What the Tariff Actually Bought

American steel mills ran at 78.3 percent of capacity in the week to the third of October, producing 1,801,000 net tons. Output is up 5.3 percent so far this year. Imports are down 19.6 percent. Hot rolled coil sells for about 1,220 dollars a short ton against roughly 800 in March of 2024.

That is the record of the most aggressive trade protection the industry has had in a generation. The Section 232 tariff was doubled from 25 to 50 percent in June of last year, and in April of this year the rule was rewritten so that the duty applies to the full customs value of a product containing steel rather than only to the steel content. It worked exactly as described. Imports fell, domestic volume rose, the domestic price went to roughly fifty percent above its recent trough.

Employment in the sector is flat. The federal payroll series for iron and steel mills has sat at about 85,400 jobs across the entire escalation, from the first tariff through the doubling and through the customs rewrite. The volume is American, the price is American, and the headcount did not move.

The Historical Echo

This has been run as a controlled experiment before, with a published result.

In March of 2002 the United States imposed safeguard tariffs of 8 to 30 percent on a range of steel products under Section 201. They were scheduled to last three years. The World Trade Organization ruled against them on the eleventh of November 2003, the European Union prepared retaliation aimed at politically sensitive exports, and the measures were terminated on the fourth of December 2003, twenty one months into a three year programme.

The International Trade Commission later assessed the whole exercise and put the net welfare loss to the American economy at 41.6 million dollars. Steel producers gained. Steel consuming manufacturers, which employ many times more people than the mills do, lost more. The tariff was not a transfer from foreigners to Americans. It was a transfer from one set of Americans to a smaller set of Americans, with a deadweight loss on top.

The structural difference between then and now is worth holding onto, because it cuts against the pessimistic reading. In 2002 the American industry was heavily integrated blast furnace capacity with legacy pension obligations and no cost advantage. Today electric arc furnaces are 71.3 percent of national output. They are cheaper, faster to build, and they run on scrap and electricity rather than on coke and iron ore. The industry being protected now is a genuinely competitive one, which raises the question of what it is being protected from.

Where Patient Capital Is Positioning

Follow the capital commitments rather than the price, because the price can be undone by a signature and the concrete cannot.

Roughly twenty billion dollars of new domestic capacity has been announced since the escalation began. Hyundai and POSCO are putting 5.8 billion dollars into Louisiana. Nucor is building a four billion dollar sheet mill in West Virginia. The Japanese acquisition of the largest legacy American producer closed in June of last year at about 14.2 billion dollars, with eleven billion of committed investment and a golden share held by the government. That last structure is the tell: the state now has a veto inside a private steel company, which is a different arrangement than a tariff and much harder to reverse.

So the honest ledger reads like this. The tariff bought volume, price and investment announcements. It did not buy jobs, because the modern mill does not need them, and it has not yet bought a single ton of operating capacity, because the plants announced since 2024 will not produce until the late part of this decade. Everything arrives at once, into a protected price, from four or five simultaneous projects that were each justified by that price.

For patient capital the useful observation is that protection and capacity expansion are a sequence, not a state. The protection comes first, the capacity follows with a four to six year lag, and the two are then briefly in the market together. The 2002 episode ended early because of a trade ruling. This one will more likely end the ordinary way, with twenty billion dollars of new domestic supply arriving at seventy eight percent utilisation and finding out what the price is without a tariff holding it up. The capacity is the durable asset. The margin that justified building it is not.

The Capital Memo