Use this URL to cite or link to this record in EThOS:
Title: The consequences of behavioural bias : bandit problems and product liability law
Author: Luppi, Barbara
ISNI:       0000 0004 2684 9504
Awarding Body: London School of Economics and Political Science
Current Institution: London School of Economics and Political Science (University of London)
Date of Award: 2010
Availability of Full Text:
Access from EThOS:
Full text unavailable from EThOS. Please try the link below.
Access from Institution:
The aim of the thesis is to explore how certain behavioural biases affect decision making. It focuses on two contexts, bandit problems and the case of legal decision making. In regard to bandit problems, the focus of interest is to examine the role of risk aversion and loss aversion, which are both excluded from the standard literature on bandit problems. We consider the standard bandit problem under alternative models of behaviour, from the standard expected utility model with risk neutral agents, to models with risk aversion, with loss aversion, and with risk and loss aversion. In Chapter 2 we show that new qualitative features of the prediction emerge in the presence of both risk aversion and loss aversion. Subjects who are moderately loss averse find it optimal to experiment more than highly loss averse or loss neutral individuals. In Chapter 3, an experimental study is conducted that involves a three-stage one-armed bandit problem. The main substantive finding in Chapter 3 is that there is a bias towards over-experimentation in stages 1 and 2, and a bias towards under-experimentation in stage 3, relative to the theoretical model, which is not explicable by reference to subjects' levels of risk or loss aversion. This remains a 'behavioural' puzzle. In regard to legal decision making, in Chapter 4 we examine the role of overconfidence in undermining the incentives provided by liability rules. Even factually informed, overconfident people tend to think that risks are less risky to materialize for themselves than for others and therefore inadequately react to legal threats and incentives such as liability rules. We examine the role of tort rules in "debiasing" overconfidence, showing paradoxically that the most effective way to correct overconfidence could be to forgive it, rather than to penalize it through liability. In Chapter 5 we examine applications to the field of product liability. Products regulations that impose liability on producers for not increasing the safety of their products in anticipation of consumer's overconfidence can be viewed as a legal strategy analogous to legal forgiveness of consumers' overconfidence.
Supervisor: Not available Sponsor: Not available
Qualification Name: Thesis (Ph.D.) Qualification Level: Doctoral
EThOS ID:  DOI: Not available