Reducing Moral Hazard Associated with Implied Warranties of Animal Health: Comment

Abstract

In a recent Journal article, Centner and Wetzstein examine efficiency losses associated with moral hazard and the assignment of cost liability arising from diseased animals.' Traditionally, a buyer of livestock had an implied warranty that the seller would compensate all ex post costs arising from diseased animals. The fear that implied warranties would provide disincentive for buyer self-protection against possible damages (moral hazard) has led sellers to lobby for abrogation of implied warranties. Centner and Wetzstein argue that abrogation of implied warranties does not remove moral hazard but simply transfers disincentive from buyer to seller. By noting that welfare under buyer liability may be lower than under seller liability, Centner and Wetzstein suggest efficiency losses can be reduced by maintaining seller liability and introducing partial insurance contracts where both parties share a burden of the ex post damages. To induce a buyer to self-protect, the partial insurance contract is assumed to be a decreasing function of self-protection. This is a restrictive assumption, however, because it is often impossible or prohibitively costly for insurers (sellers) to monitor and enforce the insured's (buyer) behavior (Spence and Zeckhauser, Holmstrom). Indeed, it is this lack of monitoring that motivates moral hazard in the first place (Arrow).2 The purpose of this comment is to illustrate that a decentralized solution where monitoring is unnecessary can be obtained by considering three points: (a) that the state of the world is dependent on the buyer's self-protecting actions, (b) the blurred distinction between self-protection that influences severity of loss and self-protection that influences probability of loss, and (c) a sufficiently large difference in the dollar benefits of diseased and disease-free animals. Given these three points, contrary to Centner and Wetzstein's claim, the buyer will have sufficient incentive to purchase optimal self-protection despite the existence of implied warranties of animal health.3 First, Centner and Wetzstein assume that the two mutually exclusive and jointly exhaustive states of nature are independent of buyer action. Let 13 and 1 p3 represent the buyer's degree of belief of selecting a diseased or a disease-free animal (using Centner and Wetzstein's notation). As pointed out by Marshall, however, independent states of nature exist only if a complete set of contingent contracts exist. While theoretically feasible, complete contingent claims contracts rarely if ever exist in reality. For example, given imperfect contingency markets, the buyer can influence the probability of r ceiving a disease-free animal by adopting assorted self-protection strategies, i.e., pretesting for disease, use of experienced sellers and markets. L t x represent buyer self-protection such that the degrees of belief of selecting diseased and diseasefree animals are P(P3*, x) and (1 P(P3*, x)), respectively, where /3P is an endowed probability. Assume that an increase in self-protection decreases the probability of selecting a diseased animal, /3,, 0.4 Subscripts denote relevant partial derivatives.