Eliciting Utilities With A Focus On Loss Of Life

Most literatures on prospect theory focus on decisions under risk involving monetary attribute.Recent studies have investigated utilities of attributes such as time in general and life value in medical decision-making.The present study aims to investigate decisions in the face of extreme events whose potential consequences are to lose the lives of people.We designed a qualitative questionnaire presenting a variety of scenarios where decision makers are responsible for rescuing a group of people(e.g.,who are trapped in fire),and also conducted a series of quantitative experiments to elicit the prospect theory components in the loss domain including the decision weights and the utility function.Our results show that people are indeed risk seeking for life loss decisions in contrast to risk averse behaviour found for losses of time and consistent with findings about losses of money.However,unlike monetary results this risk seeking behaviour continued to be observed while framing the scenario as a gain or taking away the possibility of escaping from loss.Moreover,the value that subjects assign to human life is not affected by decision characteristics of the situation(e.g.,number of people in the situation)since the human life value has a different nature than time and money.We realized that even though subjects are risk seeking,the utility function for life loss is not convex in contrary to prospect theory utility loss function for money.In fact,our findings suggest that the shape of utility for losses of lives is concave(or at most linear)which is in line with findings for loss of time.In addition,the results of our comparison of life loss decisions and its corresponded monetary values imply that despite the general findings of monetary decisions,people are risk averse in the face of losing big monetary values that encompass all their wealth.Finally,people underweight both small and large probabilities in the face of losses of lives.In other words,people are optimistic about almost all probabilities when outcomes are losing life.Hence,we obtained almost a convex probability weighting function for life loss which is in contrast with inverse S-shape probability weighting function observed in many studies for money and time.