Ricochet is the best place on the internet to discuss the issues of the day, either through commenting on posts or writing your own for our active and dynamic community in a fully moderated environment. In addition, the Ricochet Audio Network offers over 50 original podcasts with new episodes released every day.
Questions About the Semiconductor Bill
I just saw a story from Reuters (here) that a “sweeping semiconductor industry bill” has advanced in the Senate, and could be passed by both the Senate and House within the week. The story reports that the bill “provides about $52 billion in government subsidies for U.S. semiconductor production as well as an investment tax credit for chip plants estimated to be worth $24 billion.”
The stated justification, in the Reuters story, is “to make the domestic industry more competitive with China.”
Are any of you sufficiently knowledgeable about the semiconductor industry to be able to provide a helpful explanation about this bill, or about the industry in general?
My initial hypothesis is that this bill may be a wise measure, though costly, aimed at returning semiconductor production to the US. It appears that the main competition, however, is not China. At least, not the People’s Republic of China, but rather its breakaway province, Taiwan.
My general impression, probably from a podcast by Peter Zeihan, was that Taiwan is the leading semiconductor manufacturer. I found a March 2021 story from CNBC (here) setting forth market share for “semiconductor contract manufacturers” as follows:
- 63% – Taiwan
- 18% – South Korea
- 6% – China
- 13% – All others combined
I’m not sure whether this includes all semiconductor manufacturing worldwide, or only a portion performed by “contract manufacturers.” My suspicion is that other companies might manufacture semiconductors for their own use, and that those might not be included in these figures. I’m not sure about this, so I’d appreciate any additional information.
My thought about the semiconductor bill advancing in Congress is that its primary purpose is to reduce our dependence on semiconductor imports from Taiwan, not China. Again, this strikes me as a good idea.
If I am correct about this, though, I wonder if there is an ulterior motive. If I were a top US government official, and I was concerned that Taiwan might fall to China — or just be attacked and wrecked by China — I would want to take action to reduce our economic exposure to such an event. On the other hand, I probably wouldn’t want to say that this was the purpose of the bill.
For my part, I wouldn’t object to this possibility. While I wish the Taiwanese well, I’m becoming increasingly skeptical of the wisdom of defending the island from the mainland Chinese.
What do you all think?
Published in Foreign Policy
We are not that far behind. Intel makes their 14nm in the US. I think they might now be making their 10nm in the US. However, that is recent I think. But TSCM keep getting further and further ahead of all their competitors. They are just such a well-run company everyone else just can’t compete on cutting-edge manufacturing. I am not sure how far 52 billion in the bill will go when TSCM is spending over 40 billion in just capital this year. Let alone their R&D expenditures. There new Giga-Fab which is about three times the size of a football stadium, and will be making the next generation of chips is costing them over 20 billion alone.
Semiconductor company executives talk openly about their efforts to “reduce geographic risk”. On the surface, this could mean reducing manufacturing exposure to natural events like earthquakes or tsunamis, but in reality it’s mostly just code for reducing exposure to Chinese bad behavior, especially re Taiwan. It’s not for nothing that Taiwan Semiconductor is building a new advanced fab in Arizona.
Semiconductors should, and do, involve a different national strategic calculus than other kinds of products. Our national security is, in some ways, dependent on our differentiated computational capacity. There are longstanding (but half-hearted IMO) efforts to manage access to advanced designs and manufacturing technologies by our adversaries.
The lamentable enthusiasm for “off shoring” of the last few years has, predictably, gone too far and a lot of the hyperventilating by congress of late is an effort to correct for their own dereliction of the past several years. But, of course, it is important for them to lard up any bill with excesses and opportunities for graft. Hence, all of the barnacles you see on the semiconductor bill.
In short, the bill is, in principle, a good idea but the relentless stupidity of congress has turned it into a dog’s breakfast.
My own take is that we have a much larger problem, and one not often enough discussed. The cultural shift in “American” technology companies over the last 40 years is such that they are American in name only. Even (especially?) our most strategic technology companies are staffed by people from widely different cultures (diversity!), many of whom have no particular loyalty to the United States. There is also a thoroughgoing naïveté that afflicts the American techno/scientific community regarding their “colleagues” from adversarial geographies. So the semiconductor bill is fine I suppose, as far as it goes. But it doesn’t address (as far as I know) the more existential concerns regarding just who it is that is actually designing the technologies on which our national security depends.
Ban the importation of semiconductors, pharmaceuticals and weaponry. Give the market 18 months and we’ll be making everything we need.
I doubt if you could get most of the required production equipment within 18 months, let along construct the facilities, design the process flows, hire the people, and train them.
Sounds okay to me. Let the companies do it on their own dime, because they understand the threat of China, not because they’re getting slush fund money.
From Nikkei Asia today: supply chain delays for key chip-manufacturing components. Leadtimes are in months:
This is one angle I have not considered . . .
Excellent point, David.
It is a complex problem. Defenders of individual responsibility and freedom cannot come close to a rational policy by considering only markets, as I did in my first criticism (to make sure it got covered before we addressed the national security side. (Or maybe it’s because I wasn’t making the effort to read and think carefully.)
But as you point out, it is equally irrational to consider only defending the homeland, and single out one industry as security-critical, and suppose that we must ignore the benefits of markets for that one case.
What the forces of freedom and our nation must do is simultaneously (a) attack the domestic enemy—interventionism, including mercantilism, which tries to progressively undermine markets and thus human rights—and defend against the foreign enemy, which requires compromising individual human right.
A big part of the recent chip shortage is a shortage of relatively low tech commoditized chips. Many of these may be 20 or 30 year old designs made on similarly old, if not older, equipment. Much of the equipment is out of production and would be tremendously unprofitable to put back in production and use.
This is one of the reasons why Tesla was in much better shape than the legacy automotive manufacturers. If Ford designed a car in 2018, it likely went to the proverbial “parts bin“ and used things like HVAC control components that have been around since the 1990s, using 1990s chips. Tesla, on the other hand, would have likely designed the HVAC control from scratch using modern, and often custom, chips.
It’s a protection scheme much more insidious than anything the anti-tariff crew accuses tariffs of. The regulations encumber industry while the carefully tailored subsidies only benefit a small few, creating protection of businesses instead of industries. Our protection schemes should apply across the entire industry. Not one company.
Only if the State interfered with markets and artificially kept prices too low to make a profit.