Can overconfidence be debiased by low-probability/high-consequence events?

During the first half of 2008, China suffered three natural disasters: a heavy snow storm, an outbreak of hand-foot-mouth disease, and a severe earthquake. The aim of the present study is to explore how low-probability/high-consequence events influence overconfidence. In Study 1, opportunity samples were obtained by recruiting residents in three different types of disaster-hit areas to answer a peer-comparison probability judgment questionnaire about 1 month after the corresponding disaster occurred. The performance of 539 participants in disaster-hit areas was compared with that of 142 residents in a nondisaster area. The findings indicate that residents in disaster-hit areas were less overconfident than those in the nondisaster area on both positive and negative events. In Study 2, we surveyed a total of 336 quake-victims 4 and 11 months after the earthquake to examine whether the impact of disasters on overconfidence would decay with time. The resulting data indicate that the disaster victims became more overconfident as time elapsed. The overall findings suggest that low-probability/high-consequence events could make people less overconfident and more rational and seem to serve as a function of debiasing.