The Problem with Tipping

Association between tipping and corruption at the country level, by Magnus Torfason. I have no idea what's going on to the left of zero.

Association between tipping and corruption at the country level, by Magnus Torfason. I have no idea what’s going on to the left of zero.

The topic of tipping has been on my mind since a dinner in San Francisco this April, when Allan Dafoe brought up the difference between Sweden and the US. In Europe, the pattern seems to be that workers are paid more and tips are not expected. Sometimes (eg in Germany) a waiter would likely be told to keep the change from a cash payment, but the reason given for this is that making change is demeaning. In the US, most waitstaff are paid a paltry $2.13/hour and make the rest up in tips. Why has the micro-institution of tipping endured in the US, and what are advantages and weaknesses of it?

One advantage of tipping is that it helps solve the principal-agent problem between the restaurateur and the staff. It is difficult for the restaurant manager to monitor the quality of servers, but much easier for the customers to do so. Thus, having customers set the compensation level for servers encourages them to be responsive. This in turn keeps customers coming back to the restaurant, which is the manager’s goal. However, tip-pooling can undermine this effect, as can the opportunity for servers to offer lower quality service to a higher volume of tables.

A major disadvantage to tipping is that it seems only weakly correlated with service quality. In an informal, small-n poll of fellow grad students on the same San Francisco trip, most of us tip a relatively fixed percentage of the bill (20-25 percent) regardless of service. This tends to be associated with work experience in the service industry, which can also lead to higher tipping percentages. Furthermore, small “nudges” such as paying on an iPad that offers tips in round dollar amounts rather than percentages can have a strong effect on the tips received by staff.

Another problem with tipping is uncertainty about when it is or is not appropriate. In a country where it is never appropriate, there is no uncertainty. When tipping exists in some settings but not others, there can be a great deal of uncertainty about when to use it. Restaurants and taxis seem like obvious tipping situations. But what about when you order food for carry-out? (I vote yes, but a smaller percentage.) And why not tip the attendants on a flight, who provide beverage and sometimes food service?

These and other problems with tipping are well-known, yet the micro-institution seems fairly ingrained in American society. If there were one profession where I would add tipping it would be my pharmacist–they often provide fast and friendly service for something I could not do myself. On the whole, though, we would be better off without this confusing and inefficient practice.

See also:

Why Do We Tip? (Planet Money podcast)
Should Tipping Be Banned? (Freakonomics podcast)
Research by Michael Lynn (over 50 papers)

2 thoughts on “The Problem with Tipping

  1. Easily one of the worst graphs I have seen. The y-axis is not labeled? What does “prevalence” mean in this context? And indeed: Negative prevalence what?

  2. It is true that a large portion of people tip a fixed percentage, but that isn’t what matters. What matters is that there are a sufficient number of diners who make their tipping decision *on the margin*. With five years experience as a waiter, I would say that was definitely my experience. The worst servers consistently received lower tips, the better servers consistently received better tips.

Comments are closed.