Applying the contingent valuation method to value changes in traffic accident fatality risk has been criticized on the ground that people cannot understand changes in low levels of risk. This paper proposes an indirect way of testing subject comprehension of valuation tasks in contingent valuation surveys. It is hypothesized that answers to properly understood valuation tasks are based on an implicit logarithmic valuation function (Fechner's Law). If valuation tasks have not been understood, answers will be arbitrary and not display the systematic pattern implied by Fechner's Law. Evidence from eight contingent valuation studies is examined and found to be broadly consistent with Fechner's Law. While this finding indicates that valuation tasks were understood, it does not imply that subjects in general understand absolute levels of risk correctly. (A)
Abstract