Demystifying UPI’s New Merchant Discount Rate Fee
On September 15, the National Payments Corporation of India announced the Merchant Discount Rate (MDR) Framework. Most people might think that 0.4% looks like nothing more than a rounding error. Even better, for most people, the MDR won’t even apply sinceonly about 4% of UPI payments to merchants exceed ₹2,0001.
Applicable transactions however make up about 66% of UPI’s transaction value2. That is why even a small percentage matters. The fee may not seem like much, but estimates put annual MDR revenue at ₹15,000-20,600 crore3. So what is it for and where does it go?
What was announced
From implementation, merchants will pay a merchant discount rate (MDR), the fee for accepting a payment, of 0.4% on eligible UPI payments above ₹2,000, capped at ₹300 for payments4 of ₹75,000 and above. Consumers (are supposed to) pay nothing, and banks have been told to ensure merchants don't pass the cost on to them.
P2P transfers, merchant payments up to ₹2,000, and small merchants receiving up to ₹1 lakh a month via UPI are all exempt. Credit-card-linked UPI payments are also treated differently. Some sectors like railways, telecom, insurance, fuel, and agricultural inputs pay a flat5 ₹5 above ₹2,000, while mutual fund, securities, and broker payments pay6 0.02%, capped at ₹300.
Why charge for something free?
This can be explained by the free-rider problem. Paul Samuelson's work on public goods showed that when people can use something without paying for it, nobody has a reason to fund it, so it ends up under-provided. UPI behaves this way. For six years, merchants used the rails at zero cost while banks and payment apps carried the bill, which the industry puts at around ₹20,000 crore a year for infrastructure, cybersecurity, and upkeep. The MDR makes the beneficiaries contribute.
Where does the money go?
It is distributed across the ecosystem. The MDR follows7 a 40,30,20,10 split: 40% goes to the customer's bank, 30% to the payment gateway, 20% to the UPI app, and 10% to the app's sponsoring bank. On a ₹10,000 payment, the ₹40 fee splits into ₹16, ₹12, ₹8, and ₹4. Separately, 5% of total collections will fund a small-merchant adoption drive, and 18% GST applies on top of the MDR. The government does not directly receive any of the 0.4% MDR fee but does indirectly collect taxes on it.
Who really pays?
Why is it necessary that the policy says the merchant must pay the MDR and must not pass on the cost to the consumer? Which party carries the burden depends on price elasticity, meaning how much each side's behaviour changes when the price changes. The side that can least afford to walk away pays most. The MDR is a fee, not a tax, but the logic applies to any charge on one side of a transaction.
Consider a mobile retailer with extremely thin margins, who can't absorb the fee comfortably. Such a retailer will have to bear the costs of the fee more than a retailer with comfortable margins who can choose to raise prices and pass on the charge to the customers. This is exactly what the policy aims to prevent. This is the same reason why industries like petrol pumps have asked for exemptions.
Why 0.4% feels bigger than it is
There is a behavioural reason the reaction is louder than the number. In their 2007 study, Shampanier, Mazar and Ariely found that people treat zero as a special price: free is not just a very low price, and demand jumps disproportionately when a price falls from one cent to nothing. The reverse also holds, so leaving zero feels like a bigger step than the percentage suggests. UPI has been at zero for six years, so even a small fee changes how merchants perceive the whole system. The thresholds of ₹2,000 and ₹1 lakh are designed to keep most users at that comfortable zero.
References (with links)
1. Free-rider problem: Samuelson, P. A. (1954). "The Pure Theory of Public Expenditure." Review of Economics and Statistics, 36(4), 387-389. Link: https://courses.cit.cornell.edu/econ335/out/samuelson_pure.pdf
2. Cambridge Article talking about the Legacy of Samuelson’s Theory :https://www.cambridge.org/core/product/CD54F8996C2F64B8C1E137D68E9B79DB
3. Tax incidence and price elasticity: Tax and Transfer Policy Institute, ANU (2019). "Economic Fundamentals: Tax Incidence and Elasticity," Tax Fact #11. https://crawford.anu.edu.au/sites/default/files/2025-03/Tax%20Fact%2011_2019_11.pdf. For a worked example, see EconPort, Georgia State University: https://econport.gsu.edu/content/handbook/Elasticity/taxincidence.html
4. Zero-price effect: Shampanier, K., Mazar, N., & Ariely, D. (2007). "Zero as a Special Price: The True Value of Free Products." Marketing Science, 26(6), 742-757. https://scholars.duke.edu/publication/859276
Sources
Comments