The Truth About Utility Functions

John Quiggen explains.

Comments

Quiggin.

Posted by joejoejoe | October 10, 2007 4:22 PM

Qui'gon.

Posted by Julian Elson | October 10, 2007 4:58 PM

Qui'gon John, a revered yet maverick and unconventional Econ Master.

Posted by joejoejoe | October 10, 2007 5:17 PM

Who the hell is stupid enough to think that economics is vacuous?

Posted by sherifffruitfly | October 10, 2007 6:02 PM

Utility is just a useful mathematical construct. Indirect utility (the max utility you can possibly get with your resources) is the "potential function" you get by integrating demand functions (which are things you really observe). Utility is relative, sort of like "ground" is relative when you talk about electric potential. The math is very similar.

As for tipping, the mystery is why certain services have tipping and others don't. And why it varies from country to country. Of course, this is a rather uninteresting mystery, at least in my opinion.

Posted by Mr. Noah | October 10, 2007 8:54 PM

I have small disagreement with Quiggin with his argument that utility maximazation is a tautology. This is only true for revealed preferences: "what people choose reveal their preferences and preferences is defined as that which people will choose".

However normal utility maximazation without arguing from revealed preferences is not a tautology. I basically says: People maximize something (lets call that something utility) all the time. This statement might not be true, at least its not definitionally true. What it is, though, is no-falsifiable. No set of observations can disprove it because as long as the utility function is unspecified any set of observations will be able to be fit into one.

So its potentially (and properly) false, but it can never be proven.

So all economics have to fear is the Popper-Zombie.

Posted by Tomas | October 11, 2007 6:50 AM

normal utility maximazation without arguing from revealed preferences is not a tautology. I basically says: People maximize something (lets call that something utility) all the time. This statement might not be true, at least its not definitionally true.

This is simply definitionally true. And it has been known long before modern calculus or economics. Aristotle demonstrated the existence of a unitary personal utility ordering in the Nichomachean Ethics twenty-four centuries ago.

Posted by Adult | October 11, 2007 2:27 PM