Do Probability and Certainty Equivalent Techniques Lead to Inconsistent Results? Evidence from Gambles Involving Life-Years and Quality of Life

Background: Within the standard gamble approach to the elicitation of health preferences, no previous studies compared probability equivalent (PE) and certainty equivalent (CE) techniques Objective: This study aimed to explore the differences between CE and PE techniques when payoffs are expressed in terms of life-years or quality of life. Methods: Individuals were interviewed through both CE and PE techniques within an experimental setting. Inferential statistics and regression analysis where applied to process data. Order and sequence effect were also investigated. Results: On average, the elicitation technique did not affect individuals' risk attitude significantly. Individuals proved to be risk averse in gambles concerning life-years and risk seekers in those concerning quality of life. No order or sequence effect was observed. Risk premium, measuring the strength of risk attitude as the percentage variation between the individual's estimated PE or CE and the risk neutral PE or CE, was affected by the kind of gamble that the interviewee is presented with. It increased in gambles concerning health profiles, denoting a stronger risk propensity, and decreased in gambles concerning life years, denoting a stronger risk aversion. Conclusion: The choice of the elicitation technique did not affect the individuals' risk attitude significantly, which instead was sensitive to the kind of gamble.