Electronic copy available at: http://ssrn.com/abstract=2014874 - PDF document

Electronic copy available at: http://ssrn.com/abstract=2014874 Electronic copy available at: http://ssrn.com/abstract=2014874 1 M ALIGNED MISALIGNMENTS Israel Gilead and Michael D. Green * I. Introduction In his recent Yale Law Journal article, Ariel Porat argues that tort law is (and should be) governed by a principle that he titles “ the alignment principle . ” 1 This principle is presented as one of tort law ’ s fundamental principles and as a major criterion for understanding and evaluating tort doctrines, rules , and the functioning of courts. The essence of the alignment principle is that whe n a court determines that certain conduct is negligent because of the foreseeable risks of harm that it creates ex , liability ex post should be imposed for all of the harms that materialize from these risks . Doing so is an alignment -- ex post liability is aligned with the ex ante setting of the standard of care. Failing to do so is a mis alignment. In Porat ’ s words: In negligence law, the risks taken into account by courts when setting the standard of care are the same risks considered when imposing liability and awarding damages. I call this th e “ . ” 2 Porat argues that misalignments should be suspected of being inefficient in terms of deterrence and therefore require convincing justification either on efficiency grounds or on grounds of corrective or distributive justice. In t he absence of such justification , he claims, misalignments should be corrected to avoid inefficient deterrence . On this conceptual basis he identifies five different misalignments , “ which . ” 3 Porat analyze s these misalignments from the perspectives of efficie ncy , corrective justice and distributive justice , although his focus is on efficiency . Porat concludes that none of the identified misalignments is justified on efficiency grounds and that attempts to e xplain some misalignments on ground s of justice are mostly unconvin cing. Porat ’ s analysis, a grand unifying theory of his impressive. It provide s many valuable and interesting insights into tort law, its functions and operation, and the relationship among its different goals. Porat ’ s innovative and thought - provoking work has already generated interest and comment by prominent torts scholars, and we have no doubt that it will generate further attention by others. Th e e commentators to date argue that Porat ’ s misalignment analysis begs the question because it does not address the fundamental core of tort law . 4 * Gil ead is the Bora Laskin Professor of Law, The Hebrew University of Jerusalem. Green is the Williams Professor of Law, Wake Forest University . The authors are grateful to Mark Geistfeld and Brett Green for helpful comments on earlier drafts of this Article . 1 Ariel Porat , Misalignments , 121 Y ALE L.J. 82 (2011) . 2 Id. at 84 . 3 Id. 4 Mark Geistfeld disputes Porat’s normative claim that tort law should avoid misalignments. Geistfeld focuses on personal injury and argues that misalignment should be embraced for “irreparable injuries,” Electronic copy available at: http://ssrn.com/abstract=2014874 Electronic copy available at: http://ssrn.com/abstract=2014874 2 I n this Article , we respond to Porat's cr itic i sm of one of the misalignments that he det ec ts -- created by "the wrongful risk limitation , " one formulation of the tort law element of scope of liability or proximate cause. 5 T he specific version of the scope of liability doctrine adopted in the recently completed Restatement (Third) of Torts 6 employs a harm within the risk (“HWR”) standard that limits liability to harms that arise from risks that make the defendant’s conduct negligent or otherwise tortious . Porat’s claim i s that the HWR standard, as understood and applied, creates an inefficient misalignment. In contra s t, we argue that this misalignment contributes to tort law providing an ap propriate level of liability to promote e fficiency. 7 We show in this regard that th ere is an inconsistency between Por a t ’s analysis of this misalignment and his analysis of another misalignment -- the "offsetting risks" misalignment . T his inconsistency , we believe, lends support to our claim regarding the efficiency of the HWR doctrine ( wh ich Porat terms the wrongful risk limitation ) . In Part II we begin with the Pigouvian/Coasian concept of internalization of externalities , 8 which lies at the heart of the search for efficient deterrence. This core concept instructs us that to promote efficient deterrence tort law should internalize t o actors the expected effects of their activities on others that the y externalize. These effects include the creation of risks of harm to others (negative externalities) that those for which monetary damages cannot restore the victim to the status quo ante. Alignment in such cases is inadequate because it suggests indifference between payment of compensatory damages and avoiding causing the harm. Mark A. Geistfeld, The Principle of Misalignment: Duty, Damages, and the Nature of Tort Liability , 121 Y ALE L.J. 179 (2011). Jules Col e man believes that Porat ’ s misalignment analysis , grounded in appropriate incentives to avoid unreasonably risky behavior “ does not contribute t o our understanding of tort law in general or the tort of Negligence in particular .” Jules L. Coleman, Misunderstandings and M isalignments , 122 Y ALE L.J . ___, _ __ (2012). In his view “Porat misconceives the nature of liability in torts, misunderstands compensatory damages, and fails to appreciate the relationship between the e lements of a tort and the inference properly drawn in law when those elements are satisfied.” Id. at ___. 5 The Third Restatement employs the term “ scope of liability ” rather than the more popular “ proximate cause ” for reasons that are explained in a c omment to § 29. See R ESTATEMENT (T HIRD ) OF T ORTS : L IABILITY FOR P HYSICAL AND E MOTIONAL H ARM § 29 cmt. b (2010). We follow the Restatement ’ s usage in this Article. 6 See id. § 29 . 7 We recognize that others criticize economic analysis of tort law, contending both normatively and descriptively that instrumentalism has no place in private law devoted to providing compensation for wrongfully caused harms. While this is not the place to revisit that debate, we cheerfully acknowledge that tort law has and should have concerns about matters other than fashioning appropriate incentives for efficient behavior. At the same time, we would resist claims by essentialists that instrumentalism pla ys no role in tort law today. As one of us has written, “ there is far too much purpose in the pens of judges, and far too much concern for the social implications of decisions to claim that tort law is limited to correcting individualized wrongdoing. ” W. J onathan Cardi & Michael D. Green, Duty Wars, 81 S. Cal. L. Rev. 671 , 708 (2008) ; see also Gary T. Schwartz, Mixed Theories of Tort Law: Affirming Both Deterrence and Corrective Justice , 75 T EX . L. R EV . 1801 (1997) (offering a reconciliation of the instrumentalist and non - instrumentalist understa ndings of tort law) . 8 A RTHUR C. P IGOU , T HE E CONOMICS OF W ELFARE (4 th ed. 1932); R. H. Coase, The Problem of Social Cost , 3 J. L AW & E CON 1(1960). For a discussion of similarities and differences between the two, see A. W. Brian Simpson, Coase v. Pigo u Reexamined , 25 J. L EGAL S TUD . 53 (1996). For Coase’s reply, see R. H. Coase, Law and Economics a nd A.W. Brian Simpso n, 25 J. L EGAL S TUD . 103 (1996). See also Herbert J. Hovenkamp , The Coase Theorem and Arthur Cecil Pigou, 51 A RIZ L. R EV . 633 (2009) ; Harold Demsetz, The Problem of Social Cost: What Problem? A Critique of the Reasoning of A.C. Pigou and R.H. Coase, 7 R EV . L. & E CON . 1 (2011) . Electronic copy available at: http://ssrn.com/abstract=2014874 Electronic copy available at: http://ssrn.com/abstract=2014874 3 should be in ternalized by imposing tort liability on the harms materializing from these risks. In addition , these effects may also include benefits to other s (positive externalities) , and these should also be taken into account by reducing or denying liability. In P ar t II I we explain and illust r ate why the HWR rule is one o f tort law ’s instruments to internalize externalized benefits and thereby to avoid or reduce excessive liability that may result in over - deterrence . Part I V examin e s Porat ’s analysis and criticism of HWR . Here we argue, first, that although the alignment analysis of HWR appeal s to an intuitive sense of symmetry when conduct creates multiple risks that result in harm, it does not conform w ith the internalization concept : the t wo actually clash. Under the latter , it is the externalized net social cost that should be internalized, taking into account externalized benefits. In contrast, under the alignment analysis of HWR, all foreseeable risks should be internalized by imposing l iability for all harm that materialize s from any risk considered at the ex ante stage, disregarding externalized benefits. This alignment principle would often lead to exce ssive liability and consequent over - deterrence. We proceed to highlight the inconsi stency between Porat's HWR analysis, which disregards externalized benefits , and his offsetting risks misalignment analysis , which takes these externalized benefits into account and recognizes the excessive liability concern . Finally, we illustrate, using, inter alia, Porat's own cases, how the HWR principle often reduces or avoids exces sive liability while Porat's alignment analysis leads to excessive liability. In Part V we address another significant misalignment -- the pure economic loss limitation -- which is absent from Porat's analysis. We contend that the efficiency - contribution of the HWR and economic loss misalignments undermine s Porat's claim s about the inefficiency of misalignments . In P art VI, before concluding, we return to the offsetting ri sks misalignment and argue that at least in this regard there is a discrepancy between Porat's analysis and the core internalization concept. II. The Core Concept of Economic Analysis : Internalizing Externalities To understand the basics of the economic analysis of tort law and the notion of efficient deterrence, one has to return to the roots -- the Pigouvian/Coasian concept of internalization of externalized effects . At the heart of the internalization concept lies the idea is that an actor who is engage d in an activity that creates risks of harm to other s often tends to disregard all or some of these risks when decid ing on the extent of the activity and the amount of care taken in it. 9 Where such risks of harm are externalized, the private cost of the ac tivity to the actor is lower than its social cost. This gap creates incentives for the 9 For the suggestion that negative externalities may be imposed intentionally , even when they involve private co st to the actor, in order to extract payments from victims in exchange for desisting, see Daniel B. Kelly, Strategic Spillovers , 111 C OLUM . L. R EV . 1641 (2011). 4 actor inefficiently to reduce her level of car e, increase the extent of the activity, or both. The role of tort law is to close this gap by internalizing the externalize d risks of harm to the actor through tort liability. Faced with higher costs of the activity (if the risk materializes and liability is imposed ) , the actor is encouraged to set the level of care and the extent of the activity at the efficient point . This i s the essence of efficient deterrence. 10 The logic of the internalization concept dictates that where the actor ’ s conduct 11 creates not just risk/s of harms (expected harms) but also the prospects of benefits (expected benefits) to other persons , and these benefits are externalized by the actor, the expected benefits as well should be internalized to the actor. 12 Failure to do so may create an opposing gap where the private cost of the conduct to the actor (who disregards these benefits) exceeds it s true social cost (which does account for these benefits). 13 Such a gap may induce the actor inefficiently to reduce the extent of activity or increase the level of care -- l eading to over - deterrence. 14 Internalization of the expected benefits is required to eliminate this gap. 15 10 “ E conomists describe harms that are outside private agreements as externalities . The economic purpose of tort liability is to induce injurers to internalize these costs. Tort law internalizes these costs by making injurer compensate the victim. When potential wrongdoers internalize the costs of the harm that they cause, they have inc entives to invest in safety at the efficient level . The economic essence of tort law is its use of liability to internalize externalities created by high transaction costs . ” R OBERT C OOTER & T HOMAS U LEN , L AW AND E CONOMICS 325 (5th ed. 2008). 11 By “ conduct ” we mean the act or omission that is alleged to be negligent. That conduct, typically an untaken precaution , creates certain risks and has certain benefits associated with it. One of the benefits that normally exists is avoiding the private cost of taking the precaution. 12 For discussion of the different ways in which different branches of law -- restitution, tort, intellectual property, and public law -- can and do internalize externalized benefits, see Giuseppe Dari - Mattiacci, Negative Liability , 38 J. L EG AL S TUD . 21 (2009); see also Lily L. Batchelder et al., Efficiency and Tax Incentives: The Case for Refundable Tax Credits , 59 S TAN . L. R EV . 23, 43 - 44 (2006) (addressing Pigouvian tax subsidies designed to internalize externalized social benefits). 13 “Because the defendant does not reap the full benefits of hi s act, neither should he have to pay the full costs. Otherwise, there will be too little of the activity.” W ILLIAM M. L ANDES & R ICHARD A. P OSNER , T HE E CONOMIC S TRUCTURE OF T ORT L AW 180 (1987). 14 Porat, correctly in our view, observes that theoretically exc essive liability does not lead to inefficient level of care, Porat, supra note 1 , at 119. That is because D would not spend more than the efficient amount of precaution, instead bearing the lesser cost of liability. Porat, however, recognizes that, as a practical matter, excessive liability often leads to inefficient consequences. His concern about over - deterrence is discussed infra text accompanying notes 54 - 56 ; see also Israel Gilead, Tort Law and Internalization , 17 I NT . R EV . L AW & E CON . 589, 597 - 600 (1997) (explaining the distorting effects of excessive liability resulting from failure to internalize externalized bene fits). 15 That externalized benefits should be internalized in tort law has been long recognized, usually in the context of pure economic loss, see text accompanying notes 76 - 78 ; W. Bishop, Economic Loss in Tort , 2 O XFORD J. L EGAL S TUD . 1 (1981); W. Bishop & J. Sutton, Efficiency and Justice in Tort Damages: The Shortcomings of the Pecuniary Loss Rule , 15 J. L EGAL S TUD . 347 (1986); Geistfeld, supra note Error! Bookmark not defined. , at 178 n.110; Victor P. Goldber g, Recovery for Economic Loss Following the Exxon Valdez Oil Spill , 23 J. L EGAL S TUD . 1 (1994); Landes & Posner, supra note 13 , at 251 - 55; S TEVEN S HAVELL , E CONOMIC A NALYSIS OF A CCIDENT L AW 135 - 40 (1987). For externalized benefits in a broader context of tort see Gilead, supra note 14 . Dari - M attiacci argues that two additional ways in which tort law internalizes externalized benefits other than in the exclusion of liability for pure economic loss are by allowing benefactors to injure third parties in the course of assisting gainers and by pun ishing failure to produce a positive externality in cases covered by liability for nonfeasance . See Dari - Mattiacci, supra note 12 , at 49 - 53. 5 So the role of tort law under the internalization concept is to internalize to actors the expected net social cost that they externalize by exposing them to the risk of liability for that externalized net cost. To that end , the court should first inquire whether D was engaged in conduct with expected harms and other costs that exceeded the conduct ’ s expected benefits and thereby created an expected net social cost. If she was, that means that she externalized some expected harms and s hould be held liable. 16 This inquiry, at the first stage, is an ex ante inquiry in the sense that the court maps and compares the expected harms and other costs with the expected benefits at the time of D ’ s decision to engage in the risky conduct. After that is the second ex post stage. Given that some ex ante risk/s materialized ex post into actual harm /s , the court has to decide whether to impose liability for th is harm or harms and what the appropriate scope of this liability should be in order to inte rnalize to D the net social cost that she externalized. Such internalization would incentivize future actors to take into account all expected harms that otherwise would be externalized and would thereby close the gap between social cost and private cost. To better understand the second stage one should recall that i t is often the case that conduct creates a variety of risks of harm : risks to different kinds of people, risks of different kinds of harm to the same kind of pe ople or to different kinds of p e ople , risks that materialize in different kinds of ways and other distinctions among risks . So in terms of the Hand Formula, PL is often composed of various distinct PL i (“ i ” rep resenting the different risks -- X , Y , Z , etc.) that together compose the aggre gate PL, that is ∑ PL i . Each specific PL i (PL x , PL y . . . ) may have its own balance of expected harms and benefits, and its own account of whether or to what extent these expected harms and benefits are externalized or internalized. However, these separate balances are tied together by the common conduct that created the risks. Under the internalization concept, then, at the second ex post stage, when deciding whether liability should be imposed for a PL i that materialized into harm, the court should exami ne whether this PL i should be internalized and to what extent. Liability should not be imposed for a PL i that is already internalized. For example, if PL i is an expected risk to D, no liability (whatever liability to oneself might mean) should follow as th is risk is internalized by D. If producers of defective products, for another example, remove unreasonable risks in the ir products because, as suggested by the Coase Theorem, they are “bribed” to do so by consumers’ willingness to pay higher prices for saf er product s , there is no need to internalize these risks (indeed, the risks no longer exist) . 17 There are also externalized risks that should not be internalized and therefore excluded from the scope of liability. For example, an 16 The Third Restatement provides in this regard, “Conduct is negligent if its disadvantages outweigh its advantages, while conduct is not negligent if it its advantages outweigh its disadvantages.” R ESTATEMENT (T HIRD ) OF T ORTS : L IABILITY FOR P HYSICAL AND E MOTIONAL H ARM § 3 cmt. d (2010) . 17 See, e.g. , A. Mitchell Polinsky & Steven Shavell , The Uneasy Case for Product Liability , 123 H ARV . L. R EV . 1437 , 1440, 1443 - 50 (2010) (arguing that market forces (and regulation) contribute significantly to the safety of widely - sold products); A. Mitchell Polinsky & Steven Shavell , A Skeptical Attitude about Product Liability is Justified; A Reply to Professors Go ldberg and Zipursky , 123 H ARV . L. R EV . 1949 (2010). 6 externalized PL i risk shoul d not be internalized where the expected benefits specific to the conduct that created the risk exceed its expected harms. As the internalization concept requires the internalization of both harms and benefits, where the lat t er exceed the former there is a setoff. Moreover, as such risk has an expected net social value , it actually reduces the net social cost of the conduct that should be internalized. To sum up, under the internalization concept the court should ex post exclude from the scope of liabilit y harms that materialize from internalized risks and harms that materialize from risks with associated externalized benefits when the latter benefits exceed the risks. This means that liability should not exceed the amount required to incentivize D to avo id risks that generat e net social cost. 18 III . Scope of Liability as a Benefit Internalizer a. S cope of liability : E xclu ding liability for risks with a positive net social balance Section 29 of the Restatement (Third) of Torts: Liability for Physical and Emotional Harm addresses a longstanding tort law doctrine : scope of liability or proximate cause . The particular form of this doctrine adopted in the Third Restatement, HWR, provides that “ [a]n actor ’ s liability is limited to those harms that result from the risks that made the actor ’ s conduct tortious . ” 19 This doctrine distinguishes between risks that at the ex ante first stage are considered tortious, and risks that are not . Based on this first stage distinction, the doctrine restricts liability at the ex post second stage by excluding from its scope harms that materialize from risks that were characterized as non - tortio u s at the first stage. When applied to the tort of negligence, the doctrine exclu des from the scope of liability (at the ex post stage) harms that materialize from risks that are either unforeseeable or reasonable. The reason is straightforward : u nder negligence law conduct is characterized as negligent (at the ex ante stage) because it creates risks of harm that are both foreseeable and unreasonable. These are tortious risks -- the ones that are the basis for a finding negligence . Other risks, reasonable risks 20 and unforeseeable risks, are not considered tortious at the first, ex ante , stage, and 18 “A tax system which was confined to a tax on the producer for damaged caused would tend to lead to unduly high cost being incurred for the prevention of damage. Of course, this could be avoided if it were possible to base the tax, not on the damage caused, but on the fall in the value of production (in its widest sense).” Coase, supra note 8 , at 41. Applied to our context, this means that liability should be imposed not for the harm that materializes from all the risks but only for the net social cost of the activity. That liability (in the form of a tax) for all the harms caused leads to exces sive internalization, and that liability should be limited to the net social cost of the activity (“fall in the value of production”) was later reemphasized in R. H. C OASE , T HE F IRM , T HE M ARKET AND THE L AW 182 (1988). To that end not only the benefits of the D’s activity should be considered, but also the benefits of P’s activity and the benefits of P’s alternative activities. Thus, in the famous cattle - crop hypothetical, for example, regard should be had to the benefits that the farmers can derive from “ planting another crop less susceptible to damage.” Id . at 175. 19 R ESTATEMENT (T HIRD ) OF T ORTS : L IABILITY FOR P HYSICAL & E MOTIONAL Harm § 29 (2010) . 20 See D AN B . D OBBS ET AL ., T HE L AW OF T ORTS § 205, at 710 ( “Some harms that are entirely foreseeable are nevertheless not harms a reasonable and prud en t person would seek to avoid.”) . 7 therefore are not the basis for liability at the second ex post stage should they materialize into harm . 21 But which foreseeable risks are considered reasonable, and therefore non - tortious, at the ex ante stage, so that they are later excluded from the scope of liability at the ex post stage? The discussion above of the ex post stage 22 provides the primary answer. Where a given conduc t generates a variety of risks -- to different kinds of people, 23 of different kinds of harm to the sa me or different people, 24 risks that materialize in different ways -- some of these distinct risks of harm ( PL i ) that together compose the aggregate risk of harm ( ∑ PL i ) may have a positive net social balance. Where the expected benefits associated with a give n risk of harm PL i exceed its associated expected harms, the risk would be considered reasonable because its expected net social balance is positive. 25 With this background l et us now examine the efficiency of the HWR rule . b. The efficiency of HWR An analysis of the efficiency of the HWR doctrine from the internalization perspective shows that it would usually, although not always, promote efficient outcomes in terms of avoiding or reducing excessive liability. A lthough unforeseeable risks are externalized by actors , they are not included within the scope of liability at the ex post stage , 26 because actors cannot be 21 Porat agrees that liability should not be imposed for unforeseeable risks created by the D’s conduct, one of the functions of the scope of liability doctrine. Porat, supra note 1 , at 92 - 93. Thus , if a D’s negligent conduct creates a foreseeable and unforeseeable risk, the scope of liability doctrine would limit liabi lity to harms resulting from the foreseeable risk. This omission of liability is consistent with the alignment principle as unforeseeable risks are excluded from consideration at the ex ante stage. See R ESTATEMENT (T HIRD ) OF T ORTS : L IABILITY FOR P HYSICAL & E MOTIONAL Harm § 3 cmt. g (2010) . Yet, Porat apparently believes that the HWR rule cannot be justified as a means to screen liability for unforeseeable risks and that a different scope of liability rule, one that directly relies on foreseeability, could also serve the function of eliminating liability for unforeseeable risks that materialize. See Porat supra note 1 , at 128 (“But foreseeability is a distinct conception that is independent of the wrongful risks limitation and can hardly justify the latter's adoption.”). We disagree with Porat on the utility of the HWR for limiting liability to harms resulting from unforeseeable risks. As the Third Res tatement documents, courts have struggled with the alternative foreseeability test to appreciate what has to be foreseeable, including intervening acts that occur, to whom it has to be foreseeable, and the point in time when foreseeability is assessed. By contrast, the HWR rule points directly to the time at which D’s negligent conduct occurred and inquiress whether at that time the risk was one of the foreseeable risks that made the D’s conduct negligent. See R ESTATEMENT (T HIRD ) OF T ORTS : L IABILITY FOR P HYSICAL & E MOTIONAL Harm § 29 cmt. j & rptrs. note to cmt. j (2010) . 22 See supra text accompanying notes 9 - 18 . 23 See, e.g., Palsgraf v. Long Island RR Co., 162 N.E.2d 99 (N.Y. 1928). 24 Se e, e.g., In re an Arbitration Between Polemis and Another and Furness, Withy & Co., Ltd., [1921] All E.R. 40. 25 Aside from efficiency, another reason for limiting liability may be that the risk is considered reasonable on grounds of fairness. See infra no te 70 . 26 Nor are they included in the risks that determine negligence at the ex ante stage. Nevertheless, when foreseeable risks are sufficient to render the actor’s conduct negligent and unforeseeable risks materialize in harm, the HWR doctrine prevents liability for that harm. 8 incentivized to prevent risks that are not known or reasonably knowable . 27 Moreover, there are fairness and administrative cost reas ons for denying liability for risks unknown and not reasonably knowable to the actor, as Oliver Wendell Holmes explained in The Common Law . 28 A s to foreseeable but reasonable risks, to the extent that a given risk of harm PL i is considered reasonable becau se its associated expected benefits exceed its associated expected harms, whether such a risk should or should not be internalized depends on the degree to which these expected harms and benefits are externalized. Where a reasonable risk PL i is fully e xternalized, both with regard to harms and benefits, this risk should not be internalized -- i t should be excluded from the scope of liability. 29 By definition, this PL i has a positive net social value, and as this net benefit is externalized by the actor, this externality opens a gap between the net social value of the risk, which is positive , and the net private value of the risk ,which is zero . This externalized gap -- the disregard of the net social value of the activity -- inef ficiently increases the private cost of the activity to the actor and may lead to excessive liability and resultant over - deterrence even if the expected harms of this risk are not internalized. Internalization of these harms would therefore contribute to e xcessive liability. To illustrate the foregoing, a ssume, for example, that D ’ s conduct generates two risks (two PL i ) that can be preven ted at a cost of 4. One risk -- denoted PL y -- i nvolves externalized expected benefits of 10 and externalized expected costs of 8 . The other externalized risk -- denoted PL x -- involves externalized expected benefits of 4 and externalized costs of 12. Risk PL y produces a net social benefit of 2 (10 - 8). Risk PL x produces a net social cost of 8 (12 - 4). The combined social balance of both risks (∑ PL i ) is a net social cost of 6 (8 - 2) , and they should be prevented given that prevention cost s are 4. To that end , only harms resulting from risk PL x (12) should be internalized to D (12 � 4). L iability of 12 is excessive because the ex ternalized net social cost is only 6. Imposition of additional liability of 8 for harms associated with risk PL y , the reasonable risk, will exacerbate the problem of excessive liability by increasing it from 6 (12 - 6) to 14 (20 - 6). The same analysis a pplies where parts of the expected harms and the expected benefits are internalized by the actor and other parts are not, as long as this partial internalization does not produce a positive net private value that exceeds the social net value. 30 Thus only wh ere D internalizes benefits of a reasonable risk, but not its costs, 27 Imposing liability for risks that cannot be discovered with reasonable effort would be inefficient, because that implies an obligation to expend more in search costs than is justified by what is known about what is unknown. 28 See O LIVER W ENDELL H OLMES , T HE C OMMON L AW 94 - 96 (1881). 29 For example if a rescuer attempts a rescue dangerous to a person who is trapped in a mine an d will otherwise die, the rescuer should not be held liable for any harm to the person that occurs in the course of the rescue (we assume the rescue poses no risks to anyone else). 30 Assume that in the above illustration risk PL y is only partially external ized: D internalizes benefits of 3 (out of 10) and harms of 5 (out of 8). As risk PL y now involves a net private cost of 2 (5 - 3) to D, while it has a net social value of 2 (10 - 8), the outcome would be an increase of the benefit that D has to internaliz e from 2 to 4 and that would exacerbate the problem of excessive liability and over - 9 such that the risk’s private value exceeds it social value, is it necessary to internalize harms of a reasonable risk and only to the extent of the difference between private value and so cial value. 31 The conclusion, then, is that from the perspective of internalization theory, the HWR doctrine usually, although not always, promotes efficient deterrence by excluding from the scope of liability harms that materialize from foreseeable and reasonable risks, whenever such expected harms, although externalized, are less than or equal to the externalized benefits of the reasonable risk. Without the HWR doctrine, if liability were imposed for these harms, the outcome would be excessive liability due to failure to internalize the associated benefits. Thus the exclusion of reasonable risks from t he scope of liability is an indirect way to internalize otherwise externalized benefits. Given that tort law is well equipped to internalize externalized risks of harm (by imposing liability) but is much less equipped to internalize externalized benefits, the important role of the HWR doctrine as a benefit internalizer that promotes efficient deterrence should be recognized . c. The efficiency of HWR as benefit internalizer : a n illustration We illustrate, with the following case , our analysis of the HW R doctrine under the internalization concept and the way it functions as one of tort law ’ s benefit internalizers. A cholesterol - reducing drug is used by 1,000 consumers. T he market price is 5, and the average production cost is 2. Each consumer enjoys a reduction of cholesterol level that is valued , on average, at 12 , and each is aware of a n associated minor liver problem with a cost of 4 , so that t he average expected net value of the product to each consumer is 8 (12 - 4) . 32 The same causal process that reduces cholesterol also causes the liver problem. In addition to the liver problem, the drug creates a risk of a more serious heart problem about which the consumers are unaware 33 due to a different causal process, with a cost of 80 to 10 percent of users, here 100. The only available precaution is to cease marketing the drug, as there is no way in advance to identify which patients will suffer heart problems nor is there any way to mitigate the adverse effect once it occurs. The risks ( PL i ) that the marketing of the drug generate can be classified under the HWR rule according to the two sets of affected consumers: a risk to group A , consisting of the 900 consumers who are ex ante exposed only to a risk of liver deterrence if liability is imposed for the risk of 8. Assume now that the internalized benefits of risk PL y are 7 and internalized harms are 5. That would produce a net pri vate benefit of 2 to D (7 - 5), which corresponds with the social value of 2. Thus, no internalization is required with regard to risk PL y . 31 Assume further that in the above illustration D internalizes benefits of 7 (out of 10) and harms of 4 (out of 8). That would produce a net private value of 3 (7 - 4), which exceeds the net social value (2) by 1. Here there is need to internalize to D the expected harm of risk PL Y , which is otherwise externalized by her, but only to the extent of 1 and not to the exte nt of the full harm of 8. 32 The range of the net value to different consumers lies between 11 and 5 along a downward sloping demand curve. 33 We assume that the heart risk is either known to the producer or is at least reasonably foreseeable by it. 10 problem s , and a risk to group B , comprising the 100 consumers who are ex ante exposed to both risks. According to this classification , the risk to group A is reasonable as all members of group A are beneficiaries of the product: they pay 5 and are exposed to risk of 4 but their expected benefit is 12, so on average each one enjoys an expected net benefit of 3. 34 By contrast, the 100 members of group B are all victims of the drug in the sense that the expected net impact of the drug to each is - 77 on average : total cost of 89 (5 + 4 + 80) and benefit of 12. According to the internalization concept, the expected aggregate social cost of the drug ( ∑ PL i + production cost ) is 14,000: cost of production (2,000) + liver harms (4,000) + heart harms (8, 000). The overall benefits are 12,000 (1,000 x 12). The net social balance of the activity is negative (2,000) , and the conduct of marketing the product would be found negligent at the ex ante stage. At the ex post stage, liability should be internalized s o that the producer will bear the net social cost of the activity , which is 2,000. As the producer internalizes a net private benefit of 3,000 (the difference between the market price and the average production cost (5 - 2) x 1 ,000) ) , 35 liability should be set at 5,000 (2,000 + 3,000). From this perspective, the HWR doctrine promotes efficiency. By excluding liability for the reasonable risk t o group A, which is 3,600 (4 x 900) , liability is reduced from 12,000 to 8,400. Liability is st ill excessive (8,400 � 5,000), but it would be considerably more excessive if the reasonable risk were not excluded from the producer’s scope of liability (12,000 � 5,000). The HWR doctrine functions, in conformity with the internalization concept, as a be nefit internalizer that promotes efficiency through the reduction of excessive liability from 7,000 to 3,400. According to the above analysis , 36 liability should be imposed on the producer for the liver harms only whe n market conditions allow the producer to raise prices and thereby internalize the benefits of the drug to a degree that liability for the heart problem would not suffice to incentivize it to stop production. 37 The conclusion would be the same if the reasonable risk is characterized not as the risk to the liver of the members of group A but rather as the risk to the liver of all consumers -- groups A and B -- a risk different from the heart risk to t he members of group B. 38 34 Thus, the consumer who obtains the greatest utility from the drug would enjoy a net benefit of 6. The marginal consumer would enjoy a net benefit of something just positive. We assume the average benefit across all consumers is, thus 3. 35 T he 3,000 repr esent producer surplus . 36 See infra § I II .b . 37 Whether that happens and to what extent depends on the shape (elasticity) of the demand (benefits) curve and the supply (cost) curve. These shapes determine how much of the social benefits of the activity (t he economic surplus) is enjoyed by the producer (producer surplus) and thereby internalized by it and how much of it is obtained by the consumer (consumer surplus) and thus externalized by the producer. 38 The risk to the liver is the same reasonable risk to all consumers in the sense that it is caused by the same causal process that reduces cholesterol, so that the overall balance of this process to all consumers is positive. Under this characterization, no liability should be imposed for the liver harms as well, not even to members of group B, because liability for the heart risk (8,000) would also be excessive. 11 I V. The A lignment A nalysis of HWR a. Porat's critique Unlike internalization analysis , which points to the efficiency of the HWR doctrine in avoiding or reducing excessive liability , Porat characterizes the doctrine (to which he refers as the wrongful risk limitation ” ) as a misali gnment and criticizes it for the inefficient way in which it is commonly understood and applied . The essence of P orat ’ s analysis of the HWR doctrine is the following : When D ’s conduct generates a reasonable risk ( “ ordinary , ” “ usual , ” or “ non - wrongful ” in Porat ’ s terminology 39 ), as well as another risk or risks, the court should consider all foreseeable risks at the ex ante stage when deciding whether the conduct is negligent. Although there are cases in which the conduct would be characterized as neglig ent even if the reasonable risk PL y is ignored, because the other risk PL x is sufficient to render the conduct negligent, in other cases the conduct is negligent only if the reasonable risk PL y and the other risk PL x , are counted as costs of the D’s conduc t. Therefore, at the ex post stage, liability should be imposed for all foreseeable risks including reasonable risks , such as PL y . Failure to do so, Porat argues, would lead to under - deterrence in those cases in which the conduct was considered negligent b ecause of reasonable risk PL y . As the scope of liability doctrine actually fails to do just that, it actually creates a misalignment that should be rectified. In Porat ’ s words: My argument is twofold: (a) absent special policy considerations, all foreseeable risks created by the injurer should be and are considered by courts when they set the standard of care; (b) therefore, exempting the negligent injurer from liability for harm s materializing from foreseeable risks creates misalignment . 40 The argument is illustrated by a case 41 in which the non - installation of railings along a staircase at D’s workplace creates two kinds of risk. One , risk PL x affects disabled employees who may fall and another risk PL z is posed to able - bodied employees who also may fall. The expected harm from PL x is 60 and the expected harm for PL z is 30. The cost of installing railings that will eliminate both risk s is 80 (prevention costs). Porat argues that since D ’ s conduct (failing to install the railings) is negligent at the ex ante stage because of the combined effect ∑ PL i of both PL x and PL z (60 + 30 � 80), liability (at the ex post stage) should be imposed for the materialization of either of these ri sks. The exclusion of PL z from the scope of D’s liability under the HWR rule ( because this risk to able - bodied employees is 39 Porat, supra note 1 , at 90. 40 Id. at 124. 41 Porat relies on another illustration -- the dropped gun case -- which we address below, infra text accompanying note 68 . 12 reasonable ) will lead to under - deterrence as D ’s liability would provide an insufficient incentive to install the railings ( 60 80 ) . 42 Porat acknowledges that in cases in which some risks exceed the costs of prevention there is no need to impose liability for other risks. 43 For example, in the above illustration if PL x is greater than 80, rather than 60, denial of liability for PL z would not lead to under - deterrence as liability for PL x alone would incentivize D to install railings (81 � 80). Yet, Porat argu es against attempting to distinguish among cases with multiple risks, including reasonable risks, which risks should or should not be excluded from D’s scope of liability because of concerns of over - deterrence . In Porat’s view, courts can’t distinguish among cases such as the two variations in the railing illustration. 44 In addition, Porat wants to enable the negligence doctrine t o be applied uniformly in all cases, thereby avoiding the need to distinguish the two variations. 45 Porat concludes that as “ all foreseeable risks should count in setting the standard of care ” therefore “ all harms that materialized from those risks should be recoverable . ” 46 b . The problem with P orat ’ s analysis Initially, we seek to clarify by acknowledging our agreement with Porat ’ s first assumption, namely, that there are cases in which conduct that creates reasonable risk/s PL y and other risk/s PL x may be characterized as negligent because of the expected harms that are associated with the reasonable risk/s PL y . This is so because the prevention of the former risk PL x will also prevent the latter risk PL y , 47 and therefore all expected harms should be considered at the ex ante stage. All the PL i , including the reasonable ones, should be taken into account where the aggregate PL ( ∑ PL i ) is considered. This point is exemplified by the drug illustration above . Although the risk of liver problem s to membe rs of group A (3,600) is reasonable, it is the expected harm of this risk (combined with the expected harm of heart and liver 42 For further discussion of this illustration, see infra text accompanying notes 59 - 65. 43 Porat, supra note 1 , at 127 & n.129. 44 Porat writes that courts “ do not really know what the numbers are, and there is always the possibility that they could work out similarly to the original assumptions .” Id. at 127. 45 “[T] here are definite advantages to a doctrine of negligence that can be applied uniformly to all case s. This is precisely how the doctrine of negligence works: the injurer bears liability for risks he could have reasonably prevented, even if lower liability would be sufficient to incentivize him to take adequate precautions. ” Id. We take issue with Porat’ s claim for uniformity infra text accompanying notes 73 - 74 . 46 Id. We note a similar attack on the HWR principle by Heidi Hurd and Michael Moore. See Heidi M. Hurd & Michael S. Moore, Negligence in the Air , 3 T HEORETICAL I NQUIRES IN L AW 333 (2002). T h ey argued, as Porat does now, that all foreseeable risks “ enter into the calculus that determines the justifiability of the defendant ’ s conduct .” Id. at 365. These risks include “ Risks that are Individually Sufficient for Negligence, ” id. at 366 - 69, “ Risks that Are Individually Necessary for Negligence, ” id., at 369, “ Risks that are a combination of the above, ” id. at 369 - 70, and actually all other kinds of risks, id. at 370 - 74. They therefore concluded, as Porat does now, that “ there is no way to specify a limitation on the risks eligible to be used in the HWR test, ” and thus that “ [e]very harm that some defendant ’ s act causes . . . is within the risk ” , id. at 374, and that “ [i]t is time to abandon the harm - within - risk analysis. ” Id. at 411. 47 See supra note 11 . 13 risk s to members of group B) that make the marketing of the drug negligen t . If the expected harm of the liver risk is disregarded because of its reasonableness, the benefit of the drug (12,000) would exceed the expected harm of the heart risk (8, 4 00) and the production cost (2,000), and the conduct would not be characterized as negligent, which is inefficient. However, i n our view Porat ’ s second assumption, namely, that failure to impose liability at the ex post stage for harms that result from a reasonable risk would usually lead to under - deterrence, is incorre c t . Internalization analysis, as we have seen, tells us the o pposite, namely, that where a reasonable risk ’s benefits are externalized, wholly or partially, there is most often a problem of excessive liability if liability for the expected harm is imposed . E xclusion of these harms from the scope of liability would o ften reduce or eliminate excessive liability and its resultant over - deterrence. The under - deterrence Porat fears will occur only in cases where D internalizes benefits of a reasonable risk to a degree that the risk ’ s private value exceeds it s social value. Even then, it is only this difference between private value and social value that should be internalized. 48 The problem with Porat ’ s analysis, to sum up, is that it disregards the externalized benefits that are associated with reasonable r isks and therefore the major role that the HWR doctrine plays in this regard as a benefit internalizer that reduces or prevents excessive liability and resultant over - deterrence . c. The inconsistency in Porat's a nalysis Porat ’s disregard of externalized benefits and the inefficiencies of excessive liability in h is HWR analyses is in consistent with another aspect of his alignment analysis. 49 O ne of the other inefficient misalignmen ts identified by Porat is the offsetting risks mi salignment. 50 As “[i] njurers sometimes simultaneously increase risks and decrease risks ,” courts at the ex ante stage “[i] n determining whether the injurer was 48 See supra § III.b. 49 Error! Main Document Only. It is also inconsistent with other work of Porat in which he emphasizes the importance of internalizing externalized benefits and warns that failure to do so leads to inefficient outcomes. Porat recently wrote: When people promote their own interests, they often create negative or positive effects for other people’s interests, without the latter’s consent. Economists refer to these effects as “negative externalities” . . . and “positive externalities”. . . . Ideally, from an economic perspective, both the negative and positive effects should be internalized by those who produce them, for wi th full internalization, injurers and benefactors alike will behave efficiently. Ariel Porat, Private Production of Public Goods: Liability for Unrequested Benefits , 108 Mich. L. Rev. 189, 190 - 91 (2009). Porat accordingly suggests that for better internal ization the law of restitution should, under specified conditions, require recipients to compensate benefactors for unrequested benefits. 50 Porat, supra note 1 , at 115 - 22. 14 negligent . . . should consider all risks, both increased and decreased .” 51 But, at the ex post stage, “ once the injurer is deemed negligent he will be found liable for the entire harm that resulted, without being credited for the risks he decreased . . . .” 52 Therefore, “ Courts ’ failure to reduce damages by the offsetting risks creates a misalignmen t .” 53 Porat illustrates this misalignment with a case in which a defendant doctor is found negligent at the ex ante stage for deciding to administer to his patient a treatment that involves a risk of 500 rather than a treatment that involves a risk of 400 ( all other conditions equal). As the ex ante social cost of the risk generated by the negligent decision is only 100 (500 - 400), alignment requires that at the ex post stage liability should correspond with the net risk of 100 rather than with the risk of 50 0. In practice, Porat argues, courts impose ex post liability that corresponds with the risk of 500, thereby creating a misalignment. 54 Such misaligned liability, Porat warns , is excessive and therefore inefficien t . Although in an ideal world excessive liab ility does not lead to inefficiencies , Porat recognizes that [I] n the real world, with court and injurer risk of error, the outcome is completely different. When the actual liability leads to the expected liability exceeding the social risks, the injurer will take excessive precautions . ” 55 Indeed, Porat argues that this excessive liability is a major factor behind the disturbing phenomenon of defensive medicine, 56 and he warns against the costs of refusing to apply the offsetting risks principle, which i nclude defensive medicine, overinvestment in precautions, and overburdening desirable activities with a resultant inefficient decrease in their level . 57 Obviously, Porat’s offsetting risks analysis cannot be reconciled with his HWR misalignment analysis . It is unclear why Porat limits his offsetting benefit analysis of the ex post stage only to one kind of benefit -- reduced risks -- and only to one aspect -- 51 Id. at 115. 52 Id. 53 Id. at 116. Section 920 of the Restatement (Second) of Torts suggests that the offsetting risks misalignment may not be as broad as Porat claims it is. Tha t section provides : “ When the defendant ’ s tortious conduct has caused harm to the plaintiff or to his property and in so doing has conferred a special benefit to the interest of the plaintiff that was harmed, the value of the benefit conferred is considere d in mitigation of damages, to the extent that this is equitable .” Comment a to § 920 counsels courts contrary to Porat ’ s claim that decreased risks of medical treatment and other activities are not offset. It provides that “ if a surgeon performs an unpriv ileged operation resulting in pain and suffering, it may be shown that the operation averted future suffering . . . . If a surgeon has destroyed an organ of the body, it may be shown in mitigation that the operation improved other bodily functions . . . . L ikewise one who has interfered with the physical condition of land can show in mitigation . . . that the change resulted in an improvement to the land .” R ESTATEMENT (S ECOND ) OF T ORTS § 920 cmt. a (1979) ; see also Err or! Main Document Only. Gracia v. Meiselman, 531 A.2d 1373, 1377 - 78 (N.J. Super. Ct. 1987) (applying the “benefit rule” of § 920 in an informed consent case in which the unconsented medical procedure provided a benefit to the patient) . 54 Id. 55 Porat, supra note 1 , at 119. 56 Id. at 118 - 21. 57 Ariel Porat, Offsetting Risks , 106 M ICH . L. R EV . 243, 263 - 68 (2007). 15 reduction of the award. 58 If D should be held liable only for the net social cost of her activity to av oid excessive liability, as Porat acknowledges, and if externalized benefits should therefore be internalized, that should be done across the board. It should lead not only to reduction of the award when D is held liable, but also to the denial of D’s liab ility whe n the ex post harm materializes from a reasonable risk with an externalized net social value. In other words, if Porat would have extended his analysis of offsetting risks in the tort context, as he should have, to all of the benefits that D’s co nduct generates, he would have reached the conclusion, as we do , that the HWR doctrine does just that -- prevents internalization of reasonable risks with an externalized net social value. He would have realized that the common way in which the scope of liab ility doctrine is understood and applied, which he criticizes for being inefficient, is one of tort law ’s methods for the efficient internalization of externalized benefits and reduction of excessive liability. The inconsistency becomes apparent when one compares , on the one hand, Porat’s distress at excessive liability, which leads to over - deterrence in his offsetting risks analysis, and, on the other hand, his sanguinity when confronting excessive liability in his HWR misalignment analysis. There he sta tes that excessive liability “is precisely how the doctrine of negligence works: the injurer bears liability for risks he could have reasonably prevented, even if lower liability would be sufficient to incentivize him to take adequate precautions . " 59 All i n all, Po r at’s analysis of the alleged “offsetting risks” misalignment support and confirm our criticism of his HWR analysis. c . Illustrations Our critique of Porat ’ s analysis of the efficiency of the HWR doctrine is illustrated in the cases discussed below. 1. Installing railings . Porat ’ s reliance on the railings case to illustrate and support his alignment analysis and his criticism of the HWR doctrine is unpersuasive . Porat ’s analysis assumes that the risk of falling , PL z , for able - bodi ed employees in the absence of railings would be considered reasonable under the HWR doctrine and th us excluded from D’s scope of liability at the ex post stage. But this is not so. A risk is reasonable whe n it s expected benefits exceed it s expected harms. Risk PL z is not such a risk. No expected benefits are associated with this risk that exceed its expected harms. True, the costs of prevention (80) exceed the expected harms of risk PL z (30), 58 To be sure , Porat is well aware that offsetting risks are one kind of benefit a nd that in principle all externalized benefits should be internalized. In his words: The offsetting risks are, in fact, benefits created by the injurer. A broader argument than the offsetting risk argument would be that ideally, from an efficiency perspect ive, actors should internalize both the harms and benefits that they create. Porat, supra note 1 , at 118 n.108. 58 Porat, supra note 1 , at 127. 16 but this benefit (saving 80) cannot be associated exclusively wi th risk PL z -- it must be also be associated with the greater risk PL x to the disabled employees (60). Thus, neither of these risks should be characterized as reasonable for purposes of the HWR rule, and liability for risk PL z should not be exclude d from D’s scope of liability. Porat claim s that courts applying the HWR rule exclude risks such as PL z from D’s scope of liability, and he concludes that the doctrine therefore supports inefficient results. 60 Porat also claims that the Third Restatement “clearly takes the same approach.” 61 We think he is incorrect in both re spects . T he two cases 62 Porat cites in support of his claim about what courts do are entirely unconvincing and, indeed, point in the opposite dir ection about the efficiency of HWR for risks such as PL z. Both cases involved plaintiffs asserting claims based on negligence per se. In both, the court rejected negligence per se because the plaintiff was not a member of the class for whom the statutory p rovision was designed to protect. Because the scope of liability is more constrained than for common law negligence r elying on negligence per se cases is mistaken . Court confine the scope of liability for negligence per se based on deference to the legisl ature. If legislative acts are borrowed to provide the standard of care in tort, limitations imposed on the application of that standard must also accompany the standard of care. Anderson v. Turton Develoment, Inc. , 63 one of the cases relied on by Porat, re veals precisely this point: in common law negligence claims, the scope of liability is broader than for negligence per se claims. In that case, although the court de cided plaintiff could not rely on negligence per se, the court reversed the summary judgmen t granted defendant so that the plaintiff could pursue her common law negligence claim. 64 The Third Restatement explains precisely the same principle, revealing the inappropriateness of comparing negligence per se with common law negligence limitations on l iability . 65 60 Porat, supra note 1 , at 124: “Applying the wrongful risks limitation, courts would likely not impose liability on the employer toward the able - bodied employee in this example.” 61 Id. Porat refers to a provision in the Third Restatement that addresses negligence per se not comm on law negligence. 62 Porat relies on Anderson v. Turton Dev., Inc ., 483 S.E. 2d 597 (Ga. Ct. App. 1997) and Carman v. Dunaway Timber Co., 949 S.W.2d 569 (Ky. 1997) . 63 483 S.E. 2d 597 (Ga. Ct. App. 1997) . 64 For other cases similar to Anderson permitting a party to pursue a common law negligence claim despite denying the use of negligence per se because of its more stringent restrictions on scope of liability, see Freeman v. United States, 509 F.2d 626 (6th Cir. 1975) (although parachutists vio lation of regulation prohibiting parachuting into clouds was designed to protect parachutists and therefore would not support contributory negligence per se, plaintiffs could be found contributorily negligent under common law standards); Thoma v. Kettler B ros., 632 A.2d 725 (D.C. 1993); Shahtout v. Emco Garbage Co., 695 P.2d 897 (Or. 1985). 65 R ESTATEMENT (T HIRD ) OF T ORTS : L IABILITY FOR P HYSICAL AND E MOTIONAL H ARM § 14 cmt. g (2010): For example, a worksite may contain a plainly dangerous condition that proves the employer’s violation of an occupational safety regulation; even so, under this Section negligence per se cannot be invoked if the person suffering injury on account of that dangerous condition turns out to be not a worker but rath er a business invitee on the employer’s premises. However, in such a case -- and in many of the cases involving a victim who is not in the proper class -- the discrepancy between the harm that happens and the harm the regulation has sought to prevent is narrow . 17 To complete the analysis, let us modify the railing illustration and assume that risk PL z is a reasonable risk, namely, that some expected benefits associated with it (40) indeed exceed s it s expected harms (30). 66 Under this assumption , PL z has a net social value of 10 (40 - 30), the aggregated PL (∑ PL i ) is 50 (60 - 10) and D ’ s conduct is not negligent as the prevention cost (80) exceed s the net combined risk (50). Let us a ssume further that prevention costs are only 40. In that case the cond uct would be considered negligent (50 � 40) at the ex ante stage. Yet, the exclusion of liability for a reasonable risk PL z under the HWR doctrine will not lead to under - deterrence as Porat claims. The internalization of risk PL x (60) would not only be su fficient to induce D to install railings (60 � 40) but it would also lead to excessive liability (60 � 50). Imposition of liability for risk PL z , as advocated by Porat, would exacerbate the problem of excessive liability (90 � 50). 2 . Cholesterol - reducing drug. The analysis of the cholesterol - reducing drug demonstrates that although the marketing of the product is ex ante negligent because of the expected harm associated with the reasonable risk ( to the liver of members of group A ) , these harms should not be internalized to the producer at the ex post stage because such liability will exacerbate or create a problem of excessive liability . T he exclusion of these harm s may not lead to under - deterrence -- on the contrary, their exclusion may prevent excessive l iability . 67 Under Porat ’ s alignment analysis, by contrast, which sanctions the imposition of liability for the reasonable risk to the liver, the problem of excessive liability would be exacerbate d (or create d ). 3. Loaded gun and injured toe . Another illustration, borrowed from the Third Restatement, employed by Porat to illustrate and support his analysis and conclusions 68 actually does the opposite. A hunter D hands over a loaded gun to a child P who drops it on P ’ s toe and breaks it. Let us assume th at the handing over of an unloaded gun to a child is a reasonable risk because the gun is not heavy and the two parties derive some benefit from it that outweigh the risk of its dropping . T he question becomes whether liability should be imposed for the har m arising from the reasonable risk of dropping it in a case in which the gun was loaded , and the risks associated with handing over a loaded gun to a child make the D negligent . Porat argues that liability should be imposed , but we do not agree . Handing ov er a loaded gun to a child creates ex ante two risks ( PL i ), one , risk PL y , that is reasonable (dropping) and another risk PL x ( the gun being fired ) . Note that in this illustration there are two distinct aspects of D’s conduct that could prevent risk : 1) n ot handing the gun to the child, which would prevent both risks; and 2) unloading the gun before handing it to the child, which would only prevent the risk of the gun being fired. Let us a ssume that risk PL y is reasonable because D and the child derive fro m it a benefit of 5 each and the expected harm of dropping is only 5 (10 � 5). Assume further that the risk of shooting is 9 and that it can be prevent ed at a cost of 1 (checking the gun Accordingly, even if negligence per se is properly not applied in these cases, the actor’s violation of the regulation may be admissible as evidence of negligence . 66 For example, assume that railings would induce able - bodied employees to rush the stair s thereby creating another and greater risk of falling. 67 See supra § III.c 68 Porat, supra note 1 , at 126 n.127. 18 before handing). At the ex ante stage the conduct would be characteri zed as negligent because of risk PL y . It is the combined effect of both risks ((∑ PL i ), including reasonable risk PL y , that renders the conduct negligent (5 + 9 � 10 : PL � B ). Yet, the ex post exclusion of risk PL y from the scope of liability would not lea d to under - deterrence as Porat claims. Liability for risk PL x (9) would suffice to induce D to prevent the harm because it exceeds the prevention cost (1) and D ’ s internalized benefit (5). Liability for risk PL y would be excessive. Only if D internalizes t he full benefit of PL y (10) would the exclusion of liability for that risk lead to under - deterrence. 4 . Landowner fails to warn not to swim in a polluted pond . A landowner fails to post a sign warning others not to swim in a pond that has been polluted with bacteria that cause a fatal disease . 69 This risk, PL x , is , a plainly unreasonable risk. Risk PL y is drowning (the pollution does not affect the risk of drowning), and is reasonable ( the pleasure to the swimmer exceeds the risk of drowning). The conduct is negligent regardless of the drowning risk PL y . When a swimmer in the pond drowns, imposing liability, which Porat advocates, 70 would be excessive . Application of t he HWR doctrine is preferable, as it would deny liability. d . A r eprise on the differences between aligning and internalizing 71 The HWR doctrine excludes from the scope of liability un foreseeable risks. To that exclusion from liability Porat agrees. 72 It also excludes from the scope of liability reasonable (and foreseeable ) risks whe n the other risks resulting from D’s conduct are sufficient to render D negligent at the ex ante stage. Porat agrees that such exclusion does not lead to under - deterrence. So Porat agrees that in these two major fields of its operation the HWR d octrine functions appropriately . Nevertheless , Porat argues that the doctrine , because of the way it is commonly understood and applied, should be 69 This illustration is based on Darby v. National Trust, [2001] PIQR P27. The court ruled that the plaintiff’s decedent’s death was outside defendant’s scope of liability. 70 Porat, supra note 1 , at 127 (“ Error! Main Document Only. all foreseeable risks should count in setting the standard of care and all harms that materia lized from those risks should be recoverable ”). 71 We add a note on fairness. Acknowledging that justice considerations (corrective or distributive or both) may justify inefficient misalignments, Porat finds no such convincing considerations to justify the HWR doctrine. Porat, supra note 1, at 128 - 29. But once the doctrine is understood as it should be -- namely as denying liability for reasonable risks -- this sort of justification becomes self - evident. We would frequently find it unfair to impose liability on D for the creation of reasonable risks just because the same conduct created other risks to P or to other people. E specially where P is a net beneficiary of a reasonable risk , it would strike us as unfair to award compensation to P who was exposed to that risk, just because D’s conduct created other risks to P or to other people. Is it fair to hold the cholesterol - reducing drug producer liable to members of group A, who are all net beneficiaries of the drug despite the liver risks, just because other peopl e, members of group B, are exposed to a different risk -- to the heart? As Judge Cardozo said, one should sue for a wrong personal to him “ and not as the vicarious beneficiary of a breach of duty to another. ” Palsgraf v. Long Island R.R. Co., 162 N.E. 99, 1 00 (N.Y. 1928). Is it fair to compensate the person whose toe was broken just because the dropped gun was loaded and created other risks? Or to compensate the estate of a drowning victim because of the fortuity that the pond was infected with a bacteria th at had no role in the victim’s death? 72 See Porat, supra note 1 , at 128. As explained supra note 21 , we believe that the HWR rule is the best one for dealing appropriately with unforeseeable risks . 19 transformed , because it also excludes from the scope of liability reasonable risks that ex ante rendered the conduct negligent, on the grounds that such exclusion leads to under - deterrence. This sin, in his view, outweighs the advantages of the doctrine . Porat seeks a blanket rule that automatically imposes ex post liability for all foreseeable harm . Relyi ng on the core concept of internalization to avoid excessive liability and over - deterrence , w e have attempted to show that Porat is incorrect in arguing that the ex post exclusion of reasonable risks that render D ’ s conduct negligent at the ex ante stage leads to under - deterrence. T he exclusion of reasonable risks (whose benefits are externalized) from the scope of liability, even if they rendered the conduct negligen t at the ex ante stage, does not lead to under - deterrence. On the contrary, exclusi on would often reduce or eliminate excessive liability and the consequent inefficiency of over - deterrence. It would lead to under - deterrence in cases in which D internalizes benefits of a reasonable risk to a degree that the risk ’ s private value exceeds it social value , and in such circumstances , we agree with Porat that liability should be imposed . Unlike Porat , who argues for uniformity, 73 we prefer to have courts attempt to determine which case is which and thereby employ liability in a more refined manne r. We think that negligence law has, to a large degree, refused to accept the tradeoff of uniformity for more accurate determinations. The “ethic s of particularism” described in the Third Restatement has suffused negligence law since Oliver Wendell Holmes ’ s idea that judges could, with experience, fashion specific rules of conduct for recurring fact patterns was rejected . 74 V . An Absent Efficient Misalignment : Pure Economic Loss Revealing of Porat ’ s disregard of externalized benefits 75 is his silence about the pure economic loss misalignment in negligence law . As many economic losses involve only private cost s and not social cost - - because they are offset by externalized e conomic benefits to other parties 76 - - these losses , from an efficiency perspective, 77 should be excluded from the scope of liability at the ex post stage although considered at the ex ante stage. G iven the prevalence of such losses , the ir exclusion constitute s a 73 See supra text accompanying note 45 . 74 Holmes’s view was the basis for a Supreme Court o pinion, authored by him, in Baltimore & Ohio R.R. Co. v. Goodman, 275 U.S. 66 (1927). In Goodman, the Court adopted a rule that a motorist crossing a railroad track who did not stop, exit the vehicle, and scan the tracks in both directions before proceedin g was contributorily negligent. Seven years later, the Court effectively overruled Goodman , because of the varying facts that might exist in railroad crossing cases that bear on the appropriate conduct of a motorist. See Pokora v. Wabash Ry. Co., 292 U.S. 98 (1934). " Tort law has thus accepted an ethics of particularism, which tends to cast doubt on the viability of general rules capable of producing determinate results ." R ESTATEMENT (T HIRD ) OF T ORTS : L IABILITY FOR P HYSICAL AND E MOTIONAL H ARM § 8 cmt. c (2 0 10). 75 Other than in his analysis of offsetting risks. 76 See supra note 15 ; Richard A. Posner, Common - Law Economic Torts: An Economic and Legal Analysis , 48 Arizona L. Rev. 735, 736 - 741 (2006). “Concerns about disproportionate liability sometimes also arise from the observation that private economic losses do not necessarily amount to comparable social losses . ” R ESTATEMENT (T HIRD ) OF T ORTS : L IABILITY FOR E CONOMIC H ARM § 1 rprts. note to cmt. c (Council Draft No. 1.1 Dec . 28, 2011) . 77 This is an instance where fairne ss concerns may point in a different direction from an efficiency perspective. 20 substantial misalignment . This misalignment , like the HWR doctrine , pose s a major challenge to Porat ’ s alignment analysis because it is an efficient misalignment , and it is efficient for the same reason that HWR is efficient -- it internalizes externalized benefits . Yet Porat refers to pure economic loss in two words and a short footnote in the Misalignments introduction 78 and then th is misalignment receives no further attention . The existence of these two significant efficient misalignments weakens Porat's claim that tort law is usually aligned (and therefore efficient) and that misalignments are exceptions that should be justified or rectified. Moreover, our impression is that misalignments are much more common and pervasive than Porat acknowledges and th at they may be the rule, rather than the exception. 79 VI . On t he per se Inefficiency of Alignments Jules Coleman observes that “[f]or all the work that Porat requires of the concept of misalignment, it is striking that he offers nothing in way of analy sis of it.” 80 Indeed, Porat provides no theoretical explanation why “ alignment , ” which in itself is just a formal, technical principle of linking the two stages of the tort process, actually promotes efficient deterrence. Instead he uses a case analysis to prove that 78 Porat, supra note 1 , at 88 n. 7. 79 Porat mentions three additiona l misalignments other than the five analyzed. The first “are cases in which policy considerations exclude liability even if the harm in question was caused by wrongdoing” Id. at 88. This is usually done under the duty rubric . Porat highlights in this regard the exclusion of liability for emotional harm from the scope of liability, arguing that this misalignment can be justified on efficiency grounds by the administrative costs that are saved and concerns about faked and exaggerated claims . Id. at 94. The second additional misalignment, which is “harder to analyze , ” is the apportionment of the ex post harm between the injurer and victim under the comparative negligence defense. It is a misalignment whe n each of them is found negligent at the ex ante st age because of the entire expected harm and therefore each should ex post be held liable for the entire harm . Id. at 94 - 95. The third additional misalignment is the imposition of punitive damages, which may also be justified on efficiency grounds . Id. at 9 5 - 96. (We note Mark Geistfeld’s intriguing claim that punitive damages are justified as filling a gap in damages when the harm suffered is irreparable, as is the case with personal injury. See Geistfeld, supra note Error! Bookmark not defined. .) Porat’s identification of these additional misalignments might create the impression that the y and others that are not discussed are also exceptions, often justified, to a general rul e of alignment. We do not believe that is so. The duty of care concept, to start with, excludes from the scope of liability for wrongdoing not just pure emotional harm but a variety of harms which together constitute a large share of the total harms caused by negligent wrongdoing or even the majority of these harms. Excluded harms include purely economic harms , see supra text accompanying notes 76 - 77 , harms caused by pure omissions , and many of the harms caused by governmental authorities. Another wide spread inefficient misalignment that is missing from Porat’s analysis is the apportionment of ex post liability among multiple tortfeasors under rules of comparative contribution . While Porat recognize s that such apportionment between injurer and victim is an inefficient misalignment, he does not mention that such apportionment among joint tortfeasors creates the same kind of inefficient misalignment. Alignment, and efficiency, require that whe n each tortfeasor is found negligent at the ex ante stage for cr eating the entire risk of harm, each tortfeasor should be held fully liable at the ex post stage. This cannot be reconciled with existing rules of comparative fault that apportion the harm among tortfeasors. 80 Coleman, supra note Error! Bookmark not defined. , at ___. 21 misalignments lead to inefficient deterrence. As we have seen, Porat fails to do this with regard to HWR doctrine . Porat ’ s analysis of the offsetting risks misalignment reveals that Porat indeed believes that alignment per se is efficient. Moreover, he adheres to this belief even though it is apparent, based on his own offsetting risks analysis , that alignment per se can be inefficient. Porat writes : “ The restoration of alignment in offsetting risks scenarios could be achieved either by taking those risks in to account when awarding damage or, alternatively, by ignoring them in setting the standard of care . ” 81 This means that if a court disregards the decreased risks not just at the ex post stage but also at the ex ante stage, al ignment is achieved, restored, and all is well. But is it? We do not think so . Porat ’ s doctor - patient example in his offsetting risks analysis refutes that all is well . 82 If at the ex ante stage the court disregard s the decreased risk of 400 when considerin g the increased risk of 500, it would surely find that the doctor was negligent in creating the latter risk and would impose at the ex post stage liability that corresponds with the risk of 500. But , according to Porat’s analysis , that is inefficient as it leads to excessive liability, over - deterrence and defensive medicine. So one is puzzled: on the one h and we have an alignment here -- reduced risks are ignored at both stages -- and Porat tells us that alignments per se are efficient. On the other hand, Porat also tells us that disregarding decreasing risks is inefficient. Both propositions cannot be true. VI. Conclusion The Pigo u vian/Coasian concept of internalizing externalities provides the foundation and the rationale of the economic analysis of tort la w as an efficient deterrent. Porat introduces the intuitively appealing the idea of alignment between t he determination of negligence at the ex ante stage and the imposition of liability at the ex post stage as a criterion to test and promote the efficienc y of negligence law. Focusing on Porat ’ s analysis of two misalignments, the wrongful risk limitation and offsetting risk s, we find that while externalized benefits and excessive liability are taken into account with regard to the latter , in accordance with the internalization concept, they are missing in the analysis of the former . This omission leads Porat to a radical ( and erroneous ) conclusion: liability should be imposed for every ex post harm that results from a foreseeable risk , ev en if a reasonable one, and that the scope of liability doctrine should accordingly be transformed . We have shown, relying on the internalization concept , acknowledged by Porat , t hat the scope of liability doctrine, although a misalignment, contributes to efficient deterrence. It does so by serving as one of tort law ’s benefit internalizers and ther e by reduces or avoids excessive liability and resultant over - deterrence . The pure economic loss limitation on liability , which is essentially absent from Porat's analysis , does exactly the same. Finally, we question Porat’s fundamental principle: the presumption that alignment s are efficient. On the contrary, our a nalysis reveals that at least in those instances when it deviates from the internalization concept , a lignment is contrary to efficient deterrence . 81 Porat, supra note 1 , at 118 (footnote omitted). 82 See supra text accompanying notes 53 - 54 .