A better way to fund UPI

UPI's costs could be shared more widely without distorting the incentives to use digital payments

The National Payments Corporation of India has announced a change to its flagship payments system, the Unified Payments Interface (UPI). From October 15, a merchant discount rate (MDR) will apply to a select range of transactions. This amount is expected to be absorbed by merchants and not formally passed on to consumers. The proposal has raised several questions and triggered protests by traders. They have also called for a “No UPI Day” on October 2. 

The proposed introduction of the MDR in the UPI regime has raised an important question: How should we understand the UPI system, and who should shoulder the cost of running and maintaining it? 

UPI is an essential and integral component of the country’s digital public infrastructure (DPI), alongside identification through Aadhaar and data exchange through account aggregators. The United Nations Development Programme defines DPI as “a set of foundational digital systems that enable secure and seamless interactions between people, businesses and governments. From verifying identities to opening bank accounts, ensuring fast and reliable digital payments to enabling safe and efficient data exchange across government services — DPI can make everyday life more connected and inclusive”. 

In India, UPI has been instrumental in the nudge towards a digital economy, which is often argued to deliver a range of gains including improved financial inclusion and a shift towards formalisation of the economy. For the financial system, the cost associated with operating a cash-based system would be reduced, alongside moderation in the cost of onboarding new customers for lending. 

Clearly, apart from the regulator, the UPI ecosystem has a large number of stakeholders — financial institutions, interface and technology providers, and end-users and merchants. The beneficiaries, too, include a broad range of agents in the economy. This forms the basis for the question of who should shoulder the cost of running and maintaining the system. 

Conventionally, a public good — which is non-rival (one person’s use does not reduce its availability to others) and non-excludable (it is difficult or impossible to prevent people from using it once it is provided) — should be financed out of taxes. UPI is non-rival but excludable. For such services, clearly it is possible to introduce a price mechanism to finance the service. The MDR proposes such a price, with elements of price discrimination thrown in.  

Another way to think about UPI is as a network with associated positive externalities. For such services, it is important for support to be made available in the initial stages until a critical stage is reached, after which the system can be self-sustaining. Under this approach too, the MDR is not entirely out of the picture once the system has reached a certain stage of maturity.

So what is the challenge now? We can think of three questions. 

One, as the ecosystem has a large number of beneficiaries, should the cost be apportioned among the different beneficiaries? As discussed above, apart from merchants and consumers, the beneficiaries include banks and other lenders who gained from a reduced cost of onboarding borrowers with access to UPI-based profiles. The cost of maintaining branches and running ATMs too has been significantly reduced. The business model of aggregators and fintech firms is based on the present design, which allows for profitable business ventures. It is also argued that widespread adoption of UPI has contributed to improved revenues, especially from the goods and services tax, as a result of formalisation in the economy. Is there a case for more widespread sharing of costs across the different beneficiaries? 

Two, if there are widespread benefits to the economy, exceeding the costs of the system itself, should one be thinking of an economy-wide optimisation as opposed to revenue realisation and efficient organisation for this activity alone?  

A study by the Department of Financial Services titled “Socio-Economic Impact Analysis of Incentive Scheme for Promotion of RuPay Debit Card and low-value BHIM-UPI Transactions” documents a range of socioeconomic effects. The Nasscom-Arthur D Little Study titled “Digital Public Infrastructure of India - Accelerating India’s Digital Inclusion” suggests that Aadhaar, UPI and Fastag together supported value creation equivalent to 0.9 per cent of gross domestic product in 2022. The study estimates significantly higher returns by 2030. Is there a case to be made for extended government support in this case? This is a question that arises for a number of other network public infrastructure services, like public transport, for instance. Each element of this network might not be self-sustaining in itself, but could add significantly to value creation in the economy. 

Three, the design of the proposed MDR introduces a levy for transactions above ₹2,000 at a rate of 0.4 per cent, with a maximum cap of ₹300 per transaction. This implies that for transactions exceeding ₹75,000, there is a flat fee of ₹300. For those between ₹2,000 and ₹75,000, there is a proportionate fee. A question that is doing the rounds on social media is whether this would introduce a behavioural change among people — a return to the question of choice of instruments for transactions. 

Can one consider an alternative — a flat rate fee on all transactions at, say, 20 paise per transaction? The number of transactions in August 2026 are suggested at 24.51 billion — that is, about 284.12 billion transactions in a year. A fee of 20 paise per transaction would amount to ₹5,882 crore. Flat rates are argued to create less distortion of behaviour. Perhaps in the interests of stability and growth in this digital public infrastructure, incentives to revisit the choice of payment instruments could be avoided.

---

The writer is director of National Institute of Public Finance and Policy