040820 UK Theories of Bounded Rationality (BA) (2014W)

Curriculum: Elective course of Bachelor in Economics, 2 SSt, 4 ECTS

Prerequisites: previously attended bachelor game theory class

At first sight "Theories" and "Bounded Rationality" seem to contradict one another.
And yet bounded rationality consists of everything that does not belong to the
category of being rational. The term rationality is reserved for the classic paradigm of
decision making in Economics where (it is as if) the decision maker has precise beliefs
about the likelihood that each possible event will occur and makes decisions by
maximizing expected utility. Thus, any deviation from this paradigm falls in the
category of bounded rationality. So bounded rationality encompasses optimization
under alternative criteria, heuristics, rules of thumb, ad-hoc behavior and even
mistakes.
In this lecture we will introduce and discuss the field of bounded rationality that lies
closest to rationality, namely where the decision maker chooses actions that are
optimal according to some specified criteria or "theories".
Theories are valuable as they identify the foundations of predictions and create
benchmarks for comparison to observed behavior.
Bounded rational theories are valuable when the assumptions underlying rational
decision making are less plausible and when one is not willing to characterize behavior
as ad-hoc just because it does not conform to the most stringent framework of rational
decision making.
Excitement rises when one is able to explain common behavioral patterns such as
imitation and reinforcement as optimal according to one of these theories. Theories
provide additional insights when one learns that particular forms "work" while others do not. For instance, a particular form of imitation can lead all to jointly learn what is
best. However, to simply imitate the more successful can lead to the opposite
outcome where everyone chooses the worst action in the long run.
Some of the results presented in the lecture have only been discovered very recently
which makes the topic interesting but makes reading cumbersone. Supplementary
reading will mostly original research articles, to be used to get a general idea, their
deeper understanding is beyond the scope of this lecture. All material will be
presented in the class in a very simple context.

Topics will be presented in the following order:
Making a choice for the first time: rational, maximin utility, minimax regret (French,
1986, Stoye, 2009)
Making the choice for tomorrow based on today: how to reinforce (Börgers et al,
2004), how to imitate (Schlag, 1998, Alos-Ferrer and Schlag, 2007)
Planning for the long run: the role of memory (Rustichini, 1999), evolution and
reinforcement (Börgers and Sarin, 1997), social learning and imitation (Schlag, 1998)