The Economics of Behaviour and Decision-Making Seminar invites you to:
"Making the Rich Domain of Ambiguity Tractable for Applications" by
Peter Wakker (Erasmus)
Tuesday February 3rd, 2009
Location: Westminster Business School
<http://www.decisionresearchlab.com/ebdm/?page_id=3> seminar room MR2
Time: 17h30-19h00
Abstract
Since Keynes (1921) and Knight (1921) we know that ambiguity ("unknown
probabilities") is central in economics, and since Ellsberg (1961) we
know that ambiguity requires fundamentally new models. It took over 60
years, until the end of the 1980s, before someone succeeded in
developing such new models (Gilboa & Schmeidler). These new models were
theoretical and they were primarily developed for normative purposes.
Hence, they assumed expected utility for known probabilities, and they
focused on aversion to ambiguity.
Given its importance in economics (we rarely know probabilities), and
given the availability of models at long last, ambiguity is nowadays
measured extensively in experimental and behavioral economics, and it is
correlated with other characteristics and behaviors. Ambiguity attitude
is usually captured through one number, an index of ambiguity aversion.
It then is assumed to play a role similar to, and side by side with,
risk aversion. (For specialists: this number is mostly taken to be the
alpha in the alpha maxmin model for multiple priors.) Such modeling is
too theoretical to work well empirically, and too crude to capture
ambiguity.
We propose a number of changes with respect to the presently common
approaches:
1 [Sources]. One index of aversion to ambiguity is as implausible as one
index of risk aversion for all nonmonetary outcomes. Both ambiguous
events and nonmonetary outcomes comprise domains too rich to be captured
by a single number. Tversky proposed to distinguish between different
sources of ambiguity, where a source is a family of events related to
the same mechanism of uncertainty and sharing the same characteristics
much as one numerical good (that may be nonmonetary). By allowing
ambiguity attitudes to depend on sources, we obtain the required
generality without losing too much tractability.
2 [Partial Ambiguity Seeking]. Even within one source there need not be
a universal aversion to ambiguity, but people are often ambiguity
seeking for unlikely events and at the same time ambiguity averse for
likely events. It is, therefore, desirable not to commit to a universal
attitude beforehand but rather let the data speak.
3 [Nonexpected Utility for Risk]. For descriptive purposes we have to
relax the assumption of expected utility for given probabilities.
This paper introduces an approach that achieves the three
generalizations listed, based on what are called uniform sources.
Because of the richness of our model we can let the data speak on what
ambiguity attitudes really are. The data show that, besides aversion,
also lack of sensitivity towards uncertainty is prominent. Because of
the tractability of our model we can display ambiguity attitudes through
graphs, and make concrete predictions about, for instance, ambiguity
premiums, reflecting how much money people are willing to pay extra
because of ambiguity. We do not only develop the theoretical model to
achieve our claims, but we also demonstrate the feasibility of our
measurements in an experiment. We, finally, show how the new tools
provided here are tractable enough to be implemented in applications.
How To Join The Email List
To subscribe to the seminar series email list, please visit
http://tinyurl.com/yvw2sr to opt in. You can easily unsubscribe anytime.
Please pass this message on to those who may be interested in joining
the email list.
The full schedule of talks at the Economics of Behaviour and Decision
Making seminar series is maintained at
http://www.decisionresearchlab.com/ebdm/
Sincerely,
Lionel Page (Westminster Business School)
Dan Goldstein (London Business School)
Emre Ozdenoren (London Business School)
A Related Talk Series in London
If you like decision-making, you will like the London Judgment and
Decision Making Group Seminar.
<http://www.psychol.ucl.ac.uk/ljdm/talks.htm>
--
Dan Goldstein | Assistant Professor of Marketing London Business School
S232 | Regent's Park London NW1 4SA | United Kingdom | +44 20 7000 8611
www.dangoldstein.com
______________________________________________________________________
This email has been scanned by the MessageLabs Email Security System
on behalf of the London Business School community.
For more information please visit http://www.messagelabs.com/email
______________________________________________________________________
|