Our findings have implications well beyond the area of punitive damages, because the
legal system is pervasively in the business of requiring people to map normative judgments
onto dollar amounts. Our analysis suggests that the mapping of attitudes onto
unbounded scales creates problems that are both predictable and severe. We suspect, for
instance, that similar difficulties exist in the setting of awards for pain and suffering, libel,
and civil rights violations, and in attempts to use measures of willingness to pay to
estimate the value of public goods.6 There is a need to develop mechanisms for eliciting
social preferences and values that do not suffer from the problems and defects identified
here. We do not expect to find an uncontroversial way of eliciting, or constructing, social
preferences and values, but some approaches to this important task are demonstrably
worse than others, because they make people perform tasks for which they are illequipped.
The problems we have identified and the solutions we have sketched here are
likely to be relevant to many questions currently facing law and policy