Once a card is swiped, dipped, or keyed in, the transaction has to reach the issuer. The question of which network carries it there sounds like plumbing. For debit, it is the single most contested economic decision in the stack, because the choice of network usually changes who pays what.
This module is about that decision. We will keep interchange and the merchant side (modules 3 and 5) in their own lanes and focus on one thing: the incentives behind routing, and the regulated-debit routing fight that shapes how every US debit card is built.
Why routing is a fight, not a setting
On a credit card, routing is mostly settled. A Visa-branded credit card runs over Visa, a Mastercard credit card runs over Mastercard. There is no second road.
Debit is different by law. In the US, the same physical debit card carries more than one network, and the merchant gets to pick which one processes the transaction. That single fact creates the conflict. Merchants want to route over whichever enabled network costs them the least. Issuers and the global networks want volume to land where the economics favor them, which is usually the signature-debit path branded Visa or Mastercard.
The reason both sides care is that fee differences between debit networks are real and persistent. The interchange the merchant pays, plus network fees, can vary meaningfully depending on which network clears the transaction. Routing is where that money is decided.
The two roads on a debit card
A US debit card typically carries a "global" network (Visa or Mastercard, the brand on the front) and one or more domestic PIN-debit networks such as Star, Pulse, NYCE, Accel, or Shazam.
These roads do not behave the same way technically. PIN-debit networks historically ran on a single-message model, where authorization and settlement travel together in one message. The signature-debit path runs on a dual-message model, where authorization happens first and settlement follows separately, which is the same pattern credit uses.
That technical split used to map cleanly onto the checkout experience. PIN entry meant the single-message PIN-debit road; a signature or no verification meant the dual-message Visa or Mastercard road. The merchant could nudge the customer toward PIN, but the customer had a vote, and so did the card.
Where PINless changed the math
The piece that makes least-cost routing practical is "PINless" debit. A PINless transaction runs over a single-message PIN-debit network without the cardholder entering a PIN. The merchant gets the cheaper single-message road without forcing the customer through a PIN pad, and increasingly without a card-present terminal at all.
This is what made routing a live decision online, not just at a physical lane. For card-not-present transactions, where there was never a PIN to begin with, PINless is the mechanism that opens a second road at all.
Least-cost routing in practice
Least-cost routing (LCR) is the merchant-side logic that inspects each debit transaction and sends it over whichever enabled, eligible network costs the least to clear. It is run by the merchant's acquirer or payments processor, not by the card brands.
The savings come from steering volume off the more expensive signature-debit path and onto a cheaper PIN-debit network when one is enabled and the transaction qualifies. For some transactions, particularly larger ones, the single-message path also carries lower per-transaction cost because it is treated as lower risk.
A worked example
Take a regulated issuer, meaning a bank with $10 billion or more in assets, whose debit interchange is capped under the Durbin amendment at roughly 21 cents plus 0.05 percent of the transaction, plus a 1 cent fraud-prevention adjustment. On a $40 purchase that is about 23 cents of interchange on the regulated cap.
The interchange cap applies regardless of road, but the network fees and incentives layered on top do not. If the global signature-debit path adds higher network assessments and the merchant's effective all-in cost there is, say, 30 cents, while an enabled PIN-debit network clears the same $40 transaction at an all-in 24 cents, LCR routes to the PIN-debit network and the merchant keeps the 6 cent difference.
Six cents looks trivial. Multiply it across millions of transactions and the aggregate is large. Industry estimates put potential savings for e-commerce merchants from card-not-present PINless routing at several billion dollars collectively, on the order of a 25 percent reduction in interchange and assessment costs on the affected volume. That is the prize both sides are fighting over.
The regulated-debit routing fight
The merchant's ability to route at all is not a market accident. It is mandated. Regulation II, the Federal Reserve's implementation of the Durbin amendment, contains two separate provisions that matter here.
The first is the interchange cap, which applies only to regulated issuers. The second is the routing provision, which applies to every debit card issuer regardless of asset size. Under it, no issuer or network may restrict a debit transaction to fewer than two unaffiliated networks, and none may inhibit a merchant's ability to choose among the networks the issuer has enabled.
This is the structural rule that puts two roads on the card in the first place. Even a small, interchange-exempt community bank must enable two unaffiliated networks. The price control and the routing mandate are decoupled on purpose.
The card-not-present gap, and how it closed
When the rule first took effect in 2011, the market had not built broad support for multiple networks on card-not-present transactions. In practice, many online debit transactions had only one viable road, which meant no real routing choice and no LCR.
The Federal Reserve closed that gap. It finalized amendments in October 2022, effective July 1, 2023, clarifying that the two-unaffiliated-networks requirement applies to card-not-present transactions, including prepaid. Issuers now have to enable a second routable network for online debit, which is what made card-not-present LCR a default expectation rather than an edge case.
What the 2025 court ruling did and did not touch
In August 2025, the US District Court for the District of North Dakota, in Corner Post v. Board of Governors, vacated Regulation II's interchange fee standard, holding the Fed exceeded its statutory authority. The court stayed its own ruling pending the Fed's appeal, which is now before the Eighth Circuit.
The case targets the interchange cap, not the routing mandate. The two-network requirement and merchant routing choice rest on the same statute but a different provision, and they were not the subject of the vacatur. Treat the cap as legally contested and the routing rules as intact while the appeal runs, and watch the Eighth Circuit rather than assuming the whole regime moves together.
Takeaway
Routing is where debit economics actually get decided, because the law guarantees a second road and hands the steering wheel to the merchant. Least-cost routing is the merchant exercising that right; PINless and the 2023 card-not-present clarification are what made it work online. The interchange cap and the routing mandate are separate rules with separate fates, so when the headlines move, check which one. We pick this back up when incumbents survive (module 9) and the model comes under pressure (module 10).