Endowment Effect

Evidence and Implications

The endowment effect has been widely replicated with many different objects in many different cultures and has even been demonstrated in nonhuman primates. In one study, students who were randomly given a coffee mug emblazoned with their university logo would not sell their mug for less than $5.25, whereas students randomly assigned to the role of buyer would not pay more than $2.50 for the mug, a price discrepancy that emerged even though both buyers and sellers had previously learned how to make efficient transactions for tokens in an experimental market. In another study, some people were given coffee mugs and asked whether they would sell the mug for various amounts of money; others were simply asked to choose between receiving the mug and receiving various amounts of money. The sellers assigned a median value of $7.12 to the mugs, whereas choosers assigned a median value of only $3.12.

The endowment effect has important consequences for everyday consumer behavior. For example, the endowment effect produces undertrading in markets even with zero transaction costs. To illustrate the concept of undertrading, consider an experimental market in which a random half of the people are given chocolate bars and asked if they would trade their chocolate for a coffee mug; the other half are given coffee mugs and asked if they would trade their mug for chocolate. According to standard economic assumptions, 50% of the people should trade, given that, on average, half of the people would have to trade to match the overall distribution of preference for mugs and chocolate. However, the endowment effect reduces trading rates to approximately 25%, presumably because the objects are more valuable to people who happen to own those objects than to those who do not.

The endowment effect also has important theoretical implications for economics and behavioral decision theory. In particular, the endowment effect violates assumptions of standard economic theory that people act on well-defined preferences that are consistent across measurement context. Such assumptions imply that people who happen to be endowed with an object should demand no more to sell their object than buyers or choosers would be willing to pay to acquire the same object and that an object's value (i.e., how much money an object is worth) should be independent of the way value is measured (i.e., whether one is asked to state a selling price, buying price, or choose between an object and money). In contrast with such assumptions, the endowment effect implies that people's preferences are constructed in the context of measurement and are contextually contingent.

...