Kluge: The Haphazard Construction of the Human Mind (Houghton Mifflin; 2008)

And so even today, there's some remarkable crosstalk between the two. People are less likely to donate money to charities, for example, if they are hungry than if they are full; meanwhile, experimental subjects (excluding those who were dieting) who are put in a state of "high desire for money" eat more M&Ms during a taste test than do people who are in a state of "low desire for money."* To the degree that our understanding of money is kluged onto our understanding of food, the fact that we think about money in relative terms may be little more than one more accident of our cognitive history.

"Christmas Clubs," accounts into which people put away small amounts of money all year, with the goal of having enough money for Christmas shopping at the end of the year, provide another case in point. Although the goal is admirable, the behavior is (at least from the perspective of classical economics) irrational: Christmas Club accounts generally have low balances, so they tend to earn less interest than if the money were pooled with a person's other funds. And in any event, that money, sitting idle, might be better spent paying down high-interest credit card debt. Yet people do this sort of thing all the time, establishing real or imaginary accounts for different purposes, as if the money weren't all theirs.

Christmas Clubs and the like persist not because they are fiscally rational but because they are an accommodation to the idiosyncratic structure of our evolved brain: they provide a way of coping with the weakness of the will. If our self-control were better, we wouldn't need such accommodations. We would save money all year long in a unified account that receives the maximum rate of return, and draw on it as needed; only because the temptation of the present so

*How do you get people to lust after money? Let them imagine they'll win a significant amount of money in the lottery, then ask them to think about how they would spend it. The more money they are asked to envision, the more lust is engendered. In the particular experiment I describe, people in the "high desire for money" condition spent a few minutes thinking about how they'd spend a £25,000 prize, while people in the "low desire for money" condition spent a few minutes contemplating how they'd spend a £25 prize. A remarkable indicator of the influence of induced money lust came from a follow-up question: people were asked to estimate the size of coins. The bigger the lust, the larger their estimate.

often outweighs the abstract reality of the future do we fail to do so such a simple, fiscally sound thing. (The temptation of the present also tends to leave our future selves high and dry. According one estimate, nearly two thirds of all Americans save too little for retirement.)

Rationality also takes a hit when we think about so-called sunk costs. Suppose, for instance, that you decide to see a play and plop down $20 for a ticket — only to find, as you enter the theater, that you've lost the ticket. Suppose, further, that you were to be seated in general admission (that is, you have no specific assigned seat), and there's no way to get the ticket back. Would you buy another ticket? Laboratory data show that half the people say yes, while the other half give up and go home, a 50-50 split; fair enough. But compare that scenario with one that is only slightly different. Say you've lost cash rather than a prepurchased ticket. ("Imagine that you have decided to see a play, and the admission is $20 per ticket. As you enter the theater, ready to purchase one, you discover that you have lost a $20 bill. Would you still pay $20 for a ticket for the play?") In this case, a whopping 88 percent of those tested say yes — even though the extra out-of-pocket expense, $20, is identical in the two scenarios.