Subscribe
India’s Unified Payments Interface (UPI) is no longer entirely free; a small transaction fee kicks in from October 15. This means that if you pay a shop via UPI for any amount more than Rs 2,000, the vendor or the shop will be charged 0.4% of that transaction, which is capped at Rs 300 for large payments.
As of now, we are told that the consumers pay nothing. Further, person-to-person transfers are said to be free of charge. Small merchants under a monthly cut-off also are said to stay exempt.
The entire conversation is now hinged on whether the shopkeeper will absorb the 0.4%, or will they add it to the consumer's bill? Will the exemption line always hold? Will the country's most-used payment app platforms like Paytm, PhonePe, and Google Pay make money off this rule?
These are reasonable questions. But by arguing only about who absorbs the 0.4%, we miss the bigger debate: who gets to charge for a piece of public payment infrastructure, who decided that, and what safeguards exist around that decision.
The economic model that is not entirely what it says it is
Let us begin by picturing a matchmaking app, a mall, or a card network — basically any business that has two different actors, say buyers and sellers. These actors can charge each side differently for a transaction. This concept has come from economists Jean-Charles Rochet and Jean Tirole (Tirole won the Nobel Prize partly for this work).
Card networks provide a simple example. Visa and Mastercard charge merchants a fee, while at the very same time, handing consumers cashback and reward points. The reason is simple: if customers want to use their cards, merchants have a strong reason to accept them too.
The same language is now being used to describe UPI's merchant fee. Here, merchants cannot walk away, and consumers are, at least as of now, said to be on the sweet free ride.
But there is more to the story.
The economic literature on such platforms does not suggest that they should simply be free to charge whatever they want. When merchants cannot easily refuse a payment method because customers demand it, the platform has an incentive to charge them more than would be desirable.
That finding became policy.
In 2015, the European Union capped card interchange fees at 0.2% for debit cards and 0.3% for credit cards, to stop platforms from over-charging. The United States capped debit-card fees after 2010 through the Durbin Amendment. Australia spent two decades leaning on its card networks to bring interchange down.
The broader lesson is straightforward: when a payment platform becomes difficult for one side of the market to avoid, governments have often stepped in to regulate what it can charge.
Should UPI even charge?
There is an instinct to call UPI a "public good," and therefore something that should never be charged for. In economics, a fully true public good is something that is both non-rival (your use of the good doesn't stop others from using it too) and also non-excludable (there's no practical way to stop anyone from using it). Popular examples are streetlights and clean air.
However, UPI is not that. Both National Payments Corporation of India (NPCI) and the banks that run the UPI can always meter every transaction if they wanted to.
So the useful question is not whether UPI has costs. It obviously does. The question is how those costs should be recovered. The sharper question is not whether UPI involves costs and how they should be recovered — the State is well within its rights to want that. It's about how.
There are, broadly, three ways to fund a piece of public payment infrastructure:
General taxation, the popular way such as the Reserve Bank of India’s (RBI) own currency-printing costs are absorbed by the exchequer but is invisible to any individual transaction.
A regulated flat fee tested against audited costs, the way electricity bills work. A regulator checks the utility's numbers before it is allowed to charge you.
Handing it to private operators and letting them take a cut of whatever flows through it, which is what has just happened with UPI.
The question is which of the three instruments even makes sense for a UPI transaction in the first place. This is the argument we should be having, instead of directly asking "will the merchant pass it on to consumers”.
There is also a deeper question here. Markets do not simply appear; governments decide which parts of economic life should be organised through markets, and which should be treated as public infrastructure. UPI sits somewhere close to money itself: it is the system through which millions of Indians conduct everyday transactions. Deciding how that system should be funded is therefore not merely a pricing decision. It is a political choice.
The sustainability argument
The strongest argument for the fee is that UPI costs a lot of money to run, and that the government has funded little of that cost directly. This however needs to be examined.
The Payments Council of India told the government in March 2025 that running and expanding UPI cost roughly Rs 10,000 crore a year. Eighteen months later, a Parliamentary Standing Committee put the number at close to Rs 20,700 crore, against a budget allocation of just Rs 2,000 crore for UPI and RuPay incentives.
So, in about a year and half, the number doubled. Both figures cannot be right, and neither of these figures have been verified.
That uncertainty itself matters. If the argument for a new fee is that UPI needs to recover its costs, the public should be able to see what those costs actually are. Further, the actual costs scale roughly with the number of transactions processed. But here the problem is that the new fee scales with the value of each transaction. As citizens in the country make bigger value in digital payments over time, the gap will keep widening between what UPI genuinely costs to run and what this fee collects.
That raises a basic question: if the fee is meant to pay for the infrastructure, why should the amount collected automatically rise with the value of transactions rather than with the cost of processing them?
PhonePe and Google Pay together handle over 80% of all UPI transactions. NPCI's own proposed cap limiting any single provider's market share to 30%, first floated in 2020, has now been deferred twice, most recently to the end of 2026. This delay is usually justified on the grounds that enforcing the cap could disrupt the market.
Yet the fee that will flow largely through these same dominant platforms has moved much faster. The contrast deserves scrutiny: a competition measure has repeatedly been delayed, while a new revenue stream has moved ahead.
The issue, therefore, is not simply that private companies are involved in UPI. They already are. The issue is whether a public payment system should generate a percentage-based revenue stream for dominant private intermediaries without a transparent, audited link between the fee and the cost of providing the underlying infrastructure.
The real distribution fight
The economic cost is not whether a shopkeeper absorbs Rs 12 on a Rs 3,000 sale. Instead, that happens at the edges of the exemption, what economists call a notch — a hard line where crossing it by even a rupee changes how much you owe.
Here, there are two such lines: one based on the size of a single transaction, and one based on a merchant's total monthly turnover, below which small merchants stay exempt. Economists Henrik Kleven and Mazhar Waseem studied similar tax notches in Pakistan and found that people don't cut back their activity when they approach a line like this. Instead, they “bunch” just under it. India has already seen this at home, with businesses staying just under the GST registration threshold to avoid entering the tax net.
In ordinary language, the problem is simple: when crossing a threshold suddenly creates a new cost, people have an incentive to stay just below the threshold.
The same can be expected here, where people will begin to split bills, baskets will be kept artificially small, and large purchases will be steered away from UPI altogether. The real harm lands on the shopkeeper sitting just under the monthly exemption line, who now has a very concrete reason never to cross it and grow.
This lands hardest on exactly the businesses the State spent the last decade trying to pull out of cash and into the digital, taxable economy through demonetisation, Jan Dhan bank accounts, and an incentive scheme that paid out Rs 1,389 crore in FY22, Rs 2,210 crore in FY23, and Rs 3,631 crore in FY24 to keep small merchants on UPI. Having spent that much money pulling small businesses across the line, the State has now built in a quiet reason for them to stop just short of it. A government advisory telling merchants not to pass the fee on to customers does not undo any of this.
What got quietly handed to a committee
What is missing from this whole debate is not merely economic, but also constitutional. In August 2026, the Indian Parliament amended Section 10A of the Payment and Settlement Systems Act, 2007 to permit charges on digital payments, which has reversed a statutory bar that had stood since UPI's merchant fee was scrapped in 2019.
As a result, Parliament eventually left the decision of who pays, how much, and from when to a "steering committee" headed by NPCI. The important question is what kind of accountability applies when pricing decisions for a dominant national payment system are delegated to such a body.
NPCI as an institution does not answer to Parliament for its pricing choices in the same way that Parliament controls taxation, nor does it operate as a conventional utility regulator required to demonstrate that each charge reflects audited costs.
Delegating the price of the country's dominant payment system to a body with no parliamentary accountability for its tariffs is a big step. The debate, however, has largely returned to the narrower question of who absorbs the 0.4% charge.
What the debate should actually be about
Instead of asking “will the merchant pass it on to the consumer,” we should be asking which of the three funding options – general taxation, a regulated and audited flat fee, or an unaudited percentage cut handed to private players – actually fits a system that settles the nation's everyday payments. We should ask why the least accountable of the three was the one chosen. Why does a charge on a national payment infrastructure have no audited cost behind it, when your electricity bill does? Why has a competition fix that has waited six years still not arrived, while the fee it was supposed to come before has already gone through? And what does it mean that the “fiscal dividend” from formalisation, a wider tax base and more visible businesses, exists precisely because this system was free, and is now being treated as if that had nothing to do with who should fund it?
India spent a decade telling the G20 that Digital Public Infrastructure was a genuine alternative to extractive private payment systems. The UPI fee debate is a test of what India means when it calls digital infrastructure “public”: who pays for it, who sets the price, who captures the revenue, and who is accountable for those decisions.
The question is not simply whether the merchant or the consumer will eventually pay. It is whether we have designed a transparent and accountable way to fund the infrastructure on which both depend.
Ubaid Mushtaq is an Assistant Professor in the Department of Economics, Easwari School of Liberal Arts, SRM University-AP.
Views expressed are the author’s own.
