Ambiguity: Theory and Experiments

Posted On September 20, 2012

September 20, 2012 – September 21, 2012

Georgia State University

Download Program

The theoretical and empirical characterization of ambiguity has advanced in recent decades, aided by rich specifications of models of ambiguity aversion and tighter experimental evidence. This workshop brings together theorists and experimenters to tighten the connection between the two. We are not looking for narrow specialist papers so much as continuous discussion around several broad topics and papers. To that end, we will only have three paper presentations, and instead have four panels where we “charge” theorists and experimenters to talk to each other.

How should theorists explicitly revise their approaches to any well‐defined paradoxes or implausible parameter estimates uncovered by experimenters? How should new experiments be

conducted to explicitly test remaining foundational axioms of the respective models, and estimate remaining parameters? How should experiments designed to measure preferences and perceptions be evaluated by theorists in terms of construct validity? What is the role of laboratory versus field experiments, if one accepts the tradeoff between internal validity and external validity that comes from conducting experiments in the field? How can theorists do their part in designing, and experimenters their part in designing and conducting, out‐of‐laboratory studies of data likely to reveal perceptions of and reactions to ambiguity? The type of settings of interest include insurance against frequent events such as car crashes versus insurance against infrequent events such as earthquakes, where the frequency presumably correlates with actuarial ambiguity. Or the evaluation of financial risk premia for

new products versus expansions of standard products. Although there already exists isolated work along these lines, the theme of the conference would be to make the bridge between theory and experiments as well‐defined as possible.

The focus of the workshop will be on positive models of behavior towards ambiguity, rather than normative issues and the extent to which models of ambiguity embody anyone’s definition of “rational behavior.”