High-tech manufacturing in India, backed by billions of dollars in government subsidies, has attracted both cheerleaders and critics.

Optimists point to Apple Inc.’s decision to build an alternative supply chain to hedge against China. A quarter of all iPhones now roll out of factories near Chennai, employing tens of thousands. Skeptics counter that much of this manufacturing is mere final assembly — screwdriver work that adds little value. The fact that a much smaller economy such as Vietnam exports more merchandise, they argue, shows that Prime Minister Narendra Modi’s Make in India initiative remains more hype than reality.

One way to settle the debate is to examine mobile phones — arguably the biggest manufacturing success story of the Modi years — and tally what India exports and what it imports. The verdict is more nuanced than either side suggests. Indian factories are no longer mere box packers: They plug electronics into circuit boards and assemble chassis, capturing a growing share of the production process.

But those gains come at a price. For every additional dollar going to workers, taxpayers are probably still paying over $1 in subsidies. Unless localization can survive without that level of state support, India’s manufacturing success will remain vulnerable to any weakening of the political resolve behind it.

To see the opportunity and risk, take apart a phone. Printed circuit board assemblies, camera modules, touchscreens, and glass covers make up the bulk of a handset’s bill-of-materials cost. If India is genuinely moving up the value chain, exports of finished phones should outpace the imports of the components that go into them.

Measuring that progress, however, is tricky. Phone exports are easy to identify in trade data; component imports not so much. Microchips may end up in phones, electric vehicles, solar panels or smart meters. Cameras, chargers and adapters have multiple uses, too. When former Reserve Bank of India Governor Raghuram Rajan and co-authors Rahul Chauhan and Rohit Lamba crunched the numbers in 2023, they dealt with this ambiguity by testing three situations.

They assumed that 40%, 60% or 80% of imports in categories containing such dual-use components went into mobile phones. Under every scenario, India’s mobile-phone trade balance remained negative. “In sum, we certainly cannot claim the rise in exports of finished cell phones is evidence of India’s prowess in manufacturing,” they concluded.

Three years later, my analysis shows that the trade balance is still in deficit. However, something important has changed. Under reasonable assumptions about how many imported parts end up in phones rather than solar panels, EVs or other products, the gap between component imports and handset exports has narrowed. The expansion of exports is no longer draining as much hard currency from the country as it used to. That makes it harder to dismiss “Make in India” as glorified screwdriver assembly. India is inserting itself deeper into the global electronics supply chain. The more relevant question is a different one: Has that progress been worth what taxpayers have paid for?

Indian suppliers have dug deeper into sub-assemblies, as Rajan and his colleagues predicted. But just as they warned back then, without domestic memory, processors, lenses, displays, and batteries, the scope for local value addition remains limited.

Evaluating that value addition requires more than just trade data. India is one of the world’s largest smartphone markets, so many locally made phones never cross a border. A better indication of how manufacturing capabilities have deepened comes from what factories are actually doing.

Consider Apple. With iPhone 14, India shortened its production lag behind China to two months, from six to nine months previously. The iPhone 15 was the first Apple device built in India for Day One global sales. Suppliers including Foxconn Technology Group have moved up the value chain, from importing finished circuit boards to mounting components locally. Mainboard production alone drove a $2.1 billion turnaround in the trade balance between fiscal years 2025 and 2026.

It’s hard to say where the trade balance goes from here. As data centers soak up memory-chip capacity, semiconductors used in phones are becoming more expensive. Meanwhile, India’s own handset demand is shrinking this year, led by cheaper, entry-level devices. That could reduce the component import bill, even after accounting for higher memory costs. But softer global smartphone sales could weigh on exports, too.

A bigger worry is what artificial intelligence could do to the fiscal bill. Five years ago, India began paying phone assemblers 4% to 6% of their gross handset invoice values to compensate for poor infrastructure and other cost disadvantages. Now taxpayers may also end up cushioning them against a global price shock caused by AI, the same technology that’s threatening entry-level software jobs — the very engine of the nation’s middle-class mobility. At a 4% payout rate, every $1 billion increase in invoice values driven merely by chip price inflation automatically adds $40 million to the subsidy bill.

New Delhi’s response is to double down on moving up the supply chain. It has earmarked another $13.4 billion to build a domestic semiconductor industry. Trouble is, the $10 billion in subsidies offered five years ago only spawned factories that package chips made elsewhere. Making advanced chips themselves is vastly harder, requiring cutting-edge technology, enormous amounts of capital and investors willing to endure years of uncertain returns.

India remains short of all three. Even Tata Group — Apple’s key local partner and the company behind the nation’s only semiconductor fab — is mired in a messy corporate battle for control. Its largest shareholder is questioning, among other things, the capital-guzzling nature of management’s push into electronics.

Few countries manage this transition successfully. Malaysia chased chip packaging and PCB assembly in the 1970s, yet it stalled after reaching upper-middle-income status in 1992. South Korea and Taiwan, by contrast, rode semiconductors into the ranks of rich economies. India shouldn’t bet billions on catching a fast-moving high-tech manufacturing escalator. It should aspire to be the world’s design center and engineering workshop, while directing fiscal resources toward health, education, climate change — and toward labor-intensive industries such as clothing, footwear and toys that can employ millions.

China is an especially difficult model to replicate. Apple’s success there was built with the dexterity with which Foxconn’s Taiwanese founder Terry Gou, working with Chinese provincial governments, could turn paddy fields into mega-factories. India’s cumbersome bureaucracy and its cautious tycoons will struggle to match that combination of speed, scale and risk-taking.

So no, Apple won’t turn India into the world’s next factory. It can help India acquire valuable manufacturing capabilities and move deeper into global supply chains. But paying its suppliers ever-larger subsidies in the hope of recreating China is unlikely to be a winning trade.