Applications of the analysis

Note that the first two terms of this condition resemble the two terms in Eq. 9 and have the same interpretation. Specifically, the term on the left-hand side of Eq. 19 is the marginal cost of a longer statute length in terms of increased litigation costs, while the first term on the right-hand side is the marginal deterrence benefits of an increased statute length. As was true under strict liability, greater deterrence is socially desirable because in equilibrium, negligence underdeters as captured by the fact that x N L 2 x N L . 0. What is different here compared to strict liability is the second term on the right-hand side of Eq. 19, which is positive given that ­w­L , 0. This term captures an additional marginal benefit of increasing the statute length under negligence in the form of fewer lawsuits being filed in each period up to t 5 L. Intuitively, an increase in L lowers the probability that any given injurer will be judged negligent given that ­w­L , 0, thereby making lawsuits less profitable for victims. Note that this effect was absent under strict liability because that rule did not require victims to prove negligence by injurers as a condition for recovery. The extra marginal benefit term in Eq. 19 suggests that the optimal statute length should be longer under negligence as compared to strict liability. This result cannot be established definitively because a comparison of the corresponding terms in Eq. 9 and Eq. 19 for any L is ambiguous. However, if these common effects are roughly offsetting, one would expect that the additional term in Eq. 19 would dominate, resulting in a longer statute length under negligence.

5. Applications of the analysis

The foregoing analysis suggests that one might try to test the model by asking whether the statute of limitations is longer in areas of tort law governed by negligence as compared to areas governed by strict liability. There are several problems in testing the model in this way. First, statutes of limitations are costly to change and therefore may respond only sluggishly to changes in relative costs and benefits. Second, economic theory maintains that the legal rule is itself an endogenous variable that should respond to changes in some of the same variables that affect the statute of limitations e.g., litigation costs. Third, most torts are grouped under a single statute of limitations. Thus, for example, the shift in produces liability law from negligence toward strict liability during the twentieth century may have had no effect on the general statute length for personal injury cases. Despite these problems, I offer some evidence that the statute of limitations does in fact differ depending on whether a given area of the law is governed by strict liability or negligence. This evidence takes the form of variation in the rules that determine when the statute begins to run. 10 10 I thank a referee for suggesting the second and third applications in this section. 390 T.J. Miceli International Review of Law and Economics 20 2000 383–394 Statutes of repose for products liability As noted, one area of tort law that has shifted from negligence to strict liability is products liability. 11 In response to this change, many states enacted “statutes of repose” for products liability suits Peacock, 1993. 12 In contrast to a statute of limitation, which begins to run at the time an accident occurs, a statute of repose begins to run at the time the product in question was sold Martin, 1982: p. 749. Thus, a statute of repose puts a predictable upper bound on the length of time that a manufacturer is at risk of suit from the sale of a particular product. A statute of limitation alone provides no such bound because it allows suits for accidents that occur long after sale as long as they are filed in a timely manner after the accident. In the context of the model, it is easy to show that enactment of a statute of repose serves to reduce the expected number of suits that an injurer faces. In this sense, it is equivalent to a reduction in L. To see this formally, suppose that consumers purchase a dangerous product at time t 5 0, and it causes an accident at time t . 0, which varies across consumers based on differences in use and random factors. Victims then file suit after a delay of t as described in the above model i.e., a victim of an accident at time t files at t 1 t. Figure 1 shows all possible combinations of t and t, where each point represents a potential lawsuit. In this amended model, the statute of limitations applies only to the delay in filing after occurrence of the accident. Thus, any suit with t L is allowed to proceed, regardless of t . Graphically, the statute of limitations permits all suits to the left of L in Figure 1 i.e., areas A and B. In contrast, a statute of repose of length R applies to the length of time following sale of the product, and therefore allows only those suits for which t 1 t R. Thus, victims in area B of Figure 1 will no longer be allowed to proceed, leaving area A as the set of products liability cases that satisfy both the statute of limitations and the statute of repose. What this argument shows is that adding a statute of repose on top of a statute of limitations has the same effect as simply shortening the statute of limitations. Thus, the enactment of these statutes following the shift from negligence to strict liability for products liability suits seems to be consistent with the predictions of the model. The discovery rule for negligence Historically, a cause of action under negligence law began to run at the time that an injury actually occurred, or time t the “damage rule”. Beginning in the area of medical mal- practice, however, courts and legislatures began to replace this with the “discovery rule,” under which “the statute does not begin to run until the negligent injury is, or should have been discovered.” 13 This rule has now become the majority rule for negligence cases. The impact of the discovery rule is clearly to lengthen the statute of limitations by starting 11 This evolution has been gradual, but Landes and Posner 1987: p. 307 roughly date the shift to Henningsen v. Bloomfield Motors, Inc. , 32 N.J. 358, 161 A.2d 69 1960. 12 Also see Martin 1982, who notes that as of 1982, as many as twenty-two states had enacted statutes of repose, ranging from six to fifteen years in length. Since then, however, many of these statutes have either been repealed or found unconstitutional Warfel, 1991; Werber, 1995. 13 See Teeters v. Currey, 518 S.W.2d 512 1974. Also see Keeton, et al. 1984: pp. 166-167. 391 T.J. Miceli International Review of Law and Economics 20 2000 383–394 it at a later date relative to the time of the injury. It therefore works in the opposite direction of a statute of repose, which we saw above begins the statute prior to the time of the injury. Indeed, Keeton, et al. 1984: pp. 166 –167 observe that the enactment of statutes of repose was in part a response to the widening use of the discovery rule. In combination, therefore, these two rules work in the predicted direction of effectively lengthening the statute of limitations for cases governed by negligence as compared to strict liability. Trespass versus nuisance The final application concerns the distinction between the torts of trespass and nuisance. Under trespass, the plaintiff is entitled to strictly enforce his or her right to exclusive possession of a piece of property. In contrast, the plaintiff can only assert a claim under nuisance law if the harm is both substantial and unreasonable, where the latter is based on a balancing of “the gravity of the harm and the utility of the conduct” Keeton, et al., 1984: p. 630. Thus, trespass is governed by strict liability and nuisance by a negligence standard Landes and Posner, 1987: p. 49. Fig. 1. The effect of a statute of repose. 392 T.J. Miceli International Review of Law and Economics 20 2000 383–394 It follows from the theory that the statute of limitations should be shorter for trespass as compared to nuisance, and this in fact appears to be the case. Specifically, under trespass, the statute of limitations begins to run “from the time of unlawful entry” 14 and “without regard to harm” Keeton, et al., 1984: p. 70, whereas under nuisance, the statutory period does not begin to run “until some injury has been sustained.” 15 Once again, the law seems to conform to the predictions of the theory.

6. Conclusion