Free Trade

America's Semiconductor Industry Doesn't Need $52 Billion in New Subsidies To Stay Ahead of China

Industrial policy is the wrong answer to a problem that mostly doesn't exist.


Spooked by a recent but temporary shortage of semiconductors and by China's plans to spend the equivalent of about $150 billion to bolster its own computer chip industry, the Senate is now considering a proposal to throw $52 billion in new subsidies at American semiconductor manufacturers to spur domestic production.

It's an idea that dovetails nicely with President Joe Biden's pivot toward China as the post-pandemic villain that will justify future expansions of government, as well as with the emerging nationalist economics and anti-China sentiment on the political right.

But what it really amounts to is a massive handout to a successful industry that doesn't need government aid, delivered under the guise of a national security argument that doesn't stand up to scrutiny. Rather than countering a perceived threat from China, lawmakers risk bogging down one of the most innovative and successful parts of the American economy with an industrial policy that will force chipmakers to care more about what makes Washington happy than what is best for their own businesses.

That perceived competition with China is fundamental to the United States Innovation and Competition Act of 2021, an omnibus bill that encompasses Senate Majority Leader Chuck Schumer's (D–N.Y.) Endless Frontier Act and several other proposals from lawmakers on both sides of the aisle. The very first page of a fact sheet released last week by Senate Democrats points out that America's share of global semiconductor manufacturing has fallen from 37 percent in 1990 to just 12 percent last year. Government action is necessary "to preserve our competitive edge," the document argues, before going on to warn that nothing less than America's "economic and national security" could be at risk if Congress doesn't hand over $52 billion in taxpayer cash to a handful of successful, deep-pocketed chipmaking companies.

Companies that, by the way, admit they don't need the cash to be competitive. Intel, one of the world's biggest chipmaking companies, is in the process of building a $20 billion fabrication facility in Arizona. In March, CEO Pat Gelsinger said the project "would not depend on a penny of government support or state support." (Though he immediately followed that comment by saying that "of course…we want incentives" and it appears that Congress is prepared to dutifully provide them.)

If there is some sort of problem with American semiconductor manufacturing—and there's really not, as we'll see shortly—it certainly isn't a lack of money. The New York Times reported earlier this month that equity investors have "plowed more than $12 billion into 407 chip-related companies" during 2020, which is more than double what they invested in 2019. Revenue for global chip manufacturers was up 10 percent in 2020, despite a pandemic-induced slowdown in demand, the Times reported, and NXP Semiconductors, which makes chips for automobiles and industrial equipment, saw revenue climb by 27 percent.

Those numbers don't suggest an industry in dire need of government aid.

Concerns about America's share of global semiconductor manufacturing are similarly misplaced. According to the Semiconductor Industry Association, a trade group, American-based firms control 47 percent of the global share of the semiconductor industry—a far cry from congressional concerns about the U.S. losing its competitive edge.

(Source: Semiconductor Industry Association

The trick that lawmakers are trying to pull here is to focus on where semiconductors are made. But this doesn't really matter. It's true that a smaller share is manufactured in the U.S. today than 30 years ago, but that's the result of natural shifts in the market, not evidence of a collapse in American technological prowess.

Indeed, American companies are still at the forefront of semiconductor development—earlier this month, American-based IBM announced a breakthrough in the development of the world's first two-nanometer chip.

"It is true that America has slipped to a 12 percent market share in semiconductor manufacturing, but it doesn't follow that firms need government help not to slip further," wrote T.J. Rodgers, CEO of Cypress Semiconductor Corporation and a former chairman of the Semiconductor Industry Association, in The Wall Street Journal this week. "Around 60 years after the commercialization of the integrated circuit, most chips have become commodities with little strategic value, and their manufacturing has been pushed offshore by relentless demand for lower cost."

Ah, but what about the recent supply chain issues that left automakers and tech companies without access to the semiconductors they need? "It is not an exaggeration to say at the moment that we have a crisis in our supply chain," Commerce Secretary Gina Raimondo told the Senate Appropriations Committee in April.

Except, well, it kind of is. It takes a long time to make semiconductors—up to 26 weeks, in some cases—and production is still ramping up again in the wake of last year's disruptive pandemic. This isn't a nationalist issue in which some evil foreigners are cutting off America's share of semiconductors, but a market-based issue that will be resolved as chipmakers increase production capacity to catch up to increasing demand.

But what about China? Yes, the Chinese government is investing heavily in semiconductor-making technology, but it remains far behind America in terms of technological know-how. A recent Nikkei report shows that China mostly manufactures nothing smaller than 14-nanometer chips, which are several generations behind the most advanced chips being made elsewhere—remember, IBM just announced plans for a two-nanometer chip. Closing that gap will be difficult now that America has banned the sale of semiconductor-manufacturing equipment to China (and enforced that ban even when the sale involved other countries).

If there is one major worry for the global supply chain of semiconductors, it is the island of Taiwan. The majority of the world's semiconductors are made in Taiwan, which is home to the Taiwan Semiconductor Manufacturing Company, by far the world's largest chipmaker. There are obviously many complicated geopolitical issues involving Taiwan that America and the world's semiconductor industry will have to navigate in the coming years—but it is downright foolish to believe that $52 billion in subsidies will make a meaningful impact in that complex situation, or in a global market that was worth $425 billion last year alone.

No, the United States Innovation and Competition Act of 2021 won't meaningfully change the fact that most of the world's semiconductors will continue to be produced in one of the world's biggest geopolitical hot spots. It won't do anything to reconfigure global supply chains that major chipmakers and semiconductor consumers aren't already doing—probably by copying the strategies that Toyota used to successfully weather the recent semiconductor shortage. It won't provide a needed boost to American companies' research and development efforts, which are already running at an all-time high. And it won't do much to stop companies from trying to find the cheapest places to manufacture their goods, whether those goods are T-shirts or the world's most advanced computer chips.

All it will do is shovel $52 billion of taxpayer money (some of it probably borrowed from China, ironically enough) to successful businesses flush with cash.

"We must invest in R&D, innovation, and manufacturing to ensure the U.S. continues to lead the world in science and technology," says Schumer. But private companies are already doing that in record amounts, and America is the world's leader in science and technology, no matter what the China hawks might want you to believe.

The semiconductor industry doesn't need a $52 billion stimulus package to keep doing what it is doing. It mostly just needs the federal government to stay out of the way.