The International Longshoreman’s Association (ILA) is on strike. The ILA is not America’s largest union, but it is among the most impactful for the simple reason that its workers dominate freight loading and unloading at 36 U.S. East and Gulf Coast ports. If they stop those ports from operating, Americans will face meaningfully higher prices, people will lose their jobs, and in a very general sense, the country will suffer for it.
The ILA’s strike is motivated, like all strikes, by demands for increased compensation. The ILA wants better healthcare coverage for its members, 61.5% (down from 77%) higher wages, continuation of touch fees, container royalties, and—most importantly—no port automation. That last demand is critical because the ILA is a Luddite union: port automation has made it possible to replace practically all of its members jobs.1
What We Have and What We’re Missing
America has terrible ports. This isn’t a judgment that’s based on a counterfactual where America tops the leaderboard and it should be better, this is a judgment that’s based on comparisons with existing ports in countries like Saudi Arabia, China, Qatar, Japan, and even backwaters like Tanzania and the Congo. America objectively has terrible ports. The reason America’s ports lag behind the rest of the developed and so much of the developing world so decisively is, in a word, unions. Nothing else is quite so key to America’s failure to automate its ports. What America misses out on because of those unions is nothing short of stunning.
The video below is basically Chinese propaganda, but not lies; what it portrays is bragging, bragging about the reality that China has automated ports and America mostly does not.2
Frankly, China deserves to brag. It should be depressing for Americans that a single Chinese port can outmatch the throughput of almost all American ports combined3, and that’s not the only port China has that they could use to make that claim.
To put the benefits of automation into perspective, consider if America also had automated ports and the increase in throughput those developments would generate. Broadly, America’s ports would be more efficient, they would have lower greenhouse gas emissions, reduced labor costs, and global trade volumes would increase and be able to increase greatly with benefits that would redound to everyone. The benefits of automation would be both quotidian and exceptional.
The daily and direct benefits of port improvements that every American would feel would be lower prices and faster delivery times. Automated ports work more quickly, they don’t get tired, and they work all the time, so the time it takes a shipment to get from anywhere to the U.S. will fall meaningfully. That means less waiting and fresher fruit. We can get some idea about the size of this impact by looking at transport ship willingness to pay data generated through analyzing the effects of port infrastructure improvements on bulk (>10,000 ton) carriers.
Those researchers noted that bulk carriers spent a third of their time at average ports simply waiting, and they estimated freight carriers would pay $0.90 per deadweight ton of their vessel—where the average vessel has a 50,000 deadweight tonnage—to cut those port times down by a single day. Just one day is worth $45,000 to the average ship, equal to a bit over three times the average $14,000 daily cost of operation! Because of this, 83% of shippers opted to go to ports that were further away from the final destinations of their goods because doing that meant they could spend less time at port. This means it takes longer for goods to go where they’re supposed to and as a result, America’s roads suffer more wear-and-tear, meaning more infrastructure building becomes necessary and your car takes more damage too.4
You should now be able to infer that, when ports invest in infrastructure that makes them operate faster, there are numerous diffuse benefits. In fact, the authors estimated, when one port expands it capacity by just one ship, it sees a staggering 42% additional trade volume and 4.1% less congestion, while all other ports in the U.S. see their trade go down by 0.19% and their congestion falls by 0.6%.5 In aggregate, through direct effects at ports that expanded and through spillovers at other ports, expanding port capacity by one ship increases trade by 0.67%, with aggregate welfare increasing by 0.5%, an effect equal to upping import volumes by roughly $16.75 billion in 2019 or about $25.5 billion in 2023, and that’s just one port getting one ship better. The benefit that’s actually on the table if every port is automated is equivalent to increasing port capacity by hundreds or thousands of ships, or even more, really.6
These impacts might seem unreasonably large, but considering how far behind America’s ports are, they’re actually not unrealistic. If America gets port automation underway and catches up to the efficiency frontier established in places like China and the Netherlands, it will promote growth, making Americans better off; it will make it possible for more Americans to have jobs, and it will help to sustain struggling regions of the U.S. for the foreseeable future.7 Beyond these benefits for daily life, the exceptional benefits to port automation are also notable, and everyone should be able to understand them through our recent, shared experiences during the pandemic.
Remember when the ports got backed up in 2021 and everyone couldn’t get their Amazon shipments? America’s supply chains were shocked, and ports couldn’t be cleared. As one paper’s authors noted, during the pandemic, wait times at some ports exploded from a few hours to two-to-three days, with congestion aggregately rising to 37% by mid-2021 and returning to normality two years later, in mid-2023. With automated ports? There never would’ve been that big a problem. Robots do not get sick and they do not demand time off, and they don’t care about vaccine mandates.8 In China, the robots’ overseers could also be rapidly relocated to working from home.
With port automation, pandemic-era shipping issues would have shifted from ports to truckers, and they would’ve been far more minor, not the major contributors to inflation and unemployment that they ended up being. Port automation can make these episodes more ephemeral and generally less likely going forward too, increasing robustness and preventing Americans from suffering as a result of unnecessary shipping crises.
How Do We Get Automated Ports?
You have to kill the ILA. To end this pernicious union9, take some advice from Curtis Yarvin: treat them like an enemy and pay them to retire. The ILA is an infinite grift machine historically run by mobsters and its participants are just trying to make money, so if you want the machine to break, you can buy them out.
The benefits to just paying off the longshoremen to leave are huge and the costs of doing so will probably be very small. One source estimated the strike would cost about $4 billion a day, and that’s one of the more conservative estimates I’ve seen. As the details above reveal, the benefits in the long run are much larger than that, and if I had to guess, paying the longshoremen to accept automation would cost a pittance by comparison.
With 45,000 members and assuming annual wages of $150,000 a piece, the government could afford to set up a pot of $20 billion pot to pay out about three years of wages over what people are making during the transition. It could even spend a bit more and help people to line up jobs for after they’re through working the ports. Up this to ten years of wages and it’s still worth it because automation is that good.
Worried the union won’t accept? Don’t be. The leadership is filled to the brim with mobster antics, and Mafiosi are famously short-sighted. You can pay them out, and there are many ways the deal could be structured to encourage the union membership to avoid strikes—make payouts contingent on avoiding breaches—and to accept the deal—sweeten the pot enough and unions do splinter in the face of deals. And if this really won’t work—which I doubt—, then you might have to do a round-about attack on the union’s integrity through at least prosecuting the criminal activity of the leadership and, even better, making it possible for a major company like Amazon to build a port and dredge a harbor with a targeted series of land use reforms that would disempower the union, paving the way for automation.10
Since I think a deal is possible, I want to mention one interesting proposal I saw Yale economics professor Jason Abaluck suggest on Twitter. You could set up a dockworker’s annuity fund that would pay salaries out of a fixed pot over 20 years, encouraging the union to dissolve and become disunified, because the fewer workers that remain, the larger the individual payouts would be. The leadership also has every reason to accept this in the first place if they believe the deal will be honored, because it means they could potentially get enormous payouts, well exceeding the size of anything they could ever earn otherwise.
Because the cost of peacefully eliminating the unions is likely to be small relative to the potential wealth there is to gain, we should just pay off the unions now and be done with them forever.
They are a union of the ‘horses against cars’ variety.
Want a less annoying video that makes the same point? Here’s one from Guangzhou:
The video is wrong that the port outmatches all of America’s ports combined, but it’s not that far off, and if you include the throughput from another Chinese port—like Guangzhou, Ningbo-Zhoushan, or Qingdao—into the equation, then America is outmatched by just two ports.
This inefficiency creates ample opportunities for criminals to skim money off the top, too.
Incidentally, the places estimated to get the biggest bang for their buck by expanding capacity by a single ship were located primarily in the Gulf Coast, which the ILA’s ongoing strike is currently holding up.
There will of course be diminishing returns: port improvements will not continue pushing up trade volumes by the absurd amounts they do right now after all the major ports have been brought to frontier levels of productivity.
Here are some case-studies to support my point.
Remember Obama’s tire tariffs? He hyped them up as creating jobs for Americans, but due to diffuse impacts, they cost the U.S. jobs. Remember Trump’s steel tariffs? They might’ve saved a few thousand jobs in the steel industry and prevented some jobs from going overseas, but they destroyed tens of thousands of jobs in the process. Tariffs, more generally, tend to decrease American export volumes, contrary to their intended purpose. And the Rust Belt? It declined in part due to a lack of competition, rather than due to foreign competition .
Incidentally, one of the frequently-touted benefits of port automation is that it reduces rates of accidents, making both cargo and workers safer.
To see its harms, just look at how pathetic growth in water-borne shipping labor productivity has been. It’s recently worse than famously-bad transit productivity!
Incidentally, disempowering the ILA is like permitting reform, because they are a barrier to the introduction of productivity-enhancing technologies.
Hard agree. Just pay them off.
I was thinking the same thing today. The ILA is an absurd aberration on the face of American economic efficiency.