The duty concept in negligence law: A comment

IMAGE TABLE 4

I. CONFLICTING CONCEPTIONS OF DUTY

As critics John Goldberg and Benjamin Zipursky see it, the draft version of the Restatement (Third) of Torts: General Principles (Discussion Draft) ("Discussion Draft")

that is the occasion for this Symposium has relegated the duty issue in negligence law to a relatively minor, nay-saying role.' More particularly, duty is not directly mentioned by the Reporter in Section 3, which provides that "[a]n actor is subject to liability for negligent conduct that is a legal cause of physical harm."2 And in Section 6, when the Reporter does get around to addressing the subject of duty, it is in arguably backhanded, no-duty terms:

Even if the defendant's negligent conduct is the legal cause of the plaintiffs harm, the plaintiff is not liable for that harm if the court determines that the defendant owes no duty to the plaintiff. Findings of no duty are unusual, and are based on judicial recognition of special problems of principle or policy that justify the withholding of liability.3

So reads the black-letter. To Goldberg and Zipursky, this treatment radically understates the key role that duty plays as a determinative element in many negligence cases. In Goldberg and Zipursky's view, duty is a doctrinal tool for providing affirmative content to the range of accidental harm scenarios in which causally-- related unreasonable conduct leads to liability.4 By contrast, as they view Section 6, duty serves simply as a limiting device that sometimes kicks in to defeat liability in cases of unreasonable conduct that causes accidental harm.5

More particularly, Goldberg and Zipursky argue that duty deserves pride of position along with injury, breach, and causation in a comprehensive, four-element framework of negligence laws And they proceed to enunciate a "primary sense" of duty as "asking the question of whether defendant was obligated to be vigilant of the type of harm suffered by the plaintiff."7 This primary sense of duty-which they also refer to as an "obligation sense" of duty8-in turn needs further clarification, according to Goldberg and Zipursky, because courts sometimes use the duty concept in alternative ways that "invoke at least three other ideas."9 These instances of duty in its "alternative sense" are:

(1) whether there is an appropriate nexus between the defendant's breach and the duty the defendant owed to the plaintiff: in short, whether the breach of duty is a breach of duty owed to the plaintiff; (2) whether the plaintiffs case on breach is such that a court ought to rule that there is no breach as a matter of law; and (3) whether the broader policy implications of permitting plaintiff to recover justify creating an exemption from negligence liability.10

Nor is this the full extent of their critique. The Discussion Draft explicitly adopts a self-imposed limitation to cases of physical harm, as distinguished from claims for emotional distress and economic loss; and in addition, it eschews consideration of wellestablished categories of negligent harm that raise special issues, including professional responsibility (medical malpractice is a prominent example) and landowners liability. These limitations on scope, which also for the most part can be brought under the umbrella of duty-related limits, similarly come under critical attack from Goldberg and Zipursky.11 To them, a Restatement of negligence, so limited, is arbitrarily truncated. By contrast, their conception of duty would extend across-the-board to "how the defendant was obligated to conduct himself, the type of harm he was obligated to guard against, and the class of persons to whom that obligation extends."12

Having come this far in sketching out conflicting doctrinal perspectives of the Discussion Draft and its present critics, it is time to pause and take a broader look at the lay of the land, before moving ahead to identify particular features of the landscape in a more discerning fashion. The threshold question is whether such attentiveness to doctrine is a meaningful enterprise. My answer is a highly qualified yes.

Goldberg and Zipursky are not oblivious to this question, and they offer a more forthright answer than I do, affirming the importance of doctrine. Near the end of their article they provide a "rejoinder to realists,"13 in which they take on the question of "whether anything of significance really hangs on the taxonomic and organizational choices that we have discussed."14 In my view, just as they overstate the position of objectors to this preoccupation with doctrinal concerns-asserting that legal realist critics will see their views as "mere pretension"15-so too do they exaggerate the case for taking conceptualism quite so seriously.

Let me be concrete. Goldberg and Zipursky offer the recent case of Lauer v. City of New York16 as their prime illustration of why it really matters that the duty issue is framed in their "primary sense" of obligation, rather than as a Section 6-type "exemption" from liability based on concerns about opening the floodgates of litigation.17 In Lauer, the court dismissed a claim against the city medical examiner for covering up his earlier mistake in attributing the death of a child to abuse rather than natural causes-a coverup which led the plaintiff parent to have allegedly suffered serious emotional distress from having been treated for months as the chief suspect under criminal investigation. 18

According to Goldberg and Zipursky, viewing the case in Section 6 "policy" terms rather than their "primary obligation" terms may have led to the dismissal of the case: "[W]e see no reason to doubt that the framing of the question bore on how it was resolved."19 No reason to doubt? Can one really believe that the majority somehow missed the point of plaintiffs suffering because of a rote adherence to doctrine, rather than simply giving the father's emotional distress less weight than the dissent attached to it? Why wouldn't one simply conclude that, rightly or wrongly, the majority simply struck a different balance between floodgates concerns and plaintiff-protective concerns, and as a consequence, decided that the concept of duty-which embraces both sets of considerations-weighed in favor of no liability in the case?

This skepticism on my part about conceptualism will be one theme in the comments that follow. At the same time, however, because I do think doctrine has some salience, I will offer my thoughts on why I view Goldberg and Zipursky's critique of the Discussion Draft as being on target, and yet part company with them on their own analytical approach to the duty issue. In short, I will engage in taking doctrine seriously, but nonetheless my discussion will remain grounded in the conviction that ultimately accident law leaves vast play for judicial discretion-discretion that, I will argue, does not inexorably dictate findings of duty and no duty, let alone fitting neatly into pigeonholes of "obligation" and "policy."

II. DUTY UNBOUND: BEYOND THE RESTATEMENT

Why does doctrine-and, in particular, analytical clarity about the role of duty-have some salience? Goldberg and Zipursky identify one reason in observing that at times courts conflate no-- duty findings with findings of no negligence as a matter of law.20 In fact, the Reporter makes precisely the same point in his comments to Section 6.21 A moment's reflection indicates why organizational clarity about the elements in a negligence case is important. Take the question of whether a baseball club should be responsible for injuries suffered by fans from being struck by batted balls. To begin with, courts have traditionally disposed of these claims on the basis of an assumed risk defense, namely, that a fan assumes the risk of being hit by a batted ball when attending a baseball game.22 But this is pure confusion, since a claim of affirmative defense is premised on a plaintiff having made out a prima facie case of negligence-why else would one need to interpose a defense? And surely a defendant is not prima facie unreasonable in staging a ball game because on rare occasion unfortunate fans are beaned by home runs or by foul balls. No one would argue, I assume, that analytical confusion of this sort is a matter of indifference. 23

Confusion aside, the question is whether dismissal of these claims is properly seen as based on no duty or no negligence as a matter of law. This question has to be addressed in light of the standard formulation that duty issues are for the court, and breach issues, so long as reasonable jurors might disagree, are for the jury. That allocation of decision making authority, in turn, rests on the premise that duty issues are categorical in nature; more specifically, they raise issues that are grounded in general policy considerations related to a category of harm that has occurred, while questions of breach are grounded in the particularities of the case at hand.

In the baseball situation, courts do not engage in a case-bycase analysis of what the cost of netting or Plexiglas screening of the entire spectator area would be as compared to the injury cost imposed on fans.24 Indeed, for the sake of argument, it may be that the annual injury cost in monetary terms exceeds the annualized cost of some form of screening. But the trump card-the critical variable in this situation-is an intangible social good: the open-air nature of the game and the pleasure traditionally derived by fans from this ambience, including a sense of connectedness with the players on the field. These are precisely the kinds of considerations, drawing on customary expectations and values, that go into a categorical determination of policy signaled by treating the issue as one of duty.

It would be possible, of course, to argue that under a Carroll Towing breach analysis the burden of adequate protection includes not just the pecuniary costs of screening but also the social costs of diminished experience for spectators.25 But that simply reinforces my point-doctrine does matter here, because it determines the focal point of the judge when deciding which issues should be left to the jury (the breadth of the jury's inquiry, in other words) and which decided as a matter of law. When these issues get muddled, as in recent cases involving land occupier liability for the violent acts of third parties,26 great uncertainty exists about how the next case down the road-involving a different constellation of costs of protective devices and risks to innocent land entrants-will be decided.

As Leon Green observed long ago, the allocation of decision making authority between judge and jury is a dominant theme in accident law, subject to judicial manipulation by classifying issues as duty rather than proximate cause, for example, or in the present instance, as no-duty or no breach as a matter of law (alternative strategies for taking a case away from the jury).27 At a minimum, organizational clarity about the doctrinal elements of a tort focuses attention on what is at stake, so that even if the judge retains broad discretion in the domain of duty, and the jury similarly has substantial room for play in more particularized case-sensitive dispositions, there is at least some common ground for structuring legal arguments and offering informed criticism.

The Discussion Draft, like all Restatements, by definition takes doctrine seriously. But Goldberg and Zipursky argue that it does not take duty seriously enough. On this score, I agree with their principal objections, namely, that conceiving of duty largely in restrictive, no-duty terms undervalues the heavy-lifting duty has done in expanding the universe of tort obligations, and understates the range of risk-imposing activities and risk-related types of harm falling within the ambit of negligence law. In my view, although noduty rules do play a critical role in creating categorical exceptions to liability for unreasonable conduct-the main thrust of Discussion Draft Section 6-they also serve as the conceptual pivot for elevating former no-duty (or limited duty) situations into full-blown (or more expansive) recognition of liability for unreasonable conduct.28 Duty is, in other words, a two-sided coin that can reinforce status quo limitations or serve dynamically expansive purposes. Moreover, to convey a comprehensive sense of these alternatively restrictive and expansive impulses, it is essential to address nonphysical as well as physical harm and to avoid arbitrary exclusion of traditionally important areas of injury-generating activity.29

Let me develop what I have in mind-that is, the pivotal role of duty-through some illustrative cases, mindful as well of my earlier caveat about taking doctrine too seriously. Consider first the landmark case of Tarasoff v. Regents of University of California, in which the California Supreme Court created an obligation on the part of therapists to warn a potential victim when a patient indicates to the therapist an intention to do serious harm to an identifiable third person.30 If we grant that the costs of taking some responsive action-warning the patient, her parents, the police or whatever-are minimal when compared with the risks of doing harm posed by the patient, then breach is no issue. Rather, the question is whether the therapist should have a threshold obligation to warn.

Three observations seem critical. First, the duty question does raise a distinct issue: If the defendant had been a bartender rather than a therapist, it seems safe to say that no court would have recognized an obligation to notify third parties, despite the minimal costs from a breach perspective in doing so. In other words, duty in an affirmative sense is a salient issue in the case, and in deciding what the limits of the obligation to warn (beyond the case) might be. The Discussion Draft is not entirely unmindful of this point; a comment to Section 6 discusses affirmative duties,31 and Section 16 addresses negligent failure to warn.32 But the combined effect of relegating affirmative duties to a comment and subsequently addressing the topic from a breach perspective (negligent failure to warn) is to understate their importance. After all, beyond Tarasoff, affirmative duties are the gateway to addressing the pervasive tort theme of misfeasance-nonfeasance-a theme that raises the foundational concern of how proactive the state should be in compelling private action.

Second, returning to Tarasoff, there is the other side of the coin. Perhaps it is stating the obvious, but it is critical to note that duty would have remained just as salient a factor if the case had been dismissed, in other words, if it had been registered as a no-- duty case. One way or the other, duty is the mechanism for assessing the importance of concerns about individual autonomy, confidentiality of therapist-patient relations, costs of therapeutic services, and public safety in the Tarasoff context. This is more than a mere make-weight role.

Having said this, however, is to raise the "realist" point that conceptual analysis takes one only so far. A Restatement that encapsulates the Tarasoff principle in black-letter law would not in itself address the question of what motivated the California Supreme Court. It would not offer real guidance to whether, in the next case, a therapist would owe a duty to the victim's child who happened to be present and injured when the patient acted out his violent impulses,33 or to an identified class of potential victims rather than an identified individual.34 Similarly, it would not tell us whether revelations to a family doctor, as distinguished from a bartender, come within the Tarasoff principle, or revelations from a suicidal patient rather than a homicidal patient who threatens others.35 Rather than belabor the point further, I would simply assert, once more, that doctrinal analysis is essentially static-an organizing tool but little more-unless it is attentive to the policy concerns that channel discretion in one direction or another. And even then, it frequently plays a more important role in organizing juristic thinking-clarifying analysis-than in determining outcomes.

A second illustration of the pivotal role of duty comes from the area of intrafamily tort liability, where traditionally the common law recognized a doctrine of parental immunity.36 In recent years, this immunity has been largely abolished, but it does not follow that parental status is now inconsequential. In New York, for example, the courts continue to recognize a limited area of no duty in claims characterized as "negligent supervision."37 Thus, in a line of cases involving situations such as an unsupervised young child being injured by scalding water in a bathtub, or falling off a playground slide, the New York courts have refused to engage in a costbenefit type analysis of whether a reasonably attentive parent could have prevented the injury.38 Once again, as this unwillingness suggests, these cases cannot be explained in breach terms; rather, duty is treated as a threshold inquiry before the question of reasonable parental conduct is even entertained. For policy reasons related to judicial regard for religious and ethnic diversity on the topic of child-raising practices, the court creates a zone in which it will not second-guess parental decisions on how much autonomy ought to be encouraged in children at an early age.

Not all courts would create as expansive a zone of autonomy as New York has adopted; some courts would, by contrast, recognize a "reasonable parent" approach.39 Other states recognize still other limitations.40 But the critical point is that in these cases the duty issue is the paramount consideration. Whether parents should be responsible in tort, for example, for injuries suffered by their children from parental inattentiveness while crossing a street together, or parental smoking in their presence, seems first and foremost a duty issue, whether a New York approach is followed or some other.

Here, too, however, formal conceptual analysis takes us nowhere beyond pigeon-holing particular cases or identifying doctrinal sub-categories. The only way to make sense of these cases is to get beyond formal analysis to questions of how much autonomy seems to be at stake in a given situation and how high a value ought to be accorded to parental prerogatives. Black-letter will not determine the hard cases, and indeed, can lead to dubious results if it masks the underlying policy considerations.

My final illustration of the central role played by duty in negligence law is the bystander emotional distress cases. Here, duty considerations take on a different cast from the affirmative action and parental supervision scenarios discussed above. In Tarasoff and in the parental supervision cases, the linking element that makes the duty issue so central is status: In Tarasoff, it is the defendant's status as a therapist and his relational connection to a patient; in the parental supervision cases, it is the parent-child status relationship. Similarly, in other areas where duty issues are of paramount importance-areas that cannot be discussed in detail here-such as governmental liability for negligent acts and land occupier liability, status is a critical variable in both expanding and restricting responsibility for negligent acts.41 Suffice to say, governmental liability is unintelligible without reference to immunities for discretionary functions that have been spelled out in duty terms.42 Similarly, land occupier liability has traditionally developed a texture through duty characterizations; in land occupier cases, featuring an elaborate set of categories for entrants onto land and uses of residential and commercial premises.43

By contrast, the emotional distress cases raise duty issues because of the type of harm occurring: fright, loss of companionship, eyewitnessing a horrific injury (my present illustration). Here, if status relationship is a factor, as it is in loss of companionship and bystander eyewitness cases, it is only secondarily so.44 The key question, also presented in economic loss cases, is whether there should be recovery at all for negligence causing this type of harm, as distinguished from physical injury.45 Notice that here, too, the case law assumes breach: a driver who has driven unreasonably, for example, but who causes the plaintiff "only" emotional trauma as an eyewitness to injury and no broken bones. Some modern courts, as in California, require a constellation of qualifying characteristics, including that the eyewitness to another person's injury be a close family member, that the injury was directly and proximately perceived, and that the perceived injury to the third person was serious or fatal.46 Other courts, following the lead of New York, tack on another element: that the eyewitness was, in fact, in the zone of danger.47 Whatever the view, these factors line up under the banner of duty.

In my view, the Discussion Draft is seriously underinclusive in leaving out these types of harms. Emotional distress and economic loss are in fact standard components of recovery for accidental harm when suffered as a consequence of physical injury; to read them out of a Restatement of basic principles of negligence when they occur as discrete injuries, rather than arising in combination with physical injury, seems unwarranted, even misleading.

More fundamentally, from a social welfare perspective, accidental harm to a physically injured individual more often than not imposes secondary costs of a relational character: costs to family members, an employer, perhaps others with more attenuated commercial and personal relations. Few if any of us live in isolation from such a network of relationships, and tort law acknowledges these "ripple effects" of physical harm under doctrinal heads-dutyrelated doctrinal categories-of emotional distress, economic loss, and loss of society (consortium). Acknowledgement of these harms does not necessarily translate into recovery, of course, but that is the essence of the matter: These secondary relational costs of physical injury, and the limited duties that they entail, are simply too central to the core of accident law to be left out of an enterprise aimed at spelling out general principles of negligence law. And this is without reference to "direct" claims for emotional distress and economic recovery that are at the cutting edge of current tort activity: AIDS and toxics-related fear-of-contraction claims and medical monitoring recovery efforts.48

III. SOME RESERVATIONS; CRITIQUING THE CRITICS

On these matters, presumably Goldberg and Zipursky would concur. Where I part company from their approach, apart from the apparent outcome-determinative impact they would assign to doctrinal analysis, is on the hyper-refined duty analysis that leads them to create categories of duty "in an alternative sense."49 In particular, their last category of policy-based "exemptions" from tort liability, which they differentiate from what they regard as proper duty analysis, seems to me a departure from the way courts think about duty.50 I previewed this disagreement in my discussion of Lauer, above, but let me amplify my thinking.

On occasion, courts are very candid about their concern that recognizing a type of claim will open the floodgates to excessive litigation, and this plays a key role in their determination that no duty will be recognized.51 Goldberg and Zipursky point to Strauss v. Belle Realty Co., in which the New York Court of Appeals denied the claim of a tenant injured in the common area of his apartment building by a fall in a darkened stairway, allegedly due to defendant Con Ed's responsibility for a city-wide blackout, as an illustration of this phenomenon.52 The court based its dismissal on the lack of contractual privity between plaintiff and Con Ed.53 Goldberg and Zipursky seem to think that some major gains would have accrued if the court had reached its decision fully cognizant that it was engaged in creating a policy-based "exemption," to again use their terminology, rather than considering the duty issue in their "primary" sense. I confess that this strikes me as deeply puzzling. Nowhere do they say that the Strauss court should have reached an opposite result; indeed, they expressly conclude that at times judicially-fashioned "exemptions" of this sort are perfectly appropriate.54

A principal concern of theirs is that the creation of such exemptions undermines "rule-of-law values," because they "do not engender predictability or stability. They undercut the capacity of the law to guide."55 In fact, quite the opposite is the case: Privity limitations such as Strauss adopts, for better or worse, create bright-line rules that offer about as much certainty as one finds in tort law.56 Consider, on the same score, the zone of danger test that New York has adopted in bystander emotional distress cases,57 or the limitation on loss of consortium recovery to spouses adopted by some courts;58 again these are bright-line rules clearly adopted in response to floodgates concerns.59 I am not for a moment expressing unequivocal agreement with these particular invocations of hardand-fast no-duty (or limited duty) rules; perhaps some offer greater predictability at too high a cost. But predictability is in fact their raison d'etre.

From a historical perspective, floodgates concerns have been around at least as long as the landmark case of Winterbottom v. Wright, decided more than 150 years ago.60 These floodgates concerns have never overwhelmed other duty considerations, which seems to be another source of Goldberg and Zipursky's anxiety.61 And indeed in some instances, such as the familiar narrative of products liability from Winterbottom to MacPherson v. Buick Motor Co., floodgates concerns have been overcome as other duty considerations became more compelling.62

But the most basic problem with creating a special policybased "exemption" category apart from other duty considerations is that such an effort suggests a compartmentalized mode of judicial thinking that simply does not square with real world decision making. Take Dillon v. Legg, for example, the California multifactor duty test, discussed earlier, that is widely used in bystander emotional distress cases.63 Goldberg and Zipursky claim that "what is at stake in these cases is not only (or even chiefly) the predictive floodgates question, but the legal issue of whether or not to suspend the nexus requirement so as to allow a particularly sympathetic class of plaintiffs to pursue negligence claims."64 But that is precisely the point. These cases don't reflect compartmentalized thinking in which floodgates "policy" is of a different order than the "sympathetic class of plaintiffs." If sympathetic victim considerations-or indeed foreseeability of distress-were the determinate factor in eyewitness cases, best-friend bystanders and non-present parents would be allowed to recover as well. The California court framed its duty rule in terms of status and situational considerations tempered by concerns about excessive liability.65 The New York court, which in Bovsun v. Sanpieri added a zone of danger requirement for bystander recovery66-a requirement which makes no sense logically when the claim is for distress related to viewing an injury to another-gives greater weight to floodgates considerations and lesser to victim sympathy.67 These differing perspectives are not a reflection of failure to think in a properly organized fashion; rather, they reflect different views of how the competing interests should be put together-differences that get resolved under the umbrella of duty.

IV. CONCLUDING THOUGHTS

There is a more general underlying point here. "Policy" consideration, if that is the label to be attached to a floodgates concern, is a fairly common theme in duty cases, but only rarely determinative in itself. Thus, in governmental liability cases like Riss v. City of New York, in which the police were found not to owe a duty to respond to a request for protection from threats of violence, the court's reluctance to establish a tort obligation was in part motivated by floodgates considerations.68 But only in part, for it also turned on a mix of institutional competence and separation of powers concerns about judicial second-guessing of exercises of discretion by the police.69 In land occupier liability cases, involving issues of whether to afford protection to victims of third party violence,70 or to abolish the traditional status limitations on recovery for premises-related injuries to entrants,71 the duty inquiry again turns partly on estimations of increased litigation if a legal obligation is recognized. But again only partly, for considerations of foreseeability, fairness, compensation, and deterrence, among others, also weigh in.

And then, of course, there are a wide variety of duty of affirmative action cases involving, for example, failure to assist a needy victim, in which a floodgates concern plays virtually no role-- but, in which concerns about the appropriate limits on individual autonomy and the claims of communal responsibility are often in tension.72 To regard these non-floodgates considerations as something other than "policy" questions, because they involve resolution of basic philosophical dilemmas or issues of comparative institutional competence, rather than concerns about excessive litigation, certainly cannot be based on a traditional understanding of the role of courts in fashioning legal obligations in tort cases. Traditionally, courts have addressed all of these considerations under the guise of duty, often in tandem, without resort to categorical refinements reflecting a fixed hierarchy of interests. One can certainly take exception to this polycentrism from a normative perspective, but it is nonetheless the reality of duty jurisprudence.

FOOTNOTE

1. John C P. Goldberg and Benjamin C. Zipursky, Restatement (Third) and the Place of Duty and Negligence Law, 54 VAND. L. REV. 657 (2001). Although the authors primarily critique the Restatement draft prepared by Professor Gary Schwartz, see RESTATEMENT (THIRD) OF TORTS: GENERAL PRINCIPLES (Discussion Draft, Apr. 5, 1999) [hereinafter Discussion Draft], they also briefly address the separate sections later drafted by Professor Harvey Perlman when he was appointed co-Reporter. See RESTATEMENT (THIRD) OF TORTS: GENERAL PRINCIPLES (Preliminary Draft No. 2, May 10, 2000) [hereinafter Preliminary Draft No. 2]. I limit my comments to the Discussion Draft.

2. Discussion Draft, supra note 1, 3.

FOOTNOTE

3. Id. (sec) 6.

4. Goldberg & Zipursky, supra note 1, at 698-723.

5. Id. at 664-674.

6. Id. at 698.

7. Id. app. (sec) D cmt. a; see also id. at 716.

8. Id. at 699.

9. Id. at 709.

10. Id. at 698-99.

FOOTNOTE

11. Id. at 667-72.

12. Id. app. D cmt. a.

13. Id. at 732-36.

14. Id. at 732.

15. Id.

16. Lauer v. City of New York, 733 N.E.2d 184 (N.Y. 2000).

17. Goldberg & Zipursky, supra note 1, at 733.

FOOTNOTE

18. Lauer, 733 N.E.2d at 186.

19. Goldberg & Zipursky, supra note 1, at 734.

20. Id. at 703-10. In their typology, this is duty in an alternative sense, type two. As my comments will indicate, I agree that this judicial usage is a confusion of the meaning ordinarily given to the duty concept.

FOOTNOTE

21. See Discussion Draft, supra note 1, 6 cmt. h.

22. See, e.g., Davidoff v. Metro. Baseball Club, Inc., 463 N.E,2d 1219, 1220 (N.Y. 1984).

23. For a more general discussion of the confusions engendered by the assumed risk defense, see Stephen D. Sugarman, Assumption of Risk, 31 VAL. U. L. REV. 833 (1997).

24. This assumes that the defendant has taken advantage of the "safe harbor" requirement of a modest amount of screening behind the backstop. In that sense there is a limited duty requirement imposed on ball club operators. See Davidoff, 463 N.E.2d at 1220.

FOOTNOTE

25. The reference is to Judge Learned Hand's famous negligence calculus in United States v. Carroll Towing Co., 159 F.2d 169, 173 (2d Cir. 1947), in which he framed the issue in the context of the contributory negligence of a barge owner as "a function of three variables: (1) The probability that [injury will occur]; (2) the gravity of the resulting injury, if [it] does; and (3) the burden of adequate precautions."

26. See, e.g., Sharon P. v. Arman, Ltd., 989 P.2d 121 (Cal. 1999) (involving a plaintiff who was attacked in a commercial parking garage).

27. See LEON GREEN, JUDGE AND JURY (1930).

FOOTNOTE

28. A classic instance is the landmark case Rowland v. Christian, 443 P.2d 561, 564 (Cal. 1968), in which the California supreme court abolished the traditional duty limitations on landowners to entrants, and looking to a number of duty factors, including "the closeness of the connection between the defendant's conduct and the injury suffered.... the moral blame attached to the defendant's conduct .... the policy of preventing future harm, and ... prevalence [and availability] of insurance," decided that a general duty of due care under the circumstances extended to trespassers and licensees as well as invitees. See also Robert L. Rabin, Enabling Torts, 49 DEPAUL L. REV. 435 (1999) (discussing a variety of situations in which courts have extended duty obligations to those who facilitate risk-related conduct by others that results in harm to a third-party victim).

FOOTNOTE

29. This conviction is spelled out in MARC A. FRANKLIN & ROBERT L. RABIN, TORT LAW AND ALTERNATIVES: CASES AND MATERIALS (7th ed. forthcoming 2001), in which we devote consecutive chapters (chapters three and four) to duty requirements-physical injuries and non-physical harms, respectively; and in the third chapter address discrete categories such as landowner and occupier liability, governmental entity liability, and intrafamily duties.

30. Tarasoff v. Regents of Univ. of Cal., 551 P.2d 334 (Cal. 1976).

FOOTNOTE

31. Discussion Draft, supra note 1, at 6 cmt. b.

32. Id. 16.

33. See, e.g., Hedlund v. Superior Court of Orange County, 669 P.2d 41, 43 (Cal. 1983).

34. See, e.g., Thompson v. City of Alameda, 614 P.2d 728, 736-37 (Cal. 1980).

35. See, e.g., Bellah v. Greenson, 146 Cal.Rptr. 535, 539-40 (Cal. Ct. App. 1978).

FOOTNOTE

36. See generally Broadbent v. Broadbent, 907 P.2d 43 (Ariz. 1995) (discussing the doctrine of parental immunity).

37. See, e.g., Holodook v. Spencer, 324 N.E.2d 338 (N.Y. 1974).

FOOTNOTE

38. See, e.g., Zikely v. Zikely, 470 N.Y.S.2d 33 (N.Y. App. Div. 1983), affd, 467 N.E.2d 892 (N.Y. 1984). In each of these cases, the court could have concluded that a jury entertaining the breach issue might have found that the benefits of greater parental attentiveness exceeded the risks of serious injury to the child from inattentiveness.

39. See, e.g., Gibson v. Gibson, 479 P.2d 648, 653 (Cal. 1971).

40. See generally Broadbent, 907 P.2d at 46-50 (adopting a "reasonable parent" standard in a case involving serious injuries to a child from a near-drowning incident, allegedly due to parental careless supervision, after carefully analyzing the policy reasons leading other courts to recognize a more limited duty).

FOOTNOTE

41. See FRANKLIN & RABIN, supra note 29, ch.3.

42. See, e.g., Friedman v. New York, 493 N.E.2d 893, 898-901 (N.Y. 1986). 43. See DAN B. DOBBS, THE LAw OF TORTS 232-237 (2000).

44. By this I mean that status considerations often still count as one factor in shaping the limits of the legal obligation, see, e.g., Borer v. Am. Airlines, 563 P.2d 858 (Cal. 1977), and the bystander emotional distress cases next to be discussed in the text, treat close familial relationship as a necessary condition to recovery. See, e.g., Dillon v. Legg, 441 P.2d 912 (Cal. 1968). But the principal concern in these categories of cases is not the status relationship but a range of concerns centering on the particular type of loss itself. In emotional distress cases, the intangible character of the loss raises concerns about fraud, valuation, and floodgates, and in economic loss cases, the character of the harm itself raises concerns over establishing meaningful boundaries on recovery, and in some instances on whether contract rather than tort is the appropriate paradigm for viewing the case.

FOOTNOTE

45. For discussion in the context of intangible loss, see, for example, Falzone v. Busch, 214 A.2d 12 (N.J. 1965), recognizing a claim of fear for safety without associated physical impact, but also discussing the reasons why no such claim had been allowed in the past. See also Tobin v. Grossman, 249 N.E.2d 419 (N.Y. 1969) (rejecting an obligation to compensate for bystander emotional distress), overruled by Bovsun v. Sanperi, 461 N.E.2d 843 (N.Y. 1984)).

46. See, e.g., Dillon, 441 P.2d 912.

47. See, e.g., Bovsun, 461 N.E.2d at 847 n.6, 848.

FOOTNOTE

48. On AIDS-related claims, see, for example, KA.C. v. Benson, 527 N.W.2d 553 (Minn. 1995). On toxic exposure-related cancerphobia and medical monitoring claims, see, for example, Metro-North Commuter R.R. v. Buckley, 521 U.S. 424 (1997).

49. See Goldberg & Zipursky, supra note 1, at 709-23.

50. See id. at 749. As mentioned above, I have no disagreement with Goldberg and Zipursky's second category, in which courts conflate duty with breach as a matter of law. As to their initial category, which they refer to as a "nexus requirement," see id. at 709-12, I fail to understand in what way it is duty in "an alternative sense." They define the category as follows: "there must be a certain 'nexus' between the breach and the duty, or the breach must be a breach of a duty to the plaintiff. The phrase 'no duty' is sometimes used to convey that the required nexus between duty and breach is missing." Id. at 712. They then offer a number of examples, including, as one might expect, the renowned Palsgraf v. Long Island R.R. Co., 162 N.E. 99 (N.Y. 1928). Goldberg & Zipursky, supra note 1, at 712-14.

In Palsgraf and the other cases they discuss, courts essentially seem to be concluding that there is no duty because there is no "primary obligation" in Goldberg and Zipursky's sense. That is, the plaintiff was not within "the class of persons afforded protection by the obligation" (quoting Ward v. K-Mart Corporation, 554 N.E.2d 223, 226 (Ill. 1990) to develop their own expanded definition of duty in a primary sense). Id. at 701. How is this an "alternative sense" of the concept, rather than a case in which duty in the "primary sense" is simply not made out? The distinction eludes me.

51. See, e.g., Metro-North, 521 U.S. at 433-34 (discussing the reluctance of courts to recognize a duty in emotional distress cancerphobia cases because of floodgates concerns).

FOOTNOTE

52. Strauss v. Belle Realty Co., 482 N.E.2d 34 (N.Y. 1985).

53. Id. at 36-38.

54. Goldberg & Zipursky, supra note 1, at 720-26. I put "exemptions" in quotes because surely this is not a discrete category, apart from duty, recognized in the case law.

55. Id. at 726.

FOOTNOTE

56. Along parallel lines, in their "rejoinder to realists," Goldberg and Zipursky argue that a "policy-intensive view"-presumably one recognizing bright-line rules based on excessive litigation concerns such as in Strauss-threatens the legitimacy of common law courts in the eyes of state legislatures and the federal judiciary. Id. at 736-39. What an odd argument this is to make in an era of wholesale cutback on plaintiffs' rights by legislators and preemption-minded judges precisely out of concern about a litigation explosion!

57. Bovsun v. Sanperi, 461 N.E.2d 843, 848 (N.Y. 1984).

58. Borer v. Am. Airlines, 563 P.2d 858, 866 (Cal. 1977).

59. Consider also the rejection of recovery in emotional distress cancerphobia cases. See, e.g., Metro-North Commuter R.R. v. Buckley, 521 U.S. 424, 430 (1997).

60. Winterbottom v. Wright, 152 Eng. Rep. 402 (1842).

FOOTNOTE

61. In a thoroughly confusing passage, Goldberg and Zipursky argue that:

Courts often do not engage in all-things-considered assessments of institutional and policy considerations when analyzing the duty question. Instead, at different times, they often raise under the heading of duty the obligation, nexus, and breach-as-a-matter-of-law inquiries we have discussed above. These are the contours of duty doctrine.

FOOTNOTE

Goldberg & Zipursky, supra note 1, at 724. If this means that most duty issues don't involve floodgates-type considerations-in other words, if it is a quantitative assertion-I can't imagine who would disagree. On the other hand, if this means that there is a move afoot to recast other types of duty issues in floodgates-type "policy" terms, I can't imagine what the basis is for the claim.

62. MacPherson v. Buick Motor Co., 111 N.E. 1050,1054 (N.Y. 1916). 63. Dillon v. Legg, 441 P.2d 912, 920 (Cal. 1968).

64. Gb & Ziourskv. supra note 1. at 714.

FOOTNOTE

65. A note of clarification. Analytically, Dillon-type cases do not raise floodgates concerns, but rather excessive liability concerns. Dillon-type cases would not involve much additional litigation; presumably a broader-ranging duty of bystander emotional distress or loss of consortium would entail more claims being consolidated with the basic claim for physical injury from which they were derived, not that the number of discrete cases would be greatly increased. By contrast, recognition of Strauss-type claims for city-wide physical injuries from a blackout, or a wide-ranging obligation to compensate cancerphobia or fear-of-AIDS cases, would result in a true

FOOTNOTE

floodgates concern-that is, a concern that the number of separate but like cases would be greatly increased.

FOOTNOTE

66. Bovsun v. Sanperi, 461 N.E.2d 843, 848 (N.Y. 1984).

67. Bovsun, of course, still suggests some measure of concern for a limited category of emotional distress victims. Compare id. (limiting recovery under the "zone of danger" rule to victims whose emotional distress is "serious and verfiable" and tied to the event by proximate cause), with Tobin v. Grossman, 249 N.E.2d 419, 424 (N.Y. 1969) (recognizing no duty to compensate for intangible loss).

68. Riss v. City of New York, 240 N.E.2d 860, 860-61 (N.Y. 1968). 69. Id. at 861.

70. See, e.g., Sharon P. v. Arman, Ltd., 989 P.2d 121 (Cal. 1999). 71. See, e.g., Heins v. Webster County, 552 N.W.2d 51 (Neb. 1996).

FOOTNOTE

72. Compare Harper v. Herman, 499 N.W.2d 472, 474 (Minn. 1993) (finding in favor of boat owner against victim of diving injury who claimed affirmative duty to warn of dangers beneath the water), with Farwell v. Keaton, 240 N.W.2d 217, 222 (Mich. 1976) (reinstating verdict in favor of injured "companion in a common undertaking" for failure of his friend to render assistance).

AUTHOR_AFFILIATION

Robert L. Rabin*

AUTHOR_AFFILIATION

A. Calder Mackay Professor of Law, Stanford Law School.