Unfiltered: the interchange fees illusion

The July Unfiltered column from Dwayne Gefferie, Payments Strategist, is a commentary on the interchange case involving Visa, Mastercard, and US merchants and why the settlement doesn’t address the payments aspects that truly matter.

A federal judge spent June ending the longest-running fight in payments, and three weeks later nobody can agree on who lost. Merchants call it a giveaway, while the networks involved refuse to call it a win. Both sides are wrong in the same way.  

On 9 June, Judge Brian Cogan gave preliminary approval to a revised USD 38 billion settlement between Visa, Mastercard, and roughly 12 million US merchants, in a case first filed in 2005. And although everyone’s keeping score, nobody’s asking whether what got settled ever even mattered. 

Interchange is two-dimensional, yet the industry keeps approaching it as if it’s one-sided. The first is the acceptance cost, the fee a merchant pays for a card transaction. The second is the acceptance architecture, which controls the transaction: what network routes it, whether cards travel bundled, and which credentials are the default.

The 21-year trial was over interchange costs, as were the Durbin Amendment and Europe’s caps. Although this settlement ends the fight in the US, price was never that relevant for the value economics.

Card networks vs merchants: what got settled

The settlement terms are modest: the card networks remove 10 basis points from the average credit rate for five years, and hold standard consumer cards at 1.25% for eight; premium cards have higher rates, and commercial cards can run past 4%. The cuts barely touch on the end where value sits.

The real change comes from the acceptance rules, and for the first time since the networks built the honour-all-cards rule, this is no longer viable. Merchants can decline whole categories, premium and commercial consumers, as well as surcharge and steer. There’s a permanent change in acceptance rules, whereas the one on rates expires.

The networks also need to show merchants the acceptance costs for each card, category by category, and while that may sound minor, it’s not. Transparency makes a bundled system start coming apart.

Just look at the narrowness of the terms: merchants can turn away a category, not an issuer, so accepting a tier implies accepting every card in it. Nobody refuses 85-90% of their credit volume, which is what dropping rewards cards would mean. The National Retail Federation called it ‘window dressing’, and restaurant and convenience-store groups agreed. They’re right about the mechanics.

Why you can’t litigate a default checkout setting

Those keeping score miss that in a two-sided network, the cost lies in what’s visible, while the actual product is the architecture. Sure, you can cap the fee, but the power is untouched, belonging to whoever is already the default. Realistically, a merchant can’t walk away from Visa and Mastercard, and a cardholder won’t – the product locks both mutually, and no fee schedule rewrites this.

Europe experimented similarly and ultimately proved this very point. Regulators capped consumer debit interchange at 0.2% and credit at 0.3%, a fraction of the US level, and rewards programmes weakened. The card remained the default at almost every checkout. Card architecture in Europe is under pressure from something other than the cap: account-to-account (A2A) payments (bank-to-bank transfers that skip the card rails altogether).

Two decades of merchant litigation prove the same thing in reverse: even when you win on price, the architecture still absorbs it. Merchants spent a generation and a fortune in legal fees on the one axis that couldn’t move their position: while a court can lower the price, you’re not owning the rail.  The checkout default is also the economics owner, regardless of what the posted rate says.

Where did the real payment economics fight move?

If price is settled but not architecture, the fight that matters is in three other places, none being a courtroom:

1. The checkout page: surcharging is spreading, and it’s important to see how it affects incentives. A customer who earns 3% back and pays a 3% surcharge sees the reward cancelled out. If you do that at enough registers, the rewards model falls apart because fewer people accept it, not because of associated fees. 

2. The payments rails: in the US, consumer A2A payments could surpass USD 200 billion by 2027, growing around 19% a year. They show up first where cards are weakest – bill pay, big-ticket items, recurring charges, and the verticals that already surcharge or refuse plastic. Every dollar moved with A2A is one that interchange never sees.

3. Routing: the only factor directly targeting architecture. The Credit Card Competition Act would force large issuers to enable a second network on their cards, changing who controls the transaction instead of its pricing. It returned in January with President Trump’s backing, then failed to attach to a housing bill, failed again on the stablecoin bill, drew no new Senate sponsors, and lost a hearing when the networks’ chief executives didn’t show up. The bill addressing the architecture can’t move, which is no accident.

Australia shows the opposite: from October, the country regulator scraps surcharging and cuts interchange caps outright, which presumably removes about AUD 660 million a year from issuer interchange revenue. Like it or not, it took a regulator willing to touch the price and architecture simultaneously; litigation reaches only the price.

A decade-long interchange fight: what does the settlement mean for card networks and merchants?

None of this is a verdict on the networks. Visa and Mastercard did what any rational operator of a two-sided network would do: closed a two-decade legal exposure on terms they can absorb and then moved on, now building the next architecture themselves. Mastercard publishes guidance for merchants on A2A payments, the rail bypassing its own interchange. It’s not a company defending a fee, but one that knows the fee was never the advantage.

Merchants, who spent 21 years fighting, face a harder truth. Although they won price round after price round, they remain where they started, having sued over fees instead of architecture. Issuers are now the ones at risk, as the rewards system needs very little visibility at registers to change behaviour. The premium model bets that interchange stays high enough to fund the rewards points, but if you chip at acceptance long enough, the argument dismantles with each surcharge.

Nobody is a villain here, as they’re all just rationally responding to a structure most commentary still won’t name. The settlement is a receipt, not a merchant win or network loss, signifying the end of the price fight (the easy half), and that the half not settled (the default owner) is still open.

Future card acceptance repricing will come from checkout pages that offer 2% off for paying with A2A and from customers who calculate their reward points and decide the card isn’t worth it. A judge or Senate will have nothing to do with how this evolves.

 About the author

Dwayne Gefferie is a payments strategist, data scientist, and advisor with over 22 years of experience in the global payments industry. Unfiltered is his monthly column for The Paypers.