250 LA-LB Security Officers Win Strike Vote as Nov 4 Talks Cover 85% of Long Beach Boxes

The Analysis

Two hundred and fifty people voted. The number I would write on the whiteboard is 17:00.

About 250 port security officers of ILWU Local 26 at Los Angeles and Long Beach voted to authorise a strike, originally set for 17:00 on 1 November. They are roughly 2% of the two ports' ILWU workforce. They guard eight core container terminals that carry about 85% of Long Beach volume and 33% of Los Angeles volume. The Wall Street Journal reported on 29 October that the employers' side, the PMA, was finalising a contract extension to 4 November under federal mediation, so the walkout is postponed.

Postponed is not settled. The vote does not lapse on its own, and the next session is 4 November.

The arithmetic looks wrong until you know what the job is. Security officers do not move boxes. They hold the gate.

On these terminals the function is not decorative. The facility security plan requires a security presence. Drivers get badged at the lane. Credentials get checked against the federal credentialing system. Lanes get opened and closed, escorts get assigned, the yard and the rail interface get patrolled. Take that function away and the terminal does not run a slow gate. It has no lawful gate at all, and a gate that cannot legally open is not a delay, it is a shutdown of the only door the cargo uses.

That is why 2% is the wrong denominator. Two percent of the workforce is one hundred percent of the access control, and access control is the single point every truck, every driver, every empty and every chassis has to pass through. If the crane gangs stop, you lose the ship's productivity and the vessel sits at berth. If the gate stops, you lose the truck capacity, and the truck capacity is what feeds the rail ramp, the transload warehouse and the customer's appointment book. One of those is expensive. The other one is a queue that re-forms for a week.

So the exposure is not proportional to headcount. It is binary, and binary is a far worse property than small.

And these are not eight ordinary terminals picked at random. They are the core ones, which in this harbour means the berths that take the big ships and the ones with on-dock rail behind them. That is the combination that matters, because on-dock rail is where a container stops being a truck problem and becomes a train, and a train load is the only thing that clears a yard at scale. Eight terminals with on-dock rail and 85% of the volume is the part of Long Beach you cannot work around. The remaining terminals are not a substitute. They are a spillway, and a spillway only takes what fits.

The calendar does the rest of the damage. The first week of November is the week the pre-holiday discharge peak lands, and it is the week the rail cuts to the inland distribution centres get tight. A container that misses a Wednesday rail cut does not arrive at the DC on Thursday. It arrives the following Tuesday, and by then the DC's appointment book has been rebuilt around the assumption that it would not.

There is a second layer to the timing that people miss because it is boring. November is when the last of the holiday inventory is supposed to be in the building and the first of the spring programme is supposed to be booked. A gate freeze does not just delay the holiday goods. It pushes the spring bookings back into a week that was already full, so one bad week in November shows up as a bad February.

The timing also explains what looks like generosity. Talks get extended when the cost of a stoppage spikes, and the cost of a stoppage spikes in the first week of November. The move to 4 November is not goodwill. It is the calendar doing the negotiating.

Walk a day at one of those eight terminals and it gets very concrete.

Security muster is around 05:30. The gate lanes open at 06:00 and the first appointment window runs to 08:00, and that window carries the heaviest queue of the day, because the drayage fleet front-loads to clear the yard before the second shift arrives. If the muster does not happen, the lanes do not open at 06:00. By 07:15 the queue is out of the terminal road and onto the street, and the street is not yours to manage. By 09:00 drivers are hitting their hours and turning around, and a driver who times out is not a delayed move. He is a move that is gone, because his slot gets rebooked and the container sits another day.

A four-hour gate closure therefore does not cost four hours of truck capacity. It costs the morning window, and in the terminals I have worked, that window is close to 40% of the day's truck capacity. That 40% is my own working figure, not a number from the article, and it is the one I would check against your own terminal before building anything on it.

Then the shape of the recovery. A terminal that loses a morning does not make it up by running the second shift harder, because the second shift is already committed to the vessel. It makes it up by pushing appointments into the next day, which pushes the day after that, and the queue does not flatten out until the weekend. Three days of disruption, one day of cause.

Then the part that outlasts the strike. Empties.

An importer cannot return an empty box when the gate is shut. The depot's empty pool drains in about three days of normal import flow, and once it drains, exporters booking for the following week cannot get equipment at all. Not late equipment. No equipment. Rebuilding that pool takes five to seven days after the gate reopens, because the empties have to come back out of the yards first, and the yards are full of the imports nobody could pick up.

The chassis pool behaves the same way and is less visible. Street turns stop, chassis stop circulating, and the pool's free time expires while the boxes are standing still. Then the per diem starts on equipment that is not moving and not earning, and that charge follows the container home to the shipper who has no idea a vote in San Pedro happened.

Then look at the first four hours after the gate reopens, because that is where the recovery either happens or does not. The lanes open at 06:00 and everybody who was cancelled wants the same window. A terminal that normally runs 350 trucks a shift gets 600 drivers trying to get into the first appointment block, the queue at the in-gate builds faster than the lanes can clear it, and by 10:00 half of them are still outside the fence. The fix is not more lanes, it is a sequencing decision: release the rail-bound boxes first, because rail has a hard cut and a truck does not, then the transload boxes, then the pure local deliveries. Terminals rarely do this on their own. Somebody has to ask for it, and the somebody is the shipper with a DC appointment on Wednesday.

This is the tail nobody prices. The visible cost of a gate freeze is trucks standing in the road. The real cost is a week of export bookings that cannot be filled, and those bookings belong to somebody else's contract with a penalty attached.

Here is the number in the story that got buried under the word pair LA-LB. Eighty-five percent of Long Beach volume, thirty-three percent of Los Angeles volume, on the same eight terminals. Those are not the same event. The headline is one harbour. The exposure is Long Beach.

That asymmetry is the most useful fact in the whole report and nobody drew the conclusion from it. If your book is Long Beach heavy, those eight terminals carry most of your volume and there is no hedge inside the port. Your options sit outside it, and every one of them costs days: divert to Oakland and add two days of drayage plus a rail rebooking, or route through a transload in the Inland Empire and eat the double handling, or let it sit and pay. If your book is Los Angeles heavy, two thirds of your volume is on terminals these officers do not guard. The hedge already exists, it is across the same bay, and it costs a truck move rather than an ocean diversion. Most people will never see it, because they plan by port and not by terminal.

The second thing left out is the bargaining table. Those 250 officers are Local 26, a separate unit from the longshore clerks and the crane operators. The coastwise PMA-ILWU agreement that everybody watches is a different negotiation at a different table, and a tentative deal there does not settle this one. When a headline says the ILWU has a deal, it is not about these people, and your file stays open. The evidence is in plain sight: what the PMA is finalising is an extension, under federal mediation. An extension is a standstill. You get a standstill when there is no agreement yet.

And here is the mechanism the coverage did not follow through on, which is the one that turns a two-day event into a week. The ships keep coming. A security strike does not stop a vessel from berthing, and the carriers do not blank a sailing because a gate is shut; the ship discharges on schedule, the boxes come off the vessel and land in the yard, and the yard has no way out because the gate is the way out. Run the space. A terminal discharging 9,000 TEU a day against zero gate throughput fills its import blocks in about 48 hours, and once the blocks are full the ship operation slows down too, because there is nowhere to put the next lift.

So the sequence is not gate closure, then recovery. It is gate closure, then yard saturation, then a ship that cannot work, and only then recovery. Every extra day of gate closure compounds against a yard that is already tight, which is why the recovery curve is steeper than the closure curve and why the people who plan for one day of disruption always end up surprised by four.

The third is the one everybody assumes away, and assumptions are what cost money. Nothing in the reporting says the authorisation lapses on 4 November. It goes back in the drawer, and an option in a drawer gets exercised at the worst possible moment, which is the moment the terminal is fullest. Nobody exercises an option on a quiet Tuesday in February.

Now the opposite case, because it is the more likely one and it deserves pricing properly.

Two hundred and fifty people, eight terminals, a gate that cannot legally open without them. Run the employer's arithmetic. One day of closed gates costs the port several million dollars in demurrage alone, on the estimate below, before a single customer relationship is dented. The cost of settling a unit of 250 is not in that league. Which means the base case is a short strike or no strike at all, and the people who will look smart are the ones who did nothing.

That is exactly the trap. A 36-hour gate freeze at peak does not appear in most contingency plans, because those plans were written for a ten-day coastwise shutdown. They have a diversion playbook, a customer letter and a force majeure clause, and all three are designed for a long event. A short event beats them, because there is no time to divert, no time to send the letter, and the only thing that helps is having already moved the boxes you cared about.

The condition that flips the judgement is what comes out of the 4 November session. An agreement in principle means stand down, and this becomes a paragraph in a year-end review. Another extension means the window has moved and the option is worth more than it was on 1 November, not less. Watch which of the two it is. Do not watch the headline, and do not watch the vote count, which has not changed since it was taken.

For the long-strike case to become real, one thing has to happen that has not happened yet: the dispute has to move from a 250-person unit to the main table. A security unit standing alone is a two-day problem with a cheap settlement. A security unit standing next to a coastwise negotiation is a different animal, because the strike stops being about security pay and starts being about leverage. That is the scenario the ten-day plans were written for, and the trigger is not a date on a calendar. It is whether the main table is still open. Right now the main contract is not the story, and that is the strongest argument for this staying short.

Numbers, with assumptions in the open where they belong.

Two hundred and fifty people at 2% of the workforce puts the two ports' ILWU headcount near 12,500. Eight terminals against 250 officers is about 31 per terminal, which is roughly one per gate lane on a normal shift with a rover left over.

For Long Beach I will use 8.5m TEU a year as a working assumption. That figure is mine, not the article's, and it is the number you should replace with whatever your own terminal publishes. At 8.5m TEU the daily average is about 23,300 TEU across the year, and 85% of it, roughly 19,800 TEU, sits on the eight covered terminals. Spread over a 16-hour gate day that is 1,240 TEU an hour, or about 620 FEU.

Freeze the gates for 12 hours and 7,400 FEU do not move. Clear the backlog at 15% above normal throughput per day, which is optimistic for a terminal already running at peak, and the tail is five working days.

Cost it. Assume those 7,400 FEU sit four extra days at an average of 150 dollars a day of demurrage and detention. That is my assumption, the coastal range is wide, so read it as an order of magnitude and not an invoice. The result is 4.4m dollars across the port for one gate day, before a single rail cut is missed and before anyone counts the export bookings that could not be filled for lack of empties.

Now bring it down to one shipper. If you move 500 FEU a quarter through Long Beach, that is about two containers a day, and a three-day disruption puts six boxes behind. Six boxes is not a catastrophe. Six boxes is six customer calls, six revised delivery dates, and one retailer chargeback, and the chargeback is the expensive part.

What I would do between now and 4 November.

Begin by re-cutting the open book by terminal, not by port. If your file says LA-LB, you cannot make a decision with it, because the answer for Long Beach and the answer for Los Angeles are different answers.

Next, list every container that has to be at a DC before the third week of November and pull the ones on the eight covered terminals forward by one rail cut. One cut of margin costs two days of yard time. Missing the cut costs a week and a December conversation you do not want to have.

After that, book your empty returns for the next ten days into depots outside the affected gates and get written confirmation of the return location per container. The default return instruction sitting in your system was written for a normal week and nobody has updated it.

Then hold a 30-minute call with your drayage carriers on 4 November, after the session and not before. Two questions only: which lanes they will not run if the gate closes, and what it costs to pre-position chassis in your yard for 48 hours. Both answers are cheap this week and expensive next week.

And keep one habit through all of it. Do not stand your plan down because a headline says the strike was averted. It was postponed. The authorisation is live, the next date on the calendar is not the last date on the calendar, and the difference between a plan you kept and a plan you cancelled is about six hours of work. Six hours is a morning on the phone with three carriers. A cancelled plan is a December you spend explaining why the freight was late at the exact moment everyone was watching.

↑ Back to the key points

— By Tomás Ferreira