Paying $1 to Lose $2 : Misperceptions of the Value of Information in Predicting the Performance of Others

Traditional economic and decision-making models allow for "free disposal" of information, meaning that more information will always make a decision maker (weakly) better off. This implies that those faced with decisions should always place non-negative value on information. Building on previous research on the "curse of knowledge," we explore situations where this might not be so. In three experiments, we document situations in which subjects place positive value on information, even when learning that information hurts their performance and earnings. In the first experiment, a significant number of subjects pay for information - the solution to a puzzle - that hurts their ability to predict how many others will solve the puzzle. In the second experiment, a majority of subjects choose to "hire" informed - rather than uninformed - agents, leading to lower earnings. The third experiment reveals that the phenomenon is not reduced with experience, but that also that there are individual differences in the degree to which subjects fall victim to the bias. We discuss implications of our results for the role of information and informed decision makers in real economic situations