Score contribution per author:
α: calibrated so average coauthorship-adjusted count equals average raw count
It is widely argued that individuals have biased perceptions of health and safety risks. A reconsideration of the best-known evidence suggests that this view is the erroneous result of a failure to consider the implications of scarce information. Our findings imply that the hypothesis that people make unbiased estimates of hazard rates fails to be rejected by the very data that were initially used to reject it. Thus, we are able to reconcile the alleged existence of widespread bias in risk perception with other findings that such bias is less apparent in the case of job-related hazards. The seeming bias in estimating population-average death rates and the lack of such bias in assessing job risks are two manifestations of the same behavior, which is the optimal acquisition of costly information. Copyright 1997 by Kluwer Academic Publishers