The “Closing In” Method: An Experimental Tool to Investigate Individual Choice Patterns under Risk

Abstract

In view of the considerable evidence of systematic violations of expected utility by individual subjects, a number of alternative models generalizing expected utility have been developed. As P.Fishburn put it once [1988], we have entered “a new era” in the domain of decision under risk, but “we shall have to wait and see” during the “time of shakedown and sifting” ahead of us, i.e. until one or possibly several of the models put forward until now can attract a clear consensus. Most of these models weaken the independence axiom. But some of them retain some linearity properties, like Chew’s weighted utility theory [1983] or Fishburn’s skew-symmetric bilinear utility theory [1988]. Machina provided a quite general frame of reference, by dispensing altogether with the independence axiom and allowing one to envision different alternative hypotheses — among which his Hypothesis II [1982] is only one possibility. Other models emphasize the idea that decumulative probability distributions straightforwardly undergo a (necessarily nonlinear) cognitive transformation, like Quiggin [1982], Yaari [1987], Allais [1988], Segal [1989], Wakker [1993].

The authors want to express their gratitude to D. Bouyssou, Ph. Delquié, J-Y. Jaffray, M. Machina, whose comments were very stimulating, as well as to several colleagues who attended the FUR VI conference. The usual caveat holds.