Unilateral Accident Model

Initiated by Ronald Coase (1960) and developed by Calabresi (1970), Brown (1973) and especially Shavell (1980), (1982), (1987b), the standard unilateral accident model constitutes the backbone of the economic analysis of tort law. Initially, created by lawyers and judges, tort law legally forces the wrongdoers to compensate their victims’ losses. Beyond compensation, the economists showed that the tort law challenge is to understand how the prospect of paying heavy repairs to victims can motivate potential tortfeasors to provide the highest prevention level against risk. Hence, the model is simple: a potential wrongdoer performs activities likely to cause harm to other agents (victims). The prospect of losses due to compensation after an accident

Indeed, in recent years, various contributions have extended the unilateral model of accident to radical uncertainty . The Teitelbaum (2007)’s pioneering representation formalizes Knightian uncertainty applied to tort law. He substitutes to the classical Savage Expectation Utility, the new developments of ambiguity theory. In this model, uncertainty leads to ambiguous choices as in the Ellsberg’s paradox .To formalize ambiguity, Teitelbaum(2007) assumes that the polluters' utility function takes the form of a so-called “neo-additive capacity” that allocates specific weights to extreme earnings (maximum and minimum payoff), and the expected gain (expectation). He shows that both a strict liability and a fault-based regime do not achieve a socially optimal prevention level but negligence seems more efficient than strict liability. Other contributions confirm the relationship between uncertainty and care level inefficiency. For instance, Franzoni (2012) considers the case of ambiguous risk where ambiguity is graded because of the existence of alternative distributions on the accidents likelihood. He analyses unilateral and bilateral accident models. He shows that, under strict liability, damage increases with rising ambiguity. In contrast, under negligence, safety standards increase but only when the injurer’s perceived ambiguity reduces.

However, it clearly differs from the above papers because it shows, first, that, under condition, uncertainty does not prevent the formation of a socially optimal care level. This is the case under a strict liability regime. Second, strict liability and negligence can only be compared under special conditions. In our model, enforcing a liability regime rather than another one has a strong influence on the injurers’ behavior. We show that implementing negligence involves allocating to the Court a higher status than under strict liability. The injurers will prefer considering the judges’ decisions rather than the regulator’s assessment. Consequently, this issue introduces a strong asymmetry between both regimes. This issue comes from the nature of the regulator’s utility function. The latter is benevolent, omniscient but of Negishi (1960)’s type. This means that the social utility function aggregates the agents’ preference conversely to the usual models that consider a neutral to risk