August 30, 2013

Why Do We Tip?

An economist argues that restaurant patrons and waiters would be better off if restaurants adopted no-tipping policies.
Nachum Sicherman

Last month in New York City, the famous Japanese restaurant Sushi Yasuda sent restaurant patrons and servers into a tizzy when it decided to stop accepting tips.

Economists, like myself, have long wondered why people tip at all when they eat in a place they do not intend to visit again, or where they at least believe they will not be remembered when they return. The simplest answer is that tipping is expected and stiffing a server is generally considered to be the equivalent of stealing, thus it’s not really a voluntary payment.

When viewing this exchange as a pure business transaction, with no social norms attached, the implicit promise to pay a tip in exchange for good service is what game theorists call a non-credible promise. Once the service has been provided and the parties will never meet again, there is no rationale for paying a tip. A creative take on how to convert a non-credible promise into a credible promise is portrayed in the movie classic, The Apprenticeship of Duddy Kravitz. Duddy, played by a young Richard Dreyfuss, works as a waiter at a Jewish resort hotel in the Laurentian Mountains in southern Quebec, Canada. One of the guests calls Duddy to his table, takes out a $100 bill (note that the movie is set in the 1940s), tears it into two pieces, gives one to Duddy, and declares: “We are the Farbers. We are here for the season. I like snappy service. You give it to us, you get the other half.” If you think about it for a moment you realize that Mr. Farber’s promise is totally credible. Indeed, Duddy’s service becomes extremely snappy. For obvious reasons this effective and credible incentive device is not commonly employed.

The only possible justification for the current practice of tipping is that it encourages better service in restaurants. This is the standard rationale for pay for performance. The reality, however, is that tips are not tied to performance and servers are, therefore, not incentivized. In New York City, for example, most diners pay a tip very close to double the tax, and, indeed, at least one online travel guide advises that, “it is customary to tip 20 percent and with a NYC 8.25 percent local tax, a good estimate is to simply double the NYC local tax and round up.” On top of that, in most restaurants, tips are pooled and later divided among various employees — not to mention the practice of including a gratuity charge in the bill for large parties. All of these norms and practices make the link between pay and performance very weak and raise serious doubts around the rationale of maintaining such a meaningless social practice.

What is likely to happen if we adopt the European or Japanese model of no-tipping? For one thing, menu prices will have to increase by about 15 to 20 percent to make up for the server’s forgone tips. There is one caveat to consider — economists Raj Chetty, Adam Looney, and Kory Kroft conducted research that indicated that when commodity taxes are included in the posted prices seen by consumers, it has a greater effect on supressing demand.The implication here is that diners might not fully comprehend the full cost of dining when a component of cost (the tip) is not explicitly listed. In other words, restaurant owners are able to fool diners into believing that the cost of the meal is lower than the true cost.

Abandoning the practice of tipping will not allow them to do this anymore and the menu will, at last, reflect the full cost of the meal. Will the restaurant industry suffer, overall, from the disappearance of the mirage of lower prices? My intuition suggests that, if at all, such an effect will be short lived. Even during a time of elevated unemployment, it is unlikely that an increase in menu prices would scare away most diners. The budgeting process for a meal doesn’t happen at the restaurant — it happens before. Consumers — whether consciously or subconsciously — incorporate the tip into the cost of a meal in advance of dining at a restaurant. Besides, the total cost patrons pay would not change. The only difference would be that the tip is included in the cost of the meal, as opposed to an add-on after the meal.

Furthermore, the restaurant industry is highly competitive, and dining out is a repeated experience. Diners, therefore, will most likely adjust and learn to calculate the full cost of the meal regardless of any artificial division between menu cost and service charges. As for waiters, shifting from tip-based to wage-based pay would offer a more stable compensation structure. For instance, the volatility associated with changing conditions, such as a good shift or bad weather is likely to disappear. Such a change is welcome for two reasons. First, most employees dislike the risk and uncertainty associated with their current pay structure. Second, it can be extremely frustrating when variations in earnings are the result of forces beyond the worker’s control. Indeed, one of the common concerns of tying pay to performance is how much of the variation in output is due to the worker’s behavior and how much to other forces.

No tipping does not necessarily mean uniform payments to all waiters and no incentives for performance. If restaurant owners care about the quality of service, they can easily identify and reward high-quality waiters for good performance. After all, owners and shift managers can easily observe their employees’ behavior directly. Diners, on the other hand, will provide the incentive for a good performance in the same way they do for the quality of the food or other aspects of the dining experience — by voting with their feet and wallets and in online reviews. And if restaurant owners still want to solicit direct feedback from diners, how about simply asking them to rate their waiter on a 1-to-10 scale? (Read about one restaurateur’s experience eliminating tips here.)

Another benefit of abandoning the practice of tipping is greater compliance with the law. Under the current system, waiters might find it easy to avoid paying taxes on cash tips, something a no-tipping regime would make impossible. While the federal minimum wage is currently set at $7.25 per hour, that of tipped employees is only $2.13 per hour. The law, however, explicitly states that “If an employee’s tips combined with the employer’s direct wages of at least $2.13 an hour do not equal the federal minimum hourly wage, the employer must make up the difference.” Given the difficulty in enforcing such a requirement, my guess is that, at least in low-price restaurants, servers could be easily exploited and compensated below the minimum wage. Such exploitation would not be feasible with salary-based compensation, and a no-tipping minimum wage.

Let the mayor of New York City lead the country with a public statement declaring a new voluntary social contract, suggesting a default no-tipping regime in restaurants, and in the case of exceptionally good service, follow the current practice in many European countries: a customary gratuity of $1 to $5.

Nachum Sicherman is professor of finance and economics at Columbia Business School.