In order to understand how and what to regulate, we borrow from the rationale followed for other two-sided markets that exhibit cross-platform externalities. The recent discussion paper by the RBI on charges in payment systems has triggered widespread public debate, especially on the zero-charge framework for UPI transactions. Many, including market players and the National Payments Corporation of India, have come out in support of moving away from the zero merchant discount rate (MDR), explaining how the lack of commercial compensation discourages new players from entering the market and deters investments in the upgradation of consumer protection and security. We add some more nuance to this debate, explaining key considerations and the approach to price regulation of UPI.
A good starting point is to identify reasons why a regulator might want to intervene in the price setting of the payments market. Two important reasons stand out. One, goals of financial inclusion or viewing digital payments as a public good and two, addressing market failures such as the presence of dominant firms or externalities that may arise due to the two-sided nature of this market. For both objectives, regulators might want to cap or set to zero the MDR (paid by merchants to their payments service provider) or the interchange fee (paid by the acquiring bank to the issuing bank), or both.
In the case of UPI, the government subsidises the operational costs of facilitating UPI transactions, which is reportedly inadequate. In January 2022, the Payments Council of India reported that the industry expected a loss of Rs 5,500 crore. Even with a public good motive, in the absence of evidence, one cannot assume this to be the best allocation of limited government resources. As per the Indian Digital Payments Report (second quarter of 2022), the average ticket size of P2M transactions (person to merchant) on UPI is Rs 820. RBI’s estimated cost of Rs 2 for processing a Rs 800 transaction, is 0.25 per cent of the transaction value, much lower than the MDR cap set at 0.9 per cent for debit cards and an MDR of 2 per cent being proposed for RuPay credit cards on UPI. A floor MDR of 0.25 per cent is, therefore, not unreasonable. Arguably, these are substitutable services competing for the same pool of merchants. Policymakers must also bear in mind behavioural challenges in moving from zero MDR to a positive MDR. Anchored at a zero MDR since January 2020, merchants, especially ones with thin margins, may hesitate to accept an increase in MDR, even if they benefit on net terms.
In order to understand how and what to regulate, we borrow from the rationale followed for other two-sided markets that exhibit cross-platform externalities. Simply put, consumers benefit more if the size of the merchant network accepting a payment instrument (for example, debit cards) is larger and, at the same time, merchants benefit more if many consumers use debit cards. Card networks like Visa and Mastercard compete for banks, usually not too many, to issue their cards. Since the acquiring bank must pay the interchange fee, they recover these costs from merchants. In most jurisdictions, the interchange fee is regulated to prevent banks from charging exploitative rates and the MDR is left to be commercially determined. This is also done for administrative ease, since banks are fewer, while monitoring bank-merchant contracts can be onerous.
In the UPI parallel, involving payment service providers of payers and payees, the remitter and beneficiary banks as well as NPCI, RBI could either regulate the interchange fee between payment service providers or the merchant discount rate charged by them. The market for merchant acquisition is usually more competitive and can be left unregulated, and if necessary the interchange fee between the two payment service providers can be regulated. If both markets are sufficiently competitive, regulation could mean establishing a floor/ cap charge. The decision on what to regulate is, therefore, crucial. A related example is available in the telecom industry where facilities provision is regulated through the interconnection fee, while retail prices for the relatively competitive telecom services segment are left to the market. For externalities of the two-sided market to be internalised, the choice of instrument must be carefully evaluated.
The next step is to determine the price level, which is a lot more tricky. Drawing from economic theory, the optimal level would depend on whether the regulator cares only about consumer welfare (as opposed to total welfare), and whether the issuing and acquiring banks make positive margins on each transaction. In general, benefits of regulatory intervention should outweigh the costs of intervening. The costs of intervening not only include the administrative costs, but also potential costs arising from setting the wrong interchange fee or cap, as well as any costs arising from the impact of the intervention on future investment and innovation in the market.
Pix, a two-year-old interoperable digital payments system in Brazil, provides a good comparison of how price setting might be considered in the UPI context. Pix does not regulate MDR, payment service providers have the freedom to set MDR, though in practice most banks currently don’t charge an MDR, largely to onboard more merchants on their platforms. The indicated cost is R$ 0.01 for each 10 transfers, or 16 paise in Indian rupees for every 10 transactions. This is substantially lower than the costs estimated for India and is also perhaps the reason why payment service providers are not immediately inclined to recover costs through MDR.
Policymakers must collect more data on costs of transfer, user preferences, both merchants and consumers, as well as undertake a thorough analysis of substitutability and competition in the digital payments sector, to put our best foot forward in helping achieve the potential of UPI in India.
Vatsala Shreeti is an economist at the Bank for International Settlements, Mansi Kedia is Senior Fellow at ICRIER.
Views are personal