Working Paper: NBER ID: w30353
Authors: Pedro Bordalo; Giovanni Burro; Katherine B. Coffman; Nicola Gennaioli; Andrei Shleifer
Abstract: How do people form beliefs about novel risks, with which they have little or no experience? A 2020 US survey of beliefs about the lethality of Covid reveals that the elderly underestimate, and the young overestimate, their own risks, and that people with more health adversities are more pessimistic, even for others. A model in which people selectively recall frequent and similar past experiences, including from other domains, and use them to imagine (simulate) the novel risk, explains our findings. An experience increases perceived risk if it makes that risk easier to imagine, but decreases it by interfering with recall of experiences that fuel imagination. The model yields new predictions on how non-Covid experiences shape beliefs about Covid, for which we find empirical support. These findings cannot be explained by conventional experience effects, and highlight memory mechanisms shaping which experiences are recalled and how they are used to form beliefs.
Keywords: COVID-19; belief formation; memory; risk perception
JEL Codes: D01; D84
Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.
| Cause | Effect |
|---|---|
| memory similarity (C59) | risk perception (D81) |
| memory interference (Y50) | pessimism about COVID risks (E71) |
| personal experience of having COVID (I12) | belief formation about COVID risks (D80) |
| non-COVID health adversities (I12) | pessimism about COVID risks (E71) |