On September 16, a writ petition reached the Supreme Court seeking to quash the two Gazette notifications through which the Centre reintroduced pricing on UPI merchant payments. The petition has been framed on a familiar ground, that a charge nominally imposed on merchants will, in practice, reach the consumer. That framing is intuitive. It also sits a step away from where the real difficulty in the new framework actually lies.
The Finance Ministry’s notification of September 14, issued under the newly amended Section 10A of the Payment and Settlement Systems Act, 2007, does one precise thing. It lists the instruments such as RuPay debit cards, and UPI transactions up to ₹2,000, on which “no bank or system provider shall impose, whether directly or indirectly, any charge upon a person making or receiving a payment by using one or more electronic modes of payment as the Central Government may, by notification, specify.” That is a statutory prohibition. It binds banks and system providers. The words “directly or indirectly” are there for a reason: to stop a bank from dressing up a barred charge as a processing fee, a platform fee, or anything else. Breach it, and the payer has a straightforward, judicially enforceable claim, the notification says the charge cannot be levied, full stop.
Above ₹2,000, a different instrument takes over. NPCI’s circular sets a Merchant Discount Rate of 0.4% on person-to-merchant UPI transactions crossing that line, capped at ₹300, which means a ₹40,000 payment carries ₹160, and anything above ₹75,000 caps out at ₹300. Railways, telecommunications, insurance, fuel and agricultural inputs get a flat ₹5 instead, and capital-market transactions a reduced 0.02%, also capped at ₹300. On the government’s own figures, this reaches a small share of merchant transactions by count, roughly 4%, but a much larger share by value, since it is exactly the larger, discretionary payments that clear the threshold.
It is worth noticing what did not happen here. Categories the government evidently wanted to treat gently like railways, telecom, insurance, fuel, agricultural inputs, were not at all added to the Section 10A notification, where they would have enjoyed the same statutory, charge-free status as UPI payments up to ₹2,000. They were instead given a concessional rate through the NPCI circular, the same second-tier instrument that governs every other transaction above the threshold. The government had the tool to extend statutory protection further. It chose, for these categories too, an administrative rate instead of a notified exemption. That is a considered design choice, not an oversight, and it makes the absence of any statutory safeguard for ordinary merchant transactions above ₹2,000 harder to read as accidental.
The government is plainly aware that leaving this category unprotected carries a risk. Banks have reportedly been told to see that merchants do not fold the MDR into what the consumer pays, and UPI app providers have separately been barred from levying platform fees or hidden charges of their own. Neither instruction does what Section 10A does, and the difference is not a matter of degree. An instruction to banks addresses the wrong party: banks process the MDR, but they do not set the price of what a merchant sells. Whether to absorb a ₹160 or ₹300 cost, itemise it, or quietly build it into the marked price is the merchant’s call, and the merchant is not who the advisory is addressed to. The separate bar on UPI app platform fees solves an adjacent problem, a charge levied by the app on the user, and has nothing to say about how a shopkeeper prices what he is selling.
This is not really a drafting gap so much as a limit written into Section 10A’s own text. The section binds a “bank” or “system provider.” A merchant who raises a price to recover a cost the payment ecosystem has imposed on it is neither, it is a seller pricing what it sells, which is not conduct the provision was ever drafted to reach, however widely “indirectly” is read. Between 2020 and now, that limit never mattered, because there was no merchant-side cost to pass through in the first place. The moment a real cost enters the chain, the space between what Section 10A prohibits and what the notification actually protects becomes visible for the first time, and it is precisely that space the new framework has opened up.
It is also worth being exact about what would, and would not, close it. Card networks have long relied on no-surcharge conditions in merchant agreements to stop a separate surcharge line appearing at checkout. Even assuming an equivalent condition eventually applies to UPI QR merchants, it reaches only that one method of pass-through. It says nothing about a trader simply building the MDR into the marked price rather than itemising it, which, for a merchant on thin margins with little bargaining power against the payment ecosystem, is the quieter and considerably more likely response. A recent piece in this publication anticipated exactly this outcome, framing it mainly as an economic irony, cash making a comeback once UPI acquires a price tag. The legal point is narrower, and arguably sharper: an advisory addressed to banks cannot reach a merchant’s pricing decision, and nothing in the current framework can.
The competition structure of UPI sharpens rather than dilutes this. A large share of merchant-side UPI volume runs through a small number of dominant applications, and the small merchants transacting through them have limited leverage over either the ecosystem’s pricing or an aggregator’s own commercial terms. A trader without bargaining power is, almost by definition, a trader who passes a fixed cost on rather than absorbs it, precisely the outcome the government says it wants banks to prevent, using an instrument that is not addressed to the trader at all.
None of this argues for restoring a blanket zero-MDR regime, or for reading Section 10A as though it protects every UPI payment regardless of value; the infrastructure genuinely costs money to run, and the government’s underlying case for a sustainability mechanism is not, in itself, unreasonable. The narrower point is that having decided some transactions may legitimately carry a cost, the government protected the payer’s interest in the untaxed category with a statute, and the payer’s interest in the taxed category with an advisory, as though the two were equivalent safeguards. They are not. One can be enforced in a court of law against a defined class of regulated entities. The other carries no stated consequence for non-compliance, no notified enforcement mechanism, and is addressed to a party that does not control the outcome it is meant to prevent.
The Anjan Datta petition gives the Supreme Court an early opening to test that distinction, though its own framing, that the new charges carry a wider impact on the public which is broader than the narrower question the notifications actually present: whether Section 10A’s language was ever capable of reaching a merchant’s pricing decision at all, and whether an administrative advisory can lawfully stand in for a statutory prohibition once the government itself has concluded, by issuing that advisory, that some safeguard is necessary. If it cannot, the fix does not require reopening the MDR. It requires locating the missing safeguard in an instrument that actually binds how a merchant prices a payment, rather than in a request addressed to the one party that never decided the price in the first place.
Until that gap is closed, ₹2,000 remains the one point on India’s new UPI pricing map where the payer’s protection comes from a statute rather than an assurance.
