The Commons · a hidden transfer everyone is in on

Who Really Pays for the Points

Credit-card rewards feel free. They are not free: they are funded by a fee baked into the price of everything, a fee the cashier charges you whether you pay with a rewards card, a debit card, or a fistful of twenties. A major 2026 study puts the yearly transfer at about $30 billion, flowing from people who pay cash and debit to people who pay with rewards cards, and it runs uphill, from poorer households to richer ones.

Here is the trick. When a store accepts a card, it pays an interchange fee: typically about 1.9% of the sale, most of which funds the cardholder's rewards. The store doesn't post two prices. It folds that cost into one shelf price that everyone pays. So the fee is already in the number on the tag before you decide how to pay. Pull out cash and you don't escape it; you just forfeit the rewards it was paying for.

Set your own numbers below. The one thing you operate is how you pay; everything recomputes from the same arithmetic the checker at the bottom re-runs offline.

Notice the break-even. You come out ahead only if your reward rate beats the markup: only if r > 0.95%. Below that line every shopper is paying more in hidden markup than they get back in points, and the difference is pure subsidy to whoever is above the line. That's why the game is stable: for you, individually, carrying the best rewards card is the rational move, which is exactly why nobody stops, and the transfer never unwinds.

The part that stings: paying cash doesn't get your money back

The intuitive escape ("fine, I'll just pay cash") mostly doesn't work. To recover the fee you'd need the store to charge you less for cash. Almost none do: in 2022 only about 3% of cash payments came with any discount (Atlanta Fed). So the cash user pays the full baked-in markup and gets nothing back: the worst seat in the house.

Which way it runs: uphill

If everyone used cards the same amount, the markup would wash out; you'd pay it and get it back. It doesn't wash out because card use rises steeply with income. Lower-income households pay cash far more often, so they pay the markup and collect fewer of the points it funds. In the Federal Reserve's 2024 diary, households under $25k paid cash for 24% of payments; households over $150k, just 9%. That gap is the engine of the regressive transfer.

Bars show each tier's yearly net at the current markup: red = pays in, green = collects. Direction is what's robust; the two endpoint cash shares (24% / 9%) are Fed-published, the middle three are interpolated between them and labelled below.

The honest haircut: it's smaller than a naïve model says

A crude version of this argument assumes everyone shops at the same store and pays the same fee. Reality is messier and it cuts against the alarm: cash-heavy shoppers sort toward lower-fee merchants (discount stores, cash-friendly corners), and merchant fees vary. The study finds this consumer sorting and fee heterogeneity shrink the transfer by about 25%: real, but nowhere near enough to erase it.

Policy and product design already moved this

Two forces reshaped who wins. The 2010 Durbin Amendment capped debit interchange, and the study finds regulated-debit users lost about $9.6B in vanished rewards and free checking, a transfer from middle-income to higher-income households. Meanwhile premium cards grew from 15% of credit volume in 2006 to 60% by 2022, and their richer fees widened the gap: premium holders gained about $7.9B. Both the cap and the card were found regressive. The incidence isn't fixed by nature; it's shaped by rules and rewards, and it has moved before.

The study's measured incidence, in basis points of purchasing power. These are the coefficients the calculator's per-user math is calibrated against; our simple μ = fee × share reproduces the cash line to within a basis point.