Article — "Creative Coercion: Halved Aluminum Tariffs for Onshoring, Defense De-Risking, and a Section 338 Strike on Canada": Article — "Creative Coercion: Halved Aluminum Tariffs for Onshoring, Defense De-Risking, and a Section 338 Strike on Canada": The tell is not that Washington reached for a statute nobody has used since the 1920s. The tell is that it reached for the weakest of the three statutes it invoked that day—and did so on purpose.
On July 20, 2026, three legal instruments were signed within hours of each other. An aluminum incentive under Section 232. A defense-sourcing clampdown under 10 U.S.C. 4872. And a 50% tariff on nearly $20 billion in Canadian goods under Section 338 of the Tariff Act of 1930, an authority untouched for a century.
Most coverage treated these as three headlines. They are one machine with three levers.
Most coverage treated these as three headlines. They are one machine with three levers—and the interesting question is which lever actually bears load.
And the interesting question isn't the politics of the Canada strike. It's which lever actually bears load, and which one is legally hollow.
Start with the mechanism each authority is built to run.
Section 232 is national-security industrial policy with a paper trail. The aluminum proclamation doesn't just tax imports—it builds an incentive lattice. Commit to construct, refurbish, or expand U.S. primary aluminum capacity, with construction starting no later than January 20, 2029, and you may import primary aluminum at half the applicable Section 232 rate.
The catch is that the discount is metered to your own promised output. Import volume at the reduced rate corresponds to the facility's reasonably anticipated annual production. For refurbishments, the benefit is capped at the value of the investment.
This is a coercion instrument disguised as a subsidy. Commerce monitors compliance, can require audited reports proving you're meeting manufacturing commitments, and can rescind benefits entirely for failure. If a company committed fraud or deliberately misled the government, rescission runs retroactively—the Treasury collects the previously owed tariffs and adds fines.
Note what that architecture buys the administration. The leverage doesn't expire when the deal is signed; it persists through the audit cycle. You don't get the discount and walk away—you get it conditionally, revocably, and with a clawback hanging over you until the smelter is actually running.
The 4872 order works the same way from the demand side. Starting January 1, 2027, the secretary of war and military secretaries must generally stop issuing waivers that let defense contractors source sensitive materials from covered nations—China, Russia, North Korea, Iran—unless the contractor submits an approved mitigation plan showing a shift toward domestic or allied supply.
Contractors now have to map every tier of the supply chain, tracing components to raw-material origins, and demonstrate they made sufficient efforts to find American-made alternatives. Concealment can terminate the contract.
Two authorities, one logic: build the domestic base or lose access to the government's favor.
Then there's Section 338, and this is where the design gets interesting.
Section 338 empowers the President to impose duties of up to 50% to offset a foreign country's discrimination against U.S. commerce. But it targets a specific kind of discrimination. The statute is built to counter differential treatment—a partner treating U.S. goods less favorably than the goods of some third country.
That single word, differential, is the whole game. Section 338 is not a general retaliation license. It requires a comparator. Someone has to be getting the better deal that Americans are denied.
Why would you choose the statute that carries the highest evidentiary burden and hasn't been invoked in 100 years?
Because it does one thing the others can't. The Section 338 tariffs apply to all covered Canadian goods regardless of whether they originate under USMCA. That is the override.
This matters against the backdrop of what the administration actually wants from Canada—autos. The U.S. has reportedly been pushing a 50% U.S.-content demand and running the auto file as a content fight, per protoniccreations.com. USMCA rules of origin are precisely the thing that shields Canadian-origin autos from ordinary tariff pressure. Section 338 is the tool that reaches through the agreement to hit originating goods anyway.
So the override is the feature. The evidentiary burden is the price. And you can see the White House paying that price in the way it framed each grievance.
Watch the framing carefully, because it's contorted to fit the statute rather than the other way around.
On dairy, the complaint isn't that Canada taxes American cheese. It's that Canada's tariff-rate quotas on U.S. cheese are more restrictive than the TRQs it grants the EU—despite trade agreements with both. That's the differential the statute demands: EU favored, U.S. disfavored.
The framing has to be EU-versus-U.S. because "Canada protects its dairy market" isn't a Section 338 violation. "Canada treats Europe better than us" is.
The awkwardness shows through the seams. The U.S. runs an average annual dairy trade surplus with Canada of around $600 million since 2021. That is a strange fact to anchor a discrimination claim on.
Autos get the same treatment. The proclamation says Canada imposes tariffs and quotas on U.S.-made cars but not on cars from other countries, and administers the quotas so as to push U.S. automakers to invest in Canada rather than the U.S. The supporting number: Canadian imports of U.S. motor vehicles fell about 22%, or $5.6 billion, from April 2025 through March 2026, with other countries' exports rising to fill the gap.
Alcohol is the cleanest fit and the most legally exposed. The complaint is that all but two Canadian provinces and territories halted purchase, distribution, or retailing of U.S. alcohol without imposing similar restrictions on others—U.S. beverage imports down about 81%, or $582 million.
But it is provinces and territories, not the federal government, doing the halting. Whether provincial retail decisions constitute a foreign country's discrimination against U.S. commerce under a 1930 federal tariff statute is exactly the kind of question that gets litigated for years.
And here is the deeper problem, stated plainly in the record: two countries chose to retaliate against Trump's tariffs rather than negotiate—China and Canada. The alcohol bans and the auto quotas are, in Ambassador Greer's own framing, reportedly part of Canada's retaliation.
That's the contradiction sitting inside the package. Section 338 requires proven, pre-existing discrimination favoring a third country. But USTR describes the underlying conduct as retaliation against U.S. tariffs. Retaliation and third-country discrimination are not the same legal animal, and a court can tell the difference.
Which is why the enforcement asymmetry is the real story.
The 232 aluminum deal is durable coercion. It has a clean national-security pedigree, audited compliance, and retroactive clawback. Its leverage is contractual and self-enforcing—the government isn't asking a court to bless a finding; it's holding a discount hostage to performance.
The 4872 order is durable for the same reason. It reshapes procurement from the inside, waiver by waiver, without needing a courtroom to validate a novel legal theory.
The Section 338 tariffs rest on something far more fragile: a sudden finding of massive discrimination, on a statute dormant for a century, applied to override an existing trade agreement. Courts will be rightly skeptical that Canadian conduct in mid-2026 warrants anything resembling reimposition of blanket tariffs.
So the Canada strike is the legally exposed flank. It's the lever most likely to break in the hands that pulled it—not because 50% on wine, hockey sticks, and cement lacks bite, but because the theory holding those tariffs up is the thinnest of the three.
One more crack runs across the whole package. The 4872 clampdown assumes a domestic supply ramp is arriving—Navarro reportedly claimed the U.S. will have germanium refining capacity at three times its own demand by year-end, enough to export to allies.
Yet the 232 aluminum proclamation exists because Commerce found the U.S. still faces an insufficient supply of primary aluminum—the reason the incentive program was needed at all.
Same administration, same day, two opposite premises about whether the domestic base is ready. In one metal, we're a net exporter to the world. In another, we're bribing companies with half-price imports to promise smelters that won't break ground until 2029.
That gap tells you the machine was assembled fast, from mismatched parts. The durable levers—232 and 4872—will keep working regardless. The Canada lever is the one built to be shown, not necessarily to hold.
If any part of this unravels in court first, bet on the one nobody had touched in a hundred years.