The Reassurance Effect : Paradoxical Preferences in Information Acquisition

Conventionally, information acquisition is motivated by its instrumental value – the value derived from adapting action to new knowledge. In this paper, we explore the notion of reassurance, whereby people seek to acquire information in order to remove lingering doubts or fears. We formulate and analyse a model of individual preferences that accounts for elation and disappointment -the emotional responses to good and bad news. We assume loss aversion (bad news loom larger than good news) and diminishing sensitivity. We find that a decision maker faced with a large potential loss of low probability will respond positively to non-instrumental information about the occurrence of the loss. We call this the reassurance effect. The same model predicts that a decision maker faced with a large potential gain of low probability will want to avoid non-instrumental information, to maintain hope. When information is weakly instrumental, the reassurance effect causes a decision maker who is less likely to face a loss to value information more than a decision maker for whom the same loss is more likely. This paradox disappears at higher levels of instrumentality. We provide empirical support for reassurance effects, first in a simple scenario-based experiment, and then in a survey about the desire to undertake COVID-19 testing, carried out at the early stages of the pandemic