What decides what a country makes well, and can that change?

India has hundreds of millions of workers willing to work for wages no European would accept. Walk into an Indian shop, though, and the toys, the shoes, and often the shirt on the rack came from somewhere else. Cheap labour alone should decide who makes what. It does not.

After 1991, Indian firms making computer-controlled machine tools, the lathes and cutters other factories use to shape their own parts, had to compete against imports priced about 50% higher. Price was the Indian makers’ only advantage, and they could not stretch it much further. Wages accounted for only a fifth of what a machine cost to build, and the computer controls and precision screws, bought from the same few world suppliers the Japanese used, accounted for half. Doubling what each worker produced would have cut the price by only a tenth. Buyers paid the premium anyway.

At the end of the 1990s, researchers placed Indian and imported machines side by side in the same plants, running the same work. Most pairs performed alike, with matching accuracy and matching hours lost to breakdowns. A minority of the Indian machines, though, performed far worse, and no buyer could tell in advance which one they would get. The premium bought near-certainty that the machine would work every time.

An Indian engineer sent to raise standards at a new car-seat plant in China said his six months there would be largely a matter of talking. He changed no machinery. The plant inspected its seats only at the end of the line, so he trained every worker to check each piece as it arrived and again as it left. Consistency of that kind is built from habits, and habits live in the people doing the work. One worker can teach the next, but a firm that loses everyone who knows must start again.

One of India’s leading car-seat makers adopted international practice in the mid-1990s, when 20,000 parts in every million left its plant defective, against the hundred per million that carmakers then counted as world class. Five years later, the figure had fallen to 200. A brand-new plant reached 65 within three years, because a fresh workforce learns faster than a settled one unlearns old habits. By 2003, about half the parts suppliers to a new carmaker in India had cleared that hundred-per-million mark, the same share as in China.

The carmakers made that happen. They kept two suppliers for every part, shifted orders to whichever performed better, and sent their own engineers into the plants worth saving. A supplier with a customer like that improves or loses the contract. A supplier without one can stay as it is for years.

Samsonite built its plant at Nashik into the largest luggage factory in the world by volume, ahead of European hubs that had held the trade for generations. When the company needed more capacity, it built there again, drawn by the suppliers that had grown up around the plant and the workforce that had stayed. Habits collect in places as well as in people, and capability gathers where it already exists.

A country, then, can change what it makes well, but firms are the ones who change it, through customers who measure their performance and can walk away. India spent forty years trying to order capability into existence by decree, and got factories that could not sell abroad. Nobody has found a faster route than the seat maker’s five years.