A coffee shop in San Francisco pays 2.9% on every card swipe. A grocery chain in Australia pays 0.5%. A merchant in India accepting RuPay debit pays zero. All three process consumer cards on established networks.

The difference isn’t negotiating skill or transaction volume. It’s the intersection of interchange economics, regional regulation and network competition that most practitioners never fully unpack.

The four-party model and its variations

The standard card transaction involves four parties: issuing bank, acquiring bank, card network and merchant. But this “four-party model” masks significant structural differences across networks.

Visa and Mastercard operate pure four-party networks. They set interchange rates but never touch funds directly. Issuers and acquirers are separate entities, creating the interchange flow that defines merchant costs.

American Express and Discover historically ran three-party (closed-loop) networks where they served as both issuer and acquirer. This meant no interchange, just a single merchant discount. Amex has since opened its network to bank issuers (OptBlue program), creating a hybrid model with interchange-like economics.

UnionPay dominates China with 90%+ market share. It operates as a four-party network domestically but functions more like a closed system given the limited competition. Cross-border UnionPay acceptance has grown, but interchange structures differ significantly from Visa/Mastercard norms.

RuPay in India runs on the NPCI infrastructure alongside UPI. The government mandated zero MDR (merchant discount rate) on RuPay debit transactions, fundamentally breaking the interchange model. Issuers receive government subsidies instead of interchange revenue.

This structural variation means “interchange” isn’t universal. It’s a specific economic model that some networks have abandoned or never adopted.

What merchants actually pay: pricing model breakdown

The merchant discount rate (MDR) bundles multiple fees, but how it’s presented varies dramatically by processor and region.

Interchange-plus (IC+) pricing passes through actual interchange costs plus a fixed markup. A merchant sees line items like “Visa CPS Retail: 1.51% + $0.10” with a processor markup of “0.25% + $0.10” added. This transparency reveals true costs but creates billing complexity. Large merchants and payment consultants prefer IC+ because it exposes optimization opportunities.

Blended/flat rate pricing (Stripe’s 2.9% + $0.30, Square’s 2.6% + $0.10) averages all card types into one rate. The processor profits when merchants process mostly debit; the merchant loses when premium rewards cards dominate. Simple to understand, expensive for most card mixes.

Tiered pricing buckets transactions into “qualified,” “mid-qualified” and “non-qualified” categories. Processors define these buckets opaquely, often downgrading transactions for minor technical reasons. A keyed-in transaction instead of swiped? Downgrade. Address verification failed? Downgrade. This model maximizes processor margin through information asymmetry.

Regional flat rates exist where regulation caps interchange. In the EU, with interchange capped at 0.2% for debit and 0.3% for credit, processors compete on flat rates around 1.0-1.5% total. Australian merchants see similar dynamics with the RBA’s 0.5% credit interchange cap.

The pricing model determines whether merchants can optimize costs or are captive to processor decisions.

Interchange rate mechanics: why cards cost different amounts

Published interchange tables run hundreds of pages. Visa US alone has 150+ rate categories. The key drivers:

Card type: Consumer credit cards pay issuers for credit risk and rewards funding. Basic credit runs 1.3-1.8%. Premium rewards cards (Sapphire, Platinum equivalents) hit 2.0-2.5%. The cardholder’s 2% cash back comes from merchant interchange.

Debit regulation: US Durbin Amendment caps interchange at 21¢ + 0.05% for issuers with over $10B in assets. Small bank debit (exempt from Durbin) runs 0.7-1.0%. This creates a two-tier debit market where merchant routing choices matter.

Transaction type: Card-present (CP) transactions with chip/PIN cost less than card-not-present (CNP) due to fraud risk differences. CP might be 1.51% + $0.10 while CNP for the same card runs 1.80% + $0.10.

Merchant category: Supermarkets qualify for “supermarket interchange” around 1.22% on credit. Restaurants pay standard rates plus deal with tip adjustment complexity. Utilities and government often get preferential rates as policy.

Data level: B2B transactions can qualify for reduced rates (Level II/III) by passing purchase order numbers, tax amounts and line-item details. This can reduce interchange by 0.5-1.0% on corporate cards.

Geography: Cross-border transactions add 0.4-1.0% in additional fees. A US merchant accepting a UK-issued card pays domestic interchange plus cross-border assessments from both the network and sometimes the issuer.

Regional regulatory approaches and their consequences

European Union: Interchange caps (0.2% debit, 0.3% credit) under IFR eliminated high-margin rewards cards. European issuers can’t fund premium benefits, so those products largely don’t exist. Merchants pay less; cardholders get less. Total card acceptance costs dropped roughly 40% post-regulation.

Australia: The RBA capped interchange and mandated least-cost routing (LCR), requiring merchants be offered the cheapest network for dual-network debit cards. Eftpos, the domestic debit network, competes with Visa/Mastercard debit on cost. Interchange-funded rewards declined significantly.

India: Zero MDR on RuPay debit and UPI transactions eliminated interchange entirely for these rails. The government subsidizes NPCI and member banks directly. This policy decision prioritized financial inclusion over network economics. Merchants pay nothing on the majority of digital transactions.

Brazil: PIX, the central bank instant payment system, charges zero merchant fees and settled $1.2 trillion in 2023. Card networks compete against a free alternative backed by regulatory mandate. Credit card interchange remains high (around 1.5%) but faces ongoing pressure.

United States: Durbin capped debit interchange for large issuers but left credit untouched. Banks responded by reducing debit rewards and increasing credit card marketing. The unintended consequence: higher credit card penetration, meaning higher average merchant costs despite the regulation.

Singapore, UAE, Hong Kong: Light interchange regulation with market-set rates. Premium card programs thrive. Merchant costs run 1.8-2.5% on credit, similar to pre-regulation US levels.

Each regulatory approach created distinct market structures. Merchants operating globally navigate fundamentally different payment economics by region.

The acquirer and processor layer

Between networks and merchants sit acquirers, processors and payment facilitators (PayFacs), each extracting margin.

Acquirers underwrite merchants, assuming liability for chargebacks and fraud. They hold merchant reserves, manage risk scoring and maintain network certifications. Traditional acquirer margin runs 0.3-0.8% depending on merchant risk profile.

Processors handle transaction routing, authorization and settlement mechanics. Some acquirers process in-house; others outsource to companies like TSYS, Fiserv or Worldpay. Processing fees add $0.05-0.15 per transaction.

Payment facilitators (PayFacs) like Stripe and Square aggregate merchants under a master merchant account. This simplifies onboarding (no individual merchant underwriting) but limits customization and often means blended pricing. PayFacs earn the spread between their negotiated interchange rates and merchant pricing.

Payment orchestration layers (Spreedly, Primer, Checkout.com) add another tier, routing transactions across multiple acquirers for authorization optimization. Sophisticated merchants use orchestration to improve approval rates by 2-5% and reduce costs through smart routing.

Merchant services providers (MSPs) and ISOs add sales and support layers, each taking margin. A small merchant’s 2.9% rate might include 1.7% interchange, 0.15% network fees, 0.40% acquirer margin, 0.30% processor fees and 0.35% ISO commission.

Why alternative rails haven’t won

Card networks survived decades of “disruption” attempts. The economics explain why.

Network effects compound: 4 billion cards in circulation. 80 million merchant acceptance points. Each new cardholder makes the network more valuable to merchants; each new merchant makes it more valuable to cardholders. Alternatives start from zero.

Credit extension: Cards provide 30-60 days of float to consumers. Alternative rails (ACH, RTP, UPI) require immediate funding. For consumers, this is a feature downgrade.

Fraud liability: Chargebacks shift fraud losses to merchants in defined circumstances, but also protect consumers. Direct bank transfers offer no equivalent protection. Merchants accepting alternatives assume more fraud risk.

Infrastructure lock-in: POS terminals, e-commerce plugins, ERP integrations and banking relationships all assume card rails. Switching costs extend far beyond payment acceptance.

Where alternatives succeeded, government intervention forced the change. UPI and RuPay in India, PIX in Brazil, Faster Payments mandates in the UK. Market forces alone haven’t broken card network dominance in any major economy.

The practitioner view

Interchange isn’t a single fee. It’s a complex system of cross-subsidies, regulatory arbitrage and market power that varies by card, merchant, region and transaction type. That 2-3% headline rate obscures:

  • Dramatically different economics by card type (rewards vs. basic, credit vs. debit)
  • Pricing model choice that can swing costs 0.5-1.0%
  • Regional regulatory environments that range from zero-fee mandates to unregulated markets
  • Multiple intermediaries each extracting margin
  • Technical factors (data levels, transaction type, MCC) that create optimization opportunities

Understanding this system explains why merchants in different contexts pay such different rates, why regulatory interventions produce unintended consequences and why decades of “disruption” left card networks stronger than ever. The fees aren’t arbitrary. They’re the equilibrium of a complex system optimized over fifty years.