Economic and Game Theory
Forum

20 5 10 20 50 100 all
Ask a question about: Other Topics
Respond to the question: uncertainty about consequences?

02/18/2008 03:03 AM by Nick van der Beek; uncertainty about consequences

Dear Sir,

Although there is much literature on uncertainty in terms of probability, I wondered whether there it is possible in the current economic models to apply uncertainty about the type of consequences.
If there is a actor facing uncertainty / risk, this is modeled by giving the agent knowledge on the possible consequences (say rain / sunshine) and limited knowledge on what will happen (40% rain, 60% sunshine) and of course some valuations of the preferences such that the expected utility rule can be utilized (v(r)=10, v(s)=5, EU= 7). An extension similar to what I'm looking for is giving the agent access to some costly but informative signal which he might purchase or not in order to make a better investment decision (should I buy a raincoat?).

But what I'm looking for is the situation where the agent can refine the information about the possible states of the world without knowing all possible alternatives at the outset. E.g. {rain,sunshine} becomes {a lot of rain, a bit of rain, sunshine}. I'm not familiar with any model which inquires into this situation. Even more pessimistic, I understand from Tirole, Incomplete contracts, 1999 that it is impossible to model a incomplete contracting situation without knowledge on all possible states of the world.

The interest I have in this topic comes from work on my Ph.D. on long term contracts: Wouldn't planning be advantaged vis a vis ex-post solutions if we allow for this kind of information? Isn't it too unrealistic? Isn't it so that Columbus never thought America existed? Wasn't it just an accident, while looking for a passage to India? Then isn't there a realistic way for modeling Columbus' choice in economics?

Thanks in advance for you answer. [Manage messages]