Riskgaming

There’s always been a global race to develop chip technology

Description

Semiconductors are ubiquitous in modern life, powering our appliances, smartphones, cars and electronics. That’s led to soaring demand from consumers, companies and governments much to the chip industry’s benefit, but its centrality to the global economy has also brought heightened scrutiny from analysts concerned by the deep dependency we have on a handful of companies around the world producing these products.

The semiconductor industry is now on the front pages of news sites almost daily, but its story and history show that this isn’t a new development, but rather a continuation of decades of globalization and competitions for international economic supremacy.

“Securities” host Danny Crichton is joined by Fletcher School professor Chris Miller, whose new book “Chip War: The Fight for the World’s Most Critical Technology” offers a panoramic global view on one of the world’s most important industries. The book has already been shortlisted for best business book of the year by The Financial Times.

In this episode, we discuss how geopolitics influenced the evolution of chips, the Soviet chip industry and how it struggled due to a lack of a consumer market, the transition from human computers to chips, the importance of the pocket calculator in chip history, the transition from Hong Kong to Taiwan in outsourcing assembly, how TSMC became the leader in lead edge fabrication, free trade and its centrality to the industry, and finally, whether interlinkages help or ultimately hurt national security for countries like the U.S. and China.

Transcript

This is a human-generated transcript, however, it has not been verified for accuracy.

Danny Crichton:
The only other thing is just because of the echo in here, just tone the volume down just slightly if you can. You don't have to whisper.

Chris Miller:
Okay.

Danny Crichton:
As you can see without noise cancellation, it's like, oh, my God. All right, let's get a show going. Hello and welcome to Securities, a newsletter and podcast devoted to science, technology, finance, and the human condition. I'm your host, Danny Crichton, and today we have Chris Miller, a professor at the Fletcher School at Tufts University, talking about his new book, CHIP WAR: The Fight for the World's Most Critical Technology, on sale and bookshelves October 4. Chris, welcome to the program.

Chris Miller:
Thanks for having me.

Danny Crichton:
So why don't we just get started by talking about what the book is and what its about.

Chris Miller:
The book is a history of the computer chip, not only the technology itself, but the broader economic, social, and geopolitical context in which chips developed. And that's important because we wouldn't have had computer chips had we not had, for example, an arms race in missile race during the Cold War. The first chips emerged out of the desire to guide rockets more accurately, and since then, there's been a really deep and integral relationship between chip firms, chip designers, and U.S. government in particular, but not only. And so if you want to understand how the technology's developed, you got to understand how that has interacted with governments and with the broader geopolitical context.

Danny Crichton:
When you look at computing, 1940s, you have John von Neumann who comes up with the Von Neumann design prior to almost any chip kind of being built. And then we sort of have this American centric story of Fairchild Semiconductor of the Silicon of Silicon Valley coming out of Project Whirlwind at MIT and a bunch of these new initiatives. But you actually write a much more global history covering everything from Russia and Japan, to Taiwan, to China. Why is that perspective so important to understanding the chip dynamics?

Chris Miller:
For two reasons. First is that there's always been a race to develop the next generation of chip technology and to have the most advanced capabilities and that race has always been global. It hasn't just been about Silicon Valley firms, although we in the United States think the most about them and they've oftentimes won the race. Second is that from the early days of Silicon Valley, U.S. firms have globalized their assembly, their packaging, and later on, a lot of their fabrication as well and they've always sold to global markets. So just from a purely economics and business perspective, this is and has always been a global industry, and that's not just a new feature. That's present from the early stage, for example, of Fairchild Semiconductor.

Danny Crichton:
When did government officials start paying attention to this space?

Chris Miller:
Well, they were paying attention to microelectronics for a long time. During World War II, for example, the focus was in radar systems, sonar, things of that nature. And there was always a desire to miniaturize computing power to put it in more types of systems. So throughout the earliest days of the Cold War in the late forties and early fifties, government officials from the Army, the Navy, what would become the Air Force were putting a lot of money towards mini strides of computing. But it wasn't until the late 1950s that we had the advances in semiconductor technology to actually put multiple transistors on one piece of silicon or germanium, which then became known as the chip.

Danny Crichton:
And when you think about the invention of the integrated circuit and a bunch of others, we think of this as a very western American story, but you're also a scholar of Russian studies, covered Russia for many, many, many years. What was it like on the other side of the Cold War?

Chris Miller:
Well, just as the U.S. was trying to guide its missiles more accurately so too was the Soviet Union. And if you rewind the clock to the late 1950s, today we now know that the U.S. won the technology race but the USSR won the arms race. But at the time, it was far from clear. 1957, right before the first integrated circuit was invented, the Soviets launched Sputnik into the earth. They followed that shortly with Yuri Gagarin and the first person in space. And so, at the time, it seemed like the Soviet Union was a really technological competitor to the United States. And it wasn't only in rocket technology, it was in nuclear weapons where sure they stole some of the technology from the U.S., but nevertheless, that was a real accomplishment. And at the time, it was possible to read any number of stories. For example, in the U.S. press of the Soviets are about to overtake the United States in technology.

And depending on what metrics you were looking at, you could find a lot of evidence for that. For example, numbers of engineers trained, the Soviets trained a ton of engineers and scientists. So there was real fear at the time that Soviets were going to overtake the U.S. and they started pouring money and resources into microelectronics as well for the exact same reasons, the desire to miniaturize computing power and put it in weapon systems. And they invented their first integrated circuit just a couple of years after TI, and then Fairchild did the same in the U.S. but they faced huge problems in actually building out their infrastructure for a number of reasons. One was that they never had a vibrant consumer market, whereas in the U.S., although chips started being almost exclusively sold to defense and space uses, by the 1970s, they were being sold widely to commercial uses for computers, for pocket calculators, things like that.

In the Soviet Union, that was not the case. There wasn't big consumer market and so the overall amount of spending on chips was substantially lower because you didn't have those consumer use cases. But in addition to that, the Soviets didn't globalize. They globalized with Romania, they globalized with Hungary, the eastern block, but the U.S. from basically day one was bringing in cheap labor from Southeast Asia, was bringing in vast markets in Western Europe and Japan and had the industrial base of the entire industrialized world on its side. And so you could, if you were TI turn to producers in Japan or Western Europe for equipment you couldn't find it in the U.S. and that was something the Soviets could never do.

Danny Crichton:
I find this so interesting because, obviously, the 1960 Kennedy campaign very focused on the missile gap. This idea that the Soviets were way ahead in that moment of the Sputnik moment. The U.S. passes, I believe, the National Defense Education Act somewhere in 1958 with this goal of massively accelerating what we now call STEM, probably not a historically accurate term. And that leads to a huge surge of folks in that category, universities research labs. I find this part around the consumer market really fascinating because when I think of computing in that era, I think of accounting firms. So Arthur Anderson, a lot of these big Deloitte now, but its predecessors were very focused on computing because when you had these large companies, you were still doing this manually. There was no such thing as a spreadsheet. And so computers, one of their first major applications was deploying, in large companies, to calculate profits and revenues and accounting. What was the consumer market like back then?

Chris Miller:
So you're right that this civilian market started primarily as a corporate computer market. You rewind back even further to the 1920s and thirties, computers was a profession. You could be a computer.

Danny Crichton:
Right. Yes. And most of those computers were women.

Chris Miller:
That's right. And they took lunch breaks, they demanded weekends, and so there was a real desire for corporations to buy computers that wouldn't demand lunch breaks and would operate more efficiently. And so that was the key market in the 1960s outside of military markets. And so by the late 1960s, most chips produced were going to the corporate computer market. Quickly thereafter, consumer markets began to emerge. At first in small ways for things like hearing aids, but the big product that made chips into a consumer product was pocket calculators, which were really kind of the iPhone of the 1970s. A small piece of plastic that everyone could put in their pocket utilizing the most advanced technology. And today we laugh at it, but that was a real big change. What's interesting about pocket calculators is that they used chips largely designed by U.S. firms produced by U.S. firms, but they were assembled and really the product was invented basically by Japanese firms.

It was one of the first big products that Japanese firms after the war made themselves a global reputation with. And so it was an interesting example of the symbiosis between Japanese consumer electronics firms using U.S. technology and both benefiting from that collaboration.

Danny Crichton:
Japan has sort of receded from our imaginations in some ways. It's still part of the story today in semiconductors, and I don't know if we're actually going to come back to it, but they're still in chemicals. They still have a large part of, I believe the actual wafer manufacturing, so the actual materials that go into chips, but they've sort of fallen out in DRAM space where Korean firms, and even now Chinese firms, are increasingly have taken the lead and so they've kind of disappeared. At the same time that this was going on in Japan though, we have Taiwan, which is a tiny island, not very rich at the time, struggling mightily given the kind of Umbra of China right across the South China Sea. How did Taiwan get involved in this market and why did it suddenly materialize so heavily in the news in the last couple of years?

Chris Miller:
The Taiwanese have been trying to break into electronics since the 1960s. They looked at what was happening in Hong Kong in the early 1960s, which was the first place that U.S. firms off-shored their assembly and packaging. And they realized this was an industry that required lots of workers required, at the start, relatively low skill levels and could provide a substantial amount of employment for their populists. So the Taiwanese government was very focused on trying to attract U.S. and also Japanese firms to build assembly facilities in Taiwan. And as a result of that, by the 1970s, the 1980s, they had a substantial chunk of workers who had been in semiconductor facilities for some time.

Taiwan also benefited because there were a lot of Chinese Americans and Taiwanese Americans, and of course, the boundary between those two categories was complicated at the time, who were educated at Stanford, or Berkeley, or Purdue, who were very talented electrical engineering had worked at Fairchild or Texas Instruments, but also had some sort of connection to Taiwan. And so they were very skilled at taking advantage of those individuals, who there were many, to transfer information and expertise back and forth. And there's great examples of the Taiwanese government setting up all sorts of working groups and committees of Chinese Americans or Taiwanese Americans to advise on the development of their industry.

Danny Crichton:
I think what's interesting here is one, at the time, obviously, very vertically integrated, most firms were producing their own chips. They wanted to own everything top to bottom, and that obviously transitions over the next couple of decades to one that people are much more willing to design chips or manufacture chips, but they don't necessarily have to do it all in house. The other piece here, which I think gets back to the USSR story is the globalization story for Taiwan. So I'm remembering a book Making Money: How Taiwanese Industrialists Embraced the Global Economy, 2018 Stanford Press, Gary Hamilton and co-authors, which argues that, basically, starting in the 1950s, Taiwan realized it had to own global markets. It had to compete on an equal playing field with Japan, Korea, Singapore, Hong Kong, and everyone else. And I think that's one of the stories that makes a slight tweak to the Japanese story where I think the Japanese were always very globally focused, but always felt like there's always this huge domestic market in a way that Taiwan never had.

Chris Miller:
If you're in Taiwan, you've got to export. There's no other market to sell to.

Danny Crichton:
Taipei is a nice city, but it's not that big.

Chris Miller:
That's right, that's right. In addition to that, I think the status of Taiwan itself made integration into U.S. supply chains in particular, not only important for economic reasons, but important for geopolitical reasons. And if you think back to where Taiwan was in the 1970s, the U.S. was in the process of losing or have lost the Vietnam War. There was great fear that following Nixon the U.S. was going to withdraw from Asia Pacific. Eventually the U.S. switched recognition from Taiwan to China in terms of the government of China. And so the Taiwanese were deeply fearful they were about to be abandoned, and they had good reason to be fearful at the time.

The U.S. was pulling back its military from the region. And so there was a deliberate strategy by the Taiwanese government to say, "We've got to plug ourselves into supply chains, not only for ergonomic reasons of which there were many, but also because we're not sure if the U.S. is willing to defend Taiwan, but we know the U.S. is willing to defend Texas Instruments. And so if TI has a plant in Taiwan that gives us an additional margin of safety in case the PRC tries to retake the island."

Danny Crichton:
And ironically, I mean we're talking about this in the seventies and that was triggered from the Civil War in the forties up through Kinmen and in some of the fights in the fifties. Obviously, tension all the way forward. And what I find fascinating is nothing in some ways has changed. We're talking about the seventies, we could be reading the Financial Times today in 2022, and in much of the same way the dynamics are the same.

Chris Miller:
I think that's right. I think there's one big thing that has changed, which is that in the 1970s, Taiwan was a sliver of the global semiconductor supply chain, whereas today it is the beating heart of the digital world. Everything we rely on smartphones, a big chunk of PCs, data centers, requires chips from Taiwan. It's sort of like the Cuban Missile Crisis, but rather the Cuban Missile Crisis. It's as if Cuba was at the center of the global economy. It would've been a different missile crisis.

Danny Crichton:
I'm trying to think of the 1960s. Is it like the toaster, it's like all GE Toasters came from Cuba from the 1960s and the Cuban Missile Crisis. So why has Taiwan become so central to the semiconductor industry?

Chris Miller:
So today, TSMC not only produces the most advanced logic or processor chips, which they've done now for at least a couple of years, they also produce an extraordinary quantity and capacity is really crucial here because as we found the past couple of years, it's very difficult time-consuming, costly, to build up new capacity. And so they're not only important because of their technological capabilities, which are really substantial, but also there's so much chip making capacity in Taiwan that if that were to go away, it would take a decade to replace.

So because of that, TSMC in particular, but Taiwan in general, do occupy this place where we would really struggle to get by without them. It would be possible, but it would be very, very hard because if you look at anything with an on-off switch today, most of those devices require some sort of semiconductor. Any sort of advanced computing today requires a chip that can be made in a really small number of facilities around the world, and a surprising number of those facilities are in Taiwan, most of which are operated by TSMC so it really is at the center of everything digital today, and it's irreplaceable, at least in the short term.

Danny Crichton:
There have been a number of folks who've written about the geopolitics of Taiwan and the importance of TSMC. It is a story that's become, I think, very popular in the last two years, but I've always had this question of like if push comes to shove and you had to make trade-offs between war with China, protecting TSMC, protecting the global semiconductor chain, what happens in that context? How do you even begin from a game theoretic perspective, from balancing different interests and national interests? What would be the value and cost to the United States for getting involved and intervening in that sort of conflict?

Chris Miller:
Yeah, it's a really complicated dynamics. On the one hand, Taiwan is important because of semiconductors. Obviously, true. On the other hand, a lot of those semiconductors are assembled into final goods in China. Think of the entire smartphone supply chain. A little bit is in India, a little bit's in Vietnam, but China's the bulk of it. PC's similar story. You can't simply say, "Oh, well, if we've got Taiwanese semiconductors, we've got what we need." You need the entire supply chain. Now, is it probably easier to reconstitute a assembly supply chain outside of China than it is to reconstitute TSMC outside of Taiwan? Probably. But either way, you're talking about tremendous disruptions of the sort that would make the disruptions of the Russia, Ukraine war seem really minuscule.

I think there's a lot of focus in our discussions of sort of D-Day style scenarios vis-a-vis Taiwan, which are the ones that I sort of worry less about, not because they're impossible, but because I think others are more probable. And I think if you look at what would happen if there were small outlying islands of Taiwan that were compromised would happen in cases of a partial blockade, would the U.S. feel compelled to respond for geopolitical reasons or would it feel fearful to respond fighting over a small island and thereby threatening the entire global electronic supply chain?

Those are scenarios that are actually more problematic than the sort of D-Day scenarios and where I think the economic and the technological and ultimately the semiconductor ramifications are probably more important in U.S. decision making.

Danny Crichton:
We recently had an episode focused on the agricultural supply commodities exchange with Chris Mahoney from Glencore Agriculture and Soren Schroder from Bunge. And one of the things they really emphasized about agricultural commodities is just how much free markets and open trade and globalization matters in that world. It's a well tuned well oiled fine machine. There are so many different actors involved. I mean, really if you include every farmer, I mean potentially millions of different actors that all aggregate to create grains. And soybean oil and whatever the case may be to sell globally. And to me, in some ways the story is much the same, which is at a certain point, any disruption, anytime you close the borders, anytime... I mean we've seen this with COVID-19 but anytime you're closing borders, stopping free trade, it's just going to cause massive disruption because the entire industry has been built around a set of assumptions, around institutions, around openness, that as soon as those assumptions go away, the industry sort of falters.

Chris Miller:
I think that's largely right with one difference. I think that's right in the sense that semiconductors are everywhere and they're cheap today because we have this tremendous well-oiled global machine, and it wouldn't be possible to do in just one country. You need expertise, you need capital investment, you need customers from the world over. That's certainly true. And it's like agriculture in the sense that there's many steps in the process. You need the farmer, you need the grain elevator, et cetera. Same thing is true in chips. You need the design, you need the capital equipment, you need the materials, et cetera. I think the one difference is that in semiconductors there are a number of choke points where one firm or a handful of firms or one country controls the entire segment of part of the supply chain. The best example is in lithography equipment where ASML is the sole producer of UV lithography machines, which are crucial for advanced ship making.

They make Saudi Arabia's position in the oil market look really minuscule. A hundred percent market share is pretty impressive. And ASML's the extreme example of this, but it's not the only one and it's basically impossible to build an advanced fab today that doesn't buy equipment from the five biggest equipment producers, ASML being one of them. There are three firms that control the EDA market today. There are basically three competitive producers of DRAM chips today. There's a lot of places in the supply chain where there are monopolistic or oligopolistic tendencies, which given the capital investment and technological expertise, there are reasons for that. But that introduces some additional vulnerabilities to the chip industry. And I think if you look the last couple of years, you've obviously seen plenty of headlines about supply chain shocks and one of the things that I think people misunderstand about the supply chain shocks of 2020 and 2021 is the extent to which it wasn't a supply side shock.

The supply of chips increased dramatically in 2020 and even more dramatically in 2021. Double-digit increases in 2021. It was a demand side shock. Demand grew even more than expected. I actually worry less about the peace time stability of the supply chain because I think it's shown that it can respond pretty rapidly to changes in demand. But I do worry about geopolitically induced shocks of which I don't think it's difficult to imagine a couple that could be really pretty dramatic for our ability to access technology.

Danny Crichton:
It's interesting to me because we had a Hondo Geurts on recently talking about the defense base, and one of the patterns that we've seen, basically, in the peace dividend and post Cold War era, in the defense base is this massive consolidation of prime contractors. There used to be two or three dozen in the 1990s, now there are five, and you see that all across the defense base where there used to be real competition, there was choice, there was resilience, and that has consolidated down to one or two oligopolistic providers. And what's interesting, particularly in the Pentagon's case, is the budgets have actually gone up since the nineties. There has been massive expansion. So in some cases we've actually doubled the budget and yet the number of companies has shrunk in some cases by 80 or 90%. And so you see more and more money going to fewer and fewer players, which is great from an efficiency point of view, and that was Hondo's point, but it's terrible from a resiliency point of view.

But that leads me to an open question because I think all these concerns about choke points, the supply chain has become a hot issue in DC these days and projecting forward into the 2020s, the senate is investigating this, intelligence community is doing estimates on the supply chain. It's actually fascinating how little we actually kind of know how all this works together mean coming from industry, we don't know. At some point, there's always a black box somewhere in the supply chain where you just don't know where the material comes from or something else. But projecting forward, obviously there's a huge focus on indigenization of technology. Bring back reshoring technology both in the United States, which is bringing fast back from overseas but we're also seeing this in China where China wants to indigenize technologies from overseas and bring it home. I'm curious, how realistic is this in a totally globalized industry in which half a dozen countries own a critical choke point in the supply chain for chips? Can one country kind of vertically integrate its own supply chain and can multiple countries do that?

Chris Miller:
I think it's a fantasy for any one country to try to do that. The country that would have the best chance would probably be the United States, but even still, it's easy to point out a dozen different places where that would be an extraordinarily expensive proposition. And if you tried to do that, you would be spending so much money building up legacy capacities in parts of the supply chain that you wouldn't have the funds you need to invest in the future. R&D spending is very intensive in the chip industry, and so you need funds to pour in R&D. So I don't think anyone can really do that. I think for certain defense applications that already exists in the U.S. and China's trying to create it domestically. That's, interestingly, only for lagging edge defense applications. For leading edge logic the U.S. relies on Taiwan to produce and so does China, which is sort of an incredible position to be in.

But both the PLA and PLA is not supposed to, but they do, and the U.S. military relies on TSMC for their leading edge logic, and that's probably not going to change in the short term. The U.S. might get some more leading edge logic onshore. Part of the goal of CHIP act to do that. At China, over time, we'll get closer and closer to the leading edge, I suspect, in terms of domestic manufacturing. But at the end of the day is that everyone will be relying on not global but international supply chains for some time. Does that mean that we're not going to see shifts? No. I think we're going to see a substantial but not complete bifurcation driven both by Beijing and by Washington. And you see plenty of dynamics in both capitals pushing for bifurcation, but that won't be a complete bifurcation. And so I think the language you hear about decoupling, for example, that's the wrong metaphor. We're going to find a new equilibrium or new balancing act that will be less connected, but there'll still be lots of connections because, ultimately, everyone relies on technology from someone else.

Danny Crichton:
I think of it as everyone will reduce their insecurity, where people will feel much more that they're in empowered, that they're in control, they have negotiating leverage and so, ultimately, it's a balancing game of like, well, I can take this away from you. You can take this away from me and so we might as well work together sort of a game theoretic positive outcome that comes out of that.

Chris Miller:
I think that's how people are thinking about it. I think there's two questions you want to ask on top of that though. So one is to what extent, if any, does inter linkage give us security, which I'm not sure the answer is yes, but I think that's an open question. Is there sort of a technological mutually assured destruction a place right now, for example, in the fact that our smartphone infrastructure depends very heavily on the U.S., China, Taiwan nexus, you need all three of them to build an iPhone, for example. Does that give us some security against the geopolitical relationship getting worse? Maybe. I'm not sure. Second is if you are thinking about your security, you also need to think about your capabilities today and in 10 years time in a military sense. And there's, I think, open questions to be asked about what's the best way to get those capabilities?

Is it by trying to domesticate as much as possible? Is it by domesticating as little as possible, but trying to take the most advantage of open global markets? There's a balance to be struck there too, but I'll give you a look at, for example, where military technology is heading. There's more and more focus not only in the U.S. but globally on taking commercially off the shelf technology plug into military systems. And you've got to do a bit of that because if you're looking at advanced memory chips, advanced logic chips, Apple gets more advanced logic chips than the military by multiple years on a regular basis. So if you're not getting commercial technology, you're behind, and that's going to be the case for a very long time.

Danny Crichton:
I feel like there's a prediction that Tim Cook becomes the McNamara. I mean, the positive sense early on coming out, McNamara, head of GM, General Motors, goes over to the Pentagon prior, ignore the Vietnam War era, but comes with the concept of saying, we have a huge logistics problem. We have a huge industrial base problem. Here's this person who's created the single best logistics empire in the world, GM in the fifties and sixties. Let's have him solve the problem. So in some ways, Tim Cook with logistics takes over the Pentagon someday. That's a prediction, not from you, but I'll throw it on there as on the board.

Chris Miller:
Well, I like to say that Tim Cook has more power over this micro supply chain than anyone in the U.S. government, which I think is true today, and I think that makes the defense industrial base very uncomfortable because they're used to having more control over the products, and in fact, they are responding to economic signals that are set by people like Tim Cook. Would he make a better defense secretary than what we've got now? I think that's an interesting question.

Danny Crichton:
We'll just leave that one on the board. But this is actually an interesting question. So I'm ripping this from the headlines. I have no idea when we'll publish this show, but there's a huge concern because Apple has sort of said that they will take DRAM trips from Yangtze.

Chris Miller:
These are NAND chips.

Danny Crichton:
NAND chips. Sorry, NAND chips. Yeah, we've moved forward a generation of memory. So NAND chips and, obviously, a lot of concern for Marco Rubio and other senators who are saying, "Look, this is a state funded enterprise in China building memory chips. It's part of the indigenization campaign for China to create capabilities against Korea and other major NAND and DRAM suppliers." I'm curious, look at Apple as a company that has to survive both in the Chinese market, which is a massive important market from the consumer perspective, as well as in the U.S. How do you glue? Are you essentially just trying to hold this all together or do you think that, at some point, there's just a bifurcation in the consumer markets as well where these products will just not be sold in both places?

Chris Miller:
Well, it's certainly Apple and Tesla's another example of a company that's trying to straddle that very, very carefully. Has the balancing act gotten harder? It certainly has. Have these companies proven very capable thus far of managing it? Thus far they have. I think if you're an Apple or a Tesla or a company like that, on the one hand, you need to make sure that you keep China's leaders happy because it's a huge market. On the other hand, you've got a lot to offer them, not only in terms of providing them with cool phones, you are directly or indirectly employing a whole lot of people in China. And Apple has, I think quite interestingly, been proving over the past couple of years that it can just as easily do what it does in China, in Vietnam and India and elsewhere, which I think is not just a resilience operation.

I think it's also a message to China saying, "You better treat us pretty well because if you don't, we don't need your workers. There are a lot of people in India too, that could do a similar job." Now it's an open question. Could India be competitive with China in terms of smartphone assembly? One of the challenges electronics assembly faces in India is that you have strikes, for example, which of course, you don't get in China, but I think Apple and a lot of other firms that are in this position have been trying to demonstrate to China that there are other options precisely to improve their negotiating position vis-a-vis China. But do they face pressure to buy Chinese? They would deny it, but I think the answer is certainly yes, and the YMTC news is very much part of that.

One of the things we haven't seen China do nearly as much as I think you might have expected is pressure foreign firms or even Chinese firms to buy domestic components for consumer electronics. I wonder if that's about to change, and it would be, I think, a very understandable move if China were to start to do that. Something that would actually be quite similar to what Japan did in its developmentalist days or Korea as well. If so, that would really upset the way that Apple and other firms devise their supply chains and I think push them to a certain extent to diversify further offshore.

Danny Crichton:
Well, that was sort of Apple's response. They're considering it, but only for the Chinese domestic market, which in some ways breaks the whole notion of the efficiency of the supply chain because the whole point is there's one global iPhone, it works everywhere. There's no indigitization, there's no changes, and that creates savings for Apple, which comes up in its margin structure. Now I want to go back to the book. So you wrote this Chip Wars, dozens of chapters, amazing stories. I mean, the chips industry is really like a locus of so many interesting dynamics, both geopolitical spy stories, intelligence agencies, industrial policy analysts. What was the most fascinating story, anecdote, fact, that you learned in the process of researching the book?

Chris Miller:
I think the most interesting characters is Morris Chang himself, who we know today as the former chairman of TSMC. He's now theoretically retired, although the time I spoke with him, it was from TSMC's offices so he doesn't seem that retired.

Danny Crichton:
I retired to my office.

Chris Miller:
But his life trajectory illustrates and shaped the chip industry itself. Born in mainland China, moving to the U.S., working with a U.S. security clearance on defense systems at TI, moving to Taiwan, having only been to Taiwan previously on Texas Instruments business trips. He had no other connections to Taiwan. Building up the industry there, not only in Taiwan, but building up the entire Trans-Pacific and cross straits supply chain. I found that his life was sort of a glimpse into how globalization happened, and globalization, as we know it today, is so dependent on these electronic supply chains that people like him, he personally helped build. And so it's so cool to see these individual stories, I think, map on so perfectly to the economic changes that we've seen over almost a hundred years.

Danny Crichton:
Well, amazing. Well, Chip Wars, October 4 on book stores everywhere. Chris, thank you so much for joining us.

Chris Miller:
Thanks for having me.

continue
listening