Listen

Description

January 13, 2026

“India has become the back office of the world. But where is the factory floor?” — Anonymous economist, circa 2005

“We wanted to set up a factory for the Nano in West Bengal. The land was acquired. The protests began. We left.” — Ratan Tata, Chairman of Tata Group, 2008

“The labor laws are so rigid that firms stay small to avoid them. This is the tragedy of Indian manufacturing.” — Jagdish Bhagwati, Columbia University

[2024: Manufacturing is 17% of India’s GDP. In China, it’s 27%. In Vietnam, 25%.]

[Part 3 of 6]

New parts publish every Tuesday and Thursday at 9AM EST

In 1991, India and China were roughly comparable. Both were poor, populous, agricultural nations emerging from decades of socialist planning. Both were opening their economies. Both had ambitions to become global powers.

Thirty years later, China is the world’s factory. India is the world’s back office.

China manufactures your iPhone, your furniture, your clothes, your toys, and your electronics, with Chinese factories employing hundreds of millions of workers who migrated from farms to assembly lines in the largest movement out of poverty in human history. India, by contrast, writes your software, answers your customer service calls, processes your insurance claims, and manages your IT infrastructure through service companies that employ millions of educated workers in air-conditioned offices while generating billions in export revenue. Both are success stories, but they’re not equivalent.

Manufacturing absorbs masses of workers with limited education. You can train a farmer to work an assembly line in weeks. Service exports require English fluency, technical degrees, years of education. India’s IT boom created a prosperous urban middle class. It did nothing for the hundreds of millions still trapped on farms.

China moved 300 million people from agriculture to manufacturing, while India moved them from farms to... other farms. This is the story of the missing factory.

The IT Miracle (And Its Limits)

The 1991 reforms unleashed something unexpected: an IT revolution.

India had advantages no one anticipated. English fluency (a colonial inheritance Nehru had tried to replace with Hindi). A time zone perfectly positioned to work while America slept. Elite engineering colleges (the IITs) that produced world-class technical talent. And crucially, once import restrictions lifted, access to computers.

To understand how transformative this was, consider the telephone. In the 1980s, the waiting list for a landline connection stretched ten years or more. A working telephone was so valuable that families listed it as an asset in marriage negotiations, right alongside property and gold. People bribed officials or bought “out of turn” connections on the black market. The state telephone monopoly was a monument to dysfunction.

By the 2000s, India had more mobile phones than landlines had ever existed. Private telecom companies, freed by liberalization, built networks faster than the government had managed in four decades. The same pattern repeated across IT: once the government stepped aside, Indian entrepreneurs moved at astonishing speed.

The combination was explosive. Indian software companies, Infosys, Wipro, TCS, began as body shops, lending programmers to Western firms. They evolved into global IT services giants. Bangalore transformed from a sleepy pensioner’s paradise into India’s Silicon Valley.

But here’s the irony: India went into software because hardware was impossible.

Narayana Murthy, who would become one of India’s richest men as Infosys co-founder, tried to import a computer in the early 1980s through a process that took three years, fifty trips to Delhi, and endless forms, officials, and bribes. By the time the computer finally arrived, it was nearly obsolete.

“We chose software,” Murthy later explained, “because software was just ideas. You couldn’t regulate ideas. You couldn’t require a license for thinking.”

The hardware barriers weren’t just bureaucratic. They were geopolitical. In 1987, India’s scientists requested a Cray XMP supercomputer for weather forecasting and nuclear research. The United States refused, citing Cold War export controls. India might use the machine for weapons development.

The denial stung. But it also motivated. Indian scientists at the Centre for Development of Advanced Computing built PARAM, India’s first indigenous supercomputer, within four years. “Thank you for the sanctions,” one scientist reportedly said. “You forced us to become self-reliant.”

The lesson was clear: hardware meant dependency, either on bureaucrats in Delhi or politicians in Washington. Software meant freedom.

India’s IT revolution wasn’t a strategic choice. It was an escape route from a system that made manufacturing impossible. The entrepreneurs who built Infosys and Wipro weren’t software visionaries who saw the future. They were pragmatists who discovered the one sector the License Raj couldn’t strangle.

By the 2000s, “Bangalored” had entered the American lexicon as a verb meaning “outsourced,” with Indian call centers answering questions about American credit cards, Indian programmers maintaining American banking systems, and Indian engineers designing components for American tech companies. The numbers were staggering: IT services grew from essentially zero in 1991 to over $200 billion in annual revenue by the 2020s, creating millionaires, building gleaming corporate campuses, and proving India could compete globally in knowledge work.

But here’s what the IT miracle didn’t do: employ the masses.

The entire IT and IT-enabled services sector directly employs roughly 5 million people out of India’s workforce of over 500 million, meaning the IT boom, for all its visibility, touches perhaps 1% of Indian workers while the other 99% needed factories.

Why Factories Never Came

The 1991 reforms freed product markets. They did not free factor markets.

To build a factory, you need three things: land, labor, and capital. Post-1991 India had capital (foreign investment finally flowed in). It had land (in theory). What it didn’t have was a labor market that allowed factories to function.

The Industrial Disputes Act of 1947 remained untouched by reform. Under Chapter V-B, any firm with more than 100 employees needed government permission to lay off workers. In practice, permission was never granted. Politicians wouldn’t sign documents that could be used against them in elections.

This created a “regulatory cliff” where firms with 99 employees faced a stark choice: stay small, or grow and lose all flexibility forever. Most stayed small, and the math was brutal. A factory that might optimally employ 500 workers would instead employ 95, staying just under the threshold, or it would hire contract workers (exempt from the law) rather than permanent employees, creating a two-tier workforce with no job security for most.

The result was India’s “missing middle.”

Indian industry consists of a few massive conglomerates (Tata, Reliance, Birla) that are large enough to absorb regulatory costs, and millions of tiny micro-enterprises that stay small to avoid regulation. The mid-sized firms that drive manufacturing employment in every successful industrial economy barely exist.

In China, the average manufacturing firm employs far more workers than in India. Chinese factories achieve economies of scale. Indian factories, constrained by labor laws, fragment into inefficient sub-scale units.

Land acquisition was equally broken.

The Land Acquisition Act of 1894 (yes, colonial-era legislation) governed how the government could take land for industrial use. The process was slow, corrupt, and politically explosive. Farmers displaced by factories became protest movements. State governments that acquired land for industry faced electoral punishment.

Building a factory required navigating years of legal challenges, compensation disputes, and political opposition. Easier to write software in an existing office building.

Infrastructure r…



This is a public episode. If you'd like to discuss this with other subscribers or get access to bonus episodes, visit tatsuikeda.substack.com/subscribe