But if you look at the research you find that a lot of it doesn’t buttress my priors. In the first place, the amount of a tip is rarely related to the quality of the service. Michael Lynn of Cornell, who is the leading scholar on the subject, finds that the quality of the service has a relatively small effect on tip size.

What matters most is the size of the check. If you want bigger tips, induce your customers to order more. The second thing that matters is the customer. A survey of 40 million Uber trips found that men tip more often than women and people in the middle of the country tip better than people on the coasts. The identity of the customer matters more than the quality of the service.

In short, the meritocratic argument for tipping falls apart. Then there are all the downsides:

Tipping inflames a sexist dynamic. Some men use their tips as leverage to harass female servers. Young blond women are tipped more than older brunettes. Male Uber riders tip female drivers 12 percent more, but only if they are young.