E-piphany

Something odd happened to me several times over our Christmas holiday trip back home, and I’ve been mulling it ever since. It feels significant, albeit on a long-horizon time scale.

At least three times, restaurants added a ‘credit card surcharge’ to our check, or had a sign on the door warning customers of the same. The surcharge was small, on the order of what the credit card processors charge the restaurateur (1-2% of the bill), and was often framed that way.

Think about that for a moment. First of all, credit card fees haven’t changed meaningfully in a long time. Probably on balance they’ve risen, but it’s fractions of a percentage point. In the running of a restaurant, it is total rounding error.

Moreover, if the cost of processing credit cards had gone up, say, 1%, then it would be much better to simply raise prices 1%. Diners aren’t particularly resistant to small changes in prices, especially after 2021 broke the skin of the milk so to speak. Adding a surcharge to use a different payment method ticks the customers off, and I saw this a couple of times.

The number of patrons who are using credit cards has probably gone up a lot in recent years. I’ve noticed myself that I so rarely use cash that I sometimes forget my ATM PIN. Apple Pay and other proximity-payment methods make it so easy that carrying my whole wallet so that I can have physical cash seems silly. It has bitten me a couple of times when I wanted to tip someone, but that’s the only time it has made a difference. Almost no one takes only cash any more. We are not yet (in the US) a ‘cashless society,’ but ease of payment has pushed us pretty far in that direction. But still, that effect is simply not big enough to make much of a difference to the restaurateur – and if it was, then it would be simply solved by changing prices just a tiny bit. Your $15 entrée becomes a $15.15 entrée. No one is walking out over that.

When you see something that seems to make little sense economically, it usually means one of two things. (1) there’s some weird behavioral bias happening, or the problem is complex and confusing, so that people are making the non-optimal decision, or (2) people are behaving rationally; you just haven’t figured it out yet. The former point seems unlikely here. The problem is pretty simple, and the behavioral biases work the wrong direction – your customers get irrationally annoyed by a 1% surcharge, so all else equal you’d want to avoid that.

So I have been mulling #2, and I have a possible answer. Why would restaurateurs annoy their customers by adding a 1% surcharge to their checks, which can be easily avoided by paying cash? Obviously, it’s because they want to receive cash. Why annoy their customers over something so small that can be addressed another way? Because the actual cost to them is not as small as it appears.

When a business receives cash, or credit, there’s a small difference in the revenue received. Why does this matter to gas stations, which have applied a discount for cash for decades and not to a service business like a restaurant? Another difference between a restaurant and a gas station is that a gas station’s costs are almost entirely the cost of goods sold – gasoline – while for a restaurant the COGS is more like 25-40% of revenues. In what way does this increase the value of cash to a restaurateur?

The answer is simple – by operating part of the restaurant on a cash basis, a very significant cost can be reduced: taxes. A gas station would have a hard time pocketing cash and not declaring the revenue, because it would be quickly obvious when the tax man looked at the books. If you’re taking in less revenue than the actual cost of the gasoline, something’s fishy. But for a restaurant, that’s harder to establish especially if you pay some of the staff in cash.

The only way it makes sense to me that some restaurants would risk ticking off customers in order to push them towards cash in a very blatant way is if the cash revenue is worth much more than a 1-2% advantage over credit card revenue, and if the number of cash-paying patrons was changing meaningfully. The former has been true for a while, but as long as plenty of people still paid cash there was no reason to risk annoying customers. Only if cash as a payment form is decreasing meaningfully – and I would bet it is – would this make sense.

I’m open to other possibilities.

The reason this is interesting to me is that the driving force here is the desire to avoid metering of revenue. But the habits of the customer base aren’t the only reason this is changing. A sign at my bank warns that anyone who transacts in cash will be subject to extra questions and ID requirements. As government deficits stay wide and taxation rates rise, incidence of avoidance should be expected to go up.

Some people are aggrieved by the movement by central banks towards Central Bank Digital Currencies (CBDCs) because they fear that authorities could abuse having absolute power over the medium of exchange. That only works if it is the only medium of exchange. But the restaurant behavior suggest that moving entirely to a cashless society could also raise prices in some ways.[1] If people simply won’t pay with cash, prices will have to go up to cover the additional taxes that business owners will have to pay on the newly-recognized revenues. Incidentally, to the extent that a movement towards contactless payments (CBDC or no) moves commerce from the cash economy to the metered economy…growth will also appear artificially higher by a small amount and tax receipts will also be higher than would otherwise be expected.

Outside of restaurants, I don’t know how prevalent cash payments for services are. I know that it is a large part of home improvement and maintenance, and I know that car dealers vastly prefer cash. If it’s just 2% of the economy, then this is merely interesting. If it’s 5-10% of the economy, it’s also significant. I don’t know that in either case I can see a trade to be made, but it’s interesting.

What do you think?

---

[1] Hey, I have to tie this back to inflation somehow.