For much of its life over the last decade, the United Payments Interface (UPI) operated on an unusual economic premise. Its value increased as more people and businesses used it, yet the transaction (cost) itself remained almost invisible in price.Consumers paid nothing, merchants were largely insulated from any additional charges, and the government helped sustain the ecosystem through policy support. Another point that merits close attention is how UPI, not just as a platform payment system, but as a mode of payment became its own currency. If one can, let’s say buy Rs 3,000 worth of goods or services by spending cash, one should be able to do so with UPI too. Not pay Rs Rs 5 here or 0.2% there for any reason.That model was remarkably effective at creating scale. It also meant that a payment system handling hundreds of billions of transactions had no conventional price for its basic service. The absence of a visible price, however, did not mean the service had no economic cost. It meant that the cost was being distributed elsewhere in the system.The 0.4% Merchant Discount Rate (MDR) on certain merchant transactions over Rs 2,000 does just that. The charge is insignificant on its own. It is important because it brings a price into a network that was partly developed without a price.Who pays?The first question is who pays? The more interesting economic question is what happens when a previously unpriced service gets a price after consumers and merchants have developed their behaviour around it. The rule of economics is at play here. The person who is legally charged is not always the person who ends up paying the cost.While the economic incidence will be determined by market behaviour and relative price elasticities, the National Payments Corporation of India (NPCI) may not allow merchants to pass the MDR separately to the consumers. The MDR would cost Rs 4,000 to a retailer who processes Rs 10 lakh of eligible UPI transactions. It might absorb the cost, change prices in other areas, cut discounts or promote another payment method.Consumers will thus never have to pay an additional UPI fee, but will still feel the impact of the economic cost in the form of price or loss of discount. This will vary depending on competition and availability of substitutes. If the demand is relatively inelastic, then the final cost will be borne by the consumers.That’s the economics of the incidence of taxes on payments. Under this transition, there is another concept, externalities. Digital payments offer more than just the transaction. They minimise the need for cash, create a record of transactions and can help businesses build a financial history.However, there are costs associated with the infrastructure of each transaction. Servers, cyber security, fraud monitoring, settlement systems and banking technology need to grow as usage grows.Zero MDR did not eliminate these costs. It shifted them to the government, banks and private payment companies. The new MDR therefore begins to put a price on infrastructure that had previously appeared free.This raises the question of whether UPI is a public good. Strictly speaking, it is not a pure public good because access is institutionally governed and transactions consume processing capacity.But it is clearly public infrastructure, created to generate benefits wider than the private return from an individual payment.That distinction matters. Roads, electricity grids and telecom networks also have public value while requiring mechanisms to recover their costs. The challenge is to recover enough to keep infrastructure viable without pricing away the wider benefits that justified building it. UPI is now confronting that problem.A classic economic problemIts scale also creates powerful network effects. Merchants accept UPI because customers use it, while customers use it because merchants accept it. Once QR codes are embedded across millions of shops, moving away from the network carries costs beyond the MDR itself. This also gives the network a degree of resilience that a conventional payment product, competing solely on price, would not necessarily possess.The Rs 2,000 threshold introduces another classic economic problem, behaviour around policy boundaries. A Rs 2,500 purchase creates Rs 10 of MDR. Where operationally possible, splitting it into smaller payments changes the fee outcome without changing the underlying purchase. Similarly, the Rs 1 lakh monthly exemption protects smaller merchants but creates an incentive for businesses approaching the threshold to pay greater attention to transaction routing.This does not mean that everyone is avoiding it. It means that the policy has established a new margin within which behaviour can react.That’s important because UPI is also a part of India’s formalisation process. Digital transactions generate turnover and cash flow records that may enable small businesses to access formal credit. A transition to cash thus involves costs not included in the MDR calculation.The holiday season may be an early test. If a merchant is selling a product worth Rs 15,000, he has to pay Rs 60 as MDR and might want to offer a cash discount to incentivise another mode of payment. Economically feasible cash substitution, but not necessarily a full switch to digital payments.The key change is in the financingThe change has a bigger fiscal justification. The zero-MDR system was not free. The ecosystem was supported by government incentives, banks took on infrastructure costs and payment companies took on losses as they scaled up. According to source estimates, the annual UPI and RuPay reimbursement is around Rs 1,500-2,000 crore, while industry estimates suggest that a 0.4% MDR on around half of the merchant value on UPI transactions could generate around Rs 22,000 crore annually by FY28.The key change is in the financing. A well-developed network can sustain fiscal support or generate a higher proportion of costs commercially. The latter puts less strain on public finances, but more on merchants and possibly consumers.This is where opportunity cost matters. Every rupee spent maintaining a subsidy is unavailable for another public priority. Subsidies that were useful during adoption become harder to justify indefinitely once a network reaches enormous scale. The question is therefore not simply whether UPI should be charged, but how its costs should be allocated once its social and commercial value have both become substantial.The structure also allows cross-subsidisation. A portion of MDR collections could potentially support financial inclusion and deployment in smaller towns, allowing commercially dense markets to help finance infrastructure where private returns are lower.The commercial implications extend to payment companies and banks. Fintech platforms have spent years building transaction volumes while operating with limited direct monetisation. MDR creates another revenue stream. Banks, which bear significant infrastructure and settlement costs, could also receive greater commercial recognition for those costs.The framework introduces price discrimination as well. Mutual fund transactions have a lower rate, recurring payments like SIPs and utility bills are exempt and rural and semi-urban QR payments still have zero MDR. From an economic perspective, differentiated pricing can be justified as users may have varying price sensitivities and social values. A uniform fee may deter transactions that policy makers would like to encourage.There’s an international aspect too. UPI has played a pivotal role in India’s digital public infrastructure (DPI) proposition to the Global South, which has been touted as interoperable and cost-effective. That doesn’t necessarily make commercialisation of part of the network a weak proposition. A system that can afford to pay for more of its own infrastructure may be easier to maintain in the long run.The real experiment with UPITherefore, the success of MDR will not be judged based on revenue. It will appear in merchant pricing, cash usage, transaction behaviour, small-business adoption and the effectiveness of the resulting revenue reaching institutions that bear the costs of the network.UPI’s strongest price indicator for years has been zero. This helped overcome the problem of digital payment adoption in India and helped network effects to set in. The problem has now changed.India is no longer trying to persuade people to use a digital payment network. It is trying to finance one that has become essential to everyday economic life.A price is never merely a number. It changes incentives, redistributes costs and reveals what society is willing to pay for infrastructure it once treated as almost free. The real experiment with UPI has therefore entered its next phase. The question is whether India can put a price on the payment without putting a price on the inclusion and convenience that made UPI so valuable.Deepanshu Mohan is Dean and Professor of Economics, O.P. Jindal Global University and a Visiting Professor at LSE and a Visiting Research Fellow at Oxford Department of International Development, University of Oxford. Ankur Singh is a Research Analyst with Centre for New Economics Studies, O.P. Jindal Global University.