Essays in experimental economics and risk elicitation (W. Kip Viscusi)

The work of Viscusi has been important in litigation and public policy. This thesis examines a survey used by Viscusi designed to detect whether individuals accurately estimate the true risk associated with smoking cigarettes and how risk perceptions influence the decision to smoke cigarettes. The question considered by Viscusi was “Among 100 cigarette smokers, how many of them will get lung cancer because they smoke?”. The survey results were then used to determine the effects of risk perceptions on smoking behavior.; Responses to this question have been used in litigation against tobacco companies. The defendants argue, based on these responses, that smokers already knew about the risks of smoking when they started smoking, so that the tobacco companies should not be held liable. In these legal proceedings liability was always established before the determination of monetary damages. Risk perception issues played no role in the award of (compensatory) damages. They could have played a role in the amounts awarded under the settlements, since they would have been a factor influencing the parties perception of the probability of winning the liability side of the case.; Responses to this question have also been important in public policy, as they affect the willingness of politicians to levy taxes on cigarettes. The tobacco companies have argued that there is no need to levy further excise taxes if the goal of those taxes is to offset the lack of information that smokers have about the risks of smoking. In other words, the tobacco companies argue that if these excise taxes are intended to correct an informational market failure, then they are already too high since smokers over-estimate the risks of smoking.; Chapter 1 covers in detail three general problems with this survey question. Chapter 2 defines an experiment to test the hypothesis that the survey used by Viscusi provided biased responses. The experiment uses an in-person survey and asks respondents to rank their perceived risk of dying from 12 different causes. Chapter 3 tests the hypothesis of hypothetical bias directly, by changing the survey from real to hypothetical.