Why steel tariffs are always bad

Not that I expect politicians to listen...

A lot of people love steel tariffs. George W. Bush wasn’t much of a protectionist, but even he slapped tariffs on foreign steel in 2002 (only to lift them in 2003). When Trump came into office, his first big tariffs were on steel and aluminum, well before he began the trade war with China. There’s just something about steel — a sense of power and strength. Joseph Stalin named himself after the metal. It also hearkens back to our history as a developing country, when Andrew Carnegie built what would eventually become U.S. Steel into an industrial titan. Steel is an essential input into practically every piece of heavy industry, which is one reason industrializing nations often strive to build up their steelmaking capacity. America was no exception. And there’s a romance about the steel mill — the tough working-class laborer straining in the hellish heat, braving death and injury. Given all this, it’s no wonder America is always trying to shield our steel industry from foreign competition.

The problem is, steel tariffs are just bad policy. It’s not just the bad optics — our import taxes hurt close allies like Canada and Japan, which hampers our ability to build strong and stable international coalitions. More importantly, it’s bad economics. Taxing foreign steel helps protect our own producers, but it hurts other sectors of U.S. manufacturing, especially autos and other heavy industry, ultimately weakening America’s industrial might. If we want to protect our steel industry, there are better ways to do it.

The theory

First, let’s talk a little bit about the economic theory here. There are two basic kinds of goods in the economy — final goods, which are things you actually consume like cars and iPhones and Tylenol, and intermediate goods, which are used to make the final goods. Intermediate goods include the chemicals used to make the Tylenol, the chips that go into your iPhone, and the steel and aluminum that go into your car.

In 1971, Peter Diamond and James Mirrlees wrote a very famous paper arguing that we should only tax final goods, not intermediate goods. The reason is that you want to produce as much stuff as you can before you tax it. Taxing intermediate goods just makes production less efficient — it reduces the amount of actually useful goods our society makes, even before we start taxing those useful goods. So this implies we should only tax final goods. Diamond-Mirrlees taxation is basically the idea behind a Value-Added Tax, which taxes only the amount of value that a company adds to the final good. Europe uses VATs, and America really should use them too.

So, steel. Steel is pretty much just an intermediate good. No one actually uses hunks of steel by themselves; we use them to make the stuff we actually want. Construction uses the biggest share, followed by heavy industries like autos and machinery. About 22% of the cost of a car is just the cost of buying the steel. Diamond-Mirrlees theory says we shouldn’t tax the steel that goes into a car. We should just tax the car itself, or have a value-added tax that works out to the same thing.

And it doesn’t matter that a tariff only applies to imports. Taxing foreign steel makes life harder for carmakers, even if they switch to domestic suppliers, because when we put a tax on imported steel, our own steelmakers raise their prices too. Remember that the number of people who have jobs in steel-using industries is about 80 times greater than the number of people who work in steelmaking itself.

So that’s the theory, but does it work in practice? The answer is yes.

The evidence

Why did Bush drop his steel tariffs in 2003, two years ahead of schedule? A business group called CITAC (Consuming Industries Trade Action Coalition) complained that higher steel prices were hurting their businesses and killing hundreds of thousands of jobs. This, on top of international anger, was enough to get Bush to drop the tariffs.

CITAC’s evidence of job loss wasn’t actually that solid. But now, with the benefit of hindsight, we can see that they were right.

A new working paper by James Lake and Ding Liu reexamines the impact of the Bush tariffs. The authors look at local labor markets in areas where industries use more steel, and compare them to places where industries use less steel. This allows them to isolate the impact of the tariffs on jobs. They find that the tariffs raise employment in steelmaking itself by a little bit, but hurt employment in steel-using industries by quite a lot. The effects seem to be long-lasting, too.

Steel tariffs also hurt overall U.S. manufacturing competitiveness. A new working paper by economics job market candidate Lydia Cox traces the impact of Bush’s tariffs, identifying the affected industries using a new data set of companies that asked for exemptions to Trump’s later tariffs. She finds:

A 1 percentage point increase in an industry’s upstream steel tariff rate causes a relative decline in the U.S. share of that industry’s world exports…Declines in the competitiveness of U.S. exports due to the tariffs are highly persistent—global market share remains depressed relative to pre-tariff levels for at least 8 years after the tariffs are lifted. Likely a result of this loss in market share, I also find that steel-intensive industries suffered persistent declines in employment in response to relatively high steel tariff rates.

In other words, taxing foreign steel does the exact opposite of what its proponents imagine. Instead of making American manufacturing great again, it weakens our manufacturing industries against their foreign competitors.

This is a high price to pay for our romantic attachment to the steel industry. Way too high.

The alternative

But, you may ask, what about periods of crisis? As we found in the recent pandemic, having supply chains that stretch overseas can be economically efficient in good times, but can cause major problems when international trade networks are disrupted. It would be very bad, for example, to find ourselves with no domestic steelmaking capacity in the event of a major war that disrupted our ability to import steel. So preserving some domestic capacity — even a substantial amount — might be worth it.

But tariffs are the wrong tool for this job. Even if it’s important to preserve domestic steelmaking capacity, that shouldn’t come at the expense of domestic auto-making capacity. In the event of a war, automakers would probably have to pivot toward making tanks and other military vehicles. If our domestic automakers are atrophied because they lost competitiveness due to years of steel tariffs, then we’d be on much weaker footing in case a major war arrived.

In fact, there’s a better tool for preserving domestic capacity in strategic industries: Subsidies. Government subsidies for domestically made steel increase the price that American steelmakers receive, and reduce the price that American automakers pay for steel. The subsidies are paid for via taxes, which are levied on workers in a broad range of industries, most of which presumably aren’t as strategically important as steel.

Subsidization is what we already do with the agriculture industry, to the tune of tens of billions of dollars a year. This ruffles surprisingly few feathers — some trade partners often grumble (though they generally do the same thing), but they put up with it, and few Americans even pause to think about farm subsidies. Americans enjoy cheaper food as a result, though their taxes are a little bit higher. And farm subsidies help make sure that the country could never be starved out in a blockade.

That doesn’t mean farm subsidies are worth it, of course — many economists view them as wasteful sops to politically powerful rural states — but if you think supporting strategic domestic industries is important, farm subsidies seem to provide a successful example of how to do it.

So there’s never a good reason to tax foreign steel. If you want to support U.S. steelmakers, use subsidies instead. Otherwise you’ll just hurt American manufacturing out of pure economic illiteracy.