I have benefited from the discussions which took place when previous versions of this paper were presented to a faculty seminar at Cornell Law School and to a session of the Law and Policy Workshop at the Faculty of Law, McGill University. I am also grateful to Paul Craig and Frank Buckley for helpful comments on earlier drafts, and to Peter Benson and Ernest Weinrib for many stimulating discussions of tort law.

1. See e.g. Posner, Richard, “A Theory of Negligence” 1972), 1 J. Leg. Studies 29; The Economic Analysis of Law(2nd ed., (Boston: Little, Brown,1977).Google Scholar

2. Dworkin, Ronald, Law’s Empire(Cambridge, Mass: Belknap Press,1986). ch. 8.Google Scholar

3. See e.g. Weinrib, Ernest, “Toward a Moral Theory of Negligence Law” 1983)2 Law and Phil 37;CrossRefGoogle Scholar “The Insurance Justification and Private Law” (1985),14.J. Leg. Studies 681.Google Scholar

4. See e.g. Calabresi, Guido, and Hirschoff, Jon T. “Toward a Test for Strict Liability in Torts” (1972)81 Yale L.J. 1055;In his later work, however, Calabresi is probably best regarded as not defending a true theory of strict liability. See n. 36, infra.CrossRefGoogle Scholar

5. See particularly Epstein, Richard, “A Theory of Strict Liability” 1973)2 J. Leg. Studies 151;(hereafter “Strict Liability”);CrossRefGoogle Scholar “Defenses and Subsequent Pleas in a System of Strict Liability”3 J. Leg. Studies 165(hereafter "Defenses and Subsequent Pleas").CrossRefGoogle Scholar These two papers have been reprinted together in Epstein,A Theory of Strict Liability: Toward a Reformulation of Tort Law(San Francisco:Cato Institute,)1980.All references in the present article are, however, to the original papers.Google Scholar

6. Nozick, Robert,Anarchy, State, and Utopia(New York:Basic Books,)1974See e.g. the reference to unintentional and accidental boundary-crossings at71.Google Scholar

7. This understanding has been adopted not only by Posner, who defends negligence on grounds of economic efficiency, but also by theorists such as Dworkin and Weinrib, who put forward justifications of the fault-based approach which are premised on theories of individual moral right. See references in notes 1-3,supra.

8. (1910),109 Minn.456, 124 N.W. 221 (S.C.Minn.).

9. Epstein, ,“Strict Liability,”supran. 5 at 158–60.Google Scholar

10. Epstein, refers explicitly to tortious acts as “forced takings” in “Causation and Corrective Justice: A Reply to Two Critics,”(1979),8 J. Leg. Studies 417 at 501.CrossRefGoogle Scholar In his recent book Takings: Private Property and the Power of Eminent Domain (Cambridge, Mass.:Harvard University Press,1985)he emphasizes at 37–41 the similarity, from the point of view of tort law, between conversion and destruction; at 74 he says that “torts themselves are a subclass of takings”.Google Scholar

11. Holmes, Oliver Wendell,The Common Law,Howe, Mark DeWolfeed.(Boston:Little, Brown,,1963)at 76.Google Scholar

12. In U.S. v. Carroll Towing Co.(1947), F.159(2d) 169 (2d Circuit), Hand J., Learned, said that a person is negligent if the cost of taking adequate precautions is less than what the cost of the injury would be, were it to occur, multiplied by the probability of its occurrenceGoogle Scholar

13. Epstein, Richard,“Intentional Harms” (1975),4 J. Leg. Studies 391 at398(emphasis added). There are also many references throughout Epstein’s work on tort law to the negligence standard as being defined by the Learned Hand test:CrossRefGoogle Scholar see, for example,“Strict Liability”supran. 5 at158–59;Google Scholar Takings,supran.10 at40–41.Google Scholar

14. Epstein, , “Causation and Corrective Justice,”supran. 10 at500–01.Google Scholar

15. Cf. Weinrib, , “Toward a Moral Theory of Negligence Law,”supran. 3 at60–62.Google Scholar

16. Cf. Honoré, A.M, “Ownership,” in Guest, A.G.ed.,Oxford Essays in Jurisprudence(Oxford:University Press,1961)107 at 129. It is true that Epstein himself often speaks in terms of owning one’s own person or body, but when anything might turn on the point he tends to be more circumspect. Thus, in “Causation and Corrective Justice,” supra n. 10 at 500, he first says that “it by no means contorts the English language to say that each person owns his own body,” but then goes on in a footnote to state that “[if] the ownership language does seem artificial, the language of ‘personal integrity’ can be substituted without change of effect.”Google Scholar

17. Epstein, ,“Strict Liability,”supran. 5 at 158;Google Scholar cf.“Intentional Harms,”supran. 13 at398.Google ScholarPubMed

18. See e.g. Nozick, supra n. 6 at 28-35.

19. Holmes, , The Common Law,supran. 11 at 76;Google Scholar cf.Englard, Izhak“The System Builders; A Critical Appraisal of Modern American Tort Theory,”(1980),9 J. Leg. Studies27 at 61–62.CrossRefGoogle Scholar

20. Cf. Epstein, ,“Strict Liability,”supran. 5 at166;Google Scholar Weinrib, “Toward a Moral Theory of Negligence Law,”supra n. 3 at58–59.Google Scholar

21. Holmes, , The Common Law, supran. 11 at68.Google Scholar

22. “Intentional Harms”, supra n. 13 at 398; cf. Epstein, ,“Strict Liability,”supran. 5 at159.Google Scholar This aspect of the argument for strict liability was noticed by Baron Bramwell in Bamford v. Turnley (1862), 3 B. & S. 66 at 85, 122 E.R. 27 at 33 (Ex. Ch.): “It is for the public benefit that trains should run, but not unless they pay their expences. If one of those expences is the burning down of a wood of such value that the railway owners would not run the train and burn down the wood if it were their own, neither is it for the public benefit they should if the wood is not their own. If, though the wood were their own, they would still find it compensated them to run trains at the cost of burning the wood, then they obviously ought to compensate the owner of such wood, not being themselves, if they bum it down in making their gains.” On the basis of considerations of justice Bramwell thus argues, in effect, for Pareto-superiority over a Kaldor-Hicks conception of economic efficiency. (Notice the similarity to Epstein’s argument concening what the hypothetical consequences of a defendant's actions for himself would have been had he, and not the plaintiff, owned the property affected). But as Ronald Coase demonstrates in “The Problem of Social Cost” (1960), 3 J. of L. & Econ. 1 at 30–34 – by means, interestingly enough, of a discussion of the same example employed by Bramwell of a railway burning down neighbouring woods - the economically efficient result will not necessarily be achieved by always making the railway liable. The criticism which will be offered here of the claim that it is necessarily just always to make the railway liable parallels, up to a point, Coase’s criticism of the Pigouvian claim that that result is necessarily always efficient. See further n. 34, infra.

23. See e.g. the quote in the text at n. 22, supra.

24. See e.g. Epstein, ,“Defenses and Subsequent Pleas”supran. 5 at169,Google Scholarwhere he rejects private necessity and compulsion by a third party as legitimate defences in tort: “The only proper question for tort law is whether the plaintiff or the defendant will be required to bear the loss. The argument here is only that it is fairer to require the defendant to bear the loss because he had the hard choice of harming or being harmed when, given what is alleged, the plaintiff had no choice at all.”

25. See Epstein, ,“Automobile No-Fault Plans: A Second Look at First Principles”(1980)3 Creighton L. Rev. 769 at 775:Google Scholar“Once, moreover, assumption of risk is confined to narrower grounds, then it is difficult to argue that the plaintiff upon the highway is never entitled to the prima facie protection that strict liability provides to the landowner. Thus the plaintiff who is injured while sitting in a parked car, or while waiting for the light to change, or while proceeding through an intersection under the protection of a green light, surely has done nothing ‘wrong’ if struck by some other driver. If the plaintiff owning land is entitled to protection on strict liability principles where (as is almost always the case) his conduct is purely passive, then the same principles should apply as well to passive plaintiffs in the highway case”.

26. Holmes, , The Common Law, supran. 11 at75–78.Google Scholar

27. Holmes’ utilitarian argument is criticized by Weinrib in “Toward a Moral Theory of Negligence Law”, supra n. 3 at 42–43.

28. Holmes, The Common Law, supra n. 11 at 76.

29. Epstein has a number of techniques for dealing with collision cases; these will be discussed at length in section 4 below

30. See Bamford v. Turnley, supra n. 22.

31. Fletcher v. Rylands (1865), 3 H. & C. 774 (Ex.) at 790, per Bramwell B. (dissenting); the majority decision was reversed in (1866), L.R. 1 Ex. 265 (Ex. Ch.); affd sub nam. Rylands v. Fletcher (1868), L.R. 3 H.L. 330.

32. (1616), Hob. 134.

33. Holmes, , The Common Law, supran. 11 at68.Google Scholar

34. The point being made here was also made by Coase in the course of his critique of the Pigouvian economic tradition of dealing with externalities by distinguishing between the “private” and the “social” product of a single party whose actions were regarded as the cause of a given harm. See Coase’s discussion in “The Problem of Social Cost”, supra n. 22 at 12- 13, of the case of Bryant v. Lefever (1878-79), 4 C.P.D. 172 (C.A.), in which the plaintiffs chimney smoked whenever he lit it because the defendant had build a wall nearby that kept the air from circulating freely: “Who caused the smoke nuisance? ... [It| was caused both by the man who built the wall and by the man who lit the fires.... Eliminate the wall or the fires and the smoke nuisance would disappear.... If we are to discuss the problem in terms of causation, both parties cause the damage”. Coase’s own approach to the economic problem is a global one, requiring the policy-maker “to compare the total product yielded by alternative social arrangements” (supra n. 22 at 43). Epstein criticizes Coase on the grounds that he describes harmful interactions “by the use of sentences that differentiate between the role of the subject of [a proposition] and the role of the object” (“Strict Liability”, supra n. 5 at 165).Why this grammatical distinction should have any bearing on the analysis of causation is not made clear. It would appear to be nothing more than a metaphor which itself calls for explanation: see further the text accompanying n. 42, infra.

35. Wright, Richard, “Causation in Tort Law”(1985),Cal. L. Rev. 1737.CrossRefGoogle Scholar

36. This conclusion should apply, of course, to a general theory of strict liability whose justification is framed in economic terms as well as to one which is defended on grounds of individual moral right. It is interesting in this regard to consider the development of Guido Calabresi's thought about liability in tort. In an early article, “Some Thoughts on Risk Distribution and the Law of Torts”(1961), 70Yale L.J.. 499 Google Scholar,Calabresi argued at 533 that "'tort' costs should be borne by the activity which causes them ..." (emphasis added), where the main justification for the resulting general theory of strict liability was a Pigouvian approach to the allocation of resources:“The function of prices is to reflect the actual costs of competing goods, and thus to enable the buyer to cast an informed vote in making his purchases”(id. at 502). In subsequent work, however, Calabresi acknowledged that “[t]here is no formula for allocating the cost of an accident among the activities involved”, and, in particular, that this problem could not be solved on causal grounds. See “The Decision for Accidents: An Approach to Nonfault Allocation of Costs” (1965), 78 Harv. L. Rev. 713 at 725. He also began to recast the concept of allocation of resources in terms of what he eventually came to call general or market deterrence, which “would require allocation of accident costs to those acts or activities ... which could avoid the accident costs most cheaply”.Google Scholar See The Costs of Accidents(New Haven: Yale University Press,1970)at 135. The result is that Calabresi can no longer be regarded as defending a general theory of strict liability, i.e. a theory which would require a given loss to be borne by the actor or enterprise which had caused it. More recently, inGoogle ScholarPubMed “Concerning Cause and the Law of Torts”()1975),43U. Chic. L. Rev.69–85,Calabresi has said that a requirement of causation in the but for sense (as opposed to a mere probabilistic linkage between the type of activity and the type of injury in question) is not even necessary in a system of general deterrence.Google Scholar

37. See, for example, Holmes, , The Common Law supra n. 11 at 81ff.;Google Scholar Winfield, P.H.“The Myth of Absolute Liability”(1926),42L.Q.R.. 37;Google Scholar Milsom, S.F.C. Historical Foundations of the Common Law(2nd ed.,London:Butterworths,(1981), at296–300.Google Scholar

38. In American law, unlike Anglo-Canadian law, liability for defective products is, in some categories of cases, strict in this sense. See generally Prosser and Keeton on the Law of Torts(5th ed.,St. Paul, Minn.: West Publising Co.,1984)ch. 17.Google Scholar

39. Rylands v. Fletcher, supra n. 31.

40. Epstein, , Takings, supran. 10 at48.Google Scholar

41. In a recent article Epstein seems to dispute this. The plaintiff in the well-known English case of Bolton v. Stone, [1951] A.C. 850 (H.L.), which will be discussed in a different context in section 5 of this paper, was struck by a cricket ball while standing in a nearby road. In, “The Temporal Dimension in Tort Law”(1986), 70U. Chic. L. Rev..1175,Google ScholarEpstein says, at 1190, that under a theory of strict liability “[i]t is necessary only to explore the interval between the time the ball left the bat and the time it hit the plaintiff....” Even in terms of his own theory, however, this is quite misleading. As will be further discussed in section 3, Epstein maintains that because of the “priority” of the paradigm of dangerous conditions over that of force, an application of force by A to a dangerous condition created by B which results in injury to B will give A a good “causal” defence. See “Defenses and Subsequent Pleas”, supra n. 5 at 175–76. It would therefore be necessary to inquire into whether the plaintiff in Bolton v. Stone had created a condition that was dangerous (for herself) by standing where she did when she was hit by the cricket ball. In terms of Epstein’s own categorization of dangerous conditions, this would be a potential instance of placing a thing not dangerous in itself in a dangerous position. (Note that it would be quite arbitrary to distinguish, for the purposes of determining which dangerous conditions give rise to defences, between those which are dangerous for others and those which are dangerous for the plaintiff alone). As I argue in n. 65, infra, and the accompanying text, this amounts to a fault-oriented evaluation of the plaintiffs conduct which it would be inconsistent not to apply to the behaviour of the defendant as well.

42. Epstein, , “Defenses and Subsequent Pleas,” supran. 5 at168.Google Scholar

43. Epstein, , “Automobile No-Fault Plans,” supran.25.Google Scholar

44. Epstein, , “Strict Liability,” supran. 5 at160–63.Google Scholar

45. Id.at 165ff.

46. Epstein, , “Defenses and Subsequent Pleas,” supran. 5 at 168, n. 10.Google Scholar

47. Epstein, , “Intentional Harms,” supran. 13 at432.Google Scholar

48. Epstein, , “Strict Liability,” supran. 5 at177–19.Google Scholar

49. Id.at 178.

50. Id.at 169.

51. Epstein, , “Defenses and Subsequent Pleas,” supran. 5 at168.Google Scholar

52. Epstein, , “Intentional Harms,” supran. 13 at402.Google Scholar

53. Id.at 410–11

54. “Defenses and Subsequent Pleas”, supra n. 5 at 174-75.

55. “Strict Liability”, supra n. 5 at 180.

56. See Englard, I., “Can Strict Liability be Generalized?,” (1982,2 Oxford J. Leg. Studies 245245 for a more general critique of Epstein’s theory of causation. Richard Wright, supra n. 35 at 1750–58, criticizes Epstein’s account of causation as really being a theory of responsibility. I shall be arguing in support of a similar conclusion here.CrossRefGoogle Scholar

57. I think that he relies on it at a number of different points, one of which is described in n. 71, infra. 58.

58. Epstein, , “Strict Liability,” supran. 5 at160.Google Scholar

59. Id.at 179.

60. Seavey, Warren,“Negligence - Subjective or Objective” 1927)41 Harv. L. Rev.1 at5–7.Google Scholar

61. Epstein says, in the passage quoted in the text at n. 59, supra, that J’s act was not dangerous as described because it did not create a store of releasable energy, but that is only one of the categories of dangerous acts which he describes; the act of parking the car, if it could be said to be dangerous, would presumably fall into the category of a thing not dangerous in itself which has been placed in a dangerous position.

62. The non-empirical, normative dimension of Epstein’s analysis of causation stands out especially clearly in the case of dangerous conditions which involve rightsof way. The status of rights of way in his theory is discussed in section 4 below.

63. Epstein, ,“Strict Liability”supran. 5 at185.Google Scholar

64. Id.at 186–87.

65. The anomaly here becomes especially clear when it is recalled thatthe plaintiffs creation of a dangerous condition which is triggered by an application of force by thedefendant gives the latter a “causal” defence: Epstein, , “Defenses and Subsequent Pleas,”supran. 5 at175ȓ76.In light of what has been said about the meaning of the term “dangerous” in Epstein’s theory, the consequence is that the theory contains, despite Epstein’s claim to the contrary (“Strict Liability”, supra n. 5 at 181), what amounts to a defence of contributory fault; the defence is, moreover, a complete bar to recovery, even, it would seem, where the defendant's conduct was faulty in the same sense.Google Scholar

66. Epstein, , “Defenses and Subsequent Pleas”supran. 5 at174–75.Google Scholar

67. The intuitive judgment that if a defendant has been compelled by a third party to hit the plaintiff then the defendant should have an action over against the third party seems clearly correct, since the third party has committed what is presumably an intentional, and therefore in the circumstances an obviously wrongful, act. Elsewhere, Epstein tries to justify this conclusion about ultimate responsibility in what are presumably meant to be purely causal terms by saying that the paradigm of compulsion “permits us to link and differentiate the roles of the parties to the suit. A compelled B to hit C; B did not compel A to hit C. Hence it follows that, prima facie, B should prevail over A” (“Strict Liability”, supra n. 5 at 176). The argument is not sound, however, without the addition of further premises concerning the wrongfulness of (intentionally) compelling another to do something against his will, in which case it is not purely causal in character. If Epstein is not assuming the compulsion to be intentional then it is not at all clear to me what he could mean by “compel”, but whatever he means by it he needs a premise to the effect that compulsion in that sense is wrongful.

68. Epstein asserts without elaboration (“Defenses and Subsequent Pleas”, supra n. 5 at 183) that he has borrowed the expression from the Restatement of Torts 2d, s. 441, comment c, but while he might have borrowed the expression from there he has certainly not adopted its sense. Comment c distinguishes between different sorts of intervening forces, which are forces thatoperate to produce harm after the act of the person whose conduct is being evaluated has taken place.A dependent force “is one which operates in response to or is a reaction to the stimulus of a situation for which the actor has made himself responsible by his negligent conduct”. An independent force “is one the operation of which is not stimulated by a situation created by the actor’s conduct”. Epstein’s examples make it quite clear that whatever he means by dependence, it has nothing to do with these definitions.

69. Epstein, ,“Defenses and Subsequent Pleas”supran. 5 at183.Google Scholar

70. Idat 179.

71. Dependence seems in fact to involve a kind of but–for test: the hitting would not have occurred but for the blocking, whereas the speeding would have taken place whether there was an act of blocking or not.

72. Epstein, ,“Strict Liability”supran. 5 at178. The paradigm of dangerous conditions is said, of course, to have priority over the paradigm of force, “so that if B blocks A’s right of way then prima facie A is entitled to prevail in any suit between them regardless of who is injured on impact”: “Defenses and Subsequent Pleas”, supra n. 5 at 176.Google Scholar

73. Cf. Englard, ,“Can Strict Liability Be Generalized?”supran. 56 at251.Google Scholar

74. Epstein, ,“Defenses and Subsequent Pleas”supran. 56 at195.Google Scholar

75. Id at 177, n. 38.

76. Expectations are important in the context of traffic regulation because they are often associated with conventions that provide solutions to coordination problems: it does not matter whether we drive on the right side of the road or the left, or stop at red lights or green, so long as we all do, and can expect one another to do, the same thing. For an interesting discussion of the possible application of coordination theory both to the foundations of law and to particular legal problems such as the resolution of tort actions arising out of traffic accidents see Gerald Postema “Coordination and Convention at the Foundations of Law”(1982)11J. Leg. Studies 165.Google Scholar

77. Epstein, ,“Defenses and Subsequent Pleas”supran. 5 at181.Google Scholar

78. Id. at 180.

79. Id. at 181.

80. Id. at 182, n. 49.

81. Epstein, “Intentional Harms”, supra n. 13 at 402: “The sequence of pleas is: (1) A hit B, (2) B hit A, (3) A intended to harm B. Given the causal involvement of both parties, the intention to harm decides the case”.

82. Epstein, ,“Defenses and Subsequent Pleas”supran. 5 at179–80.Google Scholar

83. Hart , H.L.A. and Honoræ, Tony Causation in the Law(2nd ed., (Oxford: Clarendon Press,1985)).CrossRefGoogle Scholar

84. Wright, ,“Causation in Tort Law”supran. 35 at174550.Neil MacCormick makes a similar point, although he seems to imply that Hart and Honore intended to expound a theory of responsibilityGoogle ScholarSee MacCormick, ,“The Obligation of Reparation”in hisLegal Right and Social Democracy: Essays in Legal and Political Philosophy (Oxford:Clarendon Press 1982)212, at 220.A similar misunderstanding of Hart and Honore by Englard (“The System Builders”, supra n. 19 at 57, n. 135) is criticized by Wright (supra n. 35 at 1739, n. 11).Google Scholar

85. See Fleming, John,The Law of Torts(6th ed.,Sydney:Law Book Co.,1983)at 242–44,250-51for an analysis of how “the inveterate predilection of the Common Law mind for assigning occurrences to a single responsible cause”- described by Fleming as a “hollow pretence”- led to seriously inconsistent reasoning and the promulgation of highly dubious doctrine in the area of contributory negligence.Google Scholarsee also Maclntyre, Malcolm“The Rationale of Last Clear Chance”(1940),53Harv. L. Rev. 1225 at1226;Google Scholar Maclntyre, “Last Clear Chance after 30 Years”(1955),3333 Can. Bar Rev. 257–59.Fleming makes an essentially similar point with respect to intervening causes in The Law of Torts at 192-93.Google Scholar

86. Epstein, ,Takings supran. 10 at 40.CrossRefGoogle Scholar

87. “Toward a Moral Theory of Negligence Law”supran. 3 at52–53,Weinrib argues that the Learned Hand test represents, on a rights-based view, the correct formulation of the standard of care in negligence law. His argument depends, however, on a counterfactual proposition, concerning what would have transpired had the plaintiff and the defendant been the same person, which resembles the counterfactual relied upon by Epstein in the argument criticized in section 2(iv) above. Weinrib's argument is vulnerable to objections similar to those presented there against Epstein.Google Scholar

88. [1951] A.C. 850 (H.L.). Epstein criticizes the decision in Bolton v. Stone in “Strict Liability”, supra n. 5 at 169-71. As was pointed out in n. 41, supra, he also discusses the case in a more recent article. In neither discussion, however, does he recognize that the House of Lords characterizes negligence in a way which is very different from the balancing approach that he seems especially concerned to reject.

89. Id. at 867.

90. See id. at 858–60 per Lord Porter, 861 per Lord Normand, 863 per Lord Oaksey, and 868-69 per Lord Radcliffe.

91. In the subsequent case of Overseas Tankship (U.K.) Ltd. v. Miller Steamship Co. Pty. Ltd. (The Wagon Mound No. 2), [1967] A.C. 617 (P.C.), Lord Reid said at 643-44 that the possibility of eliminating the risk could be taken into account so as to justify liability where the degree of risk would ordinarily have been too slight to have led to that result, but only in certain circumstances: a reasonable person would not neglect a “real” risk (i.e. one which is foreseeable but non-substantial) “if action to eliminate it presented no difficulty, involved no disadvantage and required no expense”. This is clearly not a balancing test. Lord Reidis simply saying that if the defendant can achieve his own ends by following either of two separate courses of action, where neither is more costly for him and only one is risky for others, then he should opt for the safer alternative. This is just good common sense. On the facts of The Wagon Mound No.2 the defendant’s servant had discharged bunkering oil from the defendant’s ship into Sydney Harbour. The plaintiffs ship was severely damaged when the oil caught fire while floating on the surface of the water, an occurrence which was held to be foreseeable but extremely improbable. LordReid stated, at 643, that the defendant’s action not only held no advantage for the defendant but was in fact contrary to its own interests:“Not only was it an offence to [discharge theoil], but also it involved considerable loss financially. If the ship’s engineer had thought about the matter there could have been no question of balancing the advantages and disadvantages. Fromevery point of view it was both his duty and his interest to stop the discharge immediately”

92. See U.S.v. Carroll Towing Co., supra n. 12.

93. I shall be presenting such an explication and defence in a forthcoming paper entitled “The Moral Foundations of Tort Law”. I have been much assisted in arriving at the conclusion that the negligence standard should be formulated in nonbalancing terms by discussions with Peter Benson.