what the trade groups are asking the Federal Maritime Commission to do is close that seam. Their petition asks for a ruling that demurrage cannot accrue while cargo is immobilised by a government hold. The sentence looks modest. If granted it would override a contractual clock that carriers and terminals have relied on for as long as free time has existed, which is why the outcome deserves closer reading than the dollar figures.

Begin with the vocabulary, because the reported figures are aggregations and aggregations hide the argument. Demurrage is conventionally the charge for cargo that overstays its free time inside a terminal. Detention is the charge for equipment held beyond its permitted period outside the terminal. Storage is what the terminal itself bills for occupation of space. The reports describe hold and storage charges of 8,000 to 12,000 dollars, which is not one category but at least two, and the legal footing for contesting each of them is different. If an invoice presents a single line described as charges, the recipient cannot say which contract the line arises under, and a party who cannot identify the contract cannot say whether the clock should have stopped. The first thing a dispute needs is not an argument. It is a taxonomy.

The next premise to settle is the reach of the 3 June order, which is described as targeting importers with overseas headquarters. Read quickly, that sounds like a description of ownership. Read as a legal matter it points at an entity, and the entity that matters on an entry is the importer of record named on that entry. A group with an overseas parent may import through a domestic subsidiary, through an affiliate, or through a buying entity that is itself the importer of record. Whether a particular shipment falls inside the described class is therefore a question about which legal person appears where, and the answer can differ between two shipments that arrive on the same vessel for the same commercial group. I am not in a position to tell you what the order says about any specific structure. I am telling you that this is an entity question, and that it should be answered in writing rather than inferred from a group chart.

Then there is the nature of what changed. The order set a posture, and the intensified examinations are its operational expression. The volume of examinations is a matter of enforcement intensity rather than a rule with a threshold written into it, and the record in front of us shows the intensity moving while the standard stands still. The distinction matters because of what it does to remedies. A legal standard can be challenged. Intensity cannot. There is no appeal against a probability, no hearing at which a party establishes that its cargo should not have been selected, and no finding to overturn, because no finding was made. The container was pulled, and that is the whole of the decision. The only place a remedy can exist is downstream, in the question of who bears the cost of the pull, and that is precisely the question the petition puts to the Commission.

The group discussed least is the one I would spend the most time on, the importer inside the described class whose paperwork is flawless. Two different probabilities are in play and they are easy to confuse. Filing quality affects the probability of an adverse finding during an examination. It does not affect the probability of being selected for one, if selection runs off a class characteristic. A clean file inside the class is still inside the class. Practitioners tell each other that good documentation protects them, and at the level of findings it does. At the level of pulls it may not, and the storage bill is generated by the pull rather than by the finding. That gap between what documentation buys and what it is believed to buy is where a great deal of quiet money goes.

Now the contract, because the free time clock is a contractual construction and needs to be read as one. Like most standard clauses it was drafted with a particular antagonist in mind, the merchant who is slow, who cannot be reached, who has not paid. It speaks to merchant-caused delay. A government hold is not merchant-caused, and a clause written around merchant delay is typically silent on the point. Silence in a contract is not neutral. It resolves in favour of the party who drafted the clock, which is to say the clock keeps running.

That is the entire reason a petition to the Commission exists at all. If the contract already answered the question, nobody would be asking a regulator. Whether the Commission has authority to supply a term the parties did not write is a separate question from whether the result would be fair, and the material I have does not resolve it. That is an honest gap, and anyone who tells you the outcome is obvious is guessing.

Burden of proof deserves a sentence of its own, because it decides how this argument is actually run once it is run at all. The party asserting that the clock should have paused is asserting an exception to the ordinary operation of a contractual term, and exceptions are generally carried by the party that relies on them. In practice that means the file has to establish two dates, the start of the government attributed immobilisation and its end, and has to keep both of them separate from the ordinary dwell that any import operation generates on its own. An importer who reconstructs those dates eight months later from a carrier portal will discover that the portal records release dates and not reasons. Reasons live in the examination notice, in the correspondence, in the terminal own records of why a box could not be lifted. Contemporaneous documents win this kind of argument and reconstructions lose it, which is an unglamorous conclusion and the correct one.

Follow the money one step further and the parties come apart. The party liable to the terminal is the party on the terminal contract. The party who ultimately bears the figure is frequently the beneficial cargo owner, several steps down a chain of pass-through billing. Forwarders pass charges to their customers, which is what the arrangement is for. So when a charge is contested, standing and economic interest sit in different hands. The entity with the contractual relationship holds the standing and often has no motive, because it is passing the figure along. The entity with the motive has no direct contractual footing and has to act through somebody else. Anyone building a file on this should know which of the two they are, because the documents you need to assemble are not the same documents.

On exposure, start with the largest importers inside the class, the ones the order described. Their selection rate is the highest, their dwell is the longest, and their volume converts a rate into a recurring line rather than an incident. For an importer running continuous programmes, an elevated selection rate is not an emergency. It is a change in the cost of trading that has to be absorbed somewhere in the landed cost calculation. The awkward part is that it is not predictable per shipment. It is a distribution with a long tail, and a long tail is difficult to budget and easy to under-provide for, particularly when the provision has to be justified to a finance function that sees an average rather than a tail.

Consolidated shipments deserve their own paragraph, because the 45,000 dollar case is not an outlier in the statistical sense so much as a structural one. When an entire bill of lading is detained as a unit, every container on that bill shares the immobilisation, and the charge scales with the size of the bill rather than with anything wrong in any particular box. Allocation among the parties on a consolidated bill is a contractual matter between them, and the party who contracted with the terminal may end up carrying the whole figure while attempting to recover from others afterwards, if the contracts permit recovery at all. That is the mechanism by which one examination becomes a five-figure invoice for a cargo owner whose goods were never examined.

Smaller importers and those who file without an intermediary sit at the other end, and their disadvantage is not only financial. The Commission record is a public good. A ruling would help everyone in the class whether or not they participated in the proceeding. Participation, though, requires documentation, and documentation requires an administrative habit that small importers rarely have, because until now there was no reason to keep hold notices and immobilisation dates in a form anyone could use. The barrier to joining the record is not the legal argument. It is the filing cabinet.

Two illustrative assumptions make the scale concrete without pretending to know any reader book. Assume an importer moving 40 containers a month, and assume two of them are pulled in a single month. At the reported 8,000 to 12,000 dollar range per affected shipment, that month carries 16,000 to 24,000 dollars of charge which was not in the budget and cannot be recovered at the time it is paid. Assume a larger importer moving 200 containers a month, and assume its selection rate runs at the same proportion. The same arithmetic applied to five times the volume produces 80,000 to 120,000 dollars in the month, and at that size the figure stops being an operating annoyance and becomes a line a board asks about. Both numbers are illustrations built on stated assumptions rather than forecasts, and the point they carry is that the exposure scales with volume while the legal argument does not.

On timing, the first transmission is immediate and it is already over for the shipments in question. An examination pulls the box within days, the clock starts, and dwell runs from several days to several weeks. Nothing about the dispute changes that. The money is spent before any argument begins, which is why the argument has to be reconstructed retrospectively from records rather than assembled prospectively from intentions.

The second transmission runs on the contract cycle, which means quarters. Free time and government hold carve-outs are negotiated terms, and they get priced at renewal. If a carrier accepts a pause during a government hold, the cost of that concession reappears somewhere else, in the rate or in the free time allowance. If it refuses, the importer carries a risk for which no premium was ever quoted. Either way the effect arrives at the next signature rather than this month, and the practical consequence is that anyone renegotiating in the next two quarters is pricing a term whose regulatory status may change before the contract expires.

The third transmission is the slowest and the largest. A Commission ruling on whether demurrage may accrue during a government hold would function as a default term for an entire market, filling in what thousands of contracts leave blank. Default terms outlive the dispute that produced them. If the ruling goes for the petitioners, the clause becomes standard and the argument moves on to what counts as immobilisation. If it goes against them, the silence in the contracts becomes an established allocation and the next attempt has to wait for a different case. Either way the relevant horizon is measured in years, which is why the dollar figures in this week reports are the least interesting thing about the story.

Here is what the coverage has mostly skipped. The dispute is not about whether the charge is too high. It is about which of two innocent parties carries a risk that neither created. The government created the immobilisation by exercising a lawful function. The importer created nothing at all. The carrier and the terminal did nothing except apply a clause exactly as written. Everyone behaved correctly and somebody still pays 8,000 to 12,000 dollars, or 45,000, and the legal system response to that is to look for the contract that allocates it. Reading the story as a story about fees misses the point entirely. It is a story about a gap in standard contract language that a regulatory decision may or may not fill.

The second skipped point is more mundane and, for anyone who has to build a claim, more consequential. Whether a line on an invoice is demurrage, detention or terminal storage determines which contract governs, which free time applies, and which party has the right to dispute it. In a consolidated bill those categories blur, and the blurring is not accidental. A single figure described as charges is harder to challenge than three figures each tied to its own clause. Any future ruling, whatever it says, will have to be applied line by line to real invoices. An invoice that cannot support line by line application cannot benefit from the ruling, however favourable that ruling turns out to be.

The first condition that would flip the conclusion is the Commission granting the petition. If demurrage cannot accrue while cargo is immobilised by a government hold, then the 8,000 to 12,000 dollar range stops being a cost of importing and becomes, at least in part, a recoverable or avoidable item, and the 45,000 dollar case becomes the kind of figure that gets argued rather than absorbed. Note what does not change even then. Recovery still requires proof of when the immobilisation began and when it ended, and that proof lives in documents most importers are not keeping. A favourable ruling helps the documented and does nothing for the rest.

A distinction inside that first condition is easy to miss and matters to anyone holding a paid invoice. A ruling sets a rule. Whether the rule reaches charges already incurred is a separate question from whether it governs charges incurred afterwards. Regulatory decisions generally operate with some attention to reliance interests, and carriers and terminals have been billing on the existing basis for years. The realistic reading, then, is that a favourable ruling improves the position of importers going forward more reliably than it reopens the past, and that any retrospective reach will be argued case by case rather than conferred at once. Importers assuming a ruling will simply hand back last quarter storage bills are assuming a remedy the record does not promise.

The second condition is scope. If the reach of the order widens beyond importers with overseas headquarters, the selection probability spreads to a population that currently believes itself outside the conversation, and the class based reasoning that lets an importer say this is not about us stops working. Widening also changes the arithmetic of the petition, because a rule protecting a narrow class is easier for a regulator to grant than one that rewrites cost allocation across most of the import economy.

The third condition runs the other way and is the one importers least like to hear. If the immobilisation turns out to have been prolonged by something in the importer own submission, a data gap, an unanswered request, a classification question that needed resolving, then the government hold argument weakens in proportion to how much of the dwell the importer caused. The logic of the petition depends on the cargo being immobilised by the government. Dwell that the importer extended is not that. And if the examination closes inside the free period there is no charge, no dispute and no story, which is the outcome in the majority of examinations that never become invoices.

What I will not do is tell you how to file an entry or which commercial terms to buy. That is not my discipline, and the operational side of this belongs to people who do it daily. What I can say is narrower and, I think, more useful. The legal question here is not whether the fee is fair. It is whether a contractual clock drafted for a different problem should keep running during a period the counterparty did not cause and cannot end. That question now sits in front of a regulator, and the answer will arrive on a quarterly or annual schedule, long after the invoices reported this week have been paid.

↑ Back to the key points

— By Dr. Ingrid Voss