German port strike risk looms: Verdi vote could halt Hamburg and Bremerhaven from October

The Analysis

Strictly speaking, this is not a freight rate story, although it arrived dressed as one. Two things are on the record. A ballot of the Verdi union is under way on industrial action that could take Hamburg and Bremerhaven out of service from October, and the action, if it is called, would be indefinite rather than a warning strike of a day or two. In the same week, Asia-Europe container rates fell somewhere between seven and fifteen per cent. The second number is being read as an opportunity. The first number is the one that changes what your contract says.

I should be clear about what I am not going to do here. I am not going to tell you whether to book now or wait, and I am not going to tell you how to handle a declaration or a release at the port; those are commercial and operational judgements and they belong to the people who carry the risk. What I can do is set out the machinery, because the machinery starts moving before the containers do, and most of the disputes I see after a stoppage are arguments about when the clock started rather than about what happened.

Begin with what kind of event this is, because the classification decides which clause you reach for. A dock strike in Germany is not a natural catastrophe and it is not an act of the state. It is industrial action taken by a party to collective bargaining, and the legal order treats it as lawful within limits rather than as a misfortune arriving from outside. The limits are not arbitrary. The action has to be called by a union capable of concluding a collective agreement with the employer; it has to respect the peace obligation that holds while a collective agreement is in force; and it has to satisfy a proportionality test that courts have spent decades refining case by case.

The consequence for you is narrower than it sounds. Because the action is lawful, it is not automatically a breach by anybody, and because it is not a breach, the question of who bears the cost is answered by your contract rather than by a general rule of law. I would not quote article numbers from memory, and I would be cautious about anyone who does. The structure is what matters, and the structure is stable.

The ballot decides which of two quite different instruments is being prepared. A warning strike is short by design, a day or two, and its purpose is pressure rather than paralysis; the damage it causes is mostly inconvenience and it is usually absorbed inside existing buffers. An indefinite strike is the instrument that follows a successful ballot, and the word indefinite is doing legal work rather than rhetorical work. It tells you that no end date exists at the moment the action begins. Every clause in your contract that depends on duration then has an unknown input, and unknown inputs are what turn a delay into a dispute.

Whether the ballot succeeds is decided under the union's own rules rather than under a statute I could point you to, and the threshold is high by design. That detail matters when you read a headline saying a vote could halt the ports. A vote is a precondition, not a decision, and the distance between the two is exactly where a great deal of planning gets done on a false assumption.

Here is the part that tends to surprise people. An event of this kind does not usually excuse performance in the way they expect. Force majeure, where it exists, suspends an obligation; it does not allocate cost. If your supplier cannot load and your clause is triggered, the clause may push the delivery date and protect the supplier from a penalty, and it will say nothing at all about who pays for the containers sitting at a terminal in Bremerhaven or for the vessel waiting at anchor outside the river. Cost allocation normally lives somewhere else entirely: in a storage clause, a demurrage provision, a price adjustment mechanism, or a general indemnity. So the first practical question is not whether the strike is force majeure. It is which of your documents answers the cost question, because it is rarely the one people reach for first.

The list in that clause deserves a second reading for what it leaves out. Many schedules of excused events name strikes, lockouts, riots and civil commotion, and stop there. A lockout is the employer's answer to a strike and it is a distinct event; a clause that names strikes but not lockouts may not cover the week in which the gates are closed from the inside rather than the labour withdrawn from without. Then there is the question of when such an event ends, which sounds philosophical until a notice period is drafted to run from the end rather than the beginning. I have seen a clause requiring notice within a stated period after the event concluded, and two parties argue for a month about whether a stoppage concluded when the last vessel was worked, when the last container was collected, or when the settlement was signed. The clause was not ambiguous in the abstract. It became ambiguous at the only moment it mattered.

The second question is which contract you are even talking about, and this is where a great many arguments collapse. Your sale contract with the buyer is one instrument. The contract of carriage evidenced by the bill of lading is another. A terminal or warehouse arrangement is a third. Each has its own clauses, its own notice periods, its own limitation period and, quite often, its own governing law. Under the sale contract, risk passes at a defined point, and if you sold on terms that pass risk when the goods are on board, a stoppage that delays the vessel before loading is your problem and a stoppage after loading is largely not. If you sold on terms that keep risk with you until arrival, the same physical event lands on a completely different balance sheet. The uncomfortable version of this point is that the party with the strongest legal claim after a North European stoppage is sometimes the party who never paid for the freight.

Which law applies is a separate question from which contract, and it is the one that decides what your words mean. The clause you signed says force majeure, or hardship, or labour disturbance, and those three expressions do not carry the same content in every system. A contract governed by English law will read force majeure strictly as a matter of construction, because there is no general doctrine of it to fall back on; the clause does what its words say and nothing beyond them. A contract governed by German law operates against a statutory background that already deals with impossibility and with the adjustment of the basis of a transaction, so a clause that is silent is not necessarily a clause that gives you nothing.

Add the ordinary rules that decide which law applies when the parties have chosen none, and you reach a situation where two importers facing the same stoppage, on the same vessel, hold different rights because their purchase orders were drafted in different countries. I have sat in a room where three subsidiaries of one group gave three different answers to the same question and all three were correct.

There is an exception to almost everything I have written so far, and it is the one I would put at the top of the page. Foreseeability defeats the excuse. A stoppage that has been balloted, announced, reported and openly discussed for weeks before it begins is difficult to characterise as an event beyond the contemplation of the parties. The louder the warning, the weaker the defence that the warning appears to create. I have watched parties spend September drafting a labour clause and then discover in October that the same clause is hard to invoke, because their own correspondence shows they expected exactly this.

There is a second edge to it. A duty to mitigate usually attaches to the party claiming relief, and a claimant who did nothing between the ballot and the first missed sailing has a thin story to tell about being overwhelmed. None of this means the clause is worthless. It means the clause protects the party who behaved as though the risk was real while it was still only a threat.

One distinction I would underline, because it is where the calendar and the law disagree. A threat has no legal consequences of its own. Until a shift is not worked, nothing has been breached, no notice is due and no claim exists; a party that suspends performance because a stoppage looks likely may discover that it, and not the other side, is the one in breach. Yet that same period is when the mitigation duty is quietly running and when the record that will decide everything later is being created. So the stretch between the ballot and the first missed shift is legally empty and practically decisive at the same time. Companies handle it badly in both directions: some act early and manufacture a breach, some wait and lose the evidence. The competent posture is to perform, prepare and write everything down, which is dull advice and almost always correct.

Then there is the distinction that gets lost in translation, between a port that is struck and a port that is closed. If a harbour authority closes access for safety reasons, that is an administrative measure, and it sits in a different category from industrial action: different clause, possibly a different notice regime, and a different party against whom any claim runs. The practical relevance is not academic. Two shippers can suffer the same three-week delay at the same terminal and have entirely different legal positions, because one lost access to a union dispute and the other lost it to an official order.

Equally, indefinite does not mean unlimited. The peace obligation returns once a settlement is reached, emergency and safety functions are protected throughout, and the law does not permit action that is disproportionate to the objective. A stoppage that runs long enough to threaten functions the legal order regards as essential starts to generate its own constraints, and those constraints are usually what produces the settlement rather than goodwill.

There is a European layer that rarely gets mentioned, and it explains why proportionality is not merely a national question. The right to take collective action is recognised as a fundamental right in the Union legal order, and the Court's line of cases on collective action and the internal market, from the disputes that became known by the names of the companies involved, established that it must still be exercised proportionately and go no further than necessary to achieve its objective. I would not push that further than the case law supports and I would not predict how any court would apply it to a particular stoppage.

The point for a shipper is narrower and more useful. The lawfulness of the action is contestable at the margin, which means the characterisation of the event is contestable at the margin, which means a clause that borrows its trigger from a legal category is a clause with an argument built into it. Contracts that name the event and describe the consequences, without requiring anyone to characterise it, are the more robust ones.

On duration, I would separate what happens in the quarter from what happens in the year, because they are different questions with different answers. In the quarter, the operative dates are the ballot result, the notice of action, the first shift that is not worked, and the settlement. Each of those four has a clause attached in a well-drafted contract, and each starts a different clock. The first shift not worked is usually the date that matters for notice obligations, not the ballot result, and parties who count from the ballot routinely miss their own deadlines. Over the year, the event recedes and the paperwork remains. Limitation periods run. The clause you negotiated in a hurry is tested once and then either works or is rewritten in the next contracting cycle. A year from now nobody will remember how long the queue was. People will remember whether the mechanism they paid a lawyer to draft actually engaged when it was needed.

I want to say something about size, because the legal exposure is not distributed evenly and the difference is structural rather than proportional. A small importer with a handful of containers a month is exposed through a single contract and usually a single customer, and the exposure is concentrated in whatever remedy that customer has: a penalty, a discount, a cancelled order. The legal question for that firm is narrow and answerable, and it is worth answering precisely because it is narrow. A large importer with a weekly flow is exposed through a network of contracts that were drafted at different times by different people, and the real risk is not any single clause but the gaps between them. I have seen a group where the purchase terms covered labour disturbance, the logistics agreement covered strikes but not labour disturbance, and the two documents named different notice periods for the same event. Nothing in either document was wrong. The seam between them was.

A related point about the direction of trade, which people tend to skip. If you are the seller and the risk has already passed, you may have no claim at all in law, and yet your buyer will still call you, because in commercial practice the party who chose the routing is treated as responsible for it whether or not the contract says so. That gap between legal position and commercial expectation is where relationships are lost. It is not a legal problem and I will not pretend it can be drafted away. What can be drafted is the sentence that says which of you carries the consequence of a North European gateway becoming unavailable, and whether that sentence survives the event.

Two further interfaces deserve naming, because both are places where people assume a remedy exists and then find that it does not. The first is the carrier. A bill of lading is not a promise that the cargo arrives on a particular day. The usual liberty and deviation wording gives the carrier room to call at another port, to omit a call, to tranship and to route around a disruption, and a claim for pure delay generally fails unless there is an express undertaking about the date, which in container trades there almost never is. That is not a scandal, it is the allocation the trade agreed to decades ago and then stopped reading.

The second is insurance. Cargo cover is built around physical loss or damage, and delay on its own is typically excluded unless it has been bought as an extension, and even then the cover usually carries a waiting period before it responds. So the honest position is that neither the carrier nor the policy is likely to pay for three weeks of a container standing still. What remains is the contract with your counterparty and the sentence you put in it, which is why I keep returning to that document rather than to the ones that read more comfortingly.

The record is the part nobody enjoys and everybody needs. Notice obligations in these clauses are usually formal: a defined form, a defined recipient, a defined period measured from a defined event. A notice that is late, sent to the wrong address, or phrased as a complaint rather than as a reservation of rights may fail to do the work it was written to do. Keep the carrier's own omission or delay notice in its original form, keep the dates on which you passed it on, and keep the correspondence that shows what you knew and when you knew it.

I am not going to stray into how any of this is handled at the frontier; that is a different discipline with a different set of deadlines, and getting it wrong there is expensive in a way that has nothing to do with contract law. What I will say is that the evidentiary question and the operational question are answered by two different people in most companies, and the failure mode is that neither of them knows what the other filed.

Two counterfactuals, because the conclusion is conditional and I would rather state the conditions than let the piece be read as a warning.

If the ballot fails, none of this machinery engages. The containers move, the rates that fell seven to fifteen per cent stay roughly where the market puts them, and a clause renegotiated in a panic turns out to have been unnecessary. That is not an argument against the audit. A clause that was reviewed and found adequate is a result, even when the event never tests it, and the cost of the review is trivial against the cost of discovering a gap at the moment you need it. The only genuine waste in that scenario is a renegotiation conducted under time pressure, because pressure produces wording that is broad where it should be precise and silent where it should be explicit.

If the strike is called and runs for three weeks, the question changes shape. Duration clauses start to bite, long-stop dates become live, and the counterparty begins to ask whether it can terminate rather than wait. That is the moment when the difference between a suspension and a right to walk away matters more than any rate. A contract that suspends indefinitely without a long-stop gives neither side a way out and both sides an argument. If you find that structure in your own documents, it is worth knowing before October rather than during it.

I would rather be boring about this. The rates were soft in the week the ballot was announced, and soft rates are a fact about a market, not a fact about an obligation. The clause does not read the market. It reads your notice, your dates and your record, and if those three are in order, the outcome of a vote you cannot influence becomes a manageable event rather than an argument you will lose slowly.

↑ Back to the key points

— By Dr. Ingrid Voss