Skipping a Step: The Snake in India's Growth Story

There is a standard path to GDP growth that every economics textbook teaches, and it goes something like this: people move away from farms and into factories, they get paid better than in agriculture, and only once the industrial muscle is in place does the services economy grow on top of it. South Korea and China walked this 'path' and became the behemoths they are today, all thanks to the Lewis model. For most of the last century nobody argued with it, because it was simply how development was understood to happen. Newer theories emerged, but the Lewis model remained central to explaining how an economy grows. India has been breaking that model to reach its tremendous 7.6 percent GDP growth rate. But within this 'path-breaking growth' also lies a snake in the grass, one that can strike if the country does not tread carefully. That snake is India's nearly stagnant manufacturing sector.

The Production Linked Incentive programme (PLI), launched in 2020 with an outlay of nearly 2 lakh crore rupees, aimed to push manufacturing's share of GDP from around 15 percent up to 25 by paying companies for every extra unit they managed to produce. Judged on its own numbers, the outcome was spectacular. Electronics production more than doubled between FY21 and FY25, climbing from 2.13 lakh crore to over 5.25 lakh crore, mobile phone exports rose a staggering 775 percent, and pharmaceutical firms that had spent decades bleeding money on imports flipped into surplus in the span of three years. The scheme generated more than 14 lakh jobs, and realised investment crossed 2.16 lakh crore by late 2025.

At first glance, PLI looks like a resounding success. And yet the one number the whole scheme was planted to grow refused to sprout. Manufacturing's share of GDP did not climb over those five years at all. It slipped, from just over 15 percent to 14, so that the gap to the 25 percent target opened wider than it had been before the money was ever spent.

The easy explanation is bad execution, and there is some truth in it. Labour laws still punish firms for growing past certain headcount thresholds, pushing many to stay small on purpose. Vocational training has not kept pace with what a modern factory floor demands. And no amount of incentive money can change the fact that a firm afraid to hire will simply pocket the reward and stay exactly the size it was.

There is a harder explanation, though. The reason India's growth never stumbled when manufacturing did is that a brighter door had already opened elsewhere, and the economy walked straight through it. That door is the Global Capability Centre. These centres have swelled into a workforce of nearly 2.4 million people across more than 2,100 sites, up from roughly 1.9 million just two years earlier. Goldman Sachs runs its risk analytics out of Bengaluru, Rolls-Royce does core engineering in the same city, and firms like Walmart, Ford and Morgan Stanley have moved genuine high-value work into India for the depth of the talent, not the size of the discount. The whole sector is on course for 2.8 million employees and over 100 billion dollars in revenue by 2030.

But here, at last, is the snake. The GCC boom looks like abundance until someone notices how little ground it actually covers. It sits in just six cities and hires almost entirely from a narrow pool of English-speaking engineering graduates, a thin slice of the roughly 65 lakh people who join the workforce every year. The service sector was never built to swallow the country whole the way factory floors once swallowed the villages of South Korea and China. Manufacturing, for all its stumbling, was the one model that could absorb millions of ordinary hands at once.

The mistake is treating these two stories as a zero-sum trade-off, where a win in services offsets a failure on the factory floor. They are not rivals. They are a lesson and a warning sitting side by side. Services scaled so fast because it carries no physical weight, no components to import, no goods to ship, no land to acquire, just talent meeting demand down a fibre-optic cable. 

Manufacturing enjoys none of that ease, and no subsidy cheque under PLI can change the physical cost of making things. Steep tariffs on imported components, freight that eats into already thin margins, and land laws that turn a new plant into a liability are not problems money can paper over. If India wants assembly lines to absorb millions, it has to move past screwing together imported phone kits, which means cutting those tariffs so local plants can plug into global supply chains, building ready-to-use industrial zones, and fixing the freight networks that keep Indian goods slow and expensive.

The snake was never manufacturing itself. It was the belief that an elite service enclave could absorb 65 lakh new workers a year when it was only ever built to open its door to a few. A strong GDP figure can buy time, but it cannot hire the crowd waiting outside the gate, and 7.6 percent cannot hide that crowd forever. Eventually, the math catches up, and so does the snake.