Ellsberg's Paradox and the Value of Chances (Keynote)
speaker: Richard Bradley (London School of Economics)abstract: The Ellsberg paradox has gured prominently in recent debate in de- cision theory, serving to motivate several claims about the limitations or inadequacies of Bayesian approaches to rational decision making. In particular it has been claimed that the choices typically observed in Ells- berg.s experiments reveal not only that agents are not expected utility maximisers, but that they are not probabilistically sophisticated choosers either. These claims in turn have motivated a number of di erent pro- posals for alternative decision theories that explain how agents choose in situations of ambiguity. In this talk I will argue that the typical pattern of preferences revealed in Ellsberg-type experiments do not have these implications and that, to the contrary, they are both perfectly compati- ble with probabilistic sophistication and can be rendered consistent with subjective expected utility maximisation. The centrepiece of the argu- ment is the contention that the hypothesis that agents are risk averse with respect to chances of, or opportunities for, monetary gains gives a good prediction of these behaviours.
timetable:
Tue 9 Jul, 9:00 - 10:00, Sala Azzurra
documents:
<< Go back