Senators Warren and Kelly Push USTR to Reinstate Section 301 Port Fees on Chinese-Built Vessels

The Analysis

Start with what is in force, not with what was asked for. As reported, the current Section 301 port fee structure places $23 per net ton on arriving vessels, effective 17 April 2026, and raises that to $28 per net ton in 2027. Those are the operative numbers and that is the operative date. The event of 18 June is a different kind of object: a letter from Senators Elizabeth Warren and Mark Kelly to the U.S. Trade Representative, Mr. Greer, urging USTR to reinstate port fees aimed at Chinese-built and Chinese-operated vessels. The letter follows the recent meeting between President Trump and President Xi Jinping. All of that is context. None of it is an instrument.

Here is the premise I want on the table before anything else. Advocacy and rules are different classes of thing, and they behave differently. A letter asks an agency to do something. An agency decides. A rule, once it issues and takes effect, binds. When you read a forceful letter as though it were a scheduled rule, you have made a category error, and category errors are expensive precisely because they feel like prudence. So the honest statement is this: the fee structure described above is in force; whether USTR reinstates, narrows, widens or leaves alone the fee the senators are asking about is not something the letter answers, and it is not something I am going to predict. Strictly speaking, the letter tells you what two senators want. It does not tell you what will be.

I am putting this in the column on clauses and risk rather than on the operations desk because the interesting question is not what the fee is. It is what the fee attaches to, and what has to be true for it to attach to a given vessel, a given charter, a given booking. That is a boundary question, and boundary questions are answered by documents rather than by headlines. So the through-line of everything below is a single sentence: this depends on what, and I am going to tell you what it depends on.

Take the category distinction first, because it is the one that costs the most when it is missed. A letter is a request. A rule is a command. Between the two lies an agency decision that has not been made. Teams that blur the line do it in two directions, and both are expensive. One team reads a strong letter as a near-certain rule and reorganises trade around it before the rule exists; that is a bet on a document that has not been issued, staked with the asset that is hardest to rebuild, which is time and goodwill inside a trading relationship. The other team reads the same letter as noise and does nothing at all, including nothing about the fee that is already in force. Neither is a position. The disciplined middle is boring and correct: know what is in force, know what is merely requested, and keep the two clearly labelled so that nobody downstream confuses them.

The applicable premise, then, is not a general trade action. It is a fee on arriving vessels, aimed, as reported, at Chinese-built and Chinese-operated ships. I want to slow down on that construction, because it carries two tests sitting side by side and they are not the same test. Built is a fact about a hull origin. Operated is a fact about who runs the ship. A vessel can satisfy one and not the other. A vessel can satisfy neither while the cargo inside it originated in China. A vessel can satisfy both while carrying goods that originated somewhere else entirely. I am not going to tell you which of these describes your trade, because that is not knowable from a summary. What I can tell you is that each limb of the description draws its own boundary, and the boundary is where the answer lives.

Then there is a layer under that description that people rarely itemise. Ownership, operation, chartering and flag are four separate facts about the same ship, and they can point in four different directions. A vessel can be owned by one entity, operated by another, time-chartered to a third and flagged in a fourth jurisdiction. Each of those is established by a different document, held by a different party, and dated differently. Whoever has to prove that a fee applies, or that it does not, will do it by matching a defined test to documented facts about a specific hull on a specific date. That is a records exercise, and records exercises are settled by registry files and official notices, not by the temperature of the news cycle.

Which brings me to the thing to verify, and I want to be careful about how I say it. The scope language, meaning which entities, which ownership or control relationships, which chartering arrangements and which registries the fee reaches, sits in the official notice. It does not sit in a press summary and it does not sit in this column. I am not going to recite scope provisions I have not read in front of me, and I would caution you against accepting them second-hand from anyone who has not. What I can give you is the shape of the question. The notice defines the test; the records establish the facts; and the match between them is the position. If the match is made from memory, you do not have a position, you have a recollection, and recollections do not survive a dispute that is eighteen months old.

Now the conflation I would circle before anything else. A port fee on a vessel is not a tariff on your goods. They are levied on different objects: a vessel call versus an imported article. They are assessed by different authority on different bases, one on net tonnage and the other on value or on a specific rate. They land in different lines of different parties books, they generate different filings, and they are recovered under different clauses. If you merge them into one mental line item called the trade action, you will build your file on the wrong premise, and a wrong premise is the hardest thing in a compliance file to unwind, because it contaminates everything that was stacked on top of it.

The two also produce different remedies, and that difference matters more than people expect. A port fee is a charge; it produces a number, and a number can be paid, disputed or allocated. A sanctions designation is a prohibition; it produces a bar, and a bar is not about money at all. A fee on a vessel call and an entry on a restricted-party list are not the same population of facts and they are not resolved the same way. I mention it because both get filed under the heading of China measures, and once two things share a filing cabinet people start treating them as one risk. They are not one risk. They do not even work through the same mechanism.

Net ton, for the record, is a measure of a vessel enclosed volume. It is not a weight and it is not a cargo capacity. That matters for sizing and it matters for allocation. The fee, as structured, tracks the ship rather than your shipment, so a larger vessel pays more per call regardless of what is stowed in it. The practical consequence is that this is a call-level cost, and whether it ever reaches your landed cost is not decided by the fee at all. It is decided by the contract between you and whoever bears the charge, by the tariff under which the cargo moves, and by the terms of the fixture. Allocation, not the rate, sets your exposure. That sentence is the most useful one I can hand you, and note what it is: a statement about your paper, not a prediction about policy.

The two dates do a different kind of work, and I want them separated. The 17 April 2026 date puts the current rate in force now. The step to $28 in 2027 is a change scheduled for a known future date. A rate that moves on a published date is a planning input for anything that renews across that date: a charter, a service contract, a policy, a booking template. That is why I work on a quarter-to-year clock rather than a week clock. The question what is the fee today is almost never the useful one. The useful question is what does my paper say the fee is, and does my paper survive the date on which it changes.

Two things about the 2027 figure, and I will not overstate either. The first is that it is described as the current structure, which means it is what is in force or announced now; it is not a warranty of what will be in force then, and I am not going to treat it as one. The second is arithmetic and therefore solid: the step from $23 to $28 is an increase of about 21.7 percent. That percentage is the number to carry forward if anything in your contracts indexes to the fee, because a percentage is what moves a contract term, and a contract written when $23 was the figure has a gap if it renews after the step. The gap is not a matter of opinion. It is a subtraction.

I would also keep the neighbouring tracks in separate columns. USTR is described as also pursuing a Section 301 tariff expansion aimed at 60 trading partners, and modifications to Section 232 metal duties. Those are their own actions, on their own authorities, with their own premises, their own scopes and their own effective dates. Folding them into a single narrative about tariffs going up is how a team ends up applying the remedy it read about to a shipment it was never aimed at. Separate instruments, separate boundaries. That they are being pursued at the same time is a fact about scheduling, not a fact about substance, and the two should not be allowed to merge in your register.

One more boundary, and it is a narrow one. The fee, as reported, attaches to arriving vessels. That makes the arrival the relevant moment, not the load port, not the routing, not the flag history. If the boundary is the arrival, then a change in your routing does not by itself change whether the fee attaches to a call; only the vessel, the call and the rate in force do. I raise it because routing is the one thing teams reach for first when they hear a fee, and here the routing is not the lever the description points at. The lever the description points at is the identity of the ship on the day it arrives.

Let me put a size on the fee, with the assumptions written down, because a number you can see is worth more than an adjective and because an unlabelled number is exactly the sort of thing that misleads people. Assume, for the arithmetic only, a vessel call of 100,000 net tons. That is my assumption, not a figure from the reporting; change it and the answer changes with it, in a straight line. At $23 per net ton the call carries about $2.3 million. At $28 it carries about $2.8 million. The step from 2026 to 2027 adds roughly $500,000 on that one assumption.

Suppose, again only for the arithmetic, that the call moves 8,000 twenty-foot-equivalent units. Spread flat, the fee is on the order of $287 per unit in 2026 and $350 in 2027, a spread of about $63. I am not telling you to price anything and I am certainly not telling you where that money lands in your books; I am showing you the magnitude, so that the boundary clause and the allocation clause get the attention the headline will not give them.

Sensitivity is the honest half of that exercise, so let me state it plainly. Halve the assumed tonnage to 50,000 net tons and every figure halves: about $1.15 million a call at $23 and about $1.4 million at $28. Double it to 200,000 and everything doubles. The fee scales with the hull, which is a property of net tonnage, and it is why two shippers on two different vessel sizes can face the same rule and very different magnitudes. The rule does not care what fraction of the ship you book. Only the contract does. Write the assumption, keep the assumption, and the exercise stays honest even when the numbers move.

From there, the records, and I want to be clear that I mean the compliance side of records and not the operational side. What establishes a hull built origin is the registry and the build documentation. What establishes operation is the corporate and operating structure. What establishes the applicable scope of the fee is the notice. What establishes who bears the charge as between you and your counterparty is your contract. Four questions, four documents, four owners inside your own organisation, and none of them is answered by a news article. If one of the four is answered from memory, the file has a hole in it, and the hole is the first thing anyone looking for leverage will find.

So the boundaries, plainly, and this is the part I would copy into a memo. The conclusion that this fee touches your trade holds only if three things are true at once. The vessel meets the scope as finally framed. The call falls within the range the fee covers. The rate you apply is the rate in force at the relevant time. Flip any one of the three and the conclusion moves. It moves if USTR declines to reinstate the fee the letter asks about, because then what governs is the current structure, whatever that proves to be. It moves if the scope is drawn more narrowly or more widely than a summary suggests, and scope is precisely the kind of thing that gets narrowed between a request and a final notice. It moves if the effective date or the 2027 step is altered. And it moves if the measurement basis is read differently.

Here is the exception most people miss when they read this story, and I would underline it. A fee measured on net tonnage is a charge on the ship. That puts a non-Chinese vessel carrying Chinese-origin cargo on the far side of the line, and it puts a Chinese-built vessel carrying non-Chinese cargo on the near side. As reported, the cargo origin is not the test. The ship is. Be careful with that, because the instinct is to look at where the goods came from, and the description points at the hull instead. If your trade runs Chinese-origin cargo on vessels that satisfy neither limb of the description, the fee as structured does not attach to that call, and the thing to verify is the description, not your supply chain.

One more open question, and I will leave it as a question rather than resolve it, because resolving it needs documents I have not verified. The letter asks USTR to reinstate a fee. Reinstatement implies that something lapsed, while the structure described as in force implies that something is operating. I am not going to reconcile those two from a summary. What follows is narrow and useful: do not assume that what is in force and what is being requested are the same instrument at different moments. Establish from the notice which instrument governs which period, and treat the gap between them as unresolved until a document closes it. Strictly speaking, that is the correct posture whenever a request and a rule use similar words for possibly different things.

Now the part that is actually yours to run, and I will keep it on a quarter-to-year clock because that is the clock this subject runs on. Through the current quarter, confirm the applicable scope against the official notice rather than against commentary, and map it onto your vessel mix and your booking terms. Mark every contract that contains language capable of passing the fee to you, directly or through a reference to port charges, and note the date each one renews relative to the 2027 step. Set one accountable owner for the question and one register where the answer lives, carrying the notice version and the date it was read. Put a review at each quarter end. And put a specific trigger on the calendar: the publication of any notice that changes the rate, the scope or the effective date. When that trigger fires, the review runs inside a defined number of business days, not whenever somebody happens to notice.

Targets, because a target is what keeps a review from becoming a ritual. Every vessel in your regular rotation classified against the applicable scope, with the classification written down and sourced rather than carried in someone head. Every contract capable of passing the fee to you flagged, with the renewal date and the rate it was written against. One named person accountable for the file, and one date each quarter on which it is refreshed. If a notice issues that changes the rate or the scope, the refresh happens within five business days. Those are modest targets and they are the difference between a position and a hope.

So my answer to the question does this affect us is the one I give most often, and it is not evasion. It depends, and now you know on what. It depends on whether the vessel meets the scope as framed in the notice, not as paraphrased in a summary. It depends on the rate in force at the moment you are looking, and on whether your paper survived the 17 April 2026 date and will survive the 2027 step. It depends on whether what the senators requested becomes what USTR does, which is not something a letter decides. Hold those premises, keep the two rates and the one effective date in front of you, and re-run the question the moment the document moves. The letter is the ask. The fee is the fact. Keeping those apart is the whole job.

↑ Back to the key points

— By Dr. Ingrid Voss