Why Harris and Trump don’t want a Japanese company to buy US Steel

It’s hard to think of a more iconically American company than US Steel. It’s right there in the name.

But it’s more than that. The company emerged from Andrew Carnegie’s Pittsburgh-area steel mills, which produced one of the largest fortunes and were the site of one of the bloodiest and most brutal industrial strikes in US history. Its formation in 1901 was orchestrated by J.P. Morgan, who blended Carnegie Steel Company with several smaller concerns in one of the first large-scale corporate mergers the country had ever seen. At its wartime peak in 1943, the firm employed over 340,000 people. When the organized crime figure Meyer Lansky reputedly claimed in 1962 that the mafia was “bigger than US Steel,” he was making quite a brag.

Today, that wouldn’t be much of a brag. As of the end of last year, the company had fewer than 22,000 employees. In terms of employee headcount, the University of Pittsburgh is bigger than US Steel. It’s no longer even the largest American steel company (it’s third). And that’s not much competition — the US as a whole produces only 5 percent of the world’s steel, compared to nearly half in 1950.

So in the face of increasingly tough competition both domestic and foreign, US Steel started to look for a buyer. Late last year it found one in Nippon Steel, the largest steel manufacturer in Japan, which offered $14.9 billion for the company.

In many ways, it seemed like a natural fit. The world’s current leading steel producer, by a wide margin, is China, and just as a US-Japan alliance is the linchpin of efforts to contain China militarily, a US-Japan corporate merger could be a linchpin of efforts to contain China’s efforts to dominate the steel market. Letting a military rival control the production of such a crucial material (and such an important one for defense applications like warships and warplanes) comes with clear risks.

Except the deal now looks unlikely to go through. President Joe Biden, who came out in opposition to the deal in March, will reportedly announce soon he is blocking the acquisition on national security grounds. It’s in his power — the Committee on Foreign Investment in the United States (CFIUS, an interagency council controlled by Biden’s cabinet and other appointees) can veto mergers and acquisitions it deems dangerous to national security.

While CFIUS would need at least a fig leaf of a national security reason to act, security is not necessarily the reason why Biden wants to block a sale to Nippon Steel. His administration’s diplomats have reportedly told Japanese officials they need to kill the merger so Democrats win Pennsylvania this November. Kamala Harris and Donald Trump have both signaled opposition to the acquisition. US Steel is threatening to shut down multiple mills should the deal not go through, which could put thousands out of work, but the presidential candidates and Pennsylvania elected officials (who represent one of the plants likely to go) are holding firm against the deal.

According to reporting from the Washington Post, executives from Nippon Steel and US Steel were set to meet Wednesday with representatives from CFIUS in an effort to save the deal, but its chances look increasingly slim. Why? Some of it surely is the symbolism of “US Steel” being sold to “Nippon Steel,” which if included as a plot point in a late ’80s/early ’90s movie about the unstoppable economic rise of Japan, would come across as a little too on the nose. Unsurprisingly, Trump, whose form of nationalism has a distinct 1980s vintage, explains his opposition as motivated by a desire not to sell out to “Japan.”

But the bigger reason politicians are lining up against the deal is the United Steelworkers union (USW), which includes most of US Steel’s workforce among its 60,000 steelmaking members and strongly opposes it. Sen. John Fetterman (D-PA) candidly states he’ll oppose the deal as long as the union does.

At a time when a mere 6 percent of private sector workers are unionized, it’s an impressive flex of union strength. But it demands some further explanation. If the steelworkers are the biggest factor pushing against a deal, why are they so opposed?

To understand the union’s perspective, you have to know a little bit about how steel manufacturing works.

The traditional model, pioneered in plants like Carnegie’s in the late 19th century and still existing in a far more technologically advanced form today, is what’s called an “integrated” steel plant. The full process of steel production takes place in these facilities, going from raw iron mined from the earth to finished steel fit to use to make cars and household appliances and all the other steel mainstays of our lives.

Integrated mills start with iron ore and melt it into liquid pig iron in a blast furnace; the pig iron is then blasted with oxygen in a “basic oxygen furnace” to reduce its carbon content and make it a low-carbon steel; the steel is then poured into a mold and cooled into a solid; finally, it’s rolled into a finished product, like sheet or plate metal or beams for buildings or railway rails.

Integrated mills require a huge amount of energy to melt raw iron ore and to run the machines that transform it into steel and cast and roll it into useful form. The International Energy Agency estimates that a metric ton of steel produced this way requires 15 gigajoules of energy. (For comparison, the average American household uses 10,500 kilowatt hours, or about 38 gigajoules, of electricity each year.) This energy comes from burning purified coal, or “coke.”

Like any other form of coal power, this is pretty bad from an emissions standpoint. Globally, iron and steel account for about 7 percent of CO2 emissions. But integrated mills also employ huge numbers of people. The current largest mill in the US, the Gary Works in Gary, Indiana, owned by US Steel, employed over 30,000 people at its peak; after decades of decline and labor-saving technical improvements, it still has over 3,700 workers.

In the 1960s, an alternative form of steel production began to emerge: so-called “minimills” using “electric arc furnaces” (EAFs). As the name suggests, these furnaces are powered by running current through electrodes, rather than by burning coal onsite as at integrated mills. The electricity needs to be plentiful and reliable, but because it’s just electricity, it can come from sources cleaner than coal, like natural gas or nuclear or even (with sufficient reliability) renewables.

EAFs mostly recycle existing scrap steel, and when they do they require much less power per tonne than blast furnaces (only about 2 gigajoules, compared to 15 for the blast furnace route). They can also produce steel from iron ore that’s undergone a process called “direct reduction,” which normally uses natural gas and is thus cleaner than coal-based blast furnaces. Because direct reduction can also use pure hydrogen instead of natural gas, it’s a promising path for “green steel” with no carbon footprint at all. Currently, though, this is the costliest and most energy-intensive way to make steel (at least 18 gigajoules per tonne when using natural gas).

The term “minimill,” often used for facilities built around EAFs, should give you some clue as to the relative labor requirements of this kind of steel production. Minimills require fewer workers. The economists Allan Collard-Wexler and Jan De Loecker have found that the emergence of EAF-focused minimills was the principal driver of a massive increase in productivity for the US steel industry from 1962 to 2005. Over that period, output per worker grew fivefold: that is, a worker in 2005 could produce as much steel as five workers in 1962.

Not unrelatedly, steel employment fell by three-quarters over this period. About 81 percent of the surge in productivity, they estimate, was directly due either to minimills themselves being more productive, or to them driving the least-productive integrated steel mills out of business.

One can hardly blame the United Steelworkers union for being skeptical of minimills, given how the spread of the business model has decimated its membership. The problem is not just that minimills require fewer workers, but that they tend to be located in southern, anti-union states, with non-union labor.

There are a total of eight operational integrated steel mills in the US, all owned by US Steel or Cleveland-Cliffs; three are in Indiana, two in Ohio, one in Michigan, and one in Pennsylvania. The eighth, in Granite City, Illinois, idled its blast furnaces indefinitely late last year, though it continues to roll and finish steel slabs produced elsewhere. All eight of these facilities are unionized, six by the United Steelworkers.

By comparison, there are 88 electric arc furnace facilities in the US. While it’s hard to know what share are unionized, most are not; only about 23 percent of iron and steelworkers in the US overall are covered by a union contract, down from over half in 1983. Given that almost all workers in integrated mills are covered, it’s reasonable to surmise that the large majority of minimill workers aren’t in a union, making steel a majority non-union industry overall.

There are always exceptions, like a US Steel electric arc furnace facility in Alabama where workers are USW members, but for the most part, big integrated mills mean union power, and minimills with electric arc furnaces mean union decline. Nucor, the largest steel company in the US with over 25 million tons sold last year to US Steel’s 15.5 million, both pioneered minimills and is famously non-union. Even US Steel, long a center of union strength, acquired an Arkansas non-union electric arc furnace mill in 2021.

How does this tie into the Nippon Steel bid? Essentially, the steelworkers see Nippon as threatening to move US Steel toward minimill-type production and away from the conventional blast furnace/basic oxygen furnace integrated mills where the union is strongest.

In a letter to members, the union’s leadership expressed a worry that Nippon would “continue the [US Steel] business plan of running away from basic steelmaking—instead prioritizing Big River and Big River II.” “Basic steelmaking” here means integrated steel mills; Big River is the name of the Arkansas electric arc furnace plant that US Steel bought, and Big River II is a major expansion planned for it.

There’s some evidence that’s the case. Nippon has closed a few integrated mills in Japan in recent years and has proposed closing more. When the acquisition was announced, Nippon executive Takahiro Mori told investors he planned to continue the Big River II expansion and US Steel’s plans for it, which include moving production from northern union-represented mills down to the non-union workforce in Arkansas.

“The reality is that there are certain crucial products that simply cannot be made without blast furnaces, including those used in automotive, energy, and national security applications,” the union insists. They have a point. We can’t run the world economy on recycled scrap metal alone, and advanced high-strength steel (AHSS), needed for car manufacturing among other uses, tends to be made with blast furnaces, not electric arc furnaces, in part because scrap of high enough quality to make AHSS is rare. EAFs running on iron produced through direct reduction, not blast furnaces, may be able to make inroads here, but right now we need blast furnaces for cars.

There are other union concerns as well. The acquisition was announced without giving the union prior notice, which it claims violates the collective bargaining agreement reached between the union and US Steel.

Moreover, the union has another buyer in mind: Cleveland Cliffs, the No. 2 steel company in the US and the only other operator of traditional integrated mills. The company committed to the union that no union member would lose their job upon acquisition, and would continue to operate blast furnaces. Once again, the USW position emphasizes keeping traditional mills, with large union workforces, going.

However, Cleveland Cliffs only offered $7.3 billion, about half of Nippon’s $14.9 billion, for US Steel. It reportedly offered much more than that privately in response to the Nippon bid, but even then it didn’t match the Nippon offer. A Cleveland Cliffs purchase also raises major antitrust issues that would presumably bother the unusually antitrust-focused Biden administration. The Alliance for Automotive Innovation, the US auto manufacturers’ lobby, wrote to policymakers to express concern over one firm controlling 100 percent of US blast furnaces, and 65 to 90 percent of the steel used in vehicle manufacturing.

Industry press coverage of Cliffs notes quite candidly their strategy of trying to dominate blast furnace production so they can charge a higher price. In other contexts, that’s a kind of monopoly-oriented strategy that Biden appointees like Federal Trade Commission chair Lina Khan or Department of Justice antitrust chief Jonathan Kanter would normally object to.

The Biden administration has prided itself on supporting US-based manufacturing and unions, so it’s perhaps no surprise that it would throw in with the United Steelworkers. Then again, plenty of other concerns that the administration has would presumably prioritize a push in favor of the Nippon purchase. There’s the needs of Japanese allies, for one thing, and the desire to counter China’s “oversupply” of cheap steel. Japan is not the US, but it’s certainly not China, and strengthening an ally’s industrial firm makes sense.

Then there’s the environment. Some environmental groups have criticized the deal on the grounds that Nippon is committed to keeping high-emissions blast furnaces running — precisely the opposite conclusion of the steelworkers’ union. If the steelworkers are right, that’s probably good news for Nippon and US Steel’s carbon footprint.

As it stands, electric arc furnaces are far cleaner than blast furnace/basic oxygen steel production. Columbia researchers Zhiyuan Fan and Julio Friedmann describe blast furnace production as “particularly stubborn to any decarbonization technology”; electric arc furnace production from direct reduced iron or scrap metal is much more promising. Fan and Friedmann estimate that using green hydrogen in blast furnaces can reduce carbon emissions by perhaps 20 percent, but not much beyond that. Carbon capture at blast furnaces can further reduce emissions but remains imperfect and will likely never reach zero carbon.

The administration has defended high tariffs on Chinese steel (currently at least 50 percent, combining two different 25 percent tariffs) in part on the grounds that it’s emissions-intensive. But a major reason Chinese steel is so dirty is it’s made in coal-fired blast furnaces, at mills not dissimilar from those the United Steelworkers are fighting so hard to defend. And it, too, is moving toward more efficient electric arc furnaces.

All this points to a deep tension that neither Biden nor the rest of the Democratic Party seems likely to resolve anytime soon. The form of steel production that the United Steelworkers prizes and seeks passionately to preserve is likely not compatible with deep decarbonization. And the Democratic coalition cares deeply about decarbonizing the economy. In the current battle, the administration and party are siding with the union, but it’s not clear that the alliance can last indefinitely.