The industry body argues that banks, third-party apps like PhonePe and Google Pay, and PSPs have been absorbing the full cost of switching, fraud handling, chargebacks, and 24/7 uptime for a system the government mandated as free. After years of lobbying, they contend that reintroducing MDR for large merchants is necessary to make the economics of UPI sustainable.
The editorial frames the 2020 zero-MDR decision as one of the boldest public-infrastructure bets any government has made on payments, crediting it with driving $3 trillion in value and 172 billion transactions in a single fiscal year. It argues UPI is uniquely the only real-time rail globally to reach national-utility scale without a card-network oligopoly or walled-garden super-app, precisely because India mandated the price at zero and let private apps compete on UX.
By surfacing the BBC story to the top of Hacker News with 118 points, the submitter frames the potential reversal as a significant global-interest event — implicitly treating the zero-fee model as a notable achievement now at risk. The framing centers on India 'paving the way' for fees, signaling the end of an era rather than a routine policy tweak.
The editorial highlights that a 2023 IIT Bombay paper attempted to quantify the annual subsidy required to keep UPI running, underscoring that 'free' was never actually free — the cost was simply pushed onto banks and fintechs. It argues the economics never balanced and that transparency about who pays is overdue.
India's payments industry body has recommended that merchants be allowed to pay a fee — a Merchant Discount Rate, or MDR — on transactions made over the Unified Payments Interface. If the government adopts it, it will end nearly six years of a zero-MDR mandate that turned UPI into the largest real-time payments rail on earth and also, quietly, into an unfunded mandate for the banks and fintechs that operate it.
UPI processed roughly $3 trillion in value across more than 172 billion transactions in the last fiscal year, and the entire stack was, for the merchant and the consumer, free. That was the point. When the Reserve Bank of India and the National Payments Corporation of India (NPCI) killed MDR on UPI and RuPay debit in January 2020, the goal was to blow past cash and card networks in a country where informal commerce dominated. It worked. UPI is now the plumbing under everything from a chai stall's QR code to Zomato checkout to salary disbursement.
But the economics never balanced. Banks, third-party app providers like PhonePe and Google Pay, and PSPs eat the cost of switching, fraud handling, chargeback investigations, and 24/7 uptime for a system the government promised would be free. The Payments Council of India has been lobbying to reintroduce fees for large merchants for years. This recommendation is the closest they've come to winning.
The zero-MDR decision was, in retrospect, one of the boldest public-infrastructure bets any government has made on payments. It's also the reason UPI is the only real-time rail in the world that has scaled to national-utility status without either a card-network-style oligopoly (Visa/Mastercard) or a walled-garden super-app (WeChat Pay, Alipay). India built a public API for money, mandated its price at zero, and let private apps compete on UX on top of it.
The cost has been borne opaquely. A 2023 paper from IIT Bombay estimated the annual subsidy required to keep UPI running at scale in the range of ₹8,000-10,000 crore (roughly $1 billion). Some of that has come from a government incentive scheme that reimburses banks a slim margin on small-value P2M transactions. Most of it has come from banks eating the loss and cross-subsidizing from other lines. That's fine when the volumes are a rounding error. It's not fine when UPI is 80%+ of retail digital payments and still growing double digits year over year.
Contrast this with Brazil's Pix, the closest analog. Pix is also free for consumers but the central bank charges banks a small per-transaction fee, which banks pass through to merchants at a rate typically well under 1%. Pix is profitable for its operator. UPI is not. The difference isn't the technology — the rails are architecturally similar — it's that Brazil didn't try to make the whole thing free forever.
Community reaction on the Hacker News thread is split along predictable lines. One camp: fees will kill the flywheel, merchants will steer users back to cash, and India's inclusion story reverses. The other camp, mostly people who've actually built on UPI: the current model is unsustainable, fraud is rising because nobody has margin to invest in detection, and a small, transparent MDR on transactions above a threshold (say, ₹2,000) is a rounding error for merchants who currently pay 1.5-2% on card swipes anyway.
The second camp is more right. The zero-MDR mandate is a classic case of a subsidy whose costs are invisible and whose benefits are visible — the political incentive is to preserve it long past the point where it's optimal. But the operators quietly winding down feature investment, deferring fraud tooling, and lobbying for relief are a tell. Something has to give.
If you build payments in India, the near-term signal is that pricing is coming back on the table, but almost certainly tiered. The likely shape is MDR on P2M transactions above a value threshold, with small-ticket and P2P transactions staying free — which means your integration needs to handle a per-transaction fee variable that today you've hardcoded to zero. Model it now. The rate cards will be political and will change; the code path should not need to.
If you build fintech elsewhere and treat UPI as the reference architecture for real-time payments — and a lot of central banks do, from the ECB's TIPS to the Fed's FedNow — the lesson is narrower than it looks. UPI's technical design is sound and worth copying. The pricing model is not. FedNow launched with per-transaction fees for financial institutions from day one; that's the right call. Free-at-the-edge with fees at the plumbing layer is a stable equilibrium. Free everywhere is not.
For product teams outside payments, the second-order effect worth watching is what happens to the fintech apps built on top of UPI. PhonePe, Paytm, Google Pay, and CRED all monetize on adjacent products — lending, insurance, wealth — because the transaction layer itself pays nothing. If MDR returns and app providers get a cut, the pressure to cross-sell everything eases slightly, and you may see cleaner, less spammy payment UX. Or the reverse: with a new revenue line to optimize, expect A/B tests on default payment method selection.
The government hasn't signed off. Finance ministry officials have gone on record repeatedly saying UPI will stay free for consumers, which leaves merchants as the only place fees can land. Expect a compromise: a low-single-digit-basis-point MDR on P2M transactions to large merchants over a threshold, phased in, with the government continuing to subsidize the small-merchant tail. That preserves the inclusion story and starts to close the funding gap. The bigger question is whether other countries watching UPI — and there are many — internalize the pricing lesson before they mandate their own version of free.
Top 10 dev stories every morning at 8am UTC. AI-curated. Retro terminal HTML email.