In the last post, I promised answers to some “hypothetical choice questions” that informed and were informed by Prospect Theory. That theory suggests most people choose greater risk for biggere losses rather than certain smaller losses — which shows up as convexity in the lower quadrant of the S curve on the chart below (from ScienceDirect). Why? The theory says that people have diminishing sensitivity to loss.
Richard Thaler provides a nice summary of the original work by Kahneman and Tversky in chapter 4 of Misbehaving (his excellent book). “I found the idea that you could just ask people questions and take their answers seriously to be quite liberating,” he writes (p. 39). He explains that 72% of respondents chose the option of a sure gain over an unsure larger gain when the expected payoff was identical. In contrast, only 36% chose a sure loss over an uncertain 50-50 chance at a double loss.
Well, I’m afraid that conclusion only holds for certain gambles. It’s about framing, as the behavioralists would say. I’m not sure what the factors are, but my suspicion is that hypothetical losses with cash are different in important ways from losses with certain goods. A first dollar and a tenth dollar are both a dollar. But missing one finger and missing ten fingers feels different. Excuse the pun, I just couldn’t get a grip on a better line.
Anyway, I asked my readers to answer six survey questions to assess if they made the *correct* choices as predicted by prospect theory. In each of the questions, the theory suggests people will choose the riskier options, which are:
1-in-2 odds of losing $200
1-in-6 odds no eating for 6 days
1-in-5 odds of losing 5 fingers
1-in-2 odds losing hearing in both ears
1-in-7 odds for 7 straight days without electicity
1-in-3 odds of 3 favorite restaurants closing permanently
Aside from the first cash question, I predict more people will choose the certain, smaller loss because it involves something more tangible. In other words, I think losses of goods are increasingly painful.
I will report the results in a new post at the end of the week, but I think you can see the results as they roll in on the original post as well.