Apple in China

McGee's Best Interview!

Patrick McGee is the author of Apple in China: The Capture of the World’s Greatest Company. Our discussion led us through a detailed history of Apple’s relationship with China, where iPhone manufacturing became a project of nation-building.

Cohosting today is Kyle Chan of the High Capacity Substack.

Today, our conversation covers:

• Why Apple moved production to China in the 1990s, and why it struggles to leave,

• How Apple’s obsession with perfection catalyzed China’s industrial upgrading and why it bought every CNC machine in the world

• The political side of production in China, including how Apple’s relationship with the Chinese authorities has evolved over time,

• The rise of Foxconn and other partners in Apple’s network,

• A peer into Apple’s management style, including the “Divorce Avoidance Program.”

Listen now on iTunes, Spotify, or your favorite podcast app.

Job Opportunity to Research AI + China

The Centre for the Governance of AI is looking for folks to research China and AI! Check out their three-month Winter Fellowship, one-year Research Scholar position, and Research Fellow position. I like the people there and they’re looking specifically for ChinaTalk listeners—take a look at the positions!

---

“Real Men Own Fabs”

Jordan Schneider: I want to start with the near-bankruptcy moment and the challenges of manufacturing in America in the 1990s.

As you write, Patrick, there was very much a “real men own fabs” energy to hardware manufacturing, where most companies, Apple included, decided that if they wanted to compete, they had to build all the stuff themselves. But as you write in your book, that line of thinking reached its limits in the mid to late 1990s. Why don’t you start our narrative there and tell us that story?

Patrick McGee: I talked to the earliest employees at Apple, and it was part of the ethos that you built your own computers. This is how Steve Jobs and Steve Wozniak met at the Homebrew Computer Club. They’d be taking apart computers, presenting to other people, and Steve Wozniak’s genius was to take apart a computer, improve some of the circuitry himself, and then put it back together in a faster way.

Apple wasn’t doing anything different by assembling computers themselves. That’s what you had to do. I don’t think the electronics manufacturing supply chain existed, certainly not in a globalized sense of taking things from Taiwan or elsewhere. That’s all later.

The computer I give credit to for changing everything was the IBM PC in 1981. Steve Jobs largely misunderstood it because when he disassembled it, he looked inside and said, “There’s nothing special here.” He’s not wrong about that. But what he missed was that IBM had launched a whole new battleground that Apple wasn’t prepared to fight against.

The IBM PC wasn’t about the graphical user interface. It wasn’t about user design or UX. It was about making these things and building them at a massive scale. To do that, they relied on third-party assembly giants. One that I highlight is called SCI Systems, which you could think of as the Foxconn of its day. It was founded in Alabama and played a role in satellites and rockets — it played a role in the Saturn V rocket that went to the moon.

Once they had that military precision, IBM scouted them out and they ended up doing the circuit board for the first PC and then assembling computers writ large. Then they played this massive role, doing that for HP and Dell and all these other companies, along with the rise of a whole new industry of Solectron, Celestica, and Jabil— all these North American companies who outsourced on behalf of the PC giants.

They gave birth to electronics globalization because once you have standardized parts in this booming industry, then the likes of Taiwan can outmaneuver everybody else and start building them in great quality and at great scale. Then they vertically integrated. They went back upstream and did more of the computer until by the mid-1990s, you had Taiwanese companies that were putting together everything and badging it with different names, and then they were even badging it with their own names.

Flying Geese, Takagi Shirō 1972. Source.

There’s this fascinating 20-year history where Steve Jobs in particular has this DNA that he leaves to the company, which is something like, “We will build it ourselves because we’re going to design it ourselves. That’s how we distinguish ourselves, even as the PC companies go to this outsourcing boom.” When he goes to NeXT Computer — the other company he founded after he’s ousted from Apple in 1985 — he says of the computer, “My favorite thing about it is that it wasn’t built in Osaka.” Obviously Japan was playing the role that China plays now.

There’s something fascinating about that. Even when the narrative after 1997, when he came back, was that Apple had to outsource, it took Tim Cook years to convince Steve Jobs to give up manufacturing. I write about how in 2000, he handed out t-shirts to everybody that said “Mac-tories” on them. He wanted these Mac factories to play a role.

Without the dot-com crisis, you would have had Apple trying to make a push in building things themselves, which would have included building in America. But Apple stock, as people sometimes forget, lost 50% in a single day in September 2000.

They’re almost purchased — well, maybe I shouldn’t say they’re almost purchased, but Gateway thought they were such a contender that they could purchase them. That’s how bad things were at Apple — that Gateway of all companies was going to acquire them.

Anyway, my point is I have this manufacturing-focused history of Apple, that’s a new angle to report on the company, and yet I think it shouldn’t be a new angle. This is a company that manufactures — now they orchestrate the manufacturing — but their DNA was doing it all themselves. This shouldn’t be an obscure angle to have brought a new lens to, but somehow it is. I’m uncovering more than I should have been able to because I think we should have been reporting on this the last 25 years.

Jordan Schneider: It’s fascinating, Patrick, because when people think of Apple’s moat and what distinguishes them from the competition, a lot of folks first go to iOS. We have this America arc, and then we have this “go abroad to lots of different countries and try to find one or more that works” arc. What were some of the starts and stops in that arc in the 1990s and early 2000s before they ended up going to China?

Patrick McGee: This will sound less strange if the listener understands that Apple had a tri-continental strategy for building locally but at the continental level. Apple built its own computers in California and Colorado — that’s obviously for North America. For Europe, they built them out of Ireland and for Asia, they built them out of Singapore.

When they get involved with contractors in the late 1990s, they have them replicate this strategy. LG makes the translucent computer — the Bondi blue color that then becomes Life Savers colors — and LG does it in Korea. But when they expand to meet demand, they do so in Wales and in Mexico.

When Foxconn comes on as a second supplier, this is the meeting of the minds of Foxconn founder Terry Gou and Tim Cook, who’s a pretty new person at the time but is running operations. Foxconn does it in China, which is significant because Foxconn’s Taiwanese and some of the other Taiwanese companies are not in China at that point. I’m thinking of Inventec, for instance, which made the iPod in 2001, and that’s out of Taiwan.

But then Foxconn expands, and they expand to the Czech Republic for Europe and California for America. Once you are comparing costs and efficiency between these three regions, China is winning out time and time again for multiple years, such that by 2003, Apple has effectively given up everywhere else and consolidated into China.

I suppose technically the Czech operations are in effect until 2010 or so, but everything is moving to China because China is tailor-making policies for someone like Apple. Industrial clusters are forming. They’re putting up factories, not in great quality, but at an amazing scale.

I have this funny anecdote with this engineer from Apple in the early 2000s, and he’s literally counting the stairs between floors because it’s so obvious to him that it’s not the same distance between each floor. He’s even measuring the stairs and he’s finding that this is so slap-dash — everything’s put together so quickly. This does, if we’re jumping ahead, have implications for us, because of course, we would never build something that quickly because we would emphasize the quality, the safety, and everything else that we should emphasize.

We would still be doing the environmental paperwork by the time an entire factory in Shenzhen would be built.

John Rubinstein’s anecdote for the iPod is that it’s steel girders when he goes there and sees the iPod factory actually up and running because you get to the second floor and it’s all finished, but the rest of the building is not, and yet they’re still churning out iPods.

There’s this amazing sense that when Apple went to China in the early 2000s, it was very early days. China was not known for quality at the time. They are known as a manufacturer, of course, but they’re not building quality electronics. That was something that Apple played an instrumental role in.

Kyle Chan: One reason I found this book so fascinating — it’s the story of global electronics manufacturing told through one company across a number of different countries. You chart the movement of Apple’s efforts to figure out the best place to manufacture — the one place, the company, the country — that might suit their quality needs. They’re extremely demanding with rapidly changing specifications.

The way I think of it, they’re dating a bunch of countries over time, and you know that they’re going to marry China in the end. They’re going to marry Foxconn. It’s like a rom-com — “When Harry Met Sally” — and you’re going to follow the dating along the way.

Could you tell the story of searching for the right partner? In the background I have in my head the whole “flying geese” model of development in East Asia, conceived by a Japanese theorist in the 1930s about how Japan would take the lead technologically. Then, over time, different sectors would move on to South Korea, Taiwan, and then eventually to China.

Source.

The Apple story, the Apple manufacturing story, matches that so well in terms of the search. By the time you go through the process, you’re like, “Okay, come on. I know you’re going to end up in China.” But what were some of the steps along the way? Why did they end up settling on China as the place to be?

Patrick McGee: A saying I love about journalism is that you should always show, never tell. When I was submitting the book, my biggest black swan worry was that my editor would say, “I don’t understand why this book begins in 1996. China’s not in the book for the first 90 or 95 pages. Let’s get there.” But the whole point was, I needed to tell you without actually ever explicitly saying so that Apple had this totally obsessive, maniacal, perfection-oriented, no-tolerance-for-defects attitude.

You need to understand that when they were building something like the iMac G4 — the thing that people might remember because it’s anthropomorphic and looks a little bit like a Pixar lamp — I joke that it looks like it’s ready to break dance. That’s actually the first commercial that Apple used for it, grooving to the beat, nodding its head.

Nobody knew how to build that thing. Apple upended tools and machine makers all across Taiwan to get it done. Without me explicitly saying so, the reader begins to understand that Taiwan is too small to get this done. They’re literally having to get people from the Philippines to fill the factory lines, and they’re using a VCR machine maker in Malaysia. They’re relying on Singapore Airlines for a blade fabrication facility.

You realize that for them to have a meeting about various supply chain conundrums that they’re having, you’re having to get on flights and hand-carry things from one country to the next, which by the way is illegal. They’re doing so with five or six countries as far apart as Singapore and Japan.

When China begins to offer all of this in one industrial cluster around Shenzhen, it is mind-bogglingly good for anyone involved. One thing that I say in interviews — I don’t think I explicitly say so in the book — is that Tim Cook is often credited as the architect of the China strategy. He’s not the architect, and there is no architect. It was the suppliers themselves — and there are hundreds, if not thousands of them — are all choosing China because if they’re comparing Singapore versus China, the labor rates, the abundance of labor, the flexibility of that labor, the hardworking determination of that labor is all off the charts in China versus any other place.

The suppliers themselves are all moving there and Apple is able to take advantage of that movement more than anybody else. If you still want to give Tim Cook a title, I would say he was more like an admiral navigating the macroeconomic seas of supply chains. But he’s not the architect because he didn’t have to strategize anything. Terry Gou was strategizing. Tim Cook was following, but then doing more with what Terry Gou and others were coming up with than anybody else.

Jordan Schneider: Yeah, you have this line: “Nobody architected the move to China, but in one opportunity after another, Apple operations were lured into the country.” I think having a tri-continental strategy sounds ridiculous sitting here in 2025. But it’s also the case that you were scaling hardware manufacturing at an exponential rate for 20 years. If you’re making 10,000, 50,000 computers, it’s not insane.

But as you write about in your book, the number of actual items you’re making is increasing by a factor of 10 or by 50 or even by 100 on a year-on-year basis. It is getting more difficult and complex and the labor required is also rising at an exponential curve. You’re not going to find that many people in Wales and Singapore to do all this stuff.

Patrick McGee: I would emphasize another thing, which is that Moore’s law is decreasing the size of the products. Apple was not a big volume player in anything until 2003. I think in 2003, they made more iPods in a year than they made iMacs in the prior five years.

If you’re shipping things either by air or by sea, the fact that it’s the size of an iPod rather than a desktop computer is a major distinction. When SCI was building stuff in Huntsville, Alabama, it wouldn’t have made sense for them to move over to China and be early on this trend. That would make no sense at all. The computers were too big and bulky. Especially if you went back 10 years when it was a mainframe computer the size of a garage.

Once you’re into laptops, which is where Taiwan first makes its mark, it begins to make a lot of sense. Especially when you get into iPods and then dumb phones and especially smartphones, which then scale — that becomes a dramatic increase in efficiency. The cost per item — the logistics becomes de minimis once you’re doing it in the millions and let alone the tens and hundreds of millions.

Kyle Chan: Speaking of the logistics, I love the parts about United Airlines and the Boeing 747 playing a role in both directions. You’re talking about the product getting smaller and smaller to the point where you can have a Boeing full of iPhones that’s totally worth a flight, more than worth a flight. They’re helping to export this product around the world. At the same time, they’re providing the inputs in terms of the Apple engineers and the managers flying in.

It’s ridiculous — entire flights, I think you had documented, were for Apple staff to go back and forth between Cupertino and Zhengzhou or Shenzhen or other parts of China. It was astounding.

Patrick McGee: You’re finding this poetic symmetry that I didn’t make myself. Well done, Kyle. But to give that anecdote, because that’s pretty funny. I forget which cities it was and in what order, but in 2014 it’s probably Zhengzhou and in 2016 it’s probably Hangzhou, where the same thing happens where Apple convinces United Airlines: “Look, we need you flying nonstop to this place three times a week. It doesn’t matter if the rest of the plane is empty, we will pre-purchase so many of the first-class tickets that you will still make money.”

I think that’s hilarious. People who went to Shanghai and then took the bullet train to Hangzhou beforehand said it wasn’t exactly a schlep — it was pretty easy. But if you were an American engineer who had to go there regularly and you didn’t want to have to deal with the Chinese signs and all this kind of stuff, it was a nice comfort.. Because Apple had so many people going, it was worth them doing it. That was amazing.

The thing that was already out, but maybe people don’t know it, is in 2019, United accidentally leaked that 50 engineers from Apple were flying first class from San Francisco every day. They were the largest corporate client on the planet. When you talk about Apple’s size, there are many things you can point to that show they upend entire industries. But who knew that the airline industry was one of them?

Manufacturing the “Un-manufacturable”

Jordan Schneider: I want to talk about the design of the items themselves. You have this refrain over and over again where people say the contractors who don’t see the future ask, “Are you crazy? We can’t do this.” Then Apple comes up with a new manufacturing methodology that’s both labor-intensive and requires a level of sophistication that they have to teach to their contract manufacturers in order to execute.

Let’s do a case study. Pick your favorite item — which one haven’t you talked about on the other podcast that you want to discuss?

Patrick McGee: The translucent iMac is the one where the chapter is called “Unmanufacturable,” and that’s where I devote most attention. It’s also a computer. Everybody knows there’s one thing I hadn’t banked on — people say there’s a sense of nostalgia when they’re reading the early chapters because they’re going through some of the products they loved as high school or university students.

Jordan Schneider: Totally. I felt the same way. It was very tangible. I remember being 12 years old, going to my uncle’s house, seeing his candy-colored iMac and thinking, “What the hell?” It’s this object from space — but it’s not from space, it’s from these crazy engineers who were battling with Steve Jobs and the manufacturing people. Thank you for that.

These things are in design museums now and in your older cousins’ closets if they were smart enough to keep them.

For the younger members of our audience, this was the translucent iMac G3. Source.

Anyway, please continue, Patrick.

Patrick McGee: The computer was considered un-manufacturable by the product design team at Apple. The tooling engineers at Apple looked at what Jony Ive had come up with and said, “This can’t be built. I’m not saying it can’t be built at scale — we can’t build one of these in the lab.”

It’s interesting because they had created this thing that’s impossible, yet this is where you get the Apple mentality of “even though it’s impossible, we’re going to figure it out.” In the end, they don’t figure it out and the design has to be changed. But nevertheless, in hindsight, it looks like Jony Ive had tasked them with something that was impossible just for the sake of seeing who would stay on the project to get it done.

I’m not saying that was actually intentional — Jony Ive had to go back to the drawing board and make major changes to the product. But by the time they have a product that can be built, the only people left on the team are those who are willing to try and experiment with all these new things.

If people remember the computer, it has these translucent pinstripes on the front. Those used to be horizontal. What the engineers told me is that they couldn’t be horizontal because of how plastic injection molding works — the mold has to be parallel with the lines. Jony Ive had done something where they were perpendicular to the lines, and that couldn’t work.

They experimented for months. They had specialists come in from the outside to work on the product. What’s fascinating is that Steve Jobs had just come back. This was the product that, in my telling — because I’ve got all these Steve Jobs notes from the summer of 1997 — this is an interesting period where he’s given the role of CEO if he wants it, but he won’t take it. What his notes reveal is that he thinks Apple’s about to die, and he doesn’t want to oversee its demise. He is trying to help and see if it’s possible.

It’s a meeting of the minds with Jony Ive. Steve Jobs goes to the meeting expecting to fire him. Jony Ive goes to the meeting with his resignation letter in his pocket. Instead, they hit it off. What they come up with is the translucent iMac.

If you’ve seen the eMate 300, this translucent product they’d already made — it was a Newton and laptop mixed together — that was the first translucent product they made. You can see how Steve Jobs would’ve gotten excited by that and began to envision what he thought at the time was a network computer. That was their big idea.

They came up with this design. Steve Jobs got all excited about the idea of Apple making a comeback and redefining the aesthetics of the computer. But the first thing that Jony Ive comes up with literally couldn’t be made. The product design team goes through multiple iterations of who the leader is because nobody can do this, but Steve Jobs had inherited this team. He honestly didn’t know if they’re any good.

When they told him it was unmanufacturable and he goes on this tirade where he threatens to sell his “one last fucking remaining share of Apple stock,” he goes to his favorite design consultancy called Lunar. These two guys had actually designed the NeXT computer — that black magnesium cube, if you’re familiar with it. They go through all the blueprints and they agree with Apple’s team: “This cannot be built. This is not a quality design.”

I found this interesting because when I’ve read biographies of Jony Ive, even in grade seven, 12-year-old Jony is all about minimalist designs and elegant sketches.

For me, this was a real turning point where Jony Ive realized, “I need to know more about manufacturing to actually be able to give something that can truly be built.” He does have to change the lines. They do have to make the computer foggier than they wanted it to be because the inside was kind of ugly — the circuit boards were sticking out. That’s something they eventually changed by 1999 or 2000.

He had idealized this computer that couldn’t be built. But the changes that they made — I quote somebody else saying — they’re not major changes. The essence of the idea still lived. It’s still this egg-shaped thing rather than this angular square.

But it is the product that absolutely saves Apple. If they couldn’t figure out how to build it and how to change the design so that it was manufacturable — the term is DFA, designed for manufacturing — Apple wouldn’t exist. It was absolutely clear — the company was going to be bankrupt if this product wasn’t a hit, and it became America’s bestselling computer. It’s quite the dramatic narrative.

Jordan Schneider: I want to tie this to Ren Zhengfei 任正非. We’re running this episode after our series on Huawei, and this story is another one of these “leadership matters” anecdotes. The other thing that struck me about that is how far Steve Jobs was willing to go with the gimmick of, “Okay, this impossible thing — figure it out, figure it out, figure it out.”

There’s a lot of human pressure when all these people are saying, “No, you can’t do this.” Kyle, is there a Ren Zhengfei angle here?

Kyle Chan: Yeah, I think there is. Ren Zhengfei is famous for pushing his team to the limit and then some. There are all sorts of issues related to the work culture — people being driven too hard — but then that culture is carried with those personnel later on when they leave Huawei and go out to other parts of the Chinese ecosystem.

“Create advanced production methods to increase production!” Chinese poster, 1953. Source.

I see this parallel with a maniacal focus on realizing this concept, whatever it is — whether it’s going to be a cutting-edge Ascend AI chip or whether it’s going to be this candy-colored, perfect device that everyone wants to buy immediately when they see it in the store. That kind of drive is very reminiscent.

Terry Gou also plays a huge role in this story, and I see a similar parallel with his personality, although complementary in some ways, he’s so cheap, for lack of a better term. He’s focused on efficiency — he eschews the glitzy corporate HQ of Apple. But they share ultimately the same goal of producing at scale with ruthless efficiency these goods that will make all of them together a lot of money.

Patrick McGee: I’ve got to throw in the DAP here — the Divorce Avoidance Program — because so many marriages are broken up in the first five years of Steve Jobs’s comeback that the engineers are given these different policies to save their marriages. One of them is an understanding that, for example, there might be days when Jordan and Kyle aren’t going to be in to work because their marriages are on the line, and we have to give them a break.

That works well for a time, but we need Jordan and Kyle on the factory line. Instead, it becomes, “If we need to send them to Korea on a random Sunday even though they only got back two days earlier and their spouse is going to be upset, let’s give her $10,000. We need to assuage her because we don’t want to be losing engineers through attrition.”

That was happening to such a degree that all these policies needed to come out. It’s funny — well, I say it’s funny, it’s tragic as well. I’m getting the stories 20 years after the fact. People talk about the dollars being called “Danny Bucks” or “Dan Bucks” in reference to VP of Product Design Dan Riccio, because he was the person who fought for these things.

Sometimes our narrative is overly emphasizing how difficult the Foxconn workers worked, and that’s necessary. Obviously it gets to the point of suicides and suicide nets to prevent that from happening. But the Apple engineers were doing a similar number of hours while also having to fly back and forth all the time. It was causing so much strain that Apple had to institute these policies that engineers talked about. If you talk to anyone on the iMac project in the late 1990s and you mention the DAP, they haven’t thought about it for 20 years, but they immediately know what you’re talking about. Then they start telling stories.

I should say that one guy said, “Never mind the divorces, you need to look at the deaths.” That shook me. I was finding the DAP funny at times, and this person was saying there was nothing funny about it. He could rattle off names of people who had died on the production line or come back and died. I didn’t want to overemphasize that because as a journalist, I can’t determine, when someone died 15 years ago or 20 years ago, that overwork was the cause of death.

Steve Jobs himself says that the reason he thinks he got cancer is he was working so hard as CEO of both Pixar and Apple in 1997 that his immune system was weak and it allowed cancer to creep into his system. Whether that’s the medical diagnosis, I don’t know, but Steve Jobs understood that about himself or believed that about himself.

Jordan Schneider: Another interesting parallel is that the exit opportunities for these people existed. The compensation they were getting was not so different from what they could find at other firms from the Apple side that they couldn’t find other jobs. But there was something to this work that, similar to early Huawei, was not just the money. There is an aspect of team and mission and excitement that was around in this era of Apple that drove the Apple employees to give themselves and give all they had and more to this company and these products.

Patrick McGee: Jordan, let me agree on one part. I do think there’s a band of brothers, bunker mentality that would happen among the Apple engineers putting in the 18-hour days.

On the other hand, Apple later gets sued for collusion because there was an agreement among Silicon Valley giants not to hire from each other’s firms. One of the engineers I spoke with is someone who sued Apple later on because he didn’t understand at the time how being such a high-profile product design engineer at Apple wasn’t getting him job offers. He said, “I’m doing such good work. Why aren’t I being hired by Motorola or Google or whoever?”

Much later, he found out that was taboo. Steve Jobs had yelled at people and the emails came out later. I forget exactly what happened — I’m pretty sure they settled. The evidence was pretty good that there was collusion among the tech giants.

Kyle Chan: Let’s talk about the example of the CNC machines. One of the parts that struck me — I had heard a little bit of this before, but reading it in the book, I’ve got to read this short snippet:

“Apple dumbfounded the industry when it purchased more than 10,000 CNC machines in a single year, enabling a form of mass production that Steve Jobs called ‘a whole new way of building notebooks.’ Apple even made a deal with Fanuc Automation Group from Japan to purchase its entire pipeline of CNC machines for years to come, shutting out all of its competitors from access, and then scoured the globe for more. ‘There were not enough CNC machines in the world to do the machining that we needed to do,’ one person said.”

I was struck by the scale and also by the — again, this is reminiscent of Elon Musk — “We need this part now. Whatever it takes, fly it out personally, make it happen.” For this Apple story, for the aluminum single-body MacBooks, it’s what it’s going to be. If it takes 10,000 CNC machines, then so be it. We’ll fly them over. Those stories capture the “whatever it takes” mentality.

Patrick McGee: We’ve jumped ahead to 2008 here. This was an achievement of imagination. There’s nothing novel about a CNC machine — I forget how many decades they go back. But the thing is, they build prototypes. Apple had the wherewithal and the imagination to say, “Well, wait a minute. If these things are so good, why don’t we build all of our products using these things?”

My focus is on the MacBook — they got the unibody MacBook — but they’re used in all sorts of products, including the iPhone, and have been ever since. These are machines that individually cost between $500,000 and a million dollars. Who knows what Apple is paying when they’re able to buy them at scale and they have someone like Tony Blevins negotiating for them.

This is so successful that this is the first time that MD (manufacturing design) is put on the map. In other words, it’s the first time Steve Jobs, in a keynote presentation, talks about their manufacturing prowess. This is probably the time — I don’t know this for a fact — that MD becomes MD. Before then, they were a part of product design, and it’s called Supply Base Engineering. Now they’re a part of ops.

It’s interesting that MD doesn’t have their own senior vice president. I don’t know exactly why it’s a part of ops — they should be their own thing — but it’s possible that Apple doesn’t love the narrative of more people knowing about this division because they are the ones where, when my book talks about the geopolitical influence that Apple has, it’s because they’re flying MD engineers who are brilliant, usually based in California, over to Asia to train and audit and supervise and equip these third-party manufacturers with what becomes billions upon billions of dollars of machinery.

This is where I get into the argument that Apple’s influence is like that of a nation-building program.

iPhones and Nation Building

Jordan Schneider: I have a smaller question — more on the design stuff. As the design process is being incredibly creative with the CNC machines, one of the questions I have is to what extent they could have optimized for design features which would have given them more geographic flexibility?

The degrees of freedom that Jony Ive and company were able to operate under are unique in the industry. As we've discussed with the CNC Apple Unibody story and the translucent iMac, they were willing and excited to push the physical design features of their products to the point where the answer at the end of the equation had to be China and Chinese suppliers.

I'm curious whether there were other pathways where they could have introduced different constraints at the design phase. This approach might have allowed them to manufacture more products in different places around the world without losing as much capability. The alternative would be trying to manufacture something like the unibody MacBook—as it was designed to be constructed in 2008—somewhere else outside of China, which would result in significant compromises.

Patrick McGee: That’s a great question. One question that you get now is: why doesn’t Apple automate these processes so that it’s not dependent on China, where you’ve got 400,000 people building an iPhone? That sounds like a rational thing to do.

But what you would be doing is upending how Apple designs its products, which is to say that Jony Ive’s team — and I’m aware that Jony Ive has left, but we’re talking about the golden era here — they could operate with a “no constraints” mentality. In other words, other companies will say from the get-go, “This is going to be automated, so we’re going to design for automation.”

The degrees of freedom under which Jony Ive and company were able to operate is unique in the industry. As we’ve talked through with the CNC Apple unibody story, as we’ve talked through with the translucent iMac, they were willing and excited to push the physical design features of their products such that the answer at the end of the equation had to be China and Chinese suppliers.

But I’m curious if there were other pathways where they could have introduced more or different constraints at the design phase such that you could have at the end of the equation been able to manufacture more stuff in different places around the world without losing as much — without losing as much as you would if you were trying to manufacture the unibody MacBook as it was designed to be constructed in 2008 somewhere else outside of China.

I like to ask people what their favorite Dell computer from the early 2000s was. Of course, none of us has a favorite Dell computer from the early 2000s. It’s not only the money that Apple’s making and investing in China that distinguishes them from others, because they were a small player in the early 2000s.

What distinguishes Apple is what they are doing and the fact that nobody knows how to build it. It’s not a knock against China to say that Apple taught them a lot. When you think of Apple products — whether it’s multi-touch glass or the anthropomorphic metal tubing between the base and the computer for the Sunflower iMac — nobody knows how to build that stuff. I’m not saying the Chinese didn’t know how; nobody knew. The marriage of Apple and China is skill and scale coming together.

The line I use is that Jony Ive and Steve Jobs made Apple products unique. Terry Gou and Tim Cook made them ubiquitous.

Now, if you had a different mentality where you said, “Let’s make these things easier to manufacture, more automation-friendly, to give us geopolitical safety,” that absolutely could have happened. You could have had more boring products, fewer design changes year to year, and you would’ve been able to have a more resilient supply chain based on production in Mexico, for example.

In that world, the first iPhone, if it were built at all, absolutely would have a plastic screen rather than a glass screen. It was only by working with the Chinese factories at their scale and industriousness that Steve Jobs was able to make dramatic changes. Literally after he has presented the iPhone — when he famously announced the iPhone in January 2007 — he was holding a phone that has a plastic screen. It was in the period between him announcing it and it going on sale that Apple upended the entire thing.

They work with a Taiwanese supplier called TPK, which builds the invisible circuitry within the glass so that your finger actually causes an effect. You’ll notice that if you do that on your window, it doesn’t do anything. You need to have a bunch of technology in there. Then they worked with Lens Technology to shape and temper the Corning glass that everyone’s familiar with.

The glass is American-made, but then it has to be shaped, tempered and cut. There are tens of thousands of people working at Shenzhen Lens to do that. You have people like Steve Zadesky, who’s on a whole bunch of iPod and iPhone patents, literally sleeping on the factory floor to get that all done.

China offered something that nobody else has offered. The reason why the epigraph to the book is a “Made in China 2025” document from 2015 that says:

“Without manufacturing there is no country and there is no nation.” Find me another country that has that as their motto, their mantra. You don’t get that just anywhere.

Sometimes I find it amusing that I have a 90-second sound bite on X and someone will say, “This guy’s saying China couldn’t be anything without Apple.” It’s the total opposite. I’m saying Apple is screwed without China because nobody will offer the things that they offer. They were a once-in-a-century partner to respond to the demands of Jony Ive.

If I’m fast-forwarding a bit, the only thing I would say is that Apple thought they were wearing the pants in this relationship for the first two decades — from 2000 to 2020. It’s only afterwards that there’s this realization that, “Wait a minute, we were lured in by the siren call of an emerging superpower who had all their companies willing to do all these things so that they would learn all the technology transfer and put America in a position, and all other industrialized nations in a position, where none of them can compete.” Everything has gone into China.

Kyle Chan: Can we talk about learning and this process of learning and technology transfer? To me, this is such a big deal — not only for Apple, but Apple exemplifies a lot. All these foreign firms that China in many cases very deliberately tried to bring into the country, tried to attract — maybe it’s by the central government, maybe it’s by local government, maybe it’s certain incentives, maybe it’s a whole concerted effort.

“Study the Soviet Union’s advanced economy to build up our nation.” Chinese poster, 1953. Source.

The point was to bring in these foreign firms, have them manufacture in China, share know-how, and train a whole generation of not only workers, but engineers and managers. The story of Apple and China captures that fantastically.

To give you a contrast, I love the quote: When Apple was working with Sony in Japan, you had an operations executive from Apple leaving Tokyo saying, “I was in Japan about five minutes, and Apple can teach the Japanese nothing.” This was earlier in the story, and this was a difficult product, but not as difficult maybe as some of the later ones. That story was not one of technology transfer.

Whereas later on, especially for the case of China and Foxconn, there was a very deliberate effort to use Apple as a source of upgrading your entire industrial ecosystem — upgrading your suppliers, upgrading your machinery, your equipment, all of the tacit knowledge, not only the blueprints and stuff that you can license and patent. Some of this stuff was not anywhere in existence. It was pioneered — they were all operating at the forefront.

Can you say more about this process of learning? I read a lot and talk a lot about, from the Chinese side, their very active process of trying to get as much as possible from every kind of foreign firm in high-tech industries. This process of technology transfer and learning.

Patrick McGee: In the late 1990s, Apple actually struggled to find contract manufacturers for its products because they were not seen as worth the effort. We have to remember that Apple was creating things in pretty small volumes, and they were very demanding. The workers at these Chinese factories often did not like working with Apple because you’ve got these overbearing engineers who are dictating all these things. Usually they’re dictating it through translators because these are American engineers who don’t speak the language.

I have these funny anecdotes that aren’t in the book where some engineers have a free weekend, so they decide to go to this factory that they’ve been training up. They cause such hell that when they’re trying to leave the following day, there’s a problem with the car — the wheel is flat or something. As the car drove, it was causing sparks to fly. They’re thinking, “I guess we’ll have to stay here another day.” Then they get a call from the leader of the factory who demands that the taxi driver drive even with the sparks flying and everything, because God forbid these Apple engineers come back because they’re such a pain in the neck.

This is the sort of level of training that Apple is providing — the sort of stuff that makes you go mad. Yet obviously they’re very, very effective.

My narrative about Terry Gou is that he is the first person to understand: “This is not about making big volumes necessarily, and this is certainly not about making margins, but we’re getting tuition-free, on-the-ground training from these overbearing engineers who are willing to teach our cohorts all sorts of stuff.”

Apple’s orderly meets its match in Terry Gou and vice versa. This on-the-record anecdote from Tony Fadell is that he and his team would come in to train the engineers they’ve been working with and find not a face that they recognize because Terry has taken these graduates and put them on the Dell line or whatever, where they can use their newfound skill sets to actual money-making effect. He’s, in a sense, without Apple’s permission, started a new semester for these other engineers.

I don’t think the Chinese understood that until 2016. That is a key reason why you’ve had Foxconn being squeezed out the last couple of years and they’re having to go to India because Apple’s essentially telling them, “If you want to keep your market share, you need to go to India for us.”

The learning is absolutely fundamental to the relationship between Foxconn and Apple and to why all the Chinese factories are able to do so much. If you want, we can go into the wonky differences between ODM and OEM and why Foxconn does so much better than Quanta or Inventec or anybody else.

Jordan Schneider: Sure, let’s tell that story.

Patrick McGee: In the late 1990s, none of the Taiwanese wanted to be doing assembly. Assembly is how you get the orders, but it’s very low margin, it’s totally cutthroat. The likes of Inventec and Quanta and the predecessors to Pegatron are doing something called ODM work. “D” is for design, and design means more margin, and it also means more investment. It’s an investment that typically pays off.

What it means is that if you’re a Western manufacturer of a computer, you’re already offloading your manufacturing to a third party in Taiwan. But what this will allow you, if you work with an ODM, is that they’ll also do the research and development, and they’ll also do the design. You have less and less to do. It gets to the point where the Taiwanese can literally show a catalog of computer designs, and you choose one. Imagine how much that takes off your balance sheet.

It works very well. The problem, of course, is that the Taiwanese begin branding their own computers and competing against you. If there are times of scarcity of a certain product, they’re going to source it to their own companies rather than to you as the third party.

Apple never falls for this trap, in part because they’re wanting to do so much design themselves. It’s never appealing for someone to say, “We’ll do the design for you.” Foxconn maintains itself as an OEM, a manufacturer — the manufacturing-as-a-service model. They were condescended to because it’s not sexy, high-margin work. Yet Terry Gou has an understanding of Chinese politics that nobody else seems to have.

Kyle mentioned that this is what we get wrong about Chinese communism: it’s not only about the relationships with the federal officials, it’s the provincial officials because the cadres in various districts are incentivized for factory growth. They will offer tax exemptions and bonded zones and tailor-made policies and all sorts of stuff — getting in the labor from the hinterlands or whatever — so that your investment goes to that district and not some other cadre’s.

It’s worth knowing that that is one of the key distinctions between Soviet and Chinese communism. The Soviet system was top-down.

I don’t know that you’d call Chinese capitalism grassroots per se, but I compare it to federalism on steroids. It’s something that we miss all the time.

The reason why being an OEM is important is the labor intensity and the vertical integration that Foxconn introduces. Instead of doing design and R&D, they are just building. What makes them so good at building is having many clients and then allocating those resources. That allows Terry Gou to expand — literally in a real estate sense — building dormitories, building entertainment venues, and having these migrants set up shop in Shenzhen all the time. If you’re doing that, you are making the local cadres in Shenzhen look good.

He’s able to parlay his brilliant investment strategy, his labor intensity, into free machinery, free tools, more labor migrants, etc., in a way that nobody else is able to take advantage of. Apple didn’t understand Chinese politics until 2013 when they were made to, but Foxconn is their biggest partner and Foxconn understands it very, very well.

What’s tragic is Foxconn was never rewarded very much for this. Their margins fell the closer they get with Apple, but the political connections that Foxconn made were instrumental to Apple’s rise.

Made in China 2025

Jordan Schneider: I want to talk about that — the getting rewarded thing. I asked ChatGPT for the market cap of Apple and all of its near-ish competitors. We’ve got Apple at $3.25 trillion. We have Samsung at $300 billion, Xiaomi at $175 billion. Huawei, no one knows, but maybe $100-200 billion. Oppo and Vivo are rounding errors.

We have this big argument. The central thesis of your book is that this was a dangerous trade for America, but we did get a $3 trillion company out of it, even though there was all of this industrial upgrading that Apple helped China do. Setting aside the getting Chinese people out of poverty angle to this, there’s got to be something to be said for the fact that America is now home to one of the largest companies that the world has ever seen.

Patrick McGee: I’ve tried to answer this. If for whatever reason we were living in a simulation and the game ended now, then obviously Apple got more out of this relationship and it was a great deal for everybody, and we lived through a golden age of Silicon Valley focusing on software and China doing the hardware.

The problem is it’s probably not a simulation. The world continues and we’re now at a point where Apple has no Plan B. China for whatever reason has become more belligerent and wants to be self-sufficient. “Made in China 2025 中国制造2025” is a grand master plan to sever itself from the West in terms of automation, robotics, electronics, and other things including pharmaceuticals.

It seems pretty crazy to me that the world’s most valuable company doesn’t have any sort of Plan B. The Tim Cook doctrine, named in 2010, maybe 2011, was to own and control as much as possible and only enter markets where they could be number one or number two. As Ben Thompson from Stratechery constantly points out, the fact that everything’s manufactured in a fairly hostile country is the biggest violation of the Tim Cook strategy that could possibly exist.

Kyle Chan: Another parallel is the way that all of these U.S. chip designers depend on TSMC. I tweeted earlier — and it’s funny, the numbers still hold a year later — but Nvidia, Apple, Microsoft, Google, Amazon, and Meta have a combined market cap of $15 trillion, give or take some trillions. Yet they all rely on TSMC for advanced semiconductors, for AI chips, for smartphones.

It’s incredible that there’s this amount of value that’s created on the American side from these companies, purely from design. All those companies I’m talking about, maybe with the exception of Tesla, outsource all the manufacturing. For the high-end chips, that goes to TSMC. TSMC itself is probably one of the most valuable companies in Asia, but itself is an order of magnitude less.

Source.

There are two big questions. One is, how did we get to this point where we were allowing global risk to be concentrated so much on a single firm or a single country? Same thing with Apple and China. But then on the other side too, do you lose something from separating design and manufacturing?

Your stories about Apple engineers flying back and forth, deeply embedding themselves with Chinese suppliers, bringing the machinery, doing this — you have this phrase, “man-on-mirror” structure that I love. That was very interesting. That was actually for Korea, but a similar model later on for China, maybe 10x. How much can the two be separated and how much can you still be innovating at the cutting edge when you know your manufacturers are somewhere else, maybe in a different country altogether, maybe 6,000 miles away?

Jordan Schneider: I want to push back on that characterization. It’s a misnomer to say NVIDIA is a design firm. They’re constantly flying back and forth to TSMC. The innovation required to create not only the chips but these entire racks is incredibly manufacturing and hard tech intensive. It’s not only sitting with EDA tools and rearranging where circuits go. A similar story can be told from their perspective.

Patrick McGee: Andy Grove, Intel’s co-founder, gave a great interview to Bloomberg Businessweek in 2010, where he lamented that everything had left America’s shores. He said we’re going to miss out on shop floor innovation — that if you’re not in the factory working on the next product, you’re going to lose out. He was concerned about the shortsightedness of it, and that was quite a profound lesson.

I come back to this later in the book to highlight how any number of Western companies completely outsource without knowing how to build things anymore. It’s not clear to me that Facebook has any idea how to build the Portal. The Portal isn’t a device that’s out there anymore, but as I understand it, it was a product that was completely outsourced to Asian companies who knew how to do it.

Apple is in a different boat — they do know how to build things. They have the experiential know-how built on proprietary processes. We know about them doing iOS, but they also do software and operating systems for machines. We don’t know about that because we’re not on the factory lines.

The problem, as I quote from engineer Michael Hillman who has 16 years of experience, is that to execute any of those plans, they need China. The line I often use is that an iPhone has roughly 1,000 components. If you’re building a million iPhones a day, that means you’re managing the logistics, manufacturing, and production of one billion components daily. There’s one country on earth that’s able to execute that. Anyone else will take at least 10 to 15 years to reach that stage. Because China is an increasingly belligerent country, they won’t let that happen.

If you follow high-quality engineering on a map, America had total dominance at the end of World War II. They began teaching the Japanese to build radar and electronics industries. When they hit supply constraints and the yen rose in value, Japanese entrepreneurs went to their former colonies — Taiwan and South Korea — to do the same thing. When Taiwan hit its own constraints and had better relations with mainland China, the Taiwanese entrepreneurs 台商 went to China and built immense manufacturing capacity over the last four decades.

The next natural place to go would be India. Vietnam to some degree, but Vietnam doesn’t have the size. India is the next logical place if you play this game. But no one knows that better than Beijing, and they want technology transfer to be a one-way gate. The information comes in, but it certainly doesn’t go to Karnataka.

That’s where Apple has this big predicament right now. It makes perfect sense to set up production lines in India, but the machinery is now produced in India and blocked by Beijing. The experiential know-how is in the minds of Chinese engineers — from low-skilled workers to PhDs — and they can’t get visas to India. That’s why Apple is captured. I’m not using that as a line to sell books. I don’t know what strategic decision Apple can make that would untie the knot with China. The more plausible scenario is that they retie another knot and end up with a bifurcated supply chain. But that’s not in China’s interest, and if it’s not in China’s interest, then good luck to Cupertino — I don’t know how they’re going to execute that.

Kyle Chan: This is a very big topic in the scholarly and development world about whether this whole “flying geese” model — Japan, South Korea, Taiwan, China’s moment — whether that ends with China. Apple is perhaps one of the best examples, but there are many other areas where maybe the auto industry will end up in China and there’s no next flying geese destination. Maybe a little bit to Vietnam and Thailand for transshipment to get around tariffs.

India is always the question. India wants to be next and has been waiting for a long time, watching China’s rise in anguish, thinking “that should be us.” They’re making a big push with manufacturing and have their own major industrial policy, especially targeting iPhone and consumer electronics manufacturing. To some extent, there seems to be success on paper — the oft-quoted statistic that 20% of iPhones are now assembled in India.

But there are big questions about how much of the real supply chain can be brought into the country the way China had done, whether that’s bringing in foreign firms to invest in manufacturing or cultivating domestic suppliers.

Jordan Schneider: The case study I’m looking forward to seeing play out over the next few years is wearables. Meta made all of the Oculus devices in China, and a few months ago, they said they’re going to try to get GoerTek, their Chinese contract manufacturer, to do half of it in Vietnam. Good luck with that.

Google’s now getting into the game. They had a big announcement a few months ago and did a big talent acquisition of HTC, which was a Taiwanese-based VR company, and are obviously trying to price all this in and diversify.

But you have six or seven Chinese manufacturers who are ahead of the game when it comes to what’s currently on the market in the West — when you look at Meta Ray-Bans as the main thing people have been exposed to. There’s that dynamic, but beyond creating redundancies around the world, it’s clear that doubling down on China will get you to the frontier of technology faster than trying to be cute in different countries.

The Apple Vision Pro — not a huge commercial success, but an incredible technical accomplishment — give me a break if you think you can try to make that at scale or with anywhere near the level of policy and precision that Chinese contract manufacturers were able to deliver.

Patrick McGee: It’s worth noting that’s Luxshare, not Foxconn — the rise of the red supply chain in action. The other example I’d point out is Sam Altman and Jony Ive coming up with their next device. Where’s that going to be built? If you’re Jony Ive, you might demand the best, and the easy thing would be to say, “Of course we’re going to work with the Chinese partners that he knows from three decades at Apple.”

On the other hand, he must be more awake to the issues I point to in the book than anybody. He might very well know the political agenda of Trump and all the ways that could be used in his favor to build out a nascent supply chain here.

It’s much easier to do when you’re beginning to build units literally in the tens because you’re doing prototypes, rather than in the quarter billion that Apple deals with — Apple’s scale is so big that sometimes you get off by an order of magnitude.

Apple in China, Future Edition 養虎為患

Kyle Chan: The phrase I kept thinking about was “training your replacement.” Apple is different, but there’s this pattern we notice repeatedly where China brings in foreign firms attracted by China’s efficient, nimble manufacturing base. They set up shop, and then — maybe it takes decades — you end up with a whole crop of Chinese competitors that first nibble away at your market share within China, then start expanding and going global.

Source.

You see this happening right now with the EV industry and Tesla. There’s a whole debate about how much Tesla turbocharged China’s EV industry or how much was already there, with Tesla playing a catfish role.

Reading your book, Apple had some of this problem. Samsung is the most famous long-term global competitor, but also Huawei and then a whole fleet of Chinese smartphone competitors — Xiaomi, Vivo, Oppo. You talk about how Apple kept trying to innovate on both design and manufacturing to stay a step ahead of the competition, at least up until recently. They were able to maintain their position in the global market, even within China, even with all the pressures of domestic competitors, in the high-end smartphone market.

Could you talk about this problem and how Apple dealt with this feeling of potentially creating the very competitors that become existential threats in other industries?

Patrick McGee: This is a great segue into one of my favorite chapters, Chapter 36, called “Five Alarm Fire”, because this is what happens when Apple realizes the competitors in China have caught up with them.

What’s amazing about the chapter is that it’s exclusively based on court documents that were made public but were never found by another journalist. I found them 16 months ago, but I’m the only person who’s written about them, even since I published the book nobody has gone to do their own deep dive. We’re talking about more than a thousand pages — depositions of Tim Cook and others, internal emails between Tim Cook and the board of directors at Apple, between him and production staff and his underlings.

The entire issue is about how the iPhone XR (pronounced “ten R”) is a dud, and they know it’s a dud. It’s geared toward the Chinese market, yet the Chinese are all deciding to buy a competitor from Huawei. Apple knew this for several weeks and then told investors that everything was fine.

The chapter is called “Five Alarm Fire” because that’s what a VP of Sales or Finance described the situation as — fire department jargon for when at least a hundred people are needed on the scene. The quote from Tim Cook, at least a week before he talks to investors, is: “This is a disaster. We need all hands on deck now.” There are people within Apple who are panicking about this.

What happened is that the tenth-anniversary iPhone is the first one with that infinity pool design — you get rid of the thumbprint and have Face ID. That didn’t ship until November 2017. Then, by April 2018, four or five months later, all four of the major Chinese brands have an infinity pool-like rival. Cupertino was expecting to bask in the glow of this amazing device, and literally everybody caught up.

You realize the Chinese went from mimicry to — maybe they’re still in the stage of mimicry there — but they’re surprising Apple with how quickly they’re able to keep up. Now, fast-forwarding, they’re in a stage where they’re doing things that Apple isn’t doing. The example I always point to is the Mate XT from Huawei, which unfolds twice. There are also silicon carbide batteries in the latest Vivo phones that have better density, faster recharge time, and longer duration.

Apple is no longer the company to pay attention to if you’re looking for the most cutting-edge phone. There’s something disquieting about that, especially because not only are the Chinese manufacturing them, but they’re doing the industrial design and product design. They’ve taken that pyramid structure I described in the book, put it organically in their own companies, and are now outmaneuvering Apple on multiple levels of the pyramid.

Kyle Chan: Now we’re seeing this especially with AI features coming out. Huawei is doing a lot with its Harmony operating system, trying to have AI everywhere you want it — and maybe in many places where you don’t, like summaries where you don’t want summaries, maybe mimicking what iPhone users are getting used to now with Apple Intelligence.

There’s this big question mark about whether Apple will be able to catch up and be on the cutting edge in that space. There are recent issues with linking up with Alibaba, for example, for an AI partner in China. If you don’t have that kind of partnership, whether as Apple, you can still be competitive in the Chinese domestic market.

Patrick McGee: You brought this up because this is mostly post-manuscript for me — it’s not in the book, yet it’s current and dynamic as a discussion. First, Apple has done poorly with Siri. Siri might have been amazing 14 years ago, but it hasn’t kept up. It’s not even a large language model, so it’s not doing the most basic stuff.

Because of Siri’s failure, Apple has to work with a partner. Because ChatGPT or anyone else, such as Perplexity AI isn’t available in China, they have to work with a Chinese partner. In a sense, that’s the closest thing we’re going to see on the product side to Apple having a joint venture — maybe it’s not technically a joint venture, but my book is about how Apple consolidated all hardware engineering into China over the last 25 years.

Are you telling me that, over the next 25 years, they’re going to work hand-in-glove with Baidu or Alibaba and essentially do the same thing with AI?

This is already the stuff of movies regarding existential risk to humanity. We’re talking about the most complex, consequential thing in tech, if not in business or anything writ large. This is up there with fire and the computer. I don’t know how much sense it makes for Apple to be training one of these companies to make their product better, which they would inevitably have to do. They’re not going to use an off-the-shelf AI tool.

Jordan Schneider: I don’t know if they can train themselves to make their product better when it comes to AI.

Patrick McGee: That’s a great point. We are at the stage where you’ll buy your next phone based on the AI features it has, and Apple is very much failing there. We are recording this during WWDC, so who knows? Maybe we’re way wrong by the time this comes out, but I feel like that’s not going to be the case. They’re behind in AI in an embarrassing way.

Jordan Schneider: One more detailed question before we do a closing one or two. The $55 billion annual investment in China — can you deconstruct that number? Where did it come from and what is included?

Patrick McGee: Apple was on the back foot in 2013 when Xi Jinping 习近平 came to power. They’re worried that either their products are going to be blacklisted or they’re going to have to form joint ventures to continue operating in the country. This is total anathema to Cupertino — not what they want to happen.

This team of people that came into the company, either hired or appointed, called themselves “the gang of eight.” They’re the first people who are senior, living and working in the country. This is where you pass the baton from Foxconn to Apple. Apple realized: we can outsource our manufacturing to Foxconn, but we cannot outsource our political relationships.

They take ownership of this and do their own supply chain study effectively in tandem with the government affairs team. What they realized is that they’re investing $55 billion into Chinese factories — into factories operating in the country.

China has this concept called “registered capital.” Apple, in any given year, doesn’t make as much money as Walmart, and Walmart gets 60 to 70% of their products from China. But Walmart’s not making investments in the country. If you’re importing kitchenware and action toys from China, that doesn’t count as registered capital or investment — that’s just spend. You’re buying that stuff, putting it on a ship, and selling it in Arkansas.

If you were Volkswagen or GM, you’re training workers to run a production line. For a certain period — 18 to 24 months — it counts as an investment into the country. The training costs of setting up a production line count as fixed assets, the same way that putting machinery on the production line is counted.

Apple realizes their investment is more like Volkswagen or GM’s than Walmart’s. They’re not sitting in Hong Kong ordering parts from Shenzhen. They’re going into the factory and training them how to do all this stuff. Their training costs, which are effectively wages, should be counted as fixed assets, as registered capital.

But they go further. Unlike Volkswagen or GM, who set up a production line and have it running for seven years because the lifecycle of a car is about seven years, the iPhone never gets to that stage where it’s running. They’re always upending the design of every product in the portfolio. They’re able to count a whole lot of training costs and wages, in addition to billions in machinery that’s put on the production line, as registered capital.

Someone could debate whether all that spending should actually be counted as investment, but it’s not an argument with me — it’s an argument with Apple and with Beijing, which accepted the argument. That’s where the money comes from, why it’s counted as registered capital, and what distinguishes them from other companies that don’t do the training. Other companies are buying off-the-shelf parts, whereas Apple is working hand-in-glove with hundreds of factories.

That’s why they’re able to operate without a joint venture — because they can say, “You have no idea what influence we’re having on hundreds of factories across the country.”

Jordan Schneider: For Kyle and Patrick, the contract manufacturers that Apple shepherded — do you have a sense of how their relationship with Chinese handset manufacturers is similar or different?

Kyle Chan: There’s a lot of overlap. You can break it down — Lens Technology is a supplier not only to other Chinese smartphone manufacturers but to others as well. GoerTek and some of these suppliers supply across the board.

The order of operations depends, but some of them were brought up by Apple and then helped the rest of the Chinese supply chain. Some of them were already serving what were, for them, lower-tier customers like Oppo and Vivo before they got their Apple contract. It’s a mixed bag.

I also love the part where you mention Apple trying to navigate Chinese politics, because good luck to anyone trying to navigate that.

They tried to cozy up to Sun Zhengcai 孙政才, who was the party secretary of Chongqing. Apple tried to invest in data centers there, which at a local government politics level is great — you want to show that you’re bringing in investment, generating economic activity, especially when it comes to anything high-tech. That’s a bunch of gold stars.

That turned out to be a mistake because Sun didn’t end up becoming the successor to Xi — it turned out Xi was the successor to Xi. Sun didn’t even end up in the very top echelon of party leadership.

There are other ways where Apple has been incredibly adept, especially Tim Cook, at navigating Chinese politics and American politics at the same time. It’s not a job I would envy, although he’s probably pretty well compensated.

Mood Music: