Dockworkers strike at Port of Montevideo, Uruguay on July 4

The Analysis

Put the map on the table first. Follow the wide brown mouth of the River Plate, where the Paraná and the Uruguay come down out of the interior and hand their cargo over to the Atlantic, and you find a coastline that looks almost empty of ports and is not. On the fourth of July the workers at Montevideo, in Uruguay, stopped handling cargo at the gate. The strike is one line inside a risk report. The map behind it is a page of history.

Ten years ago this was another matter. The South Atlantic west coast was a scatter of shallow gates with one deep-water harbour at the bottom left of the continent, a set of river ports that fed boxes down on barges, and a trade everybody treated as a side street off the main road between Asia, Europe and North America. It is still a side street, and it still carries less than the North Pacific by a distance. What has changed is what the side street is carrying. Soy, pulp, beef, ore and refrigerated cargo for half a continent now come down through these gates, and the countries behind them have started to behave like one system rather than a row of neighbours who happen to share a coastline. That is the change that makes a labour action at a small port worth a long read.

The wait table published alongside the strike tells you where that system actually stands. Montevideo reports 2.1 days of average berth wait. Santos, the largest container port in Brazil, reports 3.1 days with high yard density already pressing on its operations. Paranaguá in Brazil and Buenos Aires in Argentina both sit at 2.2 days. Chile's San Antonio and Valparaíso were restricted from the fourteenth to the eighteenth of July, by weather rather than by men. The same fortnight produced a strike called at Barcelona for the twenty-second of July and a chemical leak at Antwerp on the fourteenth, which put most of the Belgian terminal complex back to work within a day while one of its largest terminals stayed restricted. Read the whole table and not the one line about Uruguay, and the surprise that lands first is that the worst wait in the Americas on that page belongs to Mobile, Alabama, at nine days. The South Atlantic is nowhere near the top of it.

Count the numbers before you reach for the adjectives. Workers at Montevideo stopped cargo handling on the fourth of July. Santos, the largest container port in South America, followed with a stoppage later in the same month. The waits in the report that logged both are 2.1 days at Montevideo, 3.1 days at Santos against high yard density, and 2.2 days at Paranaguá and at Buenos Aires. Those are not the numbers of a port in a squeeze. A port in a genuine squeeze sits at nine days, which is what Mobile reports in the Americas on the same page, or at 12.5 days, which is Beira, or at 14.3 days, which is Conakry.

What happened on the west coast of the South Atlantic in July was labour action against a background of ample berth. Almost every wrong conclusion about that month comes from reading it as congestion, and the wrong conclusion matters, because the tool you reach for depends entirely on which of the two problems you think you have.

Ask why it happened now and the answer is not the cargo. The agreements that govern work on these quays are renegotiated on cycles measured in years, and a round of talks that lands badly does not stay inside one terminal for long. Two gates in one fortnight is the shape of a regional rhythm rather than two pieces of bad luck, and the same fortnight shows the same pattern at other latitudes. A strike was called at Barcelona for the twenty-second of July. A chemical leak at Antwerp on the fourteenth put most of the Belgian complex back to work within a day while one of its largest terminals remained restricted.

Weather shut Chile's San Antonio and Valparaíso from the fourteenth to the eighteenth. Southern winter is also the roughest weather window on that coast, and in a normal year it arrives in the same weeks that the interior is pushing its heaviest export tonnage toward the water. A port is where two calendars meet, the one on the quay and the one in the field, and the middle of July put both of them under load at the same time.

Look at the node itself, because its size is not the point. Montevideo is Uruguay's only natural deep-water gateway, and a deep-water harbour on a coast of shallow ones is worth more than its own hinterland can pay for. It is also the sea-end of a river system that comes down through the interior, and that is the part the headline never mentions. The Paraná and the Paraguay bring cargo down from landlocked country and from the southern states of Brazil toward the estuary, and the estuary needs a gate that can take the ships which carry it away.

So Montevideo's job in the network is out of proportion to Uruguay's own trade. It is the outlet for goods that were grown, cut or mined somewhere else entirely, by people who get no vote in a dispute on its quay. A node like that is easy to underestimate precisely because its own country is small. Read the map by flag and you will miss it. Read the map by flow and it is the only deep artery on that shore.

Read the strike as a node event and the line of damage draws itself. A port is two legs and one gate. The water leg is the berth, the approach channel and the pilots. The land leg is the yard, the truck queue, the depot where an empty goes back, the barge berth, the rail spur. The gate is the seam between the two, and it is the one part of the system that a labour action can close completely while every other part of the map stays in working order. Ships do not stop arriving because a gate shuts.

They stop being able to plan, which is worse, because a plan is what a schedule is made of. A berth window slips a day, then two, and then the feeder that was supposed to take those boxes onward misses its own window at the next port down the coast. That is how a small node exports a delay it never intended to export, and it is why a five-day closure can surface as a two-week headache three ports away from the quay where it started.

Who pays for it. Start with the small and mid-sized exporters on the Uruguayan side, who have no volume to argue with and no second gate written into their commercial papers. Then the landlocked shippers upstream, for whom Montevideo or a Brazilian alternative is the only decision they ever get to make, and who pay for the disruption in days rather than in choices. Then the southern Brazilian volumes that treat the River Plate as a relief valve when Santos runs hot, and had the valve close for a few days at precisely the moment it was worth something. Then the feeder operators, whose ships are small, whose rotations are tight, and whose economics cannot swallow a two-day hole without telling the next customer a new date. Look at the shape of it: at 2.1 days of average wait, nobody is short of berth. What everybody is short of is certainty. Certainty is the first thing a strike removes and the last thing it gives back.

Time the thing, because the clock has a shape to it. In the first three days, berth windows slip and vessels bunch on the approach. Between day three and day ten, the yard fills, because the boxes that could not leave are still standing in it, and this is where Santos matters: 3.1 days of average wait with high yard density means the big node had less headroom than its name suggests. From week two to week six, the inland rhythm re-couples. Barges catch up, trucks rotate back to normal, and the depots behind the gate find their floor again. Here is the correction I would make to the ordinary reading of this news. The recovery clock is not set at the gate that shut. It is set by the yard density at the biggest node in the region, because that is where diverted cargo has to go while the small gate is closed. Watch Santos and you will know when Montevideo is finished, not the other way round.

And hold the two clocks apart, because the confusion between them is where most contingency plans fall over. The event clock is short: a gate opens or it does not, and you know inside a few days. The exposure clock is long: every box that could not sail this week has to sail in one of the next four, which means the month after a stoppage is busier than the month before it, on a quay that is still working through the backlog it just built. A plan that treats the reopening as the end of the problem will be short of space in August, and it will be short of space at exactly the moment when the whole region is trying to catch up at once. Write the recovery weeks into the schedule when you write the closure, not after it.

Now the part that the single-line version of this news leaves out, which is also the part that costs money. Everybody knows why you hold a second gateway: so that on the morning the first one shuts, you still have a door. The value of a second gate is not that it is cheaper. It is that it is open when the other one is not, which makes it an option rather than an alternative, and options are worth what they are worth on the bad day. A labour wave that reaches two ports in the same fortnight destroys exactly that value.

Assume you are the disciplined shipper holding twenty per cent of your River Plate volume pre-qualified on a second gateway. That discipline bought you nothing in the middle of July, because the second choice was shut too, and the option you paid to keep expired on the same morning as your primary. This is not an argument against holding a second gate. It is an argument for knowing which risk you are holding it against.

There is a second misreading, and it matters more for the decision than the first. Read the report as a whole and the American queue that is genuinely broken is Mobile, at nine days, four times the wait at Montevideo. The South Atlantic is not the region's problem this month. It is the region with spare berth and an argument about pay. That distinction is not academic, because the tools are different. Against congestion you reroute to a gateway with slack, you buy time with money, and the map hands you options. Against a labour action the map hands you nothing, because the whole point of a coordinated action is that every neighbouring gate is shut on the same morning. Bring a congestion tool to a labour problem and you will spend a week moving boxes from a port that is slow to a port that is also stopped, which is a very expensive way to stand still.

So do the arithmetic properly, and take every assumption here as something you are free to throw out. Assume a South Atlantic trader moving 240 boxes a month out of the River Plate basin, 150 of them loading at Montevideo and 90 at Santos. Assume the base waits are the ones in the report, 2.1 days and 3.1 days. Assume a labour action shuts a gate for five calendar days, that boxes already standing on the quay absorb the closure day for day, that boxes booked in the ten days after reopening take two extra days while the queue drains, and that one sixth of the month's volume is on the quay at any given moment, which is 25 boxes at Montevideo.

Those 25 boxes take 125 extra box-days of dwell. On the Santos stream, where yard density is already high, assume three extra days rather than two: 90 boxes at three days is 270 box-days. The month costs about 395 box-days. Spread across the 240 boxes in the book, that is roughly 1.6 extra days for every box you move, and it is the honest size of one strike. It is small, and I am not going to inflate it to make a better paragraph.

Put the same arithmetic against the thing the second gate is for. Assume you keep twenty per cent of the book, 48 boxes, positioned on a second gateway that is genuinely open. A five-day closure avoided on those boxes saves up to 240 box-days, which is more than half of the 395 that the month cost. So the option is worth more than the event it protects you against, which is exactly why disciplined shippers buy it and exactly why a coordinated action is so much more expensive than a single strike. Then take the report's whole wait table, because it draws the real structure of the map.

Montevideo at 2.1 days, Southampton 2.3, Paranaguá and Buenos Aires 2.2, Rotterdam and Valencia 1.8, Hamburg 1.6, Busan and Yokohama 1.5, Port Klang 1.4, Shanghai 3.2 with Ningbo at 2, Jeddah 3.8, Manila 4.7. Then the other cluster: Mombasa 3.6, Freetown 4.6, Tema 7.8, Abidjan 9.9, Beira 12.5, Conakry 14.3. Two groups and almost nothing in between. The world's ports are a handful of nodes that keep the trade moving and a long tail that never quite leaves the queue. In July, Montevideo was in the first group. That fact, not the strike, is the one worth ten years of attention.

So: pulse or turning point? On the numbers, a pulse. The waits say the water leg has slack, and a system with slack absorbs a five-day gate closure and has forgotten it by August. But the numbers are this fortnight's photograph, and the decade is being decided somewhere the waits cannot see. What the next ten years will settle on that coast is whether these quays automate, and on what terms with the people who work them; whether the river and the rail that reach the interior finally get the investment they have been promised since before anyone now working on those docks was born; and whether Santos stays a single dominant node or the region spreads its risk across Montevideo, Paranaguá, Itapoá and Buenos Aires. If the agreements keep landing, this fortnight was noise at an inconvenient latitude. If the renegotiation cycle starts producing a strike cycle, then the premium moves off the water and onto the calendar, and an exporter pays it every year whether or not a single vessel waits.

Take the other side, because a read without a boundary condition is not a read. If the real constraint on this coast were the navigability of the interior waterway rather than the gate at the sea end, the whole story would stand upside down, the ports would be the least of it, and the money would go to whoever can move a barge in a dry year. If Santos consolidates hard and wins the depth argument outright, Montevideo stays what it has always been, a small deep-water relief valve flying an outsized flag. And if these agreements keep settling quietly after a few days of noise, then this story will be written once a year, forgotten by the end of the month, and paid for by whichever exporter happened to be sitting on the quay on the fourth of July. The headline does not distinguish a three-day action from a three-week one. Your booking does.

Where the network grows next is not a mystery on this map. It comes down to two legs. The water leg is depth, and depth is geology plus capital, which is why the deep-water argument on this coast is a thirty-year argument rather than a five-year one. The land leg is rail and river, and that is where the region has been promising itself the same thing for a century without quite finishing it. The nodes that win the next decade will not be the ones with the newest quay cranes. They will be the ones whose cargo can leave the gate by three different routes, so that when one route shuts the cargo has somewhere else to be. That is what the report is really measuring underneath the strike, and it is why a single line about Uruguay is worth more than a single line.

For the rest of July the work is small and unglamorous. Have the gate status of every River Plate gateway you use checked daily, by one named person, and written down where the booking team can see it. Decide in advance which fifteen to twenty per cent of your volume can move to a second gateway, and check that the second gateway sits in a different labour market from the first rather than merely a different postcode; a second gate inside the same negotiation is not a second gate. Put a written threshold in the routing file: a gate reporting three days of closure or more releases the alternative routing without anyone having to convene a meeting. And ask the person who plans your sailings one plain question: when the next round of talks lands, which gate do I move to, and what does the inland leg behind it cost me in days. If nobody can answer that inside a week, the strike was never the problem.

Ports are patient in a way that nothing else in this trade is. A container is a metal box with a date on it. A ship is a mortgage with a propeller. A forwarder's year is a series of promises made on a telephone. The quay outlasts all three, and that is the whole reason a line about Montevideo on the fourth of July deserves more than a line. For a few days in July, a gate that has been doing this job longer than any company on this page has existed was, for the people who depend on it, an open question again. It shut, and then it opened, and the map did not move. The map only looked still because somebody was standing at the gate, keeping it.

↑ Back to the key points

— By Tomás Ferreira