These studies establish that most individuals believe that their driving, health and financial risks are smaller than the median individual. These results may be taken either as evidence that individuals underestimate their personal risk, or overestimate the risk of others.

In order to settle this issue, we review experimental studies which compare the subjects’ self-reported risk with their actual risk determined objectively or by external examiners.

The evidence in this experiments establishes that subjects underestimate their own personal risk. We conclude the first section of our survey by reviewing the experimental and empirical evidence on the relationship between overconfidence and age.

The main findings of this literature are that overconfidence is particularly pervasive among young adults, but does not vanish with learning and experience.

Once discussed the broad evidence that many individuals are overconfident, and underestimate personal risks, we turn to review the literature that assesses the implications of these findings with the respect to precautionary activity. The second section of our survey reports broad evidence that overconfident agents underinvest in precautions.

For example, drivers’ overconfidence is considered one of the main variables affecting of traffic safety.

Focusing on the choice of insurance, we discuss the evidence that individuals may underinsure in driving and health insurance because they misperceive their risk.

The number of individuals who choose not to purchase health or motorist insurance is significantly large, and overconfidence is one of the explanations for this phenomenon.

We devote the third section to discussing training techniques developed by practitioners of accident prevention to reduce driving overconfidence and improve traffic safety.

We conclude our survey in the fourth section.