mug endowment market experiment

The mug endowment market experiment was a market-based test of whether ownership changes the perceived value of an object. Using an attractive university-insignia coffee mug worth about $6, the experiment showed that people who possessed the mug treated giving it up as a loss, while people considering whether to acquire it did not treat spending money as an equivalent loss. The resulting gap between selling and buying prices challenged the assumption that market exchange reliably moves goods toward those who value them most.

Experimental Design

The experiment adapted a token-market method associated with Vernon Smith, in which participants publicly offered to buy or sell redeemable tokens. Because tokens had value only as exchange instruments, their prices and movement closely followed standard supply-and-demand predictions. In the preliminary rounds, the estimated number of trades was typically close or identical to the amount predicted by economic theory; when half the participants received tokens, roughly half were expected to change hands.

The researchers then replaced the purely exchangeable tokens with a good expected to be held for use: an attractive coffee mug bearing the insignia of the university where the experiment took place. The mug was distributed randomly to half of the participants. Those who received one became Sellers and kept the mug in front of them, while Buyers were invited to inspect a neighboring participant's mug and state the price at which they would acquire one using their own money. This arrangement separated possession from the desire to purchase and allowed the experimenters to compare the two valuations within a functioning market.

Results and Ownership Premium

The contrast with the token rounds was dramatic. The average selling price for the mug was about twice the average buying price, and the estimated number of trades was less than half the number predicted by standard theory. The market therefore failed to produce the orderly redistribution expected for an exchange good. Owners who valued the mug less than potential buyers did not reliably sell it, because possession itself raised the psychological cost of surrendering it.

A later version added a third group, the Choosers. Unlike Buyers, who had to spend their own money to obtain the mug, Choosers could receive either a mug or a sum of money and had to indicate the amount of cash they regarded as equally desirable. Sellers and Choosers thus faced the same final alternatives: go home with the mug or go home with money. Yet Sellers placed substantially more value on the mug than Choosers did, while Buyers and Choosers gave it similar cash values. The difference could not be explained by the long-term consequences of the options; it appeared in the emotion attached to already possessing the object.

The measured values made the asymmetry concrete. Sellers valued the mug at $7.12, whereas Buyers valued it at $2.87; Choosers gave it an intermediate value of $3.12. The Seller valuation was a little more than twice the value assigned by the other groups, closely resembling the approximate 2:1 loss-aversion ratio found in risky-choice research. The experiment thus linked a market anomaly to a broader pattern in which losses relative to a reference point carry more psychological weight than comparable gains.

Use, Exchange, and the Endowment Effect

The experiment's central distinction was between goods held for exchange and goods held for use. A merchant who gives up shoes for money ordinarily does not experience the transaction as a loss, just as a person exchanging a $5 bill for five singles does not feel deprived. The money and the shoes function as proxies for other purchases. By contrast, wine, Super Bowl tickets, leisure time, and the standard of living supported by income are normally possessed for enjoyment rather than resale. Giving them up is experienced as surrendering something already incorporated into one's life.

The mug occupied the second category. Once placed before its owner, it became part of the Seller's reference state, so an offer to buy it was mentally represented not simply as a gain of money but as a loss of the mug. Buyers did not ordinarily experience the money spent on a reasonably priced mug as a loss, because the money had been held for exchange and the mug was the intended use good. This difference in mental representation created a wedge between the price at which an owner would sell and the price at which a nonowner would buy.

The findings also accorded with evidence about emotional and neurological reactions to buying and selling. Selling goods normally used activates brain regions associated with disgust and pain. Buying can activate similar regions when the price seems excessive, as though the seller is taking more money than the object is worth as an exchange; buying at an especially low price, by contrast, can be pleasurable. These reactions help explain why the market's apparent failure was not random inconsistency but a systematic consequence of how ownership and reference points shape evaluation.

Theoretical Consequences

The mug experiment gave empirical substance to prospect theory's claim that outcomes are evaluated relative to a reference point rather than solely through total wealth. In the older Bernoullian representation, a small purchase or sale could be described as a movement between states of wealth. Prospect theory instead treats gains and losses relative to a psychologically meaningful earlier state as the effective carriers of value. Possession of the mug supplied that earlier state, making relinquishment salient even when the monetary exchange would be advantageous.

Three features of automatic evaluation are especially relevant. First, the status quo or another expected state serves as a reference point, dividing outcomes into gains and losses. Second, diminishing sensitivity means that subjective changes depend on the starting level rather than merely on their absolute size. Third, loss aversion makes the negative response to giving up the mug stronger than the positive response to receiving an equivalent amount of money. The mug's ownership premium was therefore not a failure to understand prices; it was the predictable result of a value function organized around possession and loss.

The result also exposed a conflict between descriptive psychology and normative economic principles. Standard decision theory relies on dominance and invariance: a preferable option should remain preferable when equivalent outcomes are described differently. Yet human evaluations can shift with reference points and framing, and the market experiment showed that simply assigning an object to a participant changed the willingness to exchange it. The broader lesson is that market efficiency depends not only on the formal exchangeability of goods but also on whether people experience those goods as possessions to use or as instruments to trade.

Significance

The mug endowment market experiment became a compact demonstration of how ordinary ownership can distort the otherwise elegant logic of supply and demand. Tokens behaved as theory predicted because they had no use value beyond redemption for cash. Mugs behaved differently because possession converted them into objects of use and made selling emotionally costly. The experiment's importance lies in that contrast: the same market mechanism worked smoothly for exchange instruments and faltered for an object people had begun to regard as theirs.

By producing repeated results across variants of the procedure, the study transformed the endowment effect from an anecdotal impression into a robust behavioral phenomenon. It showed why a person may reject a trade that appears profitable from an external perspective, why buyers and sellers can assign sharply different prices to the same object, and why the psychological meaning of a transaction can matter as much as its financial outcome. In this sense, the mug was not merely a commodity in an experiment; it was a visible marker of the reference point from which human value is constructed.