This episode of Hub Dialogues features host Sean Speer in conversation with Chris Miller, a professor of international history at Tufts University, about his must-read new book, Chip War: The Fight for the World’s Most Critical Technology.
They discuss why semiconductors are so important to the global economy, why Taiwan has come to dominate chip manufacturing, and how the global powers are trying to secure their own supply.
You can listen to this episode of Hub Dialogues on Acast, Amazon, Apple, Google, Spotify, or YouTube. The episodes are generously supported by The Ira Gluskin And Maxine Granovsky Gluskin Charitable Foundation.
SEAN SPEER: Welcome to Hub Dialogues. I’m your host, Sean Speer, editor-at-large at The Hub. I’m honored to be joined today by Chris Miller, a professor of international history at Tufts University and author of the must-read new book, Chip War: The Fight for the World’s Most Critical Technology. The book, which has received tremendous praise from the likes of Larry Summers, Daniel Yergin, and Niall Ferguson, tells the geopolitical history of the computer chip and its growing economic and strategic importance in today’s era of geoeconomics.
I’m grateful to speak with him about the book, including the interrelationship between our economic needs for chips and broader military and security considerations. Chris, thanks for joining us at Hub Dialogues, and congratulations on the book.
CHRIS MILLER: Well, thanks for having me.
SEAN SPEER: Let’s start with some basic facts. What’s the difference between microchips and semiconductors? They seem to be treated as interchangeable terms in a lot of popular commentary. Are they?
CHRIS MILLER: In common usage, you can use them interchangeably. If you dig into the science behind semiconductors, a semiconductor refers to a type of material that’s neither a conductor nor an insulator, so conductors like copper that let current flow through them and insulators like glass block current and semiconductor materials like silicon, out of which most chips are made, have unique properties that they let current flow in certain situations and not others since they can be turned on and off in terms of their conductivity. That is the feature that makes them so useful for computing because you can turn them on and off very rapidly.
That on and off creates circuits. When the circuits are on, they produce a one. When they’re off, they produce a zero, and that’s how we create all of the ones and zeros undergirding digital computing.
SEAN SPEER: What are some practical uses for microchips or semiconductors in today’s economy?
CHRIS MILLER: Well, they are really used all over the economy, and a typical person will touch dozens if not hundreds of chips every single day. Even though you don’t see a chip very often unless you’re in the electrical engineering business. In your car, there will be dozens if not cases, hundreds of chips.
Your PC and your smartphone will have many chips inside of them, but also, your dishwasher, your coffee maker, and microwave. Almost anything with an on-off switch today has a chip inside, sometimes very basic chips that can cost a couple of cents, sometimes very complex chips that can cost hundreds of dollars. You really can’t imagine modern life or the modern economy functioning without a whole lot of chips.
SEAN SPEER: These are what you might describe as dual-purpose inputs. What is their strategic importance? I think you cite the Persian Gulf War as a pivotal moment for the role of microchips in modern military technologies. Why, in other words, Chris, is China so determined to domesticate a chip production capacity? Presumably, it’s not just for iPhones.
CHRIS MILLER: That’s right. It’s partly about reducing China’s dependence on imported technology. Today, China spends as much money importing chips as it does importing oil. Chips are one of China’s biggest imported goods. They’re largely purchased from China’s geopolitical rivals, Taiwan, South Korea, Japan, and United States. It’s a real vulnerability for China that it has to be so reliant on potentially adversarial countries for its chips. China’s focused on this not only because of the economic ramifications and technological ramifications but also because of the military ramifications.
When you listen to both military officials in Beijing and also defence planners in Washington, what you’ll hear is in their visions of the future of military power, there’s an extraordinary emphasis placed on computing and sensing capabilities, and that’s all about semiconductors. It’s no exaggeration to say that whoever’s got access to the most advanced chips in 10 years time will have a major advantage in terms of producing the most capable military systems.
SEAN SPEER: One of the most striking insights from the book is that the vast majority of global chip production is concentrated in one firm in Taiwan. You make the observation that this isn’t a new development. You explain it’s been decades in the making. Help our listeners understand how Taiwan, in general, and the Taiwan Semiconductor Manufacturing company, in particular, have come to be so dominant in this field.
CHRIS MILLER: Well, it wasn’t so long ago when Taiwan was a really small player in the chip industry, but in 1987, a businessman named Morris Chang founded, with a lot of help from the Taiwanese government, a company called the Taiwan Semiconductor Manufacturing Company, TSMC, which today is the world’s largest chipmaker. He had a unique business model in mind that proved very successful over the subsequent three decades. Rather than both designing and producing chips in-house, which at the time most companies did, he let companies design chips on their own and take the design to his facilities for manufacturing.
What that let him do was manufacture a lot more chips than all of his competitors because he was manufacturing for many different companies and many different use cases. As he manufactured more chips, he reaped enormous economies of scale, which let him hone his technological processes and also drive down the cost per chip. This business model innovation explains much of the reason why Taiwan is the world’s biggest chip maker. If you actually break apart Taiwan’s chip industry, it’s largely TSMC, this one single company that explained why Taiwan is such a large player in the chip industry on the world stage.
SEAN SPEER: Another insight from the book is that the recent story about microchips and semiconductors isn’t really about a pandemic-induced supply constraint. It’s actually demand outstripping supply, and in fact, this development itself isn’t even that unique. There have been various cases over the past number of decades where we face supply-demand disequilibrium. Why did the United States and others let this underlying issue persist? Why, in other words, Chris, is this most recent instance producing a policy action when previous ones did not?
CHRIS MILLER: Well, this chip industry has always been quite cyclical. When the economy is doing well, people buy a lot of chips, and when it slows down, people cut orders for chips. In the past, this largely impacted just the tech sector because it was, in the past, largely just the tech sector that was buying a lot of chips. Over the past couple of decades, and really in the past couple of years, the process of other use cases for chips growing in importance has accelerated. The number of chips in a car, for example, is increasing at a really rapid rate.
Ten years ago, had there had been a chip shortage, it would have been much less impactful for the typical automaker because they used fewer chips and simpler chips in their cars. Whereas now, there’s a larger number of chips in a new car and a more diverse set of uses for them and so it’s harder for chipmakers to manage auto supply demand as a result. That’s true in auto, as that’s true in industrial machinery. That’s true in a whole different range of equipment.
In the past, if you run a manufacturing company like making cars, for example, you don’t have to think that hard about your chip procurement processes and so you didn’t really spend as much time on those processes as you ought to have. Today, you’re now catching up with the fact that you can’t produce cars unless you have all of the chips that you need. You’re often buying chips from many different companies, and many different countries, and the chips aren’t interchangeable. You have certain chips for certain purposes and certain other ones.
The auto industry, in particular, I think, has had to catch up to the reality that their procurement processes hadn’t taken into account the number and diversity of chips they’re now buying.
SEAN SPEER: In August, the U.S. Congress passed the bipartisan legislation, the CHIPS Act, to try to build semiconductor production capacity in the American economy. It represents what you might call industrial policy or a traditional political economy approach to economic policymaking. Two questions. First, what do you think of the legislation? Second, is there a reason to think that the major public investment, something like $40 billion over the next five years, will indeed cultivate greater domestic production in the United States?
CHRIS MILLER: I think it’s pretty clear there will be more production in the U.S. relative to what they would have been without the CHIPS Act, but that’s actually fairly easy to accomplish. If you pay people to build facilities, they’re more likely to build them. That’s not so hard. I think that the harder thing to accomplish, and we’ll have to see if the Act does accomplish this, is not only funding new facilities in the short to medium term but then using those facilities to catalyze more innovation and more sustainable businesses in the long run. That’s where I think there’s much more uncertainty about whether the government is going to achieve these goals.
When I think about the tools the government has, there’s been a lot of focus placed on the incentive side providing tax credits and government subsidies for building facilities, but actually, a third or maybe a quarter of the funds of the CHIPS Act actually go not towards manufacturing incentives but towards R&D. The goal is to catalyze more innovation, to turn more laboratory innovations into commercial products. That’s where I think a lot of the really difficult work is going to be because unlike in subsidized manufacturing where there’s a set number of firms that could plausibly apply for funding, everyone knows how to build a chip-making facility. That’s straightforward, handing out money to firms.
It’s a lot more difficult to build an R&D infrastructure that improves R&D outcomes. I think that’s actually where we ought to be putting our focus both in terms of making sure the government is doing a good job at it and also looking to see what are the interesting results that emerge from this because if you look at the incentive side, we know what success will look like. Success will look like the U.S., just a bit more manufacturing on shore, but it’s the same type of manufacturing that would’ve happened offshore. Whereas on the R&D, the best case scenario is that there are new technologies invented that we struggle to conceptualize today. That’s a much more difficult thing to accomplish.
SEAN SPEER: What about China? Can you talk a bit about its own efforts to cultivate a domestic supply of microchips and semiconductors?
CHRIS MILLER: Well, China’s been pouring tens of billions of dollars a year into its domestic chip industry for much of the past decade. It has done this for economic reasons because it has hoped that its firms could reach the technological frontier. It’s done it also for the geopolitical reasons that we mentioned previously. The challenge that China faces today is that although its firms have a lot of money from the national government as well as potential local governments in China, they’re being cut off from their connections with the international supply chain, largely by U.S. restrictions, also, to some extent, by Chinese government restrictions.
The effect of this is to make it very difficult for Chinese firms to learn about cutting-edge processes elsewhere. If you look at the success stories of catch-up growth in the semiconductor industry, whether it’s Samsung in South Korea, which was a minor player in the 1980s and today is one of the biggest firms in the industry, or TSMC in Taiwan, they succeeded, not primarily because they were getting government subsidies, although they were getting some government subsidies, but primarily because they were able to integrate very deeply into international supply chains, learn from their customers, learn from their suppliers, find a niche in the supply chain, and then grow from that.
It wasn’t a process of trying to build up a separate supply chain, it was a process of deep integration. Right now, that’s becoming difficult for China both because of politics in Beijing and politics in Washington. Chinese firms are increasingly trying to do innovation all on their own separate from the international supply chain. That is going to be very, very difficult to accomplish. We don’t really have a good historical example of that working in the semiconductor industry. It’s certainly not the model that Taiwan or South Korea have used successfully.
SEAN SPEER: You’ve written that Taiwan’s advantage in chip production is one of the factors behind growing Chinese interest in the island. How much should we think about a potential conflict over Taiwan as fundamentally about this technology?
CHRIS MILLER: Well, I think you can’t understand the risk for potential escalation pathways between China and Taiwan without taking the chip industry into account. Now, that doesn’t mean that when China thinks about Taiwan, it thinks solely in terms of semiconductors. Indeed, the Chinese Communist Party has wanted to control Taiwan since before the first chip was invented.
Certainly, I don’t think it’s right to look at Xi Jinping’s worldview as associating Taiwan primarily with chips, but it is the case that when the rest of the world thinks about a potential Taiwan crisis, especially the U.S., the economic ramifications are going to be really impactful precisely because of Taiwan’s role in chip production. This could both help and hurt Taiwan in different scenarios. The Taiwanese government refers to what it calls the Silicon Shield, which is the thesis that because Taiwan is such an important chip maker, obstructing production would not only hurt Taiwan, it would hurt the U.S., it would hurt China because the whole world requires Taiwanese chips.
It’s certainly the case that any attack on Taiwan would impose billions of dollars of damage to the Chinese economy as well as the rest of the world’s. I think there’s an alternate dynamic or conflicted dynamic at play in a different military scenario. If you set aside this D-Day-style amphibious assault and focus instead on what I described as Russia in 2014 and the Crimea scenario where China grabs one of the offshore islands in between China and Taiwan, which is currently controlled by Taiwan, it’s plausible the Chinese can even do this bloodlessly or without firing a shot. Then they’d look to Taiwan and U.S. and say, “All right, what are you going to do now?”
In that scenario, if the U.S. decided not to act, it would really meaningfully undermine its defence commitments towards Taiwan, which are already by law in the U.S. If it wanted to act, the U.S. president would get a briefing from a security team saying, “Dear, Mr. President, or, dear Madam President, you’ve got the choice of, A, risking an escalatory cycle that could lead to World War III, but B, guaranteed to impose hundreds of billion dollars of damage in the world economy by obstructing trade with the Taiwan Straits.”
I actually think there’s a real risk that Taiwan’s centrality in the chip industry doesn’t deter China and the most likely scenario, it actually deters the U.S. from coming to Taiwan’s help. This is something that I don’t think the Taiwanese government has really fully thought through, but I worry a lot about these grey zone scenarios and the U.S. being deterred by the economic significance of Taiwan.
SEAN SPEER: If I can pick up the point about Chinese geopolitical motives more broadly, not only are you a Chinese expert through this particular book, but you’ve done a series thinking on China more generally. Should we think about Chinese geopolitical motives as trying to achieve some ideological victory—in effect, proving the superiority of its alternative model? Is it about an economic and technological victory—it wants to reshape the global economy in its image? Is it a more conventional Great Power view about its ability to build spheres of influence around the world through a combination of territorial expansion and vassal states? In sum, Chris, what makes Chinese geopolitical thinking tick?
CHRIS MILLER: Well, my sense is that the core of the dispute between China and the U.S. is that China wants to be the regional hegemon, the most powerful player in its region. China’s view of what that entails is the ability to rewrite the status quo vis-a-vis Taiwan, certainly, but I think also vis-a-vis Korea and Japan. We’ve seen the South China Sea as well. The U.S. doesn’t want to let the status quo be changed. Right now, it’s not clear which party has the stronger hand. Both China and the U.S. believe that they’ve got enough power to achieve their goals, China in revising the status quo, the U.S. in sustaining it.
It’s empirically unclear as to which side is right. I guess we’re going to find out in a decade or two’s time, which side does have the power needed to achieve its aims. Right now, there’s uncertainty. China, I think, has every reason to pursue its current strategy, so long as that uncertainty exists, and so long as there is a chance that it can rewrite the status quo over Taiwan, and more broadly, the balance in East Asia.
SEAN SPEER: If I can shift the conversation to Russia, which is another subject that you’ve studied in great detail, how much are these issues around microchips and semiconductors a useful lens to understand the opportunities and challenges facing Russian geopolitical interests?
CHRIS MILLER: Russia really faces a major issue when it comes to semiconductors and it’s above all a military and security issue. The Russian military, we now know in great detail, thanks to the Russian equipment that was captured in Ukraine, is hugely reliant on imported and smuggled chips from abroad. There’s been a number of great tear downs of Russian missiles, for example, which they found them full of U.S.-made, South Korean-made, Taiwanese-made chips. Although some of these Russia used to be able to buy legally, now are largely illegal to purchase, the reality is there is some scope for smuggling them in but Russia has real no plausible option of producing these chips domestically.
It’s going to have to keep trying to smuggle in Western-made chips from abroad for the foreseeable future. When it fails to smuggle or fails to smuggle the right quantity or fails to verify that the chips it’s smuggled are in fact real and not tampered with, it will face delays in military production. Part of the reason why we’ve seen Russian missile production rates be so low, even in the midst of a war with Ukraine in which Russia is using a lot of missiles, is that the micro-electronics that they need to acquire are more difficult to smuggle in than they might like.
SEAN SPEER: A penultimate question for you. U.S. Treasury Secretary, Janet Yellen, and Canadian Deputy Prime Minister, Chrystia Freeland, have spoken in recent weeks and months about so-called “friend-shoring”—the idea of using public policy tools to cultivate domestic productive capacities, not just in microchips and semiconductors, but in other areas of strategic interest or so-called critical assets. Do you think, Chris, that the CHIPS Act is only the beginning? Are we going to see a greater role for what you might call an industrial policy in the United States and other Western countries coming out of the pandemic?
CHRIS MILLER: I think we certainly are seeing that already. Whether you look at the U.S., Japan, a number of different European countries, Taiwan discussing its own CHIPS Act right now. Most major governments are putting more funds towards their chip industry and beginning to devote more political attention towards their chip industries as well. I think there are pros and cons to that from the industry’s perspective.
There are some things that government probably needs to play a larger role in. And other places where I think there are real risks that government officials try to play too big of a role and make the industry less efficient as a result. There’s no doubt that this trend is well underway in basically every country that has a major or even a minor chip industry.
SEAN SPEER: Final question. Is your story one of globalization working or globalization failing in your view?
CHRIS MILLER: Well, I would say first off, that although the word globalization implies that the production of things is global, in fact, the chip industry is the exact opposite of globalization. It’s the concentration of production in a number of key locations. The entire world is dependent on chips made by just a handful of companies in Taiwan and South Korea and the U.S. Globalization doesn’t really accurately express, I don’t think, a lot of the dynamics at play.
You can understand why a country like China would be dissatisfied with the current shape of “globalization” because they have to import all of the advanced technology and have to suffer the effects of cutoffs whenever the U.S. wants to cut them off. I think globalization actually obscures, that phrase obscures some of the dynamics actually at play. I think if you were to pose the question as has economic integration with China, which has been one of the key drivers of, I think what people generally consider when they use the phrase globalization, has that been a bad policy or a mistake?
I think you need a nuanced answer to that question. It can’t be a yes or a no. There are certain places where it’s been obviously beneficial in terms of providing low-cost production of goods that lots of people consume, but I do think that over the past decade, there’s been a bit too much willingness to transfer high technology capabilities without properly considering the geopolitical and direct military impact of some of these capabilities.
That’s where I think the current rethinking is welcome because 20 years ago, U.S. policymakers didn’t have to think that hard about the military balance vis-a-vis China because the U.S. was so far ahead. That’s no longer the case. Every economic exchange with China, I think, has to be assessed, at least in part, under the rubric of, will this help or hurt the military balance, will help or hurt the Chinese military?
SEAN SPEER: Well, one way to think about that question is to read Chip War: The Fight for the World’s Most Critical Technology. Chris Miller, professor of international history at Tufts University, thank you so much for joining us at Hub Dialogues.
CHRIS MILLER: Thank you for having me.