113 Ports Hit by Coordinated Dockworker Action, Threatening 40% of Med Container Traffic

The Analysis

The notice says this. On 30 October, dockworkers acted across 113 ports, with 25-plus unions involved, and cargo was disrupted at Hamburg, Le Havre, Genoa, Tangier Med and Nhava Sheva. The organisers put the participating unions at roughly 40% of Mediterranean container traffic, and they say up to 40,000 containers were delayed. Reuters could not verify the figure. Freightos notes vessels at Genoa wait one to two days for a berth anyway. C.H. Robinson flagged delays by terminal. Marsh judged a 24-hour stoppage unlikely to disrupt trade materially, though perishables suffer first.

I read rate and space notices for a living. My first move is never to accept a conclusion. My first move is to cut the number open. What is the base. What is the scope. What is the effective range. A headline that carries 40% and 40,000 in the same breath, from the same organisers, with Reuters declining to sign off, is not a fact. It is a claim wearing a number's clothes. So let's run the numbers before anyone books anything off it.

I'm not going to tell you freight is going up and that this matters. That sentence is empty and I refuse to write it. I'm going to tell you which of these digits can go into a decision model, which ones you should throw out, and what to put in their place — your own booking data, container by container. Then I'll give you the line past which you lock November space, and the hour you lock it.

Start with the bones of the event, because those are the part you can actually lean on. 113 ports, 30 October, 25-plus unions, five named hubs. These are countable, dated, and attributable to multiple parties. Even here there is a scope problem worth flagging: 'acted across 113 ports' is not the same as 113 ports fully stopped. Some may have seen a short walkout, some a full shift lost, some a rally with no cargo impact at all. A count of ports touched is a count of geography, not a count of throughput lost. Treat it as a map, not as a measure. The five named hubs matter more than the 113, because Genoa, Tangier Med and Nhava Sheva are transfer points where a missed connection costs you a whole extra rotation, not just one late box.

One more cut at the bones before I move on, because the small numbers are sloppy too. '113 ports' counts places touched, not hours lost; a two-hour walkout and a full lost shift both score as one port. '25-plus unions' carries a plus sign, which means nobody is handing you an exact count. 'Up to 40,000 containers' carries an 'up to', which is a ceiling, not a reading. Notice how every soft number in this story arrives wrapped in a hedge — roughly, up to, 25-plus, could not verify — while the headline states them flat. Count the hedges and you have already done half the analyst's job. I do not reject hedged numbers; I reject headlines that strip the hedges off and hand them to me as facts. That is a different thing, and it is the distance between a rate desk and a news feed.

Then take the two big digits apart, because there is a problem with this number, and with the one next to it. 40% of Mediterranean container traffic. Forty percent of what, exactly. Of annual throughput, of weekly liftings, of daily gate moves, of vessel capacity calling those ports, of the unions' own members' share. The phrase says the unions 'handle' roughly 40%, which is a statement about who they are, not about how much cargo stopped on 30 October. A percentage with no stated denominator, no period, and no unit is not a measurement. It is a rhetorical device with a percent sign attached. You cannot put it in a model. I cannot put it in a model. No one can, and that is precisely why it survived into the headline.

The 40,000 containers delayed is the same problem wearing a different unit. Delayed by how much — by one hour, by one day, by a missed onward sailing. Containers or TEU. A point-in-time pile at the gates, or a cumulative count across the whole action. Counted at all five hubs or across all 113 ports. And counted by whom. Both 40% and 40,000 come from the organisers, the party with every reason to make the action look large. Reuters could not verify the figure. Read that clause carefully: it is not hedging, it is a boundary. It tells you the two most alarming numbers in the story are single-source claims, not established facts. In my work a number that cannot be independently sourced does not get a vote. It gets quarantined.

Now the part I actually enjoy, because the two organiser numbers do not agree with each other, and you can show it with third-grade arithmetic. We算一下 — let's run the numbers. If 40,000 delayed containers really represented 40% of Mediterranean container traffic, then total Mediterranean traffic in that window would be 40,000 divided by 0.4, which is 100,000 containers. One hundred thousand. For a sea that carries many times that in a normal working week, that implied total is far too small to be the whole Mediterranean box flow. So either the 40% is a share of union coverage and the 40,000 is a separate pile of delayed boxes, or the two figures use completely different bases, or one of them is simply wrong. They cannot both be describing the same thing. A headline that welds them together — 40% of Med traffic, 40,000 boxes — is doing arithmetic the source itself never did. There's a problem with this number, and once you see it you cannot unsee it.

Here is the second cut, the one that brings it down to a scale a booking desk can feel. Spread 40,000 containers across 113 ports and you get roughly 354 containers per port. Three hundred and fifty-four. A single medium hub moves tens of thousands of boxes on an ordinary day, so a per-port pile of a few hundred is a rounding error against the claim that 40% of a whole sea's traffic is threatened. The arithmetic is not an attack on the workers; it is an attack on the framing. The organisers measured their own footprint and their own headline, and Reuters rightly refused to certify it. For the importer or exporter, the practical conclusion is blunt: a sea-wide percentage has almost no information value for the one sailing you own. Stop reading it.

So what survives the autopsy. Three things, and all three are third-party, scoped, and checkable. Freightos says vessels at Genoa already wait one to two days for a berth — that is an observable market fact, a baseline that exists before any strike. C.H. Robinson flagged delays by terminal, which is the right unit of analysis, because congestion is always terminal-specific, never sea-wide. Marsh judged a 24-hour stoppage unlikely to disrupt trade materially, with perishables first in line. These you can put in a model: a baseline berth wait, a terminal-level delay flag, a duration threshold, and a cargo-priority ranking. Everything else in the headline is decoration until someone verifies it. When a number cannot be sourced, replace it with your own. That is the whole method.

Who actually gets hit, and in what order. Rank it instead of saying 'the industry is affected'. Perishables go first, exactly as Marsh said, and the reason is mechanical: a reefer needs a plug and a power supply, and a stopped terminal stops moving reefers to plugs while the free-time clock keeps running. Transit and relay cargo at Genoa, Tangier Med and Nhava Sheva goes next, because a hub delay is not one late box, it is a missed feeder connection that pushes the container onto a later mainline sailing, sometimes a full rotation later. Ordinary dry cargo on a direct call absorbs the delay and mostly catches up within a week. So the loss is not spread evenly across the mythical 40%; it concentrates on cold chain and on transfer legs. If your book has no reefers and no Med relay, the honest answer is that this event may cost you a day or two and little else. Saying so is more useful than repeating a percentage.

Follow who passes the cost on, because a stoppage does not move the base rate and that surprises people. The base ocean freight on your contract is unchanged on 31 October. What moves is the layer underneath it — congestion surcharges, demurrage, detention, reefer monitoring and plug fees, re-routing and change-of-port fees, and the rebooking spread when a box is rolled to a later sailing. Carriers pass berth-wait cost to shippers as a congestion surcharge. Terminals pass dwell to shippers as demurrage and detention. Nobody passes anything back. So when someone says 'freight went up because of the strike', cut it open: the base held, the add-on layer swelled. If you only watch the base rate, you will swear nothing happened and then get a bill that says otherwise. Split every quote into base plus surcharge before you agree to anything.

Practically, that means I read every quote as two columns and refuse to look at the all-in total until both are itemised. Column one is base ocean freight, fixed by the contract, unmoved by a strike. Column two is the surcharge stack — congestion, peak-season, reefer, plug, change-of-destination, plus detention and demurrage exposure. A strike only ever touches column two. So when a forwarder sends a revised all-in number after 30 October, I ask which line changed and by how much, and I want that answer before I approve anything. A revised all-in rate with no line-item breakdown is not a quote, it is a rounding up. If a congestion surcharge is new, that is a cost being passed to me, not a rate rise I should quietly absorb. Naming the line is naming who is paying, and that is the only reason to split the columns at all.

When does it reach you, in days. At T+0 loading stops. At T+1 to T+2 the queue builds, and here the Freightos baseline bites: Genoa was already one to two days behind before anyone walked out, so a 24-hour action lands on top of an existing backlog, not on a clean slate. Inside a week you see blank sailings, skipped ports and missed connections as carriers reshuffle. At T+14 to T+28 the demurrage and detention invoices arrive, long after the news cycle has moved on. The decision window is much shorter than the cost window. You have roughly the first 24 to 48 hours after the stoppage is confirmed to act on your own bookings; the bills from those 48 hours show up almost a month later. That gap is where money quietly leaves.

Now the point nobody in the coverage wrote down, and it is the one that actually costs you money. The story is framed as 'how many boxes were held'. The expensive thing is not a held box. The expensive thing is two clocks that keep running while the terminal is shut: demurrage free time and detention free time do not pause for a strike, and neither does a cut-off. And the sharper edge is the cut-off itself. When a port reopens after a 24-hour stop, vessels bunch up and race for berths; terminals protect themselves by compressing or pulling forward the cut-off, not by extending it.

Your cargo was never 'delayed' in the 40,000-box sense. It simply failed to make a cut-off that moved earlier by a few hours, and got rolled to the next sailing. Then demurrage accrues on a container sitting in a yard you cannot gate out of, and the invoice lands three weeks later when nobody remembers which shift caused it. A rolled box plus a running free-time clock is the real exposure. Not a sea-wide percentage. Not even the 40,000. If you remember one thing from this, remember the cut-off moved and the clock did not stop.

The disagreement in the sources is worth keeping, because it tells you where the conclusion flips. Marsh says 24 hours is unlikely to disrupt trade materially. C.H. Robinson and Freightos are already flagging delays. Both are right, because they are measuring different bases. Marsh is looking at aggregate trade flow averaged over a long period, where one lost day is noise. Robinson is looking at a specific terminal on a specific morning, where one lost day is a missed cut-off. The moment the condition changes, the verdict flips: stretch the stoppage from 24 hours to 72 hours, or let it sit on top of Genoa's existing one-to-two-day wait, and 'unlikely to disrupt materially' stops being defensible.

Flip it the other way: if Reuters later confirms the 40% and 40,000 were heavily overstated, this becomes a loud headline over a modest event. But you cannot trade on that hope, because your own cut-off is hard and your own free-time clock is hard regardless of what the organisers claimed. A verified correction helps your understanding; it does not un-roll your box.

Let's run the numbers on a real book, with the assumptions stated plainly. Assume I ship 40 containers a month through Genoa as a relay, and assume one sailing of 10 of them is reefer cargo with a customer delivery window carrying only 3 days of slack. Baseline Genoa berth wait is one to two days per Freightos; take the midpoint, 1.5 days. Add the 24-hour stoppage, 1 day. Add reopening bunching of another 1 to 2 days. That sailing lands roughly 3.5 to 4.5 days late against plan. With 3 days of slack, the reefer leg breaches the window first, exactly the priority Marsh flagged. Now set that against the organiser math: 10 of my boxes against the implied 354-box average per port is about 3%, and against a claimed 40% of a whole sea it is invisible.

That is the whole argument in one line. A percentage about a sea tells me nothing about my 10 boxes; my booking data tells me everything. So I replace the headline with the only numbers I can verify: the booking confirmation, the carrier's delay notice, the terminal cut-off by vessel name, the gate-in and VGM and SI deadlines, and the live status of each of my 10 container numbers. Those are first-party, timestamped, and disputable. 40% is none of those things.

Cross-check the same book from the other direction, because one pass of arithmetic is never enough for me. If the organisers' 40% were a flow share and 40,000 were the delayed pile, the implied 100,000-box denominator fails a common-sense test; if 40% is union coverage instead, then it never described delayed cargo at all and has no business sitting in a delay calculation. Either way the headline number cannot be an input. The only honest inputs are the ones with a timestamp attached to my own containers: cut-off, gate-in, VGM, SI, live status. Build the decision off those four or five fields and a sea-wide percentage becomes irrelevant by construction, not by argument. That is what replacing a claim with your own data actually means in practice — not a slogan, an actual column in a spreadsheet that somebody has to fill in before a booking is released.

The alternatives each have a price, and the cheapest-looking one is usually the trap. Re-routing away from Genoa to an alternate hub buys you around the queue but costs a change-of-port fee, an extra leg, extra transit days, and a brand-new cut-off you have not yet confirmed. Air-freighting the perishables to save the delivery window jumps the cost by an order of magnitude and only makes sense on high-value cold cargo. Holding the cargo and waiting for reopening looks free, but it is exactly the option where demurrage and detention accrue while you wait. The most common mistake is deciding off the 40% headline — either over-reacting into an expensive air move, or under-reacting because '24 hours won't disrupt trade'. The second most common is treating reopening as the end of the event, when the rolled cut-off and the lag-time invoice are still ahead of you. And the quiet third mistake is watching only the base rate while the surcharge layer does the damage.

So here is what I actually do, with times and owners. Within the first working day I pull the booking confirmation, the carrier delay notice and the cut-off for every November Mediterranean sailing I own, and I log status at container-number level, not port level. I get the cut-off in writing by vessel name, never by sailing week, because in a bunched reopening two vessels in the same week carry two different cut-offs and one of them will catch you out. On any reefer I confirm plug availability and the free-time clock with the terminal inside 24 hours, since those clocks did not pause.

After reopening I re-confirm the cut-off at the 48-hour mark, when bunching compresses it. Weeks later I reconcile every demurrage and detention line against the real gate-in timestamp and dispute anything billed over hours the terminal was shut. And I set one hard trigger before any of this: the moment Genoa berth wait moves from one-to-two days to three days or more, or a carrier issues a congestion surcharge, I lock November space and cap the surcharge that same day. Past this line, lock it.

One dry note to close on, because somebody has to say it. 40% of a sea and 40,000 boxes, both from the organisers, both unverified by Reuters, and both failing a division check against each other — I am not signing off on that, and neither should your procurement sheet. The number that deserves your signature is the cut-off time on the vessel you actually booked. Cold comfort, but the boxes that got rolled were the ones whose owners were busy arguing about a percentage. Watch the berth wait, watch the cut-off, and when surcharges cross the line, lock. That is the whole job.

↑ Back to the key points

— By Vivian Zhao