Apple Isn't Leaving China. It's Threading a Needle Between Two Superpowers.

In 2009, you could have built an iPhone almost anywhere and changed almost nothing. China's entire contribution to the device was the final snap-together: $6.50 of assembly labor on a phone whose parts came from Japan, Korea, the US, and Germany. That $6.50 was roughly 3.6% of the phone's gross export value.³ The country that the world would soon call 'the iPhone factory' was, at the start, the cheap finishing station at the very end of the line. Fifteen years later, 157 of Apple's 187 disclosed suppliers manufacture in mainland China⁶, and a Chinese provincial city earned the nickname 'iPhone City.' Nobody decided this in a single meeting. It was built one cost-saving decision at a time, until leaving became the hardest strategic problem the company has.

The official story now is that Apple is decoupling from China - moving its supply chain to India, getting out from between Washington and Beijing. That isn't what the numbers say. In the same fiscal year Apple was loudly diversifying, it added eight net-new Chinese suppliers and removed four - the first net increase since 2021.⁶ Decoupling is the press release. Diversification is the reality. And the two are not the same thing at all.

Here is the thesis, plainly: Apple's China problem is a dual-exposure trap that Tim Cook deliberately built for cost advantage, and the two exposures - China as a market and China as a factory - cannot be exited the same way, or on the same timeline. India can take over assembly. It cannot recreate the component ecosystem for years. So Apple isn't leaving. It's threading a needle between two superpowers, and the needle gets more expensive every year.

How $6.50 became the spine of the supply chain

The dependency was not an accident of geography; it was an operating philosophy. Tim Cook met Foxconn's founder, Terry Gou, in 2000 - seven years before the iPhone existed.⁸ The partnership came first; the product that would define it came later. That ordering matters. Apple didn't pick China because the iPhone needed China. Apple built a manufacturing machine, then poured every subsequent product into it. Foxconn's Zhengzhou complex grew so large it became a city in function as well as nickname.⁸

And as the machine ran, China climbed the value chain underneath it. In the original iPhone, every key component came from Japan, the US, Korea, and Germany. By the iPhone X, Chinese firms were supplying antennae, wireless charging systems, and circuit boards - around 25% of manufacturing costs.⁴ The per-phone Chinese component content rose from that lonely $6.50 of assembly in 2009 to roughly $104 of parts by 2018.³ This is the quiet part of the story: the dependency deepened not because Apple moved more assembly to China, but because China moved up from the finishing station into the parts themselves.

The two exposures don't exit the same door

Strip the dependency down and you find it is really two separate risks wearing one coat. The first is the market: Greater China is Apple's third-largest region, $66.95 billion of FY2024 sales, behind only the Americas and Europe.¹ The second is the factory: the place where the phones get built and increasingly where their parts come from. People treat these as one number. They behave like two.

The market exposure is shrinking on its own - and not in a way Apple welcomes. Greater China's share of revenue fell from 18.9% in FY2023 to 17.1% in FY2024, with sales dropping from $72.56 billion to $66.95 billion, driven primarily by weaker iPhone and iPad sales.¹ On the Q1 FY2025 earnings call, China revenue fell about 11% year over year, and Cook named a specific culprit: Apple Intelligence isn't available in China at all, and a reported deal with Baidu to bring AI features there had stalled.¹⁰ So the market risk is real, but it's a competitive problem - one Apple could lose to Chinese rivals selling smarter phones.²

The factory exposure is the harder one, because it isn't just about where phones are snapped together - it's about the ecosystem of suppliers feeding the line. Apple's 2023 data shows 87% of its 187 suppliers have production facilities in China, and mainland China hosts over a third of all supplier factory locations.⁵ That web is what 'iPhone City' really means. And in 2022, the world saw what it costs: a COVID lockdown of Foxconn's Zhengzhou plant triggered worker walkouts and clashes with guards, delayed iPhone 14 Pro shipments straight through the most important quarter of the year - and Apple had no adequate backup plan.⁷ One city caught a cold and Apple's holiday season fell over.

Why India relieves the pain but doesn't cure it

India is the most visible piece of the response, and it is genuinely large. Apple assembled roughly $22 billion of iPhones in India in the 12 months to March 2025 - a nearly 60% jump year over year - and production rose 53% in the first half of 2025 to around 23.9 million units.⁹ This is not a press-release gesture; it's a real second base. But notice what India is doing: assembly. It's taking over the finishing station - the same job China started with in 2009. The high-value parts mostly still flow through the Chinese ecosystem that took two decades to grow, and you cannot clone that web by signing a lease in Tamil Nadu.⁹

This is the asymmetry that makes the trap so durable. The lowest-value layer - assembly, where labor is a tiny fraction of the phone's cost - is exactly the layer that's easiest to move. The component layer, where the real dependency lives, is the layer that's hardest. Apple can hit a headline percentage of iPhones 'made in India' while its supplier list in China keeps growing. Both things are true at once, and they are not in contradiction. The needle isn't moving from China to India. It's being split - assembly going one way, components staying put.

Isn't the China problem solving itself?

The fair objection runs like this: the market is shrinking, India is scaling, the share is already down to 17%. Give it a few more years and the dependency takes care of itself. There's truth in it - the trajectory on assembly is real, and revenue concentration is genuinely falling.¹⁹ But the objection conflates the two exposures it should keep apart. The market share is falling for the worst reason - Apple is losing ground to local competition, not gracefully de-risking.¹⁰ And on the factory side, the data points the other way: Apple added more Chinese suppliers than it dropped in FY2023, the first net increase in two years, and that list covers 98% of its direct manufacturing spend.⁶ A company that was truly exiting would show a shrinking supplier count. Apple's is growing.

There's also a comforting half-truth worth puncturing: that China 'owns' Apple's manufacturing and captures the value. It mostly doesn't. Of Apple's semiconductor suppliers, only about 4% are headquartered in China, while 63% are US-headquartered; the high-value work done in China is largely performed by Japanese, Taiwanese, and American firms operating there.⁵ Chinese-owned firms have climbed, but they still cluster in the lower-value segments. So the dependency isn't on China the owner - it's on China the location, the place where the whole ecosystem happens to be physically assembled. That's a subtler hostage situation, and a stickier one. You're not negotiating with a supplier you could replace. You're tied to a piece of ground.

Apple spent twenty years building the most efficient manufacturing dependency in corporate history, and the bill for that efficiency is now arriving as geopolitics. India will keep taking the assembly layer; the supplier list will keep adding Chinese names underneath. Both will be true, and Apple will keep calling it diversification because it is - just not the kind that ends the dependency. The company is no longer choosing between China and not-China. It's paying, quarter after quarter, to stand in the narrowing gap between two superpowers and keep the line running. The $6.50 finishing station turned out to be the cheapest part of the whole story. The expensive part is leaving.