To a hammer, every problem looks like a nail. To U.S. Commerce Secretary Wilbur Ross—who owes much of his personal wealth to buying U.S. steel mills and selling them on—there’s no manufacturing problem that can’t be solved with metal tariffs.

President Donald Trump is considering actions to limit imports of steel and aluminum “in a very thoughtful and systematic way,” Ross told CNBC Thursday. Decisions on the curbs, which could include charges as high as 53 percent on steel imports, are due in April.

Unlike Ross in his previous career, The Trump Organization Inc.‘s investments in hotels and office towers mean it’s a consumer, rather than producer of the alloy. That ought to give Trump pause before taking action to increase the cost base of the family business.

To see why proposed tariffs on U.S. steel imports would be such bad news for the country’s manufacturing sector, consider how production has evolved in recent years. After collapsing during the 2008 financial crisis, output from America’s mills never fully recovered—in line with developments in Europe and Japan.

The natural response to reduced demand is to cut capacity as well. But, almost uniquely among major steel producers, the U.S. never did that.

Annual output capacity of 113.3 million metric tons in 2016 was a scant 100,000 tons less than it had been in 2007. The European Union trimmed about 14.1 million tons over the same period. As a result, while Europe’s steel capacity utilization has edged up in recent years, America’s has languished at the sub-80 percent levels where making a profit becomes a struggle.

Ross has tended to put the blame for this on the global steel trade, in particular China—but as Gadfly has argued previously, the U.S. isn’t a big user of steel from Asia’s biggest economy, which barely makes it into the list of top 10 import sources.

So why is America’s heartland suffering? The answer comes from looking not at steel’s producers, but its consumers.

Analysts typically initially examine the trade in crude steel to work out which countries are importers and which are exporters—but that’s a skewed way to look at the situation. After all, if Subaru Corp. buys a ton of steel from JFE Holdings Inc. for its Japanese factories and turns it into cars for sale in the U.S., the steel production and consumption is happening in Japan but the metal is ultimately being used on the far side of the Pacific.

Compare the World Steel Association’s figures for crude steel production to its estimates of true steel demand, and a striking picture emerges:

Almost every major steelmaking country produces more steel than it uses domestically because they’re manufacturing and exporting steel-intensive products in greater quantities than they’re importing them.

The U.S. is unique in showing the opposite pattern: Its trade deficit isn’t in coils and bars, but in the vehicles, machinery and construction products that are made out of such crude materials.

That helps explain why shares in Ford Motor Co., General Motors Co. and Caterpillar Inc. dipped last Friday after Ross outlined his tariff options, while those of Nucor Corp. and U.S. Steel Corp. soared.

In trying to help America’s metal manufacturers, Ross risks dealing a blow to the far bigger car and machinery industries that depend on their raw materials. If Trump is going to address this issue in a thoughtful and systematic way, he’d do well to reflect on that fact.

This column does not necessarily reflect the opinion of Bloomberg LP and its owners.