Mental Accounting: Why a Lost Ticket Hurts More Than a Lost Bill

In economic theory money is fungible: a dollar is a dollar wherever it came from and whatever it was meant for. In practice people keep books. Richard Thaler, who gave the behavior its name in a 1985 paper in Marketing Science, described a set of mental accounts into which income is posted and against which spending is charged, with rules about which account pays for what. The accounts make budgeting possible and also produce choices that make no sense once the labels are removed.

The lost ticket problem shows the mechanism in one move. A lost 10-dollar bill is charged to general funds and leaves the entertainment account untouched, so the play still costs 10 dollars and is still worth it. A lost ticket was already charged to the entertainment account, so a replacement makes the play cost 20 dollars against that account, which is more than most people are willing to post there. The wealth lost is identical. What differs is the ledger entry, and the ledger decides.

Thaler's 1985 paper added two further pieces. The first is hedonic editing: because of the shape of the value function described by Kahneman and Tversky's prospect theory, people are happier when gains are separated into several small ones and losses are combined into one, and they arrange their accounts accordingly, which is why a holiday package feels better than the same costs paid item by item. The second is transaction utility, the pleasure or pain of the deal itself as distinct from the thing bought. In Thaler's beer-on-the-beach survey, respondents were willing to pay a median of 2.65 dollars for a beer brought from a fancy resort hotel and 1.50 for the identical beer from a run-down grocery store.

Accounts also differ in how easily they are spent. In the behavioral life-cycle model that Hersh Shefrin and Thaler set out in 1988, current income is spent readily, current assets less readily, and future income hardly at all, which is why a bonus is spent and an equal rise in home value is not. Chip Heath and Jack Soll showed in 1996 that people set budgets for categories such as entertainment and then underconsume in a category once an expense has been charged to it, even when the expense turned out to be worth less than expected.

The lost ticket problem appears in Amos Tversky and Daniel Kahneman's 1981 paper in Science, "The framing of decisions and the psychology of choice," where it is used to introduce the idea that people organize decisions into what the authors called psychological accounts. Two groups of university students read one version each. The first, 200 people, read that they had paid 10 dollars for a ticket to a play and lost it at the door, the seat unmarked and the ticket unrecoverable. The second, 183 people, read that they had lost a 10-dollar bill on the way to buy a 10-dollar ticket. Each group was asked whether they would pay 10 dollars for a ticket.

The answers diverged. In the lost ticket group, 46 percent said yes and 54 percent said no. In the lost bill group, 88 percent said yes and 12 percent said no. Tversky and Kahneman observed that the two situations are identical in their effect on wealth, since in both the person is 10 dollars poorer and faces a 10-dollar ticket, and proposed that the difference arises from the topical organization of accounts: the ticket is posted to the account for the play, and a second ticket makes the play cost 20 dollars, whereas the lost bill is posted elsewhere and leaves the play's price alone.

Thaler's 1985 paper took the idea further and gave it its name. He set out the framework of accounts, hedonic editing and transaction utility described above, with the beer-on-the-beach survey as its most quoted example. His 1999 review, "Mental accounting matters," gathered the later evidence, including the finding that gamblers treat winnings as house money and risk them more freely than their own.

The demo on this page reproduces the 1981 wording and price. The 10 dollars has not been adjusted for inflation, and the intro says so; both versions share the figure.

Any dollar that has been labeled is a dollar that can be treated differently from its neighbors. The rows below are illustrations, not findings from the studies cited on this page.

The accounts are useful for budgeting and harmful for one-off decisions, so the counters keep the budgets while removing the labels at the moment of choice.

• Restate the decision in total wealth. The lost ticket question becomes: I am 10 dollars poorer than I expected to be; is the play worth 10 dollars? Asked that way, the two versions give the same answer.
• Treat windfalls as income. Post bonuses, refunds and gifts to the same account as salary before deciding anything, so that the label does not decide for you.
• Net your accounts. Savings held next to expensive debt, or a fuel saving left in the fuel budget, are account boundaries doing damage; look at the whole balance sheet once a month.
• Evaluate savings in dollars, not percentages. A 5-dollar saving is worth the same trip whether the purchase is 15 dollars or 500.
• Use the bias on purpose. Separate accounts for rent, savings and spending work because money posted to savings feels unavailable; the trick is to set the boundaries in advance and keep them away from one-off choices.

Mental accounting is one of a family of findings from the same research program, and its neighbors are easy to confuse with it.

• It is not the sunk cost fallacy, which is continuing with something because of what has already been spent on it. The two meet in the lost ticket problem, where the first 10 dollars is a sunk cost, but mental accounting is the broader bookkeeping that makes the sunk cost feel attached to one purchase and not to another.
• It is not the framing effect, in which the same outcome is chosen differently when described as a gain or a loss. The lost ticket problem appears in the same 1981 paper, but the two versions differ in what was lost, not in how a single loss is described.
• It is not the denomination effect, the tendency to spend small bills more readily than one large bill of the same total. That is about the physical form of money; mental accounting is about the budget it has been assigned to.