The transaction fees that businesses incur using credit and debit cards cost Americans billions of dollars each year. Despite the availability of alternative, cheaper account-to-account options that cut out the middleman, Americans have been slow to move away from card payment systems due to market mechanisms that hide the true costs of card transactions to consumers, writes Eli Orbach.
In 2025, businesses in the United States paid on the order of $200 billion in service fees to credit and debit card issuers and network owners. These fees accrue whenever a customer uses their card and translate into higher retail prices as businesses pass some of these costs onto the customer. The card network owners, in particular the duopoly Visa and Mastercard, have long been criticized for facilitating and maintaining this expense.
While competition among card networks could lower prices over time, the more consequential change would come from outside the card model entirely. Account-to-account (A2A) payment systems cut out intermediary networks by facilitating direct transfers from one bank account to another. In effect, A2A payments are digitized checks, costing a fraction of a typical card-based transaction. A2A is already the dominant form of payment for person-to-person, business-to-consumer (e.g., wages to employees), and business-to-business transfers.
Outside the U.S., it is also making headway in the consumer-to-business space. In India, the government-backed Unified Payments Interface (UPI) now rivals Visa as the single most used payment system by transaction count. The Brazilian Central Bank’s Pix system is a similar success: only five years after its introduction, it is now used by over 80% of Brazilian adults.
But despite international tailwinds and lower costs, A2A has failed to gain significant traction for American consumer-to-business transactions. After first comparing the characteristics and costs of cards and A2A, this article analyzes potential explanations for A2A’s slow diffusion. Rather than technological limitations, the main barriers to A2A adoption are market mechanisms like rewards programs and merchant surcharging restrictions that obscure the true costs of card networks for consumers.
Comparing cards and A2A
Credit and debit card networks process roughly two thirds of all consumer payments in the U.S. and roughly three quarters of in-person payments. In a transaction involving a credit or debit card, the customer’s bank will communicate with the merchant’s bank to transfer funds via the card network. Business models vary. American Express and Discover, for instance, are vertically integrated with their own banking services and often interface directly with consumers and merchants. Visa and Mastercard act more as coordinators between the participating banks. As part of the transaction, merchants pay an interchange fee to the card’s issuing bank, in addition to smaller fees for the network and payment processor. Depending on the business model and transaction type, the total fee for merchants hovers around 2-3% for credit cards and 1-2% for debit.
A2A is shockingly cheap in contrast. The Automated Clearing House (ACH), administered by the non-profit Nacha, is the dominant A2A rail in the U.S. and charges fractions of a cent for standard transfers. Just as with cards, merchants must pay additional fees for processing. The final median fee for ACH transfers is still just 0.8% according to a recent report by CMS Payments Intelligence, well below credit and debit cards. Services like Venmo and PayPal also rely on the ACH, though they can route payments through multiple different rails and add their own markups on top.
However, ACH’s slow authorization and settlement speed (usually over a day) is a dealbreaker for many merchants. Fast authorization in particular is critical, since delays can leave merchants uncertain whether the consumer actually paid until long after they left the store. Cards circumvent this problem by decoupling authorization and settlement, allowing the former to happen in seconds. That’s the loading spinner at the checkout counter.
But recently, a subset of A2A called instant payments has become available, and it facilitates both authorization and settlement just as quickly. The Clearing House, an association owned by several large banks, launched Real-Time Payments (RTP) in 2017. It processes most instant payments in the U.S., though it has recently found competition in the Federal Reserve’s FedNow, launched in 2023. Zelle, also managed by a consortium of large banks, offers a popular wrapper for instant payments. RTP and FedNow charge more than the ACH but remain cheaper for merchants than card networks—especially for higher value transactions. Both India’s UPI and Brazil’s Pix utilize instant payments.
Technological barriers to A2A’s adoption are overstated
Given the cost advantage, what is stopping A2A from spreading? Many discussions of A2A center on technological limitations, especially in comparison to highly developed American card networks, as a primary barrier to its diffusion. For example, borrowing and carrying credit is, at least for now, limited for transactions on American A2A rails. But this isn’t actually a technological restriction. Credit options are built into contractual arrangements with a bank, and they operate on top of the payment system rather than within. Adding credit on A2A rails is possible, and India’s UPI is already integrating credit. Furthermore, the existence of debit cards is proof that credit is nonessential for many transactions.
As a consequence of their intentionally lightweight architecture, A2A rails also lack many network features. Where card networks have expansive chargeback policies that allow customers to dispute charges, instant payments are essentially irreversible, and the ACH only offers limited recourse for unauthorized transactions. Card networks also provide frameworks for dispute resolution, more expansive fraud protections, and insurance against theft.
Consumers face a tradeoff between these additional features and the lower costs of A2A, which may be dampening demand for A2A services. However, consumer preferences for different features vary significantly across transaction types: you likely value insurance and dispute resolution forums when purchasing from a shady online retailer, but not when buying from your local coffee shop or from established international brands. The more expansive card networks have their uses, but network features alone cannot justify A2A’s near nonexistence in the American consumer-to-business space.
Market barriers, both natural and erected
Far more than the technology itself, market characteristics create the greatest hurdles to widespread A2A access and adoption in the U.S. Some of these market-induced barriers are commonplace across many industries. Switching costs, uncertainty among consumers and merchants, and (as mentioned previously) the lack of a credit option all likely contribute to A2A’s slow diffusion. Network effects in payment markets are another major barrier, which may be especially powerful in the U.S. due to its established card ecosystems. Cards are so ubiquitously used by consumers that almost all businesses are forced to accept card payment. Likewise, since most businesses accept cards, consumers gravitate toward them over less common payment methods.
Possibly the largest barriers of all, however, are the invisibility of transaction fees to consumers, incentives of credit card rewards programs, disparity in carried-balance interest charges, and bans on merchant surcharges that collectively obfuscate the true costs of credit and debit cards for consumers. These mechanisms distort the behavior of consumers and prevent the spread of A2A.
Service fees for both debit and credit cards are charged to merchants rather than consumers. A fee’s initial placement does not determine who ultimately bears the cost in ideal markets (since costs can be passed on to consumers), but it can have profound behavioral effects. Costs passed on by merchants are less visible for consumers than a separate line item for service fees would be, and consumers thus react less strongly.
Credit card issuers also make money by charging users monthly interest on revolving balances, averaging roughly 2% per month. Since potential future interest expenses are far more nebulous to consumers than direct checkout-counter expenses, this system succeeds in making credit cards appear cheaper than they actually are and diminishes A2A’s perceived cost advantage. Issuers then use these interest payments to fund their rewards programs, which further obscure the cost of using card networks. Card issuers’ expenditure on rewards often nears or even exceeds income from service fees on a per-transaction basis.
For credit card users who pay off their balances each month, these rewards programs significantly reduce their personal expense from using credit cards. However, for the roughly 50% of credit card users who carry a balance, even industry averages do not capture the full expense they bear. This latter group is subsidizing credit cards for the former, and because the full expense is concentrated on just a portion of the total population, the market demand for alternative payment technologies like A2A is reduced. So while A2A would most benefit the users who pay interest fees, widespread diffusion is hampered by the 50% who benefit from continuing to use credit cards and collect rewards. Furthermore, individuals who carry balances are disproportionately located in the lower half of the income distribution, precisely the group with relatively diminished capacity to experiment with new technologies.
Meanwhile, merchants have sought to recover the costs of card transaction fees by levying surcharges on card purchases, making the fees more visible to consumers in the process. However, most merchants face legal and contractual restrictions on doing so. All major card networks expressly prohibit surcharging on debit and prepaid cards, ostensibly to protect consumers from unpredictable pricing and merchant rent-seeking. Credit cards are not as tightly controlled, yet networks still place caps on and require advance notice of surcharges, and certain states enforce full prohibitions. Merchants can skirt these rules by offering discounts for cash (permitted by the Truth in Lending Act) as well as other payment methods including ACH and instant payments. However, the restrictions on surcharges still limit merchants’ ability to incentivize A2A usage via transparently pricing in the true cost, and behavioral studies have long shown discounts to be less effective at influencing consumer choice than surcharges.
Without the immediate price signal, A2A loses much of its luster. Across fee incidence, rewards programs, and surcharge restrictions, a card ecosystem emerges that can effectively block out cheaper rival technologies. India and Brazil illustrate economies where these card ecosystems don’t exist. In both countries, cash is a significantly more common form of payment than it is in the U.S., and cash was what the UPI and Pix primarily displaced. Cards did not have the presence to block A2A.
Conclusion
A2A is cheaper than cards and, for most transactions, no worse. But the design of the card ecosystem prevents most consumers from adequately observing the comparative cost. This suggests a clear policy roadmap: states and the federal government should consider mandating increased visibility of merchant-side service fees for consumers and carving out explicit allowances for surcharges instead of just discounts. Much of the cost of card service fees is already reflected in retail prices, limiting the burden a surcharge would have on consumers.
Some attention should also be paid to other emergent payment technologies. Buy now, pay later (BNPL) services, which aim to provide short-term credit options to consumers, are currently experiencing a surge of popularity as a substitute for credit cards, but they largely ride credit card rails and are even more expensive for merchants. And cryptocurrencies (especially stablecoins, which are pegged to major global currencies) have been in the discussion for some years now, though their lack of security features and central bank oversight makes them a risky prospect. Due to its relative cheapness and demonstrated success across sectors, A2A is consumers’ and merchants’ best long-term option to reduce transaction costs.
Author’s Disclosures: The author reports no conflicts of interest. You can read our disclosure policy here.
Articles represent the opinions of their writers, not necessarily those of the University of Chicago, the Booth School of Business, or its faculty.
Subscribe here for ProMarket’s weekly newsletter, Special Interest, to stay up to date on ProMarket’s coverage of the political economy and other content from the Stigler Center.
