UPI charges: the economics of a digital public good

UPI is no longer merely a payment product. It is digital economic infrastructure. In August alone, UPI processed 24.51 billion transactions worth ₹29.82 trillion, according to the National Payments Corporation of India (NPCI).

The government’s new framework, effective 15 October, imposes a 0.4% merchant discount rate (MDR) on specified person-to-merchant (P2M) transactions above ₹2,000, capped at ₹300 for transactions of ₹75,000 and above. Person-to-person (P2P) transactions remain free, as do payments up to ₹2,000 and transactions covered by the zero-MDR framework for small merchants. About 96% of P2M transactions are expected to remain unaffected.

The economics of ‘free’ UPI

India introduced zero MDR in January 2020 to accelerate digital-payment adoption and subsequently compensated ecosystem participants. Government support amounted to approximately ₹8,730 crore between fiscal year 2022 (FY22) and FY25. Yet the Parliamentary Standing Committee on Finance has cited an industry estimate of about ₹20,700 crore in annual operating costs for the UPI ecosystem, against a budget allocation of ₹2,000 crore for FY2026-27. These figures highlight the underlying problem: a zero-price transaction is not a zero-cost transaction.

UPI requires servers, bandwidth, cybersecurity, fraud management, bank infrastructure and technical support. Consumers enjoy the convenience of instant payments, while the regulator and ecosystem invest substantially in transaction security. The question is who should bear these costs.

Cash handling also has costs. Merchants must deposit cash and ensure its safe custody, while consumers face the costs and risks of carrying and accessing it. MDR can therefore be viewed partly as cost-sharing by users of the digital-payment network rather than simply as a charge imposed on merchants.

In such markets, one side can rationally be subsidized because attracting participants increases the network’s value to the other. Consumers benefit from wider merchant acceptance, while merchants benefit from a larger digital-payment base. But once adoption is sufficiently high, the marginal benefit of another subsidized transaction may be smaller than the cost of maintaining the infrastructure. India is therefore moving from adoption economics to sustainability economics.

The case for differentiated pricing

The new MDR structure targets select higher-value merchant transactions while protecting small-value payments and small merchants. This resembles Ramsey pricing, under which charges are higher where demand is relatively price-inelastic and lower where users are more price-sensitive.

Transaction value, however, is an imperfect proxy for price elasticity. A ₹3,000 transaction at a large retailer and the same transaction at a low-margin small business do not have the same economic significance. A higher-value transaction does not necessarily imply a greater ability to absorb payment costs. A better system would therefore consider turnover, transaction frequency, sectoral margins and dependence on digital payments alongside transaction value.

Although MDR is formally charged to merchants, economic incidence can differ from legal incidence. Merchants can absorb the cost, reduce discounts, alter payment preferences or incorporate some of it into prices. The extent to which consumers ultimately bear the burden depends on the relative demand and supply elasticities. The government’s prohibition on directly passing MDR on to consumers addresses legal incidence but cannot fully determine economic incidence. This matters particularly in thin-margin sectors.

UPI’s infrastructure is interoperable, but the consumer-facing application market is more concentrated. If MDR becomes a significant revenue stream, larger platforms could have greater resources for customer acquisition, merchant incentives and complementary financial services. A reform designed to improve financial sustainability could therefore have unintended consequences for competition.

Two safeguards are important: transparent distribution of MDR revenue across ecosystem participants and continued competition oversight of the application layer. Revenue should finance cybersecurity, research and development, resilience and innovation, while supporting UPI marketing and adoption in new areas—that is, uses that create social value.

What should the policy architecture look like?

Four principles follow.

First, preserve zero MDR for P2P transactions and genuinely small merchants. Their price sensitivity and the wider inclusion benefits of low-cost digital payments justify continued support.

Second, make MDR cost-based and periodically reviewable. Independent audits should separate fixed infrastructure costs, transaction-specific costs, cybersecurity expenditure and costs attributable to other payment services. Without such accounting, there is no objective basis for determining reasonable MDR.

Third, move towards differentiated merchant pricing. Turnover, sectoral margins and transaction intensity should complement transaction-value thresholds, allowing Ramsey-style cost recovery with minimal distortion.

Fourth, recognize UPI partly as public infrastructure. It generates benefits that private users cannot fully capture, including lower cash-handling costs, greater formalisation, improved payment transparency and economy-wide productivity gains. Where social returns exceed private returns, some public financing is economically justified.

The choice is therefore not between ‘free UPI’ and ‘paid UPI’.

India has demonstrated that zero pricing can generate extraordinary network effects. The next challenge is to make the network financially sustainable without weakening them. The appropriate principle is neither permanent subsidization nor revenue maximization: charge where price sensitivity is low, subsidize where externalities are high, protect competition where network effects create market power, and use public resources where social returns warrant them.

UPI succeeded because India recognized that digital infrastructure requires more than markets alone. Its next phase will require equally careful economics of pricing, incidence and institutional design.

Atul Mehta is an associate professor of economics at the University of New South Wales (UNSW) Bengaluru, and Charan Singh is a former RBI chair professor of economics at IIM Bangalore.