On Friday I had the opportunity to hear author Dan Ariely speak. Dan is an MIT professor on tour to promote his new book, Predictably Irrational. He made a strong case that people are, as his book's title implies, not just irrational but irrational in predictable ways. Some interesting tidbits:
- The herd instinct extends even to one's current self being inclined to copy actions one previously made (in this case, you can think of all of the prior "yous" as the herd). People often remember what they did before in similar situations without necessarily remembering the rationale or the outcome, and this can contribute to behavioral inertia.
- People are more likely to cheat when the benefit is abstracted, even if only superficially. The example Dan gave was an experiment in which people had the ability to cheat and claim they deserved more money than they actually did. When the payoff was in cash, people only cheated a little, but when the payoff was in tokens which could be immediately redeemed for cash, people cheated a lot more. (As a side note, I also took this as evidence that the pervasiveness of cash has made people forget that it is just as abstracted as tokens are.)
- In another experiment, people were given arbitrary numbers from 1 to 100 and each was first asked if he/she would want to buy a given item at that number of dollars, and then later was given the opportunity to bid whatever he/she wanted to on the same item. There was about a 0.5 correlation between the initial arbitrary amount and the bid amount, whereas for rational agents there should've been no correlation. Provided that the methodology was sound, such a high correlation implies that people's perceptions of value are extremely easy to manipulate.
- In another experiment, Dan let students specify their own deadlines for a series of projects during the semester, but with substantial penalties for missing their self-imposed deadlines. A strictly rational student seemingly should choose the latest date allowed for all projects, but students who spread their deadlines evenly throughout the semester turned performed better. (There are obviously potential confounding factors here, which I'm assuming Dan dealt with.) This is an interesting example of a person wanting to do what's best for his/her future selves while not trusting those future selves to behave rationally and therefore imposing restrictions on his/her own future behavior.
If you like this kind of thing, you might want to check out Dan's book. You might also like overcomingbias.com, a great site run by ultra-smart guys Robin Hanson and Eliezer Yudkowsky which covers similar ideas.