A card taps a terminal, the screen says approved, and the transaction looks like one payment moving in one direction. Understanding how do interchange fees work means seeing the second payment that happens at the same moment, in the opposite direction, that the person holding the card never sees.
The merchant does not receive the full price. A fee is deducted and routed to the bank that issued the card, and the size of that fee depends on what kind of card was used.
For debit cards it is capped by federal regulation. For credit cards it is not capped at all, and that single asymmetry explains most of what follows.
Four parties and a fee
The arrangement involves more participants than a shop and a customer, and naming them makes the money flow legible.
The cardholder has a card from an issuing bank. The merchant has an account with a different bank, called the acquirer. A card network sits between the two, setting the rules, moving the authorisation messages and settling the balances at the end of the day.
When a purchase clears, the acquirer pays the merchant the sale price minus a deduction, and the largest component of that deduction is the interchange fee, which goes to the issuer.
The network takes its own smaller cut and the acquirer takes a margin on top. Smaller merchants are often quoted a single blended rate rather than a breakdown, which is why the underlying structure stays invisible even to many of the businesses paying for it every day.
The merchant pays, and then everybody does
The fee lands on the merchant, and a business running on thin margins does not absorb a percentage of every single sale indefinitely out of goodwill.
It goes into prices. The cost of card acceptance is spread across the whole price list rather than itemised on any receipt, which means it is paid by every customer of that business regardless of how they choose to settle the bill at the end.
Someone paying cash is therefore contributing to the cost of somebody else’s card transaction, and to the rewards funded out of it. The transfer is invisible in both directions, which is precisely why it survives.
That structure — a charge nobody sees, recovered from people who had no part in it — is the same shape as the arrangement behind who pays for stadiums, where the announced payer and the eventual payer are different groups.
Debit is capped by law
In 2010 Congress instructed the Federal Reserve to set standards for debit interchange, and the resulting rule is unusually specific for financial regulation.
Under Regulation II, an issuer covered by the standard complies only if every interchange transaction fee it receives or charges for an electronic debit transaction is no more than the sum of two things: 21 cents, and 5 basis points multiplied by the value of the transaction itself.
A separate provision allows an additional fraud-prevention adjustment of no more than 1 cent per transaction, available only to issuers that meet the standards set for it.
| Component | Maximum |
|---|---|
| Fixed element | 21 cents per transaction |
| Variable element | 5 basis points, or 0.05%, of the transaction value |
| Fraud-prevention adjustment | 1 cent, if the issuer qualifies |
Work through a fifty-dollar debit purchase and the ceiling comes to roughly 24.5 cents. That is the entire amount the issuing bank may take from the merchant on that sale, however large the basket or expensive the card.
Credit is not capped at all
Nothing comparable applies to credit cards at all. The same physical gesture at the same terminal produces an entirely different regulatory outcome depending only on which card came out of the wallet.
Credit interchange is set by the card networks in their own published schedules. It varies by card type, by merchant category and by how the transaction was taken, and it is typically expressed as a percentage of the sale plus a small fixed amount per item.
The percentages are considerably higher than the debit ceiling, which is why some merchants steer customers towards debit, why others set minimum spends for card payment, and why explicit surcharging has become more common in the places where network rules and state law between them permit it.
Congress did not overlook credit cards by accident. The cap was written to cover debit deliberately, and proposals to extend competition rules to credit have been introduced repeatedly since without ever passing.
Rewards are the interchange coming back
This is the part that reframes the whole subject for most people. Premium credit cards carry higher interchange rates than basic ones, so a merchant accepting a card with generous cashback or airline points pays more for that sale than for an identical sale on a plain card.
The rewards are funded out of that difference. They are not a gift from the bank and they are not free money — they are a share of a fee the shop paid, handed back to the customer whose card triggered it in the first place.
Which makes the whole system a transfer with a predictable direction. Prices rise for everybody who shops there, and the rebate flows back to holders of the most expensive cards, who tend not to be the customers paying in cash.
None of that makes rewards a scam, and nobody is being deceived. It does mean the question “who pays for my points” has a specific and locatable answer.
The rule nobody mentions
The debit regulation contains a second half that gets far less attention than the price cap and may matter just as much. Alongside the fee standards, the rule prohibits payment card network exclusivity arrangements and restricts routing limitations for debit transactions.
In plain terms, a debit card cannot be locked to a single network. It must be enabled on at least two unaffiliated networks, and the merchant — not the bank, and not the cardholder — chooses which one carries the transaction.
That choice is worth real money at scale, because the competing networks price differently, and it introduces a form of competition into a market that would otherwise have none at all.
The regulation also explicitly prohibits evasion and circumvention, which is the drafter quietly acknowledging that a rule capping one fee is an open invitation to invent a differently named one.
Who is exempt
The cap does not apply universally, and the exemptions explain a good deal of the argument about whether it worked.
- Small issuers — banks and credit unions below the asset threshold are outside the fee standard entirely
- Certain reloadable general-use prepaid cards
- Debit cards issued under government-administered payment programmes
The small-issuer carve-out was intended to protect community banks and credit unions from a rule aimed squarely at the largest institutions. Whether it has held up in practice, or whether market pressure eroded it anyway, is one of the more genuinely contested empirical questions in the field.
The Federal Reserve publishes average interchange fees by network, separately for covered and exempt issuers, which is the closest thing to neutral evidence available.
What the fight is actually about
Search this subject and most of what appears is advocacy, from banking trade associations on one side and merchant coalitions on the other, and both are arguing about the same disputed transfer of money.
Retailers say the fees amount to a levy on every sale that bears no relation to the actual cost of processing it, and that ordinary competition cannot discipline a price the person choosing the payment method never sees and never pays directly.
Banks say interchange funds fraud protection, guaranteed payment to the merchant, credit losses and the rewards customers now expect as standard, and that capping it simply shifts the cost onto account holders through higher fees somewhere else on the statement.
Both claims are partly true, which is why the evidence is genuinely mixed and why the argument has now run for well over a decade without resolving.
The honest position is that the debit cap plainly moved money from issuing banks to merchants. The extent to which either group then passed that on to ordinary customers, in lower prices or in higher account fees, is exactly what the studies disagree about.
What is worth looking at
Three things make this legible in a way the advocacy does not. The first is which card is actually being used, since debit and credit sit under completely different regimes despite feeling identical at the terminal.
The second is that the published cap is a ceiling and not a rate. Issuers are free to charge less than the maximum, and the Federal Reserve’s own data on what is actually being charged is far more informative than the number in the regulation.
The third is that a card surcharge at the till is not something the merchant invented for the fun of it. It is an attempt to move a cost out of the general price list and onto the specific transaction that caused it, which reads as welcome transparency or as an unpleasant surprise depending on which side of the counter you are standing.
Cards are also the point where payments and deposit insurance meet in the same account, and the protections covering the money itself run on an entirely separate set of rules — as how FDIC insurance works sets out.
The useful correction is also the simplest one. Card acceptance has never been free to the shop, the cost of it was never going to stay with the shop, and the price of it is already sitting inside what you paid at the till.
