The current weakness is not a crisis of crude prices or capital flows. It is the surfacing of a sequencing error from thirty-five summers ago, when delicensing moved overnight while manufacturing waited for the factor-market reforms that never came.

The rupee has moved past 96 to the dollar. Foreign-exchange reserves have drawn down by more than $30 billion in ten weeks from their recent peak. Net foreign direct investment was negative for six consecutive months through January 2026. Gold import duties have been more than doubled, from 6 to 15 per cent. The prime minister has urged citizens to stop buying gold for at least a year, cut fuel use and avoid foreign travel.

These are not passing symptoms of a bad quarter. They are the surfacing of a structural weakness that has been thirty-five years in the making, one that traces back to a specific sequencing choice in the summer of 1991.

India has grown almost without interruption since those reforms. GDP has increased roughly eightfold. Hundreds of millions have been lifted out of poverty. A services sector has emerged that the country can legitimately be proud of.

And yet something does not add up. There is a pattern that virtually every successful industrialiser has followed, and India has not followed it.

The economist Geoffrey Crowther mapped the pattern in the 1950s. Countries begin as young debtor nations, borrowing from abroad to finance imports. As they build manufacturing, exports overtake imports. The resulting trade surpluses pay down external obligations, and over time the country graduates from debtor to creditor.

Britain followed this path. So did America, Japan, South Korea and China. The engine was always manufacturing exports, compounding over decades into external strength.

India, after thirty-five years of impressive growth, has not moved through these stages. It remains a debtor nation, running persistent current-account deficits, dependent on foreign capital inflows to pay for what it buys from the world.

That is the anomaly. And to understand it, you have to go back to how the 1991 reforms were rolled out.

Two Reforms, One Sequencing Error

The 1991 package had two pillars. The first was a sharp devaluation of the rupee, roughly 18 to 20 per cent in two steps, to make Indian goods cheaper in world markets. The second was the near-complete dismantling of industrial licensing, freeing the private sector from the permit system that had strangled Indian industry for four decades.

Together, they should have launched India on the same manufacturing-led export trajectory that had transformed East Asia. The cheaper rupee would create the demand. The freed-up private sector would build the supply. Crowther's progression would begin.

What happened instead is that delicensing moved immediately while the complementary reforms that manufacturing needed did not follow.

Labour laws made hiring above certain thresholds punitive. Land acquisition stayed broken. Electricity remained cross-subsidised against industry, at roughly twice the cost faced by East Asian competitors. Logistics stayed expensive. Procurement rules stayed hostile to complexity.

Think of it this way. The devaluation was a price signal sent to the entire economy: Indian production is now globally competitive, invest in manufacturing. But a factory needs land, reliable power, flexible labour and a functioning supply chain before it can answer that signal. That takes years. Delicensing took effect overnight.

Capital did not wait. It read the environment as it actually was, not as the devaluation intended it to become, and flowed to where returns were immediate.

Services were the rational winner. They were globally tradable, English-language, lightly regulated and required no manufacturing ecosystem to scale. IT exports became the principal foreign-exchange earner created by the cheaper rupee.

The conventional narrative holds that the reformers had no choice: they were putting out a fire and did what the moment demanded. The record is more complicated. The World Bank's own 1989 report on India had recommended gradualism.

The actual reforms went well beyond what the Bank advocated in delicensing, while in other areas - consumer goods import liberalisation, small-scale industry dereservation - there was no movement for nearly a decade.

The pace was a choice, not a constraint. The reformers chose speed because the crisis gave them political cover they might not get again. That is understandable. But it was a sequencing decision, and it had consequences.

This was not unusual advice. South Korea kept heavy tariff protection through the 1960s to the 1980s while building its industrial champions. China opened through Special Economic Zones first, keeping the domestic market shielded for decades. Japan maintained import restrictions until its industries were globally competitive.

The United States itself industrialised behind high tariffs for most of the 19th century. In each case, liberalisation was gradual, giving domestic industries time to adjust. India's was not.

The economist Dani Rodrik has documented the broader result, showing that India began to deindustrialise at a per-capita income of roughly $2,000, compared with $9,000 to $11,000 for the United States, Britain and Germany. He calls it "premature deindustrialisation."

In India's case, it was not just premature. It was policy-induced. The devaluation created the space. The sequencing of reforms ensured that services, not manufacturing, filled it.

The Ghosts Return

Through the 2000s and into the 2010s, the "services export plus manufacturing import" structure worked well enough. IT earned dollars. Remittances supplemented. The current-account deficit stayed manageable. Nobody needed to ask why India had not built a manufacturing export base, because the model was delivering growth.

But as India scales toward a $5 to $10 trillion economy, the model's fragility is exposed. The import bill grows with the economy: $130 billion in crude, $72 billion in gold, plus electronics, fertilisers, defence equipment and capital goods. Services exports cannot generate surpluses at the scale needed to finance all of this.