46.91 tonnes in 1991, 41.12 tonnes in 1993 and now a third attempt in 2026 - India is once again trying to talk its own citizens' gold out of their lockers.

THE THIRD ATTEMPT TO TRUST INDIA WITH ITS OWN GOLD

India imports roughly 700 to 800 tonnes of gold a year. That number alone explains why a new voluntary disclosure scheme is being discussed in Delhi. But the number that should worry the scheme's architects isn't the size of the gold pile sitting in Indian lockers - it's the size of the trust gap between 1993 and 2026.

THE MECHANISM THAT ACTUALLY WORKED IN 1993

The 1993 Gold Bonds Scheme didn't succeed because gold was scarce or because the interest rate was attractive. It succeeded because the government offered something specific and rare: a legal promise, backed by the Gold Bonds (Immunities and Exemptions) Ordinance, that no one would ever be asked where the gold or the money used to buy it, came from.

For a household sitting on undeclared wealth in 1993 - a much less monitored economy, with no PAN-Aadhaar linkage, no GST data trail, no automated income-tax cross-checks - that immunity was worth more than the ₹40-per-gram interest. 41.12 tonnes came in within two months because the risk of staying silent about that gold was, for many holders, higher than the risk of declaring it.

WHY THE SAME PROMISE MEANS SOMETHING DIFFERENT TODAY

Run the same mechanism through 2026's tax infrastructure and the calculation flips. A household today already sits inside a far denser web of financial visibility - bank KYC, GST invoicing on jewellery purchases above certain thresholds, income-tax data matching.

The immunity a voluntary scheme could realistically offer is narrower and the population it needs to attract - those holding gold specifically because it sits outside that visibility - has shrunk. Sajjid Chinoy's research for JP Morgan shows something telling here: half of India's current gold demand (January-March 2026) is now investment-driven, not consumption jewellery, up from just 30% in 2003.

That's a demand structure shaped by decades of formal financial-asset alternatives - mutual funds, Sovereign Gold Bonds - competing for the same household rupee, not a demand structure desperate for an amnesty exit.

THE ACTUAL TEST FOR 2026'S VERSION

The 1991 crisis worked as forced collateral because reserves had collapsed to three weeks of imports - there was no choice involved. The 1993 scheme worked as voluntary disclosure because trust in state silence was still cheap to offer and valuable to receive. 2026's proposal has neither condition: reserves are healthy at over $675 billion and the state's capacity to eventually find undeclared wealth through other channels has grown enormously since 1993.

That means this scheme cannot succeed as either a forced measure or a pure amnesty - it will only work if it's redesigned as a genuinely competitive financial product, one that beats Sovereign Gold Bonds and mutual funds on their own terms rather than relying on the promise of silence.

For the exam, this is the sharper reading: India isn't repeating a policy, it's testing whether a 33-year-old mechanism survives when the exact condition that made it work - cheap trust in state discretion - has been engineered away by the state's own tax modernisation.