Negligence rules coping with hidden precaution

Abstract

This paper investigates the implementation of negligence rules when the negligent act constitutes a hidden action in the sense of principal–agent theory and where the available evidence is generated by a signal. Any liability rule exclusively based on the available evidence comes with an incentive threshold. Agents with cost savings from the negligent act above this threshold commit the negligent act whereas the others do not. In a first step, liability rules are examined that implement a given incentive threshold at minimum error costs. As this is a linear programming problem, the present paper makes heavy use of duality theory. The multiplier of the incentive constraint if understood as shadow price allows for an intuitive explanation of the results. As a second step, a comparative statics analysis with respect to the incentive threshold is provided. Surprisingly enough, the relation between the threshold and minimum error costs need not be monotonic.