Are violations of expected utility theory just memory problems?

Or should I refer to availability heuristics?  Here is Daniel Gilbert’s MP3 on the topic.  To put it simply, lottery losers, despite their great numbers, rarely receive as much media publicity as the few winners.  Here is my previous post on Gilbert, or just check the right hand side of this blog for the Blogads entry which quotes me… 

Thanks to Boing Boing for the pointer.

Addendum: Here is Gilbert’s short essay, from his blog, on whether fatherhood makes men happy.  Happy Father’s Day to our eligible readers…


In his Nature of Rationality, Nozick supplements the traditional expected utility theory with what he calls "symbolic value" to create a rough Decision Value Formula. For any act, A

Decision Value (A)=W(EU) * EU(A) + W(SU) * SU(A)

Where W refers to the weights given to both the expected utility (EU) and the symbolic utility, (SU). Listening to Gilbert's lecture, I wonder how much something like that is in play with some of his examples. For instance, his example concerning eating a potato chip either in the presence of a piece of Godiva chocolate or a can of Spam. The symbolic cultural overtones are obvious here, signalling status and fashion. Likewise with the gambling: the daring represented by laying it all the line to overcome great odds...what kind of life is that symbolic of?

If you want some interesting insights about expected utility, it is
necessary to dig deeper than Gilbert. For starters try Harry Markowitz's
extension of the Friedman-Savage work. Want more? Check out the
extension of Markowitz's model by Coelho and McClure.

Comments for this post are closed