In this paper we analyse two ways in which liability can be reduced: caps (the "judgment proof problem"), and proportional reductions (the "disappearing defendant problem"). We show that these two problems have different incentive effects and hence yield dissimilar levels of social welfare. Moreover, when they occur simultaneously they may have offsetting effects. We also show that the negligence rule with cause-in-fact may yield lower (rather than higher) levels of social welfare than strict liability. Finally, we analyse the optimal setting of the negligence standard. Our model encompasses different precaution technologies as well as monetary "v". non-monetary precautions. Copyright (c) The London School of Economics and Political Science 2008.

Do injurers' insolvency and victims' reluctance to sue affect accident prevention in the same way? Are these circumstances less of a problem under the negligence rule than under strict liability? We argue, contrary to the literature, that the answer is, in most cases, negative and make three main points. First, the judgment proof problem and the disappearing defendant problem are shown to have different effects on injurers' behavior and hence yield dissimilar levels of social welfare. Second, when these two problems occur simultaneously they may have offsetting effects. Third, the negligence rule is superior to strict liability only under some conditions, which are not always satisfied when cause in fact is considered. In this case, we find that social welfare under negligence may actually be less than, the same as, or greater than under strict liability. ABSTRACTDo injurers' insolvency and victims' reluctance to sue affect accident prevention in the same way? Are these circumstances less of a problem under the negligence rule than under strict liability? We argue, contrary to the literature, that the answer is, in most cases, negative and make three main points. First, the judgment proof problem and the disappearing defendant problem are shown to have different effects on injurers' behavior and hence yield dissimilar levels of social welfare. Second, when these two problems occur simultaneously they may have offsetting effects. Third, the negligence rule is superior to strict liability only under some conditions, which are not always satisfied when cause in fact is considered. In this case, we find that social welfare under negligence may actually be less than, the same as, or greater than under strict l iability. Our model encompasses different precaution technologies as well as monetary vs. non-monetary precautions.