A response to Kenneth S. Abraham, Response, Insufficient Analysis of Insufficient Activity, 108 Mich. L. Rev. First Impressions 24 (2009); Richard A. Epstein, Response, Activity Levels Under the Hand Formula, 108 Mich. L. Rev. First Impressions 37 (2009); Mark Grady, Response, Another Theory of Insufficient Activity Levels, 108 Mich. L. Rev. First Impressions 30 (2009).
In our article, Negligence and Insufficient Activity, we proposed that tort scholarship has overlooked the risk that injurers will behave strategically in setting their activity levels. Whereas the standard literature has predicted that injurers who are subject to a negligence regime will often invest efficiently in care but choose excessive activity levels, we showed that they may do exactly the opposite: injurers may deliberately restrict their activity to avoid investments in socially desirable precaution. After reviewing the conditions that may give rise to the risk of insufficient activity, we examined the ways in which the legal system can minimize the costs of such behavior. We hoped that our article would spark new interest in the interplay between tort liability and levels of care and activity. We are fortunate that three prominent tort scholars have already provided important insights on both our positive argument and our policy recommendations.
These responses vary in their assessments of our contribution to legal scholarship. While some find that our article addresses an "interesting omission in tort law scholarship" and highlights a "missing paradigm" that is "central to negligence doctrine," others consider it a "novel brick on a road that runs in the wrong direction." Nevertheless, and although the responses focus on different parts of the article, they share several important objections to our analysis.
The concerns raised by Professors Abraham, Epstein, and Grady can be divided into three categories. The first concern addresses our positive claim. The risk of insufficient activity levels, it is argued, is theoretically possible but unlikely to be significant in practice. The second concern questions the feasibility of our policy recommendations. According to this line of argument, to the extent insufficient activity poses a real problem, it cannot be resolved by the legal system. Finally, the third concern asserts that, even if the legal system were able to discourage insufficient activity levels, the ensuing costs would be prohibitive. The following rejoinder addresses and responds to each objection.
I. Is There Actually a Risk of Insufficient Activity Levels?
Professors Grady and Epstein present two arguments against our prediction of injurers' strategic behavior. Grady suggests that our numerical results, illustrating injurers' incentives to restrict their activity below the socially desirable level, largely hinge on the assumption of "lumpy" precaution. In our examples, we discuss cases in which injurers can invest in a single precaution (e.g. a higher smokestack), the cost of which is unaffected by their activity levels (e.g. factory's level of production). Grady argues that when a number of precautions are available (e.g. smokestacks of several heights), or when the cost of precaution is affected by activity levels, injurers would no longer be able to derive an advantage from restricting their activity. While Grady focuses on our model, Epstein's concern is doctrinal. Epstein primarily argues that the Hand formula-and its cost-benefit analysis approach-accounts for only a narrow segment of tort liability in the real world. Many other harmful activities are not subject to the Hand formula; instead, statutes and regulations govern these activities with specific safety standards that do not rely on the injurer's activity level. Injurers subject to such statutes and regulations, therefore, cannot escape a duty to invest in care by manipulating their activity level. Accordingly, Epstein concludes that our analysis regarding insufficient activity is largely unimportant.
Our response to Grady's claim is that the cost of precaution need not be "lumpy" for the insufficient activity result to materialize. As noted, Grady discusses two variations of non-lumpy precautions. The first variation concerns cases in which increasing investment in precaution continuously reduces the level of harm. A railroad's spark arrester, for example, can come in various gradations, each slightly more expensive than the other, where a more expensive arrester enables a larger reduction in the expected harm caused by fire. Under such circumstances, Grady suggests that insufficient activity cannot emerge. As we prove mathematically in the original article's appendix, however, parties also have an incentive to strategically restrict their activity below the desirable level with non-lumpy precaution. The explanation for this result is that injurers might prefer to engage in a low activity level since they bear the cost of precaution while the victims derive the benefits of such an investment. Grady also discusses a second variation of non-lumpy precaution in which the cost of care depends on the level of activity. For example, a spark arrester could disturb the train's motion, thereby causing the locomotive to burn more fuel. In that situation, the cost of the spark arrester is directly proportional to the number of uses. Grady doubts that our result of insufficient activity would still hold in this case. We briefly addressed this important question in a variation of our polluting plant example, in which a polluting plant could eliminate the harm it caused by elevating its smokestack. As we demonstrate in footnote 47, even where the cost of elevating the plant's smokestack increases with the activity level ($120 in the low level and $125 in the high level), the plant would elect to engage in insufficient activity under a negligence regime. As long as the plant's benefit from higher activity is lower than the cost of the precaution at the high level, the plant is better off restricting its activity.
Professor Grady also notes the connection between our article and his previous work on durable and non-durable precautions. Elaborating on this distinction, Grady shows that when courts hold parties liable for their lapses, a negligence regime can lead to insufficient activity levels, particularly when a contractual relationship exists between the parties. While Grady's insight advances a different theory regarding the tort system's inability to induce efficient levels of activity, it reinforces our general claim that greater attention should be paid to the effect of liability regimes on activity levels.
As noted, Epstein's critique rests on his assertion regarding the limited significance of the Hand formula in torts adjudication. Since courts usually do not engage in cost-benefit analysis when determining negligence, parties have no incentive to behave strategically in setting their activity levels. Our response to Epstein's objection is twofold. First, irrespective of the interesting empirical question regarding the percentage of cases in which the Hand formula itself is applied, cost-benefit analysis is often used to determine parties' duty to take care under a wide range of circumstances. The cases we discuss in our article demonstrate this point. The court in Donovan-assessing whether owners of a cannery are required by the Occupational Safety and Health Act to insulate their equipment to reduce noise-compared the costs of such a precaution to its expected benefit. Similarly, in Spanglo, where the standard of liability was recklessness, the court relieved the owner of a hockey rink from liability based on cost-benefit analysis. Thus, our analysis should not be viewed merely in the context of tort cases involving the Hand formula, but rather in the context of any harmful behavior regulated on the basis of negligence-like cost benefit analysis.
Second, as we explained in our article, the risk of insufficient activity may provide a new economic rationale precisely in those contexts of harmful behaviors that are governed by statutes or regulations. Consider, for example, the circumstances in Donavan and imagine (contrary to the current legal regime) that regulations require every cannery to insulate its equipment. Efficiency-oriented scholarship often regards such regulation as undesirable. Spending large amounts on precaution may be justified only if the expected harm is substantial. Accordingly, from an economic perspective, a duty to insulate should be limited to canneries with a large number of employees. The risk of strategic behavior, however, can justify such forms of regulation. A duty to invest in precaution, irrespective of the actual activity level, removes the injurer's incentive to set his activity below the socially desirable level. Our discussion of the Entergy controversy shows how even a highly professional, efficiency-oriented organization such as the EPA can overlook the virtue of regulation that imposes standard environmental duties unrelated to the parties' actual activity level.
II. Can the Legal System Eliminate the Risk of Insufficient Activity?
In our article, we propose several ways that the legal system might address the risk of insufficient activity. Our first suggestion involves extending the cost-benefit analysis that courts currently apply in determining liability. Under the conventional approach, courts examine the efficiency of untaken precautions against the background of the injurer's actual activity level. The proposed approach would require courts to examine the cost-effectiveness of precautions, assuming the activity is carried out at the socially optimal level. Given this extended inquiry, injurers would no longer be able to avoid efficient investments in care by strategically limiting their engagement in the activity.
Professors Abraham and Epstein argue that our proposal would be difficult to implement, given courts' limited information. Under our suggested extended cost-benefit analysis, a plaintiff seeking compensation-the injurer's apparently reasonable behavior notwithstanding-would have to prove three elements:
(1) A higher level of activity is socially desirable;
(2) Given this activity level, investment in a precaution is cost-effective; and
(3) Had the injurer invested in care, the victim's harm would not have occurred.
Professor Epstein is somewhat skeptical regarding courts' competence to determine the first element. Professor Abraham doubts whether courts would be able to resolve the third element when investment in care does not render the behavior entirely safe. He argues that if the injurer's untaken precaution is of the sort that does not completely eliminate the risk of harm, courts will face insurmountable informational hurdles in determining causation.
We concede that our proposal may not always be easy to apply. Nevertheless, as we have shown, courts have overcome such informational hurdles in other contexts. Our discussion of the Esposito case demonstrates that in nuisance cases, courts often assess the costs and benefits of different activity levels to determine whether injurers engage in excessive activity. Under the regime we advocate, courts would perform a similar assessment to verify whether injurers engage in insufficient activity. As for courts' ability to determine causation under our proposed regime, this inquiry should not be substantially more difficult than in many standard negligence cases. The difficulty with regard to causation arises whenever some risk of harm would have remained even had the injurer not been negligent.
Because courts will not always possess the required information, we propose in our original article two additional policy solutions. These proposals show how legislatures, rather than courts, can reduce the possible social welfare loss resulting from insufficient activity. One proposal emphasizes the effectiveness of two types of regulations: those imposing industry-specific (rather than firm-specific) safety standards and those setting maximum-harm restrictions on activities with negative externalities. The other proposal provides guidelines as to the choice between strict liability and negligence. By applying the appropriate liability regime, legislators can minimize parties' private payoff from insufficient activity. Neither Abraham's nor Epstein's critiques, which focus on courts' informational constraints, address these proposals. Thus, even if Epstein and Abraham are correct, and courts are unable to apply the extended cost-benefit analysis, legislatures may still be able to solve the problem by following these proposals to reduce the risk of injurers' strategic conduct.
III. The Costs of Removing the Risk of Insufficient Activity: Is It Worth It?
Even if one is convinced that injurers behave strategically in setting their activity levels and that the legal system is capable of discouraging such behavior, we still must consider the costs of any proposed reform. If the costs of the necessary changes to the current doctrine outweigh the resulting benefits, a second-best solution is to simply ignore the risk of insufficient activity. In such a case, the legal system should retain the current regime despite the incentives it provides for strategic conduct.
In this regard, Professor Abraham raises two arguments in favor of retaining the current form of liability and rejecting our proposal to expand the definition of negligence. First, he contends that if courts were allowed to impose liability for parties' insufficient activity, such decisions would be perceived as an inappropriate infringement of parties' autonomy. As Abraham argues, decisions regarding activity levels (e.g. how much to produce), as opposed to decisions regarding care levels (e.g. how carefully to produce), are usually considered to be within one's private discretion. Consequently, even if liability for insufficient activity is justified from a social perspective, in practice "judges and juries will hesitate to impose it." Professor Abraham's second argument is that the imposition of liability for insufficient activity will excessively complicate the litigation process. To begin with, he argues that the determination of whether parties should have engaged in more activity is often a "polycentric" question that courts are ill equipped to resolve. However, even if courts were capable of making such decisions, Professor Abraham claims that allowing the parties to introduce arguments regarding activity levels will expose the courts to many such unsubstantiated claims. Rather than raising the issue only in cases presenting a genuine concern of the injurer's possible strategic behavior, "parties will attempt to raise these questions at trial whenever the law permits them to do so." According to Professor Abraham, when considering these two arguments together-a low likelihood of actually solving the problem and a high probability of increasing litigation costs-the proposed expansion of the definition of negligence is clearly undesirable.
We stress that this type of critique is again restricted to the first solution that we discussed. Professor Abraham's concerns address the costs and benefits of imposing liability for insufficient activity as part of the negligence doctrine. They do not apply to our analysis regarding the design of regulations and the choice between negligence and strict liability. Thus, if courts and juries indeed cannot adequately solve the problem of injurers' strategic conduct, the case for addressing the risk of insufficient activity on the legislative level is even more compelling.
However, even if Professor Abraham's arguments are examined only in the context of courts' application of the negligence doctrine, it still remains to be seen to what extent his concerns are real. Because our suggestion to expand the basis of negligence deviates substantially from current doctrine, it is impossible to provide empirical evidence on the effects this change will have on judges', juries' and litigants' behavior. Analysis of the development of negligence doctrine in the last century, however, reveals that similar objections were raised against other proposals to reform negligence liability that were later adopted. In retrospect, adopting these proposals did not lead to the detrimental results that had been predicted.
An illuminating example can be found in the debate over the shift from contributory negligence to comparative negligence. Proponents of the common law tradition opposed this change on the grounds that it would likely complicate the litigation process without providing any real advantage. With respect to potential benefits, opponents of comparative negligence argued that judges and juries would usually refuse to grant compensation to victims without "clean hands" (victims who themselves contributed to the materialization of the harm). Thus, as a practical matter, in most cases the general rule would continue to be that of the all-or-nothing contributory negligence regime rather than the more flexible comparative negligence regime. With respect to potential costs, opponents of comparative negligence argued that the new regime would only motivate parties to litigate over the precise allocation of fault, especially when the stakes are high, thereby raising the overall costs of litigation. Likewise, opponents raised doubts as to courts' competence to reach sensible decisions regarding the allocation of fault. Courts may have had little trouble determining whether a party failed to comply with the relevant legal norm. However, striking the balance between two parties who failed to uphold the legal standard would have involved "polycentric" questions that courts were ill equipped to resolve. To opponents, it was one thing to determine, for example, whether manufacturers' products were defective. It was an entirely different issue to decide if, and to what extent, manufacturers' liability for defective products should be restricted if consumers were also careless in using them. Therefore, resolving the second problem necessitated addressing thorny questions of policy.
In hindsight, these objections proved to be unjustified. The trend in the United States and abroad has been towards replacing contributory negligence with various forms of comparative negligence. Despite the above-mentioned concerns, courts have successfully integrated the new doctrine into the tort system. While there are various proposals to lower the level of compensation for different types of harm and to simplify the litigation process in tort-related suits, comparative negligence seems to only gain in popularity. To be sure, the success of comparative negligence does not necessarily mean that our proposal will work equally well. But it does suggest that questions regarding the competence of the courts might be less founded than they initially appear.
* * *
Under the current legal regime, parties are usually liable for the harm they inflict only to the extent that their behavior is unreasonable. In determining what constitutes unreasonable conduct, courts and policymakers seeking to maximize social welfare often apply cost-benefit analysis. Our theory of insufficient activity indentifies a potentially significant gap in the way in which this analysis is conventionally conducted. The insightful responses to our article suggest that further analysis may be required to assess both the size of the problem and the ways to resolve it. We intend to further explore these questions and hope to continue to benefit from the important and valuable critiques of our colleagues.
David Gilo is an Associate Professor at Buchmann Faculty of Law, Tel Aviv University.
Ehud Guttel is a Visiting Professor at Duke Law School and Cardinal Cody Chair in Law, Hebrew University Law Faculty.
Suggested citation: David Gilo & Ehud Guttel, Response, Insufficient Activity and Tort Liability: A Rejoinder, 108 Mich. L. Rev. First Impressions 64 (2010), http://www.michiganlawreview.org/assets/fi/108/giloguttel.pdf.
. Kenneth S. Abraham, Response, Insufficient Analysis of Insufficient Activity, 108 Mich. L. Rev. First Impressions 24 (2009).
. Richard A. Epstein, Response, Activity Levels Under the Hand Formula, 108 Mich. L. Rev. First Impressions 37 (2009).
. Mark Grady, Response, Another Theory of Insufficient Activity Levels, 108 Mich. L. Rev. First Impressions 30 (2009).