250 Watchmen Authorize Strike at LA-Long Beach, 85% of Cargo at Risk Nov 1

The Analysis

Two hundred and fifty people can stop a port complex that moves on the order of twenty million TEU a year. That is the whole story, and it is the only number in it that matters. ILWU Local 26, the security watchmen at Los Angeles and Long Beach, has authorised a walkout from 5pm Pacific on Sunday 1 November, the hour their contract extension runs out. The deal they are working under expired in July 2025. Sixteen months of bargaining, a federal mediator in the room, and no agreement.

Here is the arithmetic the headlines skipped. Two hundred and fifty is 2% of the two-port ILWU workforce, which puts the denominator at about 12,500. A 2% slice does not stop a port, so the number doing the real work is a different one. PMA says the terminals where these 250 are employed move 85% of Long Beach cargo and 33% of Los Angeles cargo. Add those up properly, because they are percentages of two different totals, and you land between 58% and 62% of the complex, not the "over half" the copy says. If every other local honours the picket line, the ceiling is the full 12,500.

So the floor is 58%, the ceiling is 100%, and the press printed the floor. We can work with that. Show me the number, not the feeling. The number that appears in no report is the one about the clock, and it is the one that decides whether a disruption becomes an invoice.

Start with the denominator, because every other figure in this story is a percentage, and a percentage you cannot divide into something is decoration. Two hundred and fifty watchmen is 2% of the two-port ILWU workforce. Two hundred and fifty divided by 0.02 is 12,500. That is the real headcount of this dispute, because the other 12,250 are the ones who decide whether 250 people close a port complex or close nothing at all. PMA's own statement says the rest are not expected to cross a picket line. If that holds, the effective size of this walkout is 12,500 and the 250 is only the fuse.

Now the two numbers every report printed. The terminals where Local 26 members work move 85% of Long Beach cargo and 33% of Los Angeles cargo. Those are percentages of two different totals, and nobody added them up. Do it properly. Write Long Beach's share of the two-port complex as L and Los Angeles's as A, with L plus A equal to one. The combined share is 0.85L plus 0.33A. Long Beach has been running somewhere around 48% to 52% of the complex in recent years. Put in 48% and you get 57.9%. Put in 52% and you get 60.0%.

Push it to 55%, if you think Long Beach has been taking share, and you get 61.6%. The honest range is 58% to 62%. "Over half" undersells it by eight to twelve percentage points, and on a complex moving on the order of twenty million TEU a year those missing points are somewhere between 1.6 million and 2.4 million TEU of exposure that never made it into a headline. A headline that rounds 58% down to over half is not lying, it is simply not doing arithmetic, and there is a difference.

There is a second problem with the 85%, and it is the kind that decides whether a number is worth believing at all. It is a PMA number, published by the employers' side in the middle of a labour dispute, with no independent reconciliation and no terminal-by-terminal table attached. More to the point, it measures the wrong thing. It counts all cargo at terminals where watchmen are employed. It does not count cargo that watchmen move, because watchmen do not move cargo. Their job is access control: truck gates, badge checks, perimeter, camera rooms. So the 85% is a statement about which terminals employ these 250 people, written in the shape of a statement about how much cargo stops. Those are two different claims and only the first one is certainly true. This is the number I would push back on across a table, and it is the one figure here I would not put into a contract.

That said, do not let the calibre problem talk you into doing nothing, because the operational arithmetic does not need the 85% and it is worse than it looks. The notice names eight locations, APM Terminals and LBCT among them. Two hundred and fifty across eight sites is about 31 people per site. A picket line that actually stops a truck needs four to six people at a gate. Thirty-one covers five to seven gates, and most container terminals in the complex run two to four truck gates plus a marine gate.

The coverage closes. Two per cent of the workforce is enough to put a body at every door of eight terminals, and once the door is picketed, the longshore locals decide whether anything moves behind it. That is the entire mechanism, and it works whether or not anyone believes the percentage. Divide the exposure by the headcount and you get the leverage: 58% of volume controlled by 2% of the people is a multiplier of roughly 29 times. That is what a chokepoint job looks like on a spreadsheet.

Put a daily volume on it. A large West Coast container gate processes somewhere in the range of 800 to 1,500 truck transactions a day. Use your own terminal's numbers if you have them; the point is the order of magnitude. Eight terminals losing their gates is 6,400 to 12,000 truck moves a day that do not happen. Over three days that is 19,200 to 36,000 moves. Those boxes do not disappear. They queue, and they queue in a yard that was not empty on Friday.

Why 1 November, and why 5pm. The old contract expired in July 2025. July 2025 to November 2026 is sixteen months, roughly 488 days, with a federal mediator involved for part of it and still no new agreement. Sixteen months is not a breakdown. A breakdown does not take 488 days with a mediator in the room. Sixteen months is a card being held, and then a week being chosen. Look at what sits in that week. The autumn peak tail is still landing, the January contract season gets built in November, and the complex is the front door for most of the US inland network. Two hundred and fifty people have no leverage from being 250 people. Their leverage comes from the calendar, and the first weekend of November is the most expensive weekend in the complex's year.

Then the hour, which is where every headline failed a basic check. Five pm Pacific on 1 November 2026 falls on a Sunday. Sunday is the lightest truck day of the week at LA-Long Beach, with most gates on reduced hours or closed outright. A walkout that begins at five on a Sunday afternoon first bites on Monday 2 November, at the morning gate and the night vessel shift. So your clock starts Monday morning, not Sunday afternoon. As of today that is roughly sixty hours, and sixty hours is the difference between a diversion instruction that lands before the cutoff and one that lands after it.

Here is the transmission, day by day, on the assumption the walkout starts. Monday: gates. No gate means no truck, so import boxes that should have been picked up stay in the stack and export boxes that should have been delivered for loading never enter. Exporters should note the second half of that sentence, because empty pickup goes through the same gate as loaded delivery. Tuesday: yard density. A container terminal runs cleanly up to roughly 60% to 75% yard occupancy, and past about 85% the rehandle rate goes vertical, because every box you want is under three you do not. Two days with no gate moves on a complex that was already tight puts the named terminals into that territory, and rehandle is why recovery takes longer than the stoppage itself.

Wednesday: vessel work. If clerks and longshore locals are not crossing, cranes stop or run at a fraction, ships already at berth sail with the load unfinished, and on-dock rail loading inside the terminals stops with them. Thursday 5 November: the first meaningful demurrage accrual on boxes whose free time expired inside the stoppage. That is when the money starts, four days after the strike did, which is exactly why nobody budgets for it.

Five to seven days in, the carriers decide. Three days of no work at eight terminals is a schedule problem. Seven days is a network problem. Somewhere around day five to seven, lines start omitting LA-Long Beach calls and discharging at Oakland, Seattle-Tacoma, Vancouver, Prince Rupert or Manzanillo with rail. They also start blanking sailings, and blanking does the one thing a soft spot market cannot do, which is put a floor under the rate while volume is falling.

Who pays, and how much, splits four ways. The importer with boxes on the ground pays demurrage and detention and receives nothing for it. The exporter trying to load pays a missed cutoff and, if the sailing is cancelled, a roll of seven to fourteen days onto the next string. The drayage operator pays for driver hours at a gate that is not moving, and passes that through as a waiting-time or congestion surcharge the moment his contract lets him, so read the trigger before you argue about the invoice. The reefer shipper is the one nobody writes about: reefer plugs are finite, a yard that cannot move boxes runs out of monitored plugs before it runs out of space, and a reefer nobody is cycling is a total loss, not a delay.

The thing that actually decides who writes the cheque is the Incoterm. Under FOB or CFR to a US West Coast port, destination charges including demurrage and detention sit with the buyer. Under DDP or CIP to an inland point they sit with the seller, and a large and growing share of China-to-US trade is sold DDP. A strike at Los Angeles can therefore land on a factory ledger in Shenzhen. Check the term on every open order before you check anything else, because it tells you whose problem this is, and there is a fair chance it is yours when you assumed it was not.

Run the money on one box. Assume a 40ft import, cargo value USD 60,000, landed at a named terminal on Friday 30 October, with four free days of demurrage and four free days of detention, and assume a West Coast schedule of about USD 200 a day in demurrage and USD 150 a day in detention once free time runs out. Those are my assumptions; your tariff differs by terminal and by carrier, so substitute your own before quoting this back to anybody. A three-day stoppage, Monday 2 November through Wednesday 4 November, with gates turning again on Thursday, eats three days on both clocks. Day five of the demurrage clock becomes the first chargeable day and it now falls inside the stoppage. Three days at USD 200 is USD 600. Three days at USD 150 is USD 450. Call it USD 1,050 on a box worth USD 60,000, which is 1.75% of cargo value for a delay you neither caused nor could prevent.

Scale it. Forty boxes a month is a small-to-mid importer. Forty times USD 1,050 is USD 42,000. At a 15% gross margin, gross profit on that cargo is USD 360,000 a month, so a three-day stoppage takes 11.7% of the month's gross profit. Stretch it from three days to seven and the shape changes: seven days at USD 200 plus seven at USD 150 is USD 2,450 a box, USD 98,000 across forty boxes, 4.08% of cargo value and 27.2% of gross profit. Three days is an annoyance. Seven days is a quarter of the month's margin, and that is the tail you are actually pricing when you decide whether to reroute.

Add the inland leg if boxes get discharged elsewhere. Assume a box landed at Oakland instead of Long Beach for a Los Angeles delivery: inland mileage goes from roughly 15 miles to roughly 350, and at USD 3.50 a mile all-in that is an extra USD 1,170, plus a change fee of a few hundred dollars. Call it USD 1,400 to USD 1,700 a box on top of the USD 1,050. Who pays it depends entirely on how the diversion happened. If the carrier omits the call and discharges at a substitute port under its own clause, the ocean leg to that port is usually on the carrier and only the extra inland falls on the cargo. If you ask for the change yourself, you pay both, including a change-of-destination fee. Voluntary and involuntary diversion are not the same transaction, and the gap is about USD 1,400 a box. Read the clause before you pick up the phone.

Here is the part that appears in no report I have read, and it is the one that decides whether a disruption becomes an invoice. Free time does not stop running. Neither the demurrage clock nor the detention clock pauses because a picket line showed up. Terminal tariffs generally grant relief only where the terminal itself is closed to the public, and even then it is usually a partial abatement rather than a reset. More to the point, relief is almost never applied automatically. It is applied for: a written request to the terminal operator for demurrage and to the carrier or its equipment desk for detention, normally inside a short window after operations resume. Plan around ten working days, then go and read your tariff for the real deadline, because ten is a planning number and the clause is the one that binds. Miss the window and the USD 1,050 is yours by default. Nobody is going to volunteer the credit. They are going to send the bill.

So the highest-return hour of work available between now and Monday is not rerouting a single container. It is pulling the free-time clause and the force majeure clause out of your terminal tariff and your carrier contracts, finding the filing deadline, and drafting the relief request today with the dates left blank. That is a document you can send in ten minutes the morning the gates reopen. Drafted after the fact it takes two weeks and usually gets denied. Two hundred and fifty people have handed you a homework assignment, and the homework pays better than the rerouting.

Three ways this does not happen, and one way it gets worse. It does not happen if the mediator produces another extension before five on Sunday. Four hundred and eighty-eight days of bargaining has almost certainly included previous extensions, another one is the base case, and being ready for a strike that never starts costs you a few hours of drafting. It does not happen if the ratification vote does not carry, because authorising a walkout is permission, not a strike: those are two separate votes, and the coverage ran them together. It does not happen if the employers obtain an injunction, though injunctions in West Coast marine disputes are rare and slow enough that you should not build a plan on one.

It gets worse in one specific way, which is if the picket lines spread past the eight named terminals. Everything above sits on the named-terminal floor of 58% to 62%. If other locals put up their own lines, or if PMA locks out in response, exposure goes to 100% of the complex and every number above becomes a floor instead of a forecast. One signal tells you which world you are in and it costs nothing to watch: gate appointment availability at terminals that are not on the eight-location list. If those blocks start closing on Monday morning, the stoppage has spread, and you move from wait-and-claim to reroute-everything on the same day.

What to do with the sixty hours left. Split the book into three piles. Boxes already on the ground at a named terminal: start free-time tracking today and pre-draft the relief request. Boxes on the water with an ETA inside 1 to 8 November: decide on diversion now, while it is still voluntary and still cheap, because once the carrier omits the call you take the port you are given, and by then inland capacity at the alternative ports is the shortage. Boxes not yet booked: route them away from the named terminals if your transit tolerance allows it. There is no rate premium for doing that this week, only an inland one, and there are five live alternatives.

Set the thresholds in writing before you need them, so nobody is improvising on Monday morning. Three days or less: stay put, document, claim relief, absorb the USD 1,050, because 1.75% of cargo value is cheaper than any reroute. Past day three the demurrage tail crosses roughly 2% of cargo value and the arithmetic flips, so pay the USD 1,400 to USD 1,700 inland premium and move. Seven days or more: assume the network reroutes, and lock inland truck and rail capacity early, because what runs out at Oakland and Vancouver is not vessel space, it is drayage and rail slots, and every other importer reaches that conclusion on the same morning.

One thing not to do. Do not reach for air freight on anything except the high-value slice. Asia to US air cargo is running somewhere around USD 4 to USD 6 a kilo, and a 40ft box carries on the order of 18 to 20 tonnes, so flying a full box is a six-figure joke. The same box, however, usually holds a small share of cargo that is both light and urgent. Assume 5% of the value, USD 3,000, weighing 100kg. At USD 5 a kilo that is USD 500 to protect USD 3,000 of the most urgent line, against USD 1,050 of demurrage on the whole box. Split that slice onto its own bill of lading and move it. Leave the rest where it is.

Two hundred and fifty people, two per cent of the workforce, and the inbound plan of half a country's container trade resting on whether they turn up on a Sunday evening. The demurrage clock does not care that it is only 250, and it does not stop for picket lines, mediators or headlines. Draft the relief request while it is still free; if the port opens on Monday you will have wasted thirty minutes learning where the clause lives, which is thirty minutes better spent than most of this week. Lock inland capacity at an alternative port the morning non-listed terminals start closing appointment blocks, and not a day later.

↑ Back to the key points

— By Vivian Zhao