The opponents of tipping got some important things right. They saw that tipping was an aberration in a freemarket economy, and that tips had a lot in common with gifts. They also understood that economics alone could not explain why customers were willing to tip. What they missed was that the very things they thought made tipping un-American—the waiters’ need to cultivate a relationship with the customer, and their reliance on that customer’s largesse—were exactly what would eventually make tipping a powerful social norm.

So why tip? When people are asked, they usually say that they tip to reward good service. Yet how much people tip is determined mainly by how much their meal cost, and the cost of a meal at a given restaurant is usually only tenuously connected to the work required to serve it. (It’s just as easy to open a hundred-dollar bottle of wine as it is to open a thirty-dollar bottle.) In an extensive survey of tipping studies, Michael Lynn, a professor at Cornell, found only a weak correlation between the quality of service that people report receiving and the tips they give. On average, exceptional service raised tips by about 1.5 per cent, which, Lynn argues, is too small for waiters to notice. And countries where there’s no tipping—like Australia and Japan—don’t have worse service than the United States.