Plate vi in print

The Endowment

Thaler; the endowment effect and mental accounting, a thing worth more once owned, money treated as non-fungible by the account it falls in.

The thing can stay the same while the place from which we value it moves.

— the argument of this plate, after Thaler, Kahneman and Knetsch

The thing you own can become harder to surrender than the same thing was to acquire, and the distance between those two prices is the subject of this plate. The earlier Collected Possession put the matter in the image of a mug: the mug had not changed; ownership had. A buyer asks how much having it would add to life. A seller asks what would make losing it bearable. If possession enters the reference point, an exchange crosses a boundary that acquisition did not cross, and the price demanded for crossing back can rise. The question then reaches beyond objects to money itself, which the mind sorts into earnings, savings and winnings, attaching a different permission to spend to each. Ownership changes where the reckoning begins; mental accounting changes which things are reckoned together. Both put pressure on the idea that a chooser brings one settled scale of value to every transaction.

The equality of buying and selling prices belongs to a particular benchmark, whose assumptions must survive the comparison. Consider an invented chooser with £50 and a mug worth exactly £5 of additional satisfaction, whatever cash remains. With unrestricted exchange, no costs and no uncertainty, paying £5 for the mug leaves that chooser indifferent: £45 and the mug are worth as much as £50 alone. Give the chooser the mug, with the same £50 still available, and £5 is also the least acceptable payment for surrendering it. The extra satisfaction has a constant money equivalent, so possession does not move the reservation price. Alter those premises and a gap may have other causes. A poorer buyer may be unable to pay what a wealthier owner would require; a difficult replacement may make surrender costly; learning how to use an object can increase its value. Establishing an endowment effect requires more than finding two different numbers in two different circumstances.1

1. The £50/£5 example is invented and uses a constant money value for the mug. Plott and Zeiler (2005), p. 530 n. 1, state the smooth-preference and wealth-effect qualifications to the usual comparison.

The laboratory gave ownership a cleaner test by assigning the same goods to people who had not chosen to own them. In Kahneman, Knetsch and Thaler’s 1990 experiment with forty-four students, twenty-two received mugs; an ownership-independent benchmark predicted eleven exchanges. The four mug markets produced four, one, two and two trades. In the final market the median seller required $5.25 while the median buyer offered $2.25. Earlier rounds with tokens redeemable for assigned cash amounts had traded close to their predicted volume. In a separate experiment, people choosing freely between a mug and money valued the mug much like buyers, while owners demanded more to surrender it. The authors interpreted the difference chiefly as reluctance to give up an entitlement. Possession appeared to help form the valuation that trade was supposed merely to discover.2

2. Kahneman, Knetsch and Thaler (1990), § II, pp. 1329–36, especially table 2; § IV, p. 1339. These are group medians and experimental trades, not a multiplier for every owner.
⁂

A gap can be replicated while its explanation remains disputed. Plott and Zeiler reproduced a buying–selling disparity using an earlier protocol, then found no statistically significant excess of willingness to accept over willingness to pay under procedures combining explanation, anonymity and, in two sessions, paid practice. Their results made the elicitation procedure part of the question: does a participant understand that stating a larger selling price can forfeit a welcome exchange, or answer as a bargainer trained to ask high? The authors also acknowledged that their procedures did not directly establish the removal of every misconception, and could not exclude every alternative interpretation. The disagreement leaves an exacting lesson. A measured gap is evidence about decisions made under specified conditions; attributing it to loss aversion requires an argument about what those conditions have isolated. Naming the effect cannot perform that work.3

3. Plott and Zeiler (2005), §§ II–IV, pp. 535–44. Their finding does not establish exact equality or the universal absence of an ownership effect.

Money acquires a related power over choice when its label becomes a boundary on its use. Thaler’s mental accounting names the classification of funds and expenditures, and the rules for keeping those accounts. The earlier Possession pictured grocery money, a holiday fund and winnings guarded with different zeal. Make that picture an invented comparison: two unrestricted envelopes contain £40 each, and a £30 outing would leave £50 in total. Writing holiday on one envelope changes neither the price nor the remaining purchasing power. If the outing becomes acceptable only when charged to that envelope, the label has changed the decision. A genuinely restricted grant would be different, as would money held for somebody else; there the boundary changes the available uses or obligations. Mental non-fungibility begins where the financial alternatives remain the same and the chooser’s partition gives them different treatment.4

4. Thaler, Mental Accounting Matters (1999), introduction, pp. 184–85, and “Budgeting,” pp. 193–97. The envelopes and amounts are invented.

A boundary that changes spending can also serve the person who erects it. The earlier plate called these compartments the planner’s defence against the doer: savings left deliberately inconvenient to raid may survive a succession of tempting afternoons. Thaler treats budgeting as both a way of making trade-offs and a device for self-control. That possibility changes the welfare question. The useful comparison is between what the person can accomplish with the rule and what that same person would accomplish without it, including the effort of supervising every purchase. A household may reasonably accept some missed pleasures in exchange for keeping rent money intact. Yet a rule can outlive its purpose, blocking an urgent use while protecting an obsolete category. A label supplies discipline only through the choices it changes; whether those changes help must be judged against the person’s ends and the alternatives actually available.5

5. Thaler (1999), “Consumption categories” and “Self-control and gift giving,” pp. 193–95; discussion, pp. 202–3, weighs the costs of correcting one accounting habit against the problems correction can create.

The price you would sell for can therefore exceed the price you would pay because the two questions begin from different possessions, resources or reference points, and finding which difference matters is part of understanding the answer. When ownership itself changes a valuation, the initial assignment of a good can help determine who will keep it; an account of exchange that holds preferences fixed has then omitted something consequential. When a money label changes spending, a single budget has been divided into a private system of permissions. Neither observation alone selects the price or spending rule that ought to prevail. An owner’s attachment may deserve weight, a misunderstood selling task may need correction, a savings boundary may help sustain a chosen future. The difficulty is to establish what the departure does before assigning it a verdict. And the account has one more hold over its keeper: even after the money is gone, closing it at a loss can feel like the act that caused the loss. That is the next plate’s inheritance.