Wayne, a loyal MR reader, writes to me:
I am going to start tipping based on an estimate of the number of checks
a waiter presents per hour divided into my estimate of what his services are
worth per hour without regard to the amount of the bill.
Of course the way it works now we tip as a percentage of the bill, paying a kind of flat tax unrelated to labor effort. Was it really that hard to cover over the $100 bottle of wine?
Let’s assume everyone behaves that way. In essence we would tipping on the basis of how many plates are carried and not how much value is on the plate. The end result would be better service in poor restaurants and worse service in more expensive restaurants. People who patronize lower-price restaurants, or order lower-price entrees, would pay more in percentage terms in the form of tips. To some extent the price of food in these restaurants would fall to compensate and waiter wages in the restaurants would fall too. Waiters in the fancy restaurants would become more like fixed-price servants and in fact this already has happened in some fancy resorts.
If I wanted to defend Wayne’s view, I would invoke the following claim: maybe we tip the fancy waiters to feel fancy ourselves but could there be a greater potency of tipping at lower price ranges, where waiter quality is harder to monitor? Note also that more people eat in the lower price ranges, so shifting your tipping convention in that direction might bring a greater positive externality for society as a whole.