In Negligence and Insufficient Activity: The Missing Paradigm in Torts, David Gilo and Ehud Guttel argue that negligence law encourages inefficiently high and low levels of activity because negligence law ordinarily does not take activity levels into account. They suggest that the law should impose liability for failing to take safety precautions-even where precautions would not be cost-justified-whenever the threat of this liability negates the incentive for an actor to choose an insufficient level of activity. Until now, the literature on the interaction between liability standards and activity levels has failed to recognize the possibility of inefficiently insufficient activity. I commend Gilo and Guttel for both their insight and their explication of it.
Despite the novel analysis of an interesting omission in tort law scholarship, the proposal for insufficient activity liability ("IAL") would greatly complicate tort litigation, and the authors offer no evidence that it would deliver significant benefits in return. Rather, this theory of liability would create several new problems, for which the authors propose no solutions. First, IAL is riddled with causal uncertainties that the authors do not fully acknowledge. Second, juries may impose IAL unreliably because they would consider IAL an unwarranted infringement on autonomy. Third, the adoption of IAL would create fact-finding difficulties that conventional negligence and traditional strict liability do not face. In short, the authors' analysis of how IAL would work in practice is insufficient.
I. Causal Responsibility and the Scope of Liability
Gilo and Guttel recognize that IAL will require proof of a causal connection between the defendant's failure to take a precaution that is not cost-justified and the plaintiff's injury. But they too easily dismiss the uncertainties that this requirement will generate. They use a number of different terms for this form of liability, sometimes calling it "strict liability" and elsewhere calling it "negligence liability" for "failures to invest in precautions whose cost is higher than their benefit in reducing harm." It is true that imposing liability for failure to take a precaution that is not cost-justified is a version of "strict liability," in that it is liability without fault. But the use of the various terms is problematic because strict liability and negligence apply different causation requirements.
Different forms of IAL would have different implications for proving causation. In negligence cases, a defendant is not liable unless the plaintiff would not have been injured but for the defendant's failure to take a cost-justified precaution. In contrast, traditional activity-based strict liability requires no causal connection between the precautions the defendant did or did not take and the injury the plaintiff suffered. All injuries arising out of the characteristic risks of conducting the activity result in liability.
The proposal for IAL appears to contemplate a negligence-like causation requirement. IAL would be imposed only if the plaintiff would not have been injured but for the defendant's failure to take a particular precaution. As Gilo and Guttel suggest, liability might be imposed for the failure to install noise insulating material in a factory, the failure to use safety railings in bleachers, the failure to supply a bartender at a party, or the failure to install a sprinkler system in a classroom-even though these precautions would not have been cost-justified. In each instance, the plaintiff would have to prove a causal connection between the defendant's failure to take the precaution and the plaintiff's injury. Thus, although IAL would create a new form of liability, IAL would also retain a causation requirement analogous to the traditional cause-in-fact requirement of liability in negligence.
But exactly how would a plaintiff prove causation in such cases? Noise insulation does not prevent all hearing loss, safety railings do not prevent all falls, and the presence of a bartender at a party does not prevent all guests from becoming intoxicated (to use Gilo and Guttel's own examples). Yet the success of IAL depends, in part, on the possibility of proving which hearing losses, falls, or drunken mishaps the precautions in question would have prevented.
Adducing such proof would actually be extremely difficult in most cases of IAL. In contrast, traditional tort law does not pose problems of causal uncertainty in cases such as these. Strict liability depends only on the fact that the defendant engaged in the activity. Factual causation is also a simple matter, as it is proven by demonstrating the simple causal connection between engaging in the activity-operating a noisy factory, holding an event with bleachers, or sponsoring a party and serving alcohol-and the plaintiff's injury. In these cases, liability attaches regardless of whether greater precautions would have prevented the injuries at issue. And in traditional negligence cases, the fact that taking a precaution would substantially reduce the probability of causing harm often provides the decisive proof that the defendant acted negligently by failing to take that precaution. The "P" in Judge Learned Hand's B < PL negligence calculus refers to this probability. Consequently, plaintiffs in negligence actions seldom face the vexing problem of proving which injuries were caused by the failure to take a cost-justified precaution-though proving causation can become a serious issue where failure to take the precaution does not substantially increase the probability of harm.
But under IAL, almost by definition, the failure to take the precaution in question would not substantially increase the probability of harm-because otherwise, that failure probably would be negligent. As a result, plaintiffs will seldom have affirmative evidence of factual causation. In the unusual case where there is evidence relevant to causation, it will likely show that failure to take the precaution did not substantially increase the probability of harm-because, otherwise, failure to take the precaution would be negligent. Plaintiffs would only have a reasonable chance of proving causation in cases that reveal a substantial increase in the probability of causing a very minor harm or a disproportionately large cost of preventing a substantial increase in harm. Paradoxically, then, in the very set of cases in which it might be feasible to impose IAL because there is relevant causation evidence, IAL will not be imposed because the same evidence would show a low probability of causation. As a consequence, there will actually be very little IAL in practice.
Unfortunately, Gilo and Guttel barely acknowledge this problem. They relegate it to a footnote, conceding only that "where optimal investment in precaution does not entirely remove the risk of harm, proving causation may present some difficulty . . . ." This is insufficient treatment of the very problem posed by most of their own examples.
II. Optics and Autonomy
Law has an expressive function. It sends a message to those it governs, to those who enforce it, and to the public at large. Sometimes a message is just a message. But sometimes the nature of the message influences how faithfully the rule described by the message will be enforced. The message IAL sends would likely undermine its enforcement, because IAL ignores-or at the least appears to ignore-a traditional sphere of autonomy.
At points, Gilo and Guttel contend that IAL would send the message that defendants should be liable for non-negligent injuries where they have deliberately restricted their activity levels so as to avoid liability in negligence. That is a bit like holding an actor liable for engaging in otherwise-legitimate liability avoidance, and is itself likely to provoke an ambivalent reaction by judges and juries.
But this is not the only reasonable interpretation of IAL. An alternative interpretation maintains that IAL strongly invades a potential defendant's autonomy. This interpretation suggests a judgment that a defendant simply did not engage in enough of the activity in questi on-that it did not hire enough employees, admit enough patron s to its stadium, hold a large enough party with alcohol, or (as the authors suggest) pollute enough. Whether or not this is an entirely fair interpretation, it is certainly the message that many defendants will contend IAL has sent to them, as well as to judges and juries.
Yet this message will likely undermine the normative force of the IAL rule because the law does not ordinarily compel actors to engage in more of a particular activity. Since IAL crosses into this traditional sphere of autonomy, judges and juries will hesitate to impose it. In an effort to address this concern, Gilo and Guttel contend that IAL is not liability for nonfeasance, but for misfeasance, and that IAL therefore does not risk infringing on autonomy in the way that liability for nonfeasance infringes on autonomy. Calling IAL misfeasance, however, does not neutralize this autonomy concern. Even if IAL stands for something more than nonfeasance, it still constitutes liability for failing to engage in more of an activity than the actor has chosen to engage in. And no matter how often economic analysis explains that liability can be understood as a mere price rather than a sanction, the authors' very invocation of the term "misfeasance" (otherwise why not simply call it "feasance?") to characterize the IAL defendant's conduct implies that, although the defendant is legally free to incur liability rather than to increase its activity level, it would be wrong to do the former rather than the latter.
This would be mere semantics were it not for the fact that the misfeasance in question results from failing to engage in a higher level of activity. The difference between wrongfully failing to do something and failing to do something entirely optional is significant. Most forms of insufficient activity would undoubtedly fall into the latter category. Failing to buy a larger house than one wanted, or failing to buy more stock than one wished to buy, for example, would not be considered wrongful or worthy of criticism, whether inefficient or not. For the same reason, it is prima facie an infringement of autonomy to impose a "duty mandating adjustment of one's activity levels" and to enforce that duty with civil liability, if the adjustment is upward, not downward.
III. Complicating Trials with Unmanageable Questions of Fact
The general unwillingness of negligence law to assess the optimality of activity levels avoids the difficulty of weighing the third-party social costs and benefits of the defendant's activity. If negligence turned on whether the defendant engaged in excess activity, then courts would often have to pursue this difficult fact-finding exercise. Although third-party social costs and benefits are implicated, in theory, in safety-level disputes, the impact of this consideration can usually be ignored because it is likely to be small. For example, the marginal social cost or benefit that accrues to third-parties when the defendant exceeds the speed limit-faster delivery of goods carried by the defendant-is ordinarily so small that it does not figure in determinations of negligence.
It is far more difficult to ignore third-party costs and benefits, however, when the focus is on activity levels-whether allegedly excessive or insufficient activity. We need only to look to Gilo and Guttel's own examples to see why. The question of whether it was negligent for a commercial establishment to fail to provide a bartender at a party is circumscribed and manageable. The question whether to impose liability for failing to admit more patrons to the party, however, implicates issues regarding modes of recreational enjoyment and social intercourse that tort law is ill-equipped to resolve. The same would be true of the authors' fire sprinkler example, which would require an assessment of the educational costs and benefits of having larger classrooms that could hold more students, in order to determine whether the defendant should be held liable under IAL for the failure to have used larger classrooms. As was demonstrated decades ago by James Henderson in Expanding the Negligence Concept: Retreat from the Rule of Law, case-by-case adjudication is not well suited to the determination of "polycentric" questions such as these.
Gilo and Guttel imply that courts can avoid these problems by focusing on the set of cases most amenable to the application of IAL. These are cases in which, among other things, the cost of a precaution at a higher activity level exceeds the private benefit derived from increasing the activity level, but that cost is lower than the social gain from the additional activity and the diminished risk of harm. These cases, however, will not come to the courts in pre-identified form. Merely suggesting that the cost of obtaining relevant evidence "might be prohibitive" and that there should be "evaluation of activity levels" where this is "feasible" is not enough, because the parties will attempt to raise these questions at trial whenever the law permits them to do so.
Consequently, courts will often have to deal with factual disputes over whether the defendant's activity level was sufficient. Parties will introduce evidence of the social benefits associated with the defendant's allegedly insufficient current activity level and the higher, allegedly sufficient level. Judges and juries will then have to resolve polycentric factual disputes over the net social costs and benefits of different possible activity levels, and these disputes will pose even greater problems than questions of optimal social activity levels where the parties do not dispute the underlying facts.
Finally, Gilo and Guttel neglect another important question: whether the decision to impose IAL poses a question for the court or for the jury. The jury decides the question of liability in a standard action for negligence. In contrast, the court resolves the question in a strict liability action, although it may submit certain factual questions to the jury. This division of responsibility has significant implications, because in most instances, plaintiffs will allege other theories of liability in addition to IAL, such as conventional negligence. If the question of whether to impose IAL is up to the court, then the jury will likely hear evidence regarding activity levels that will be irrelevant to the negligence question it must decide. On the other hand, if the jury must decide both negligence and IAL, then the court must instruct the jury to find the defendant liable if it was negligent, and also liable if it was not negligent but has engaged in an insufficient amount of the activity otherwise conducted using reasonable care. To the typical jury, this sort of instruction will seem confusing, counterintuitive, and bizarre.
All things considered, then, IAL is a prescription for complicating tort litigation without any evidence, let alone assurance, that the advantages of IAL will outweigh its significant disadvantages.
Kenneth S. Abraham is the David and Mary Harrison Distinguished Professor of Law, University of Virginia School of Law.
Suggested citation: Kenneth S. Abraham, Response, Insufficient Analysis of Insufficient Activity, 108 Mich. L. Rev. First Impressions 24 (2009), http://www.michiganlawreview.org/assets/fi/108/abraham.pdf.