Business & Economics

China’s Threats Make Taiwan’s Dominance of Chip-Making Risky for the Rest of the World

David Masci and Christine McDaniel talk to George Calhoun about Taiwan’s crucial role in semiconductor manufacturing and the need to diversify chip production

800 pound gorilla. A Taiwanese Semiconductor Manufacturing Corporation (TSMC) chip fabrication plant in Taichung, Taiwan. TSMC makes many of the world’s most advanced chips. Image Credit: Sam Yeh/ AFP

Chinese leader Xi Jinping made clear in his speech opening the country’s 20th Party Congress last week that Taiwan is firmly in China’s military crosshairs. Saying his country would “never promise to renounce the use of force” in order to bring Taiwan back under Beijing’s control, Xi reminded the world that China might be willing to go to war with the U.S. and the West to achieve “reunification” with the autonomous island.

But Taiwan is more than a major geopolitical hotspot. The island of 24 million people hosts the most advanced semiconductor fabrication plants in the world. Currently, 92% of the most advanced chips are made in Taiwan, and roughly two-thirds of all contract chip-making worldwide occurs on the island, which is only slightly larger in area than the state of Maryland.

Given that chips are integral to all but the simplest manufactured goods, the world’s reliance for semiconductors on an island situated less than 100 miles from a hostile superpower is nothing short of reckless, says George Calhoun, a professor at Stevens Institute of Technology. The world, particularly the West, must quickly redistribute semiconductor manufacturing capacity to avoid economic catastrophe if China were to succeed in subduing Taiwan, he says.

Calhoun comes to these views after a lifetime of experience in and around the technology sector, and not just as an academic. For 30 years, he worked in the wireless technology industry, co-founding and helping to run a number of public companies including Interdigital Communications Corporation, Illinois Superconductor Corporation and GeoTek Communications. At Stevens, where he’s been since 2003, Calhoun heads up the Quantitative Finance Program. He is also a regular contributor to Forbes magazine and was recently interviewed by Discourse on China’s chip-making capacity.

Calhoun sat down with Discourse editor-in-chief David Masci and Mercatus senior fellow Christine McDaniel to discuss Taiwan’s central role in manufacturing semiconductors and the implications of China’s threats, both to the semiconductor industry and to the world. To listen to the full, unedited interview, please click below.

Can China Build On Taiwan’s Expertise?

MCDANIEL: In the event of a geopolitical incident in the Taiwan Strait, what happens if Taiwan is subdued and forced to reunite with China?

If the factories survive whatever conflict might precede the event, would the Chinese be able to fully use them and then build on Taiwan’s expertise? Would the Taiwanese or others destroy the actual Taiwan Semiconductor Manufacturing Corporation (TSMC) plants before the Chinese got to them? The head of TSMC, Chair Mark Liu, has said that if China were to invade Taiwan, the most advanced chip factory in the world would be rendered “not operable.” What does he mean by that?

CALHOUN: I don’t think he means that they would blow it up. But I think what he means is that the people that run it today, that have the decades of experience in this very knowledge-intensive industry, a lot of them will not be available to a Chinese communist regime. They’d either leave or they would in various ways resist.

We can look at what’s happening in Hong Kong in the finance sector that’s slowly atrophying, or maybe not so slowly atrophying. And I think in another few years it’s going to be clear that the prominence of the Hong Kong financial sector has been lost as a result of being taken over by the regime in Beijing.

In the case of the semiconductor world, it’s so dependent on so many super high-quality bits of know-how, that if you lose a few of those and you can’t run it anymore. And I think that’s what he’s talking about. It’s the inoperability. Without speculating on military damage or anything like that, you’d lose the ability to continue to operate. TSMC does not just sit on their mountaintop, take orders and then ship product. They have a very intensive cooperative relationship with their customers, with sharing of intellectual property, with sharing of design, with cooperating in every step of the process.

If that breaks, again, it’s another pathway to inoperability. I guess I assume that China must know this. There’s a theory that this is the protective shield for Taiwan—that China, for all their bluster, must know that they couldn’t come in and take it over and run it and maintain it. They’ve made so many blunders lately that it could be another one in the making if they go that way.

The Taiwan Hedge

MASCI: Following up on Christine’s question: Have we put too many of our chip-making eggs in one basket? If Taiwan was subdued, even if the factories weren’t destroyed, but as Mr. Liu said, they could very easily make it impossible for them to duplicate what they’re doing here. The world would be in a very, very tough bind, because so many of the advanced chips are being made there.

This goes back to the market stresses that you were talking about in our first conversation about China and semiconductors.  You said that sometimes government has a positive role to play in this. Is this one of those stressors that our government and Western governments should be attending to?

CALHOUN: The short answer I would give is yes. We have put too many eggs in that basket. I think a lot of people have suddenly realized that. By a lot of people, I mean major customers like Apple, Foxconn and even Nvidia and AMD. I think the industry is realizing it has to hedge its exposure to this geopolitical risk in Taiwan. Even the Taiwanese recognize that. It’s going to take a few years, because this is an industry that doesn’t shift on a dime, as we know.

MASCI: You can’t just throw up a factory.

CALHOUN: Can’t just throw up a factory. There are questions wherever you’re going to place it. We see Apple talking about moving production to India. We see Foxconn talking about moving some of its production out of China. These are the signs of this Taiwan hedge beginning to develop. I think, if the time is granted here in the sense that mainland China doesn’t do anything for another three, four, five years, I think the industry is going to de-risk a lot of that in the coming half-decade.

MASCI: I also understand that manufacturing chips in Asia is still cheaper than it is in Europe or the United States. Will the hedge ultimately involve just shifting, like some of the apparel makers that have moved out of China, to places like Vietnam, Thailand or Malaysia? Or will it entail bringing a lot of this capability back to the U.S.?

And, as a follow on, you mentioned Apple. About 93% of their products are now assembled in China. Which means if China invades Taiwan next year, and all of a sudden there’s some embargo of China because of the military situation, Apple basically shuts down. What can they do?

CALHOUN: Assembly is easier to set up and duplicate than the fabrication of the chips, but no doubt Apple and a lot of people would be in deep trouble if China made the wrong military move at this point. I think that they’re worried about that. That’s why they’re now talking about moving production to India and moving production outside of China. I don’t think they’re going to talk about it too loudly because it’s a delicate relationship in China that they need to maintain.

George Calhoun

I think you’re going to see, over the next five years, that the industry is going to look a lot different in its geographical deployment. The signal has been sent and received that there are too many eggs in the Taiwanese basket, and so I think we will see the redeployment.

Whether they go to other parts of Asia, whether some of it comes back to the U.S., I think that’s where you have to break it down in terms of what kinds of chips you’re making. For instance, the industry structure is different in the memory sector than it is in the microprocessor or the graphic-processing units. There are so many differences in the structure of the ecosystem. There’s not going to be one single type of answer to that.

The common thread, I think, is going to be to start to move away from China. I think that’s happening even now because people have realized that we got too far out on that ledge with Taiwan.

Allocating Resources

MCDANIEL: Staying with that description of the chip ecosystem: Suppose you were given the lead on, say, a new office in one of the U.S. government agencies that will be charged with handing out or deciding who gets the $50 billion as well as the tax credits allocated in the CHIPS Act. Knowing what you know about the industry, who should get these funds? How should the government decide that?

CALHOUN: That’s a good question. I think a lot of people are going to argue over the best policy answers to that for probably a considerable period of time. I wouldn’t be surprised if some of the funds are held up and there’s some friction about where they should be spent. What they should do is harder to answer than what they probably will do.

What they probably will do is give subsidies to the major players in the industry today that will be connected to the goal of repatriation of manufacturing capability. I think that’s what is likely to happen. Hopefully, they won’t get too specific in terms of this kind of chip or this kind of technical goal. I think that’s where, at the end of the day, it will be simple enough for the policymakers to say, “Maybe we don’t understand the technological game board in detail, but we would like to emphasize domestic production, and so we’re going to subsidize that.” I think that’s what’s going to happen.

What should happen is probably a much more nuanced policy than the U.S. government is capable of deploying, to be honest with you. I don’t think the government should be in the business of trying to design technology pathways in this kind of industry.  Maybe the CHIPS Act was misconceived in some respects, but I think the money is going to flow, and it’s going to flow in support of domestic manufacturing, just as people have said. But what manufacturing and what market failures are they trying to fix?

Another area that I’ve talked about has been in the defense world. The needs of the defense sector for semiconductors of course are acute, but it’s also very specialized and very small-scale. I saw a statement in the press not long ago, that the Stinger missile is based on chips that no longer exist, that no longer can be procured. Replenishing the Stinger equipment that’s now been directed to Ukraine and elsewhere cannot be done because those chips don’t exist anymore. They were legacy chips. They were small-scale. No major chip manufacturer is going to look at that and say, “Gosh, I will build you a plant to make another 10,000 of these particular chips,” when I can build a plant that will make 100 million of a much more profitable chip.

These are the kinds of things that I think the government, through the CHIPS Act, could help to correct in the market system, and are things that probably should be highlighted a little bit more than they have been.

MCDANIEL: It sounds like you’re saying that money could be better spent on helping, say, the Pentagon in areas where there are no economies of scale. The market failure is, with economies of scale, not enough money goes into these areas because they can’t make enough profit. Is the CHIPS Act money necessarily targeted toward defense spending?

CALHOUN: As I read the bill there’s a little bit of it that’s targeted in that direction but a lot of it that’s not. I don’t know how these policy decisions are going to be made. They’re going to create several new government-sponsored research centers and institutes, and give them billions of dollars to come up with the policies. But we don’t know yet how that’s going to work out. Regarding some of the statements from the political side—the charitable thing to say would be is that they’re exaggerated. Probably the more accurate thing to say would be that they’re just wrong, and maybe willfully wrong in some respects.

As you say, economies of scale are going to favor where the private sector makes investments, and where they may disfavor it, the government might have a role to play putting its thumb on the scale a little bit—as we do with orphan drugs. The Orphan Drug Act, was all about this concept, that the major pharmaceutical companies pursuing their best opportunities would not have focused on some of the diseases that are quite rare and that need the research and development of new therapies. That’s what the orphan drug initiative is designed to address, and in a way, that’s what we’re talking about here. There are some orphan market segments in the semiconductor world. Defense is clearly one of them. I was shocked to learn that the Stinger is no longer manufacturable. They will redesign a new one, but then that adds time, and other risk factors.

Future Location of the Chip Industry

MASCI: You talked about how we’re definitely entering a period now where there’s been a rethink about overreliance on Taiwan. When you look ahead a decade—and of course, it’s impossible to know for sure, and there are so many variables like a possible invasion of Taiwan—what’s the most likely scenario in terms of the placement of the chip industry?

Will there be some movement to Europe and the United States, but will the locus of the chip-making world will still remain in Asia? Or will there be a new more diverse placement of the industry in all advanced or many advanced countries?

CALHOUN: I think there will be a correction back in the direction of not being quite so driven to find the absolute least-cost manufacturing location, especially when that location is in a country that is subject to political risks and geopolitical risks that those decision-makers traditionally haven’t really thought too much about. They’re private-sector guys and they’re looking at the labor costs in Vietnam versus the labor cost in Arizona, and perhaps not taking into account some of the risk factors that we’ve touched on here. But I don’t think it’s going to be a case that everything flows back to the United States.

MASCI: But things will be more evenly distributed.

CALHOUN: I think you’ll see more even distribution. I think there’ll be an additional question that will be asked around the table when these decisions are being made that wasn’t being asked before, that relates to the risk factors associated with supply chains that extend to areas of high political risk. I think that’s, again, where the Ukraine situation has come along and even underscored that. It’s made it clear that you can’t make assumptions about some parts of the world being as stable and rule-of-law governed as you might have been willing to assume for the sake of an extra 20% labor cost reduction.

MASCI: History has not quite ended, in other words.

CALHOUN: History is not over yet.

MASCI: Good to know. Just a quick follow-up to that. I know that sometimes people will say if you re-shore a factory a whole manufacturing ecosystem crops up around it. Will that also happen with chips, or is that industry different from a lot of other industries?

For example, Samsung, has talked about investing$100 billion here. Will that create a knock-on effect, and will a lot of other manufacturing then will crop up in the wake of that and other big investments? Or, again, is chip-making different from say, an auto maker or some other more traditional manufacturer?

CALHOUN: No, I think you’re right. In fact, I would say that’s one way to understand the industry today. Geographically, we do have an ecosystem here. We dominate the chip design. We dominate the electronic design automation, or the EDA segment. We dominate much of the semiconductor equipment manufacturing segment. I think the ecosystem is well established in the United States. We’re talking about rebalancing some of the parts of it, but we don’t have to start from scratch with that. We have many of the key cards.

To flip over to your first question, in China, they really have none of those things. None of them. They don’t have the semiconductor manufacturing equipment. They don’t have the software tools that are used. They don’t have the design industry. Even if they started with all the right policies tomorrow, which they are not going to do, it’s going to be decades to build up that capability, and that would be operating in a friendly global economy where they’re able to cooperate with the major Western customers.

Another example here is the kneecapping of Huawei, that took place when they were denied western-made semiconductors and other important technologies. Huawei basically has lost its mobile phone market even in China. They were a national champion and one of the national champions that was playing ball with Beijing. And for them to decline so dramatically when the of the technology access spigot is turned off, is I think another indication that China is not in a position to address this problem anytime soon.

MCDANIEL: I think that is a good place to end on.

MASCI: Sounds good to me. Thank you, Christine, for your excellent questions. And thank you, George, for coming and for your wonderfully insightful answers. We really appreciate it.

MCDANIEL: Thank you, George.

CALHOUN: Thank you. It has been fun.

Submit a Letter to the Editor
Submit your letter
Subscribe to our newsletter