6 min readAug 12, 2026 12:41 PM IST
First published on: Aug 10, 2026 at 03:42 PM IST
By Arya Roy Bardhan
The Parliament has begun to dismantle the free-UPI regime through the Taxation and Other Laws (Amendment) Bill, 2026, which, for the most part, is a measure about direct taxation. However, a single clause within it repeals a rule that has stood since January 2020 — the prohibition on levying a merchant discount rate (MDR) on UPI payments. The MDR is the fee a merchant pays its bank to accept a digital payment, ordinarily one to two per cent on a card transaction. On UPI, it has been zero, mandated by statute.
The Bill itself imposes no fee. It withdraws the automatic prohibition and empowers the government to decide, by notification, which payment modes remain free of charge. It specifies no rate, no threshold, and no exemption for small merchants or ordinary users. The government has given assurances that only large merchants will be charged, and never consumers. Those assurances may well be sincere, but they are not guaranteed by the law. The statute now permits the executive to decide the matter later, without returning to Parliament.
The Bill has, moreover, cleared only the Lok Sabha; it is not yet law. The correct interpretation, at this juncture, is therefore that UPI may be charged, on terms yet to be decided.
What drove the change?
UPI was never costless. The infrastructure has always been paid for either by banks, the payment applications, or the government, which reimbursed them for the fee they were required to forgo. That subsidy has been unevenly disbursed over time. The figure oscillates from year to year, and even at its height it met only a fraction of industry’s estimated operating costs. The budget has proved to be an unreliable means of keeping UPI free. The amendment is the other half of that predicament: To relieve the exchequer, and to permit a fee in its place.
The most immediate concern is that of incidence. If large merchants are charged, will they not simply pass the cost to their customers? Economists term this pass-through, and the presumption that firms with market power will transfer the entire burden to consumers holds only under restrictive assumptions. The share of the cost that reaches the final price depends on the elasticity and curvature of demand, and may range from half the cost to more than its whole. The experience of card-fee reforms in the United States, the European Union and Australia suggests that pass-through is both partial and asymmetric: Increases in cost reach consumers more readily than reductions, and the effect is most pronounced on small, frequent purchases. A modest fee levied on large, high-value transactions is thus tolerable, unlike a fee imposed on neighbourhood stores.
A second and subtler concern is what a transaction cost does to the structure of money supply and demand. A charge does not remove UPI from the money supply; UPI merely moves bank deposits, which are already money. The fee alters the ease with which that money is used. A significant part of money’s usefulness lies precisely in its trivial transaction cost. Once a cost is introduced, the instrument becomes marginally less money-like, and theory anticipates a modest shift towards the medium that bears no fee: Cash.
The evidence is reassuring. Throughout the expansion of UPI, cash did not recede. Banknotes in circulation continued to rise, reaching approximately Rs 42 lakh crore in 2026. If free UPI did not displace cash, a small fee would scarcely revive it. The genuine concern is the exposure of low-value payments. Were a fee ever to extend to the transactions of Rs 10 and Rs 20 that constitute the core of UPI’s use, the retreat to cash would become likely. If, according to public statements, the cost is limited to large merchants, the monetary risk is largely notional.
A third concern is that of timing. The Bill has arrived in the midst of trade negotiations with the United States, and the coincidence raises suspicion. The United States Trade Representative’s 2026 report does register objections to India’s payments regime but not to zero MDR. Its complaints are that American firms cannot compete with RuPay, that they are excluded from credit on UPI, that no single application may exceed 30 per cent of the market, and that payment data must be stored within India. Free UPI has worked against Visa and Mastercard. Fees paid to foreign networks leave the country while UPI retains that value at home. The Bill does not constitute a surrender, but an overdue domestic reform enacted in the course of a trade negotiation could be mistaken for one.
The way forward
The infrastructure must be paid for by someone. But, by whom, and will that decision be taken openly? Ordinary users and small merchants should be kept free by law, not by assurance. Any fee should be capped low and confined to large merchants. Its proceeds should be ring-fenced for the security and reliability of the UPI system. It should be determined transparently, through the payments authority, and held apart from the trade talks. It should be reviewed periodically as the system matures.
None of this requires abandoning the core UPI principle. It only requires that the terms of any charge be fixed in law, deliberated, and justified by the cost of the system rather than the exigencies of a trade negotiation. A public good sustained for six years on the promise that it would cost nothing cannot now be priced on terms its users are asked to take on trust.
The writer is junior fellow, ORF
Merchant costs are rarely absorbed silently; they tend to surface as price adjustments, minimum-transaction thresholds, or a quiet preference for cash, particularly among small retailers with thin margins. 
Share
