Anyone ever heard of this? “The Ellsberg paradox is a paradox in decision theory and experimental economics in which people’s choices violate the expected utility hypothesis.” http://en.wikipedia.org/wiki/Ellsberg_paradox
This and most other similar experiments are only odd if you assume that people have perfect decision-making capability and always engage in the behavior that actually maximizes their utility ex post. This sort of thinking leads to the same apparently abberant behavior as assuming efficient markets and then recoiling in surprise when the real world does not behave according to efficient market models.
People and markets do not behave in mathematically efficient manners because people (and, therefore, markets) never have all of the assumed perfect information. Which is why the more Austrian insights of a priori and ex post differentiation is important.
I guess it’s a sort of “researcher’s strawman” where people aren’t bad at math and mathematically model probabilities of success in their heads, or, as in the case of the 1 urn experiment, are able to accurately compare the relative probabilities of of two different gambles and so know that if they choose A the first time they should choose C the second time or they aren’t making sense.
It’s only a paradox when you consider the means (expected values) of the aggregate probability distributions and neglect their deviations (widths). In both cases, it’s perfectly humanly rational to go for the narrower distributions around the expected means (i.e. narrower ranges of possible outcomes; smaller repercussions if your assumptions about those means are wrong).
If one did 1000 draws each in cases A, B, C, and D whereby the black-yellow portions were randomly changed anywhere between 0-60 and 60-0 after every draw, the income stream (equity curve) of case A would slope upwards at the same rate ($33/draw) as case B but with much less volatility around that slope. Also, the income stream (equity curve) of case D would slope upwards at the same rate ($66/draw) as case C but with much less volatility around that slope. Perfectly rational to prefer A over B and D over C.
As usual, academics teach, practitioners do. There’s no paradox here.
Z.
Expected utility? According to whom? I have a family member who complains of being broke. This same person has not looked for a job in 2 years. Expected utility would show that a job could cure financial destitution. Sadly, this person finds not working a more valuable utility than working for a living. The end result is that most people find it easier to wait for the easiest utility rather than the most benificial.