An analysis of the determinants of tipping behavior: a laboratory experiment and evidence from restaurant tipping.

Summary

---

Statistical table - Statistical data

See the full content of this document

Extract

---

An analysis of the determinants of tipping behavior: a laboratory experiment and evidence from restaurant tipping.

1. Introduction

Restaurant tipping is a significant part of the economy. In 2003, sales at full-service restaurants totaled approximately $151 billion (U.S. Census Bureau 2005). Assuming a tipping norm of 15%, waiters and waitresses in the country earned roughly $22.7 billion in tip income in 2003. Restaurant tipping is also puzzling, at least from the point of view of standard neoclassical economic theory. Why do consumers voluntarily give money to their server after the service has been rendered? Interestingly, consumers still leave their server a tip even when they plan never to visit the restaurant again (Kahneman, Knetsch, and Thaler 1986).

In this paper, I explore several determinants of tipping behavior using survey and laboratory experimental data. First, I examine social norms related to reciprocity and letdown (guilt) aversion. Reciprocity refers to the idea that people reward kind actions and punish unkind actions, and implies a positive relationship between the size of the tip and service quality. According to Fehr and Gachter (2000), there is considerable evidence that suggests a strong role for reciprocity in motivating human behavior. (1) The theory of letdown (guilt) aversion asserts that decision-makers avoid letting others down. Charness and Dufwenberg (2002) show that letdown aversion implies that a consumer's tip depends positively on what the consumer believes the server thinks the consumer will tip.

Although the relationship between service quality and tip size has been studied by several authors (Lynn and McCall 2000a provide a survey and meta-analysis), it is typically explained using either a buyer monitoring story (Bodvarsson and Gibson 1992; Lynn and Graves 1996) or equity theory (Lynn and Grassman 1990; Lynn and Graves 1996). (2) Both Conlin, Lynn, and O'Donoghue (2003) and Azar (2004) mention norms in their recent work on tipping. However, the work by Conlin et al. makes assumptions regarding, rather than trying to determine as I do here, norms that operate in a tipping environment. Azar, on the other hand, is interested in how norms evolve, a topic that I do not address in this paper.

I also examine three aspects of the tipping situation that influence how much consumers tip in restaurants: table size, the sex of the customer, and method of bill payment (cash or credit card). Looking first at table size, at least two factors are at issue, and they work in opposite directions. First, tippers may tip a higher amount in the presence of others at the table in order to assert social status. Status considerations play a nontrivial role in real-world interactions (Ball et al. 2001), and thus might induce consumers to tip more as a form of status acquisition or display. Alternatively, tippers may "free ride" on the tips of others at the table, and thus tip less. However, the common restaurant practice of adding an automatic service charge onto bills at large table sizes (usually six or more customers) suggests that incentives to free ride may dominate. Previous research on this topic is mixed, with some studies finding a positive relationship between tip size and table size (e.g., Lynn and Grassman 1990; Boyes, Mounts, and Sowell 1998; Conlin, Lynn, and O'Donoghue 2003), some finding a negative relationship (e.g., Lynn and Latane 1984; Bodvarsson and Gibson 1997), and others finding no relationship (e.g., Lynn and Latane 1984; Lynn and Grassman 1990; Lynn and Graves 1996; Rind and Strohmetz 1999, 2001a, b).

There is considerable evidence that men and women differ in the economic decisions that they make (Eckel and Grossman 2005). Women tend to be more generous than men in experimental settings such as the Dictator Game, where one player has the opportunity to share an endowment with another person or organization. This paper addresses whether this result carries over into the decision to tip. Because men on average earn about 35% more than women (Council of Economic Advisors 2002), higher income levels could result in higher tips in restaurants by men. However, in both the laboratory experiments and with the survey data, income effects can be separated from sex-related propensities to give. Previous research on sex differences in tipping suggests that men tip more than women (e.g., Lynn and McCall 2000b).

Finally, I look at ...

See the full content of this document

---