Harvard Law Review Harvard Law Review Harvard Law Review


Thoroughly Modern Tort Theory

Funny what passes for “modern.”1×1. Catherine M. Sharkey, Modern Torts: Preventing Harms, Not Recognizing Wrongs, 134 HARV. L. REV. 1423, 1425 (2021). It was around 1967 that the phrase “cheapest cost avoider” first appeared in the work of then-Professor Guido Calabresi.2×2. Guido Calabresi, Changes for Automobile Claims? Views and Overviews, 1967 U. ILL. L.F. 600, 608. Yet we learn from our friend Professor Catherine Sharkey that this half-century-old idea remains so completely “ascendant”3×3. Sharkey, supra note 1, at 1424. that our book’s sustained effort to articulate a distinctive conception of tort law4×4. John C.P. Goldberg & Benjamin C. Zipursky, Recognizing Wrongs (2020). has essentially no practical value. For the practice of “modern” tort law, she insists, really the only important question is how best to implement Calabresi’s principle.5×5. Sharkey, supra note 1, at 1454.

Sharkey’s supposition that a form of economic analysis provides the only way for judges to reason cogently about modern tort problems mainly serves to confirm one of our book’s claims, which is that law professors at elite schools have for far too long embraced the dogma that economic approaches are the exclusive marker of seriousness and sophistication in tort scholarship.6×6. Goldberg & Zipursky, supra note 4, at 44–50, 73–80. In any event, her narrow perspective on the field can only spell doom for our book. Though crediting us for recognizing that a “theory of torts must not only constrain the possible ways of fleshing out tort law’s wrongs, but also must capture and help guide judicial reasoning about what shall count as wrongs,”7×7. Sharkey, supra note 1, at 1429 (quoting Goldberg & Zipursky, supra note 4, at 231). Sharkey concludes that “none of the conceptions of wrongs [we] put forth ultimately delivers.”8×8. Id. Perhaps our “inner morality of law” intuitions exert some “gravitational pull” with respect to intentional torts that operate at the margins of contemporary practice.9×9. Id. at 1437. But they are of no help in core areas, such as products liability.10×10. See id.

This Response aims to counter Sharkey’s charge that we have failed to provide guidance on how today — at the inception of the third decade of the twenty-first century — judges think about or should think about tort law. Along the way, we’ll make the now familiar point that cheapest cost avoider analysis, in part because of its indeterminacy and its ill-suitedness to judicial application, is not likely to generate better decisions in hard cases.

To do so will put us in the awkward position of pointing out that an admired colleague seems to have glossed over an entire chapter of our book — one that carries the transparent if unimaginative title “Applications” and that is directly devoted to anticipating and rebutting her main charge against us.11×11. See Goldberg & Zipursky, supra note 4, at 293–339. In particular, it provides practical guidance on three live issues: psychotherapists’ liability to third parties for emotional distress, sellers’ liability for injuries caused by defectively designed products, and platforms’ liability for internet defamation.12×12. Id. Sharkey deems our work unable to illuminate topics of this sort yet never bothers to address our treatment of them. Perhaps the intellectual chasm that separates her from us is so vast that she concluded there was no point in engaging our analysis. Resisting that depressing thought, we shall try once again to explain why we think Recognizing Wrongs provides an accurate and useful account of “modern torts.”


High academic tort theory has for decades been dominated by two broad schools of thought. As noted, since the late 1960s, many scholars have touted economic analysis as the way to think rigorously about torts. Judge Posner’s early work claimed that one can understand torts only if one appreciates that its key concepts, though dressed up in legal jargon, are economic in substance.13×13. See generally William M. Landes & Richard A. Posner, The Economic Structure of Tort Law (1987). Judge Calabresi, by contrast, wielded economics as a tool of progressive reform — as pointing the way to a version of accident law superior to that provided by settled doctrine.14×14. See generally Guido Calabresi, The Costs of Accidents: A Legal and Economic Analysis (1970).

Sharkey, for her part, combines Posnerian interpretivism with a Calabresian analytic framework, claiming that the cheapest cost avoider principle best explains the reasoning and outcome of tort cases, or at least those decided by sophisticated courts such as the Supreme Court of the United States. Thus, she commences her Review by celebrating the majority and dissenting opinions in Air and Liquid Systems Corp. v. DeVries,15×15. 139 S. Ct. 986 (2019). a negligence case that reached the Court via its maritime jurisdiction.16×16. Id. at 991. According to Sharkey, both the majority and dissenting opinions in that case are exemplary because each asked the right question: Were the defendants the cheapest cost avoiders of the plaintiffs’ injuries?17×17. Sharkey, supra note 1, at 1424. For more on DeVries, see infra pp. 195–98.

The second influential strand of tort theory — corrective justice theory — emerged partially in response to economic analysis. Enlisting Aristotle and Kant, philosophically oriented scholars such as Professors Jules Coleman and Ernest Weinrib18×18. See generally Jules L. Coleman, The Practice of Principle: In Defense of a Pragmatist Approach to Legal Theory (2001); Ernest J. Weinrib, The Idea of Private Law (2012). have argued that tort law is all about holding persons who wrongfully cause losses to their moral responsibility to repair the losses. As such, tort law instantiates a distinct brand of justice. Although varying in their particulars, each of these accounts is largely formal in nature, emphasizing the importance of grasping tort law’s “bipolar” structure. While corrective justice scholars claim that these considerations have implications for judicial decisionmaking, often the emphasis has been on negative implications — on the sorts of considerations that judges should not attend to when making their decisions.19×19. See, e.g., Weinrib, supra note 18, at 36–38 (arguing that judges would render tort law incoherent by including considerations of a defendant’s ability to spread losses as a ground for liability). Other corrective justice theorists have devoted more effort to explaining and justifying substantive tort doctrines. See generally, e.g., Arthur Ripstein, Private Wrongs (2016).

To the bemusement of a very different kind of torts scholar, the corrective justice versus law and economics debate is just so much spilled ink. “A pox on both your houses,”20×20. See William Shakespeare, Romeo and Juliet act 3, sc. 1, l. 90 (“A plague o’ both your houses!”). say the adherents of this theory-skeptical “social welfare” approach.21×21. This is the label that we somewhat hesitantly adopt in our book. See Goldberg & Zipursky, supra note 4, at 44–47. Because its adherents resist the thought that they have adopted any general approach to tort law other than a suitably down-to-earth and skeptical approach, it is not easy to isolate a canonical scholarly statement of this position, though there are countless particular instantiations of it in treatises and scholarly articles. See generally John C.P. Goldberg, Twentieth-Century Tort Theory, 91 Geo. L.J. 513, 521–31 (2003). Tort law, on this view, is a messy business through which courts deliver some compensation, provide some deterrence, dispense some justice, and do some other stuff such that, if all goes well, they will impose liability in a manner that contributes to social welfare, broadly understood. When courts face difficult cases, they muddle through, making what are essentially indeterminate, multifactor policy judgments. To the extent that torts professors writing articles in law reviews can help, it is by providing a candid account of the impact that can be expected from different decisions, thus providing a relatively sober counterbalance to the aggressive spin provided by litigators and amici.

Because she sees everything through her economic lens, Sharkey cannot help but see our book as an attack on economic analysis. This framing miscasts the project. Although it recounts and clarifies previously published criticisms of efficient deterrence and corrective justice theory,22×22. See generally, e.g., John C.P. Goldberg & Benjamin C. Zipursky, Accidents of the Great Society, 64 Md. L. Rev. 364 (2005); Benjamin C. Zipursky, Civil Recourse, Not Corrective Justice, 91 Geo. L.J. 695 (2003). Recognizing Wrongs is not principally a critique of either. Rather, it is the social welfare view that is the book’s main target. This view — which is far and away the most pervasive in the field23×23. See Goldberg & Zipursky, supra note 4, at 44. — relies on theoretical ideas too, and these ideas have for too long escaped serious scrutiny because of the cloak of modest skepticism stylishly worn by those who adopt it. Specifically, the social welfarists offer the theoretical claim that tort law has no integrity but is instead simply the conferral of power on judges to use liability, and the threat of liability, to achieve goals they think are worth achieving.24×24. Id. at 44–46.

The era in which one could, along with the likes of Dean William Prosser, maintain a cheerfully skeptical view of tort law as a welcome conferral on judges of a roving authority to use liability to make the world a better place is long gone. Today, such skepticism only aids a cause — tort reform — that many social welfare theorists purport to deplore.25×25. Id. at 7–10. But our main point is interpretive, not political. Social welfare scholars, we contend, are mistaken to suppose that an immersion in the history, case law, and practice of tort law entails skepticism about its coherence or the embrace of an instrumental, judge-as-quasi-legislator approach to the adjudication of tort cases.26×26. Id. at 11–12. Quite the opposite, we demonstrate that an immersive approach permits one to grasp tort law’s core concepts and animating principles, to see how this body of law fits into our legal and political system, and to appreciate why there are good reasons to retain tort law while relying on courts regularly to revise it.27×27. Id.


Tort law is a law of wrongs and redress. While it might seem banal, this proposition is complex and worthy of careful consideration. Parts I and II of our book unpack its meaning, argue for its interpretive accuracy, and make a case for including such law within our legal system. Part III draws out its implications for modern practice and academic analysis.

Tort law identifies ways in which a person must refrain from treating another (or, more rarely, must assist another). Each of the torts — not just old, musty torts like battery, but also negligence, products liability, and other “moderns” — contains a relational directive that simultaneously specifies duties not to injure others through wrongful conduct and a right in those others not to be so injured. In turn, tort law confers on right-holders who can prove that they have been so mistreated a legal power to obtain the assistance of a court in ordering the duty-bearer to redress his or her wrong.

In a liberal democracy, we claim, government owes it to members of the political community to provide law of this sort.28×28. Id. at 146. We are grateful that Sharkey credits this aspect of the book as making a philosophical contribution,29×29. Sharkey, supra note 1, at 1429. though our gratitude is tempered by the recognition that she likely regards it as more a matter of “soaring prose” than substance.30×30. Id. at 1428. Perhaps greedily, we insist that we have accomplished something more. Our defense of tort law, understood as a law for the redress of wrongs, is not “mere” philosophizing. It is a philosophically informed account of how this well-established component of our legal system fits within the history, law, and positive political morality of our liberal democracy. As much as we have benefited from reading Rawls and Scanlon, it is not in their works that we encountered the idea of government owing community members law for the redress of wrongs. Rather, it was in Blackstone’s Commentaries;31×31. 3 William Blackstone, Commentaries *23. in Jefferson’s account of the Declaration of Independence as an “appeal to the tribunal of the world”;32×32. 7 The Writings of Thomas Jefferson 407 (H.A. Washington ed., Philadelphia, J.B. Lippincott & Co. 1869). in the Fourteenth Amendment’s language of “equal protection” and “due process”;33×33. U.S. Const. amend. XIV, § 1. in Judge Cardozo’s insistence that a tort plaintiff “sues in her own right for a wrong personal to her, and not as the vicarious beneficiary of a breach of duty to another”;34×34. Palsgraf v. Long Island R.R. Co., 162 N.E. 99, 100 (N.Y. 1928). and in the recognition by modern courts of feminist lawyers’ compelling arguments for treating workplace sexual harassment as a distinct legal wrong.35×35. Goldberg & Zipursky, supra note 4, at 30–42, 127–29. Don’t believe us? How about the U.S. Supreme Court? See Mo. Pac. Ry. Co. v. Humes, 115 U.S. 512, 521 (1885) (“It is the duty of every State to provide, in the administration of justice, for the redress of private wrongs . . . .”).

As Part II of the book underscores, appreciating what it means to identify tort law as a law of wrongs and redress is crucial to making sense of the field. Indeed, this understanding permits us to explain various features of the law — from “golden oldies” such as causation requirements and the objectivity of the negligence standard to “contemporary classics” such as constitutional limits on punitive damages — that have flummoxed even experts.36×36. Another example we can’t resist mentioning: Sharkey claims to be at a loss to explain how Judge Cardozo could have embraced liability-friendly cheapest cost avoider analysis in MacPherson v. Buick, 111 N.E. 1050 (N.Y. 1916), only to forego it in Palsgraf, 162 N.E. 99. Sharkey, supra note 1, at 1436 n.46. Yet she never engages the explanation that we offer in our book and elsewhere. See Goldberg & Zipursky, supra note 4, at 200 n.32; John C.P. Goldberg & Benjamin C. Zipursky, The Moral of MacPherson, 147 U. Pa. L. Rev. 1733, 1817–21 (1998). If one avoids the mistake of reading MacPherson as a bit of reductive instrumentalism hidden behind lawyerly verbiage, one will have no trouble seeing how the two decisions hang together. Still, some of our critics have suggested that a tort theory, and especially a theory rooted in the claim that torts are wrongs, needs more. They say it needs an account of which ways of injuring others count, and should count, as “wrongs,” and why.

In this respect, as in many others, our work bears a close resemblance to that of the social welfare scholars, for we too are pragmatists about the content of tort law.37×37. Goldberg & Zipursky, supra note 4, at 73–81 (discussing the strands of pragmatism at work in our account). Before one provides an account of which wrongs are or should count as torts, one must ask why such an account is needed. Our answer to that question is that we need it to help judges who are charged with applying, refining, and redefining the contours of tort law in a manner that is principled and practical.38×38. The phrase “help judges” here is admittedly misleading, for we do not see a lot of judges out there seeking help from academics. The point is rather that, in educating lawyers (some of whom go on to be judges), scholars must bring to bear an understanding of the field that is true to the law and that can sustain the sort of reasoning that judges and lawyers should apply when deciding tort cases. The way judges usually do so, and should do so, is to expound the common law. This is not a matter of finding the objective truth as to which acts really are wrongs, but of constructively carrying forth ideas, principles, and norms that are already in the law. In short, our claim is that we do not need a substantive theory of wrongs if we have a positive account of what the wrongs of tort law are and a normative theory of adjudication appropriate to this domain of law.

Sharkey claims that we make a “fatal concession” by denying ourselves the ability to rule out, in advance, the possibility that all the various wrongs of tort law — assaulting another, defrauding another, or injuring another through the sale of a defective product — boil down to the single wrong of failing to take cost-justified safety precautions.39×39. Sharkey, supra note 1, at 1430. Here she mistakes a feature of the argument for a bug.40×40. In keeping with our pragmatism, when we criticize Judge Posner’s economic account of the wrongs of tort law, we do not do so by ruling it out on philosophical grounds. Rather, we reject it because it is unsustainable as an interpretation of the concepts and rules that we have, and of the ways in which most judges go about the job of deciding tort cases. See John C.P. Goldberg, Torts, in The Oxford Handbook of the New Private Law 261, 264–65 (Andrew S. Gold, John C.P. Goldberg, Daniel B. Kelly, Emily Sherwin & Henry E. Smith eds., 2021) (offering as a partial explanation for the academic influence of economic analyses of tort law that they latch onto but then distort some of its key features). If anything, it is legal economists who bring excessively strong philosophical premises to their accounts of judging. For it is they (among others) who assume that, in hard cases, courts have no choice but to reason instrumentally by determining whether the imposition of liability for a certain kind of harm resulting from a certain kind of conduct will produce more social benefits than costs.

In adopting a pragmatist-inspired approach to explaining tort law’s substance, we are hardly alone. In fact, it is the general approach adopted by some of the most productive theory-skeptical tort scholars of the twentieth century, including the great treatise-writer Prosser.41×41. See William L. Prosser, Handbook of the Law of Torts (1941). Unfortunately, when it has come time to spell out the relevant theory of adjudication, Prosser and his heirs have falsely equated pragmatism with instrumentalism — the idea that the job of judges in hard tort cases is to ascertain the ruling that promises to balance goals such as deterrence and compensation in an optimal manner.42×42. See id. at 10–18. For reasons we explain, instrumentalist theories of adjudication fail to capture the actual substance of our tort law.43×43. Goldberg & Zipursky, supra note 4, at 209–28. Still, it is the right kind of theory, and thus, in its place, we propose a pragmatist alternative that we dub “dual constructivism.” Fancy label aside, this is an account of how judges decide and should decide tort cases. Dual constructivism will look familiar to readers, given that it resembles and indeed builds on accounts such as the one famously offered by Judge Cardozo in The Nature of the Judicial Process.44×44. Id. at 246–55.


By the time intrepid readers reach Part III of Recognizing Wrongs, they will have been presented with arguments for why tort law is best understood as a law of wrongs and redress; why, as such, it is more coherent and defensible than critics have supposed; and how judges have gone about and should go about specifying the content of the various wrongs for which redress is available. The last part of the book aims to demonstrate precisely what Sharkey’s Review denies — namely, that our theory of tort law has practical payoffs for the analysis of issues in contemporary tort law.45×45. Sharkey’s commitment to cheapest cost avoider analysis is so fervent that she claims that pretty much all of tort law’s greatest hits (Palsgraf excepted) are applications of it. This, for example, is her take on Tarasoff v. Regents of the University of California, 551 P.2d 334 (Cal. 1976). See Sharkey, supra note 1, at 1433–34. Her misreading of that case is characteristic of her tendency to regard any judicial mention of “deterrence” as proof that the court is relying on cheapest cost avoider analysis. It is true that the Tarasoff majority supposed that the imposition of liability on therapists for failing to warn third parties of threats posed by their patients promised to make the world a bit safer. See 551 P.2d at 347–48. Yet there is little evidence that it proceeded on the assumption that therapists are the cheapest cost avoiders of violent attacks by their patients. To make the latter determination, the justices would have needed to assess the relative ability of therapists to reduce the incidence of such attacks as compared to various other actors who might (in the case itself and generally) be in a better position to do so, including police, mentors, friends of the assailant, members of the victim’s family, inter alia. See generally Deborah Blum, Bad Karma: A True Story of Obsession and Murder (2013) (providing a dramatized but records- and interview-based account of Tarasoff’s assailant’s interactions with friends, therapists, campus police, and the victim’s brother). In the end, what Tarasoff primarily demonstrates is that any invitation for judges to deploy cheapest cost avoider analysis — given the indeterminacy of that concept, the structure of tort litigation, rules limiting the admissibility of evidence, judicial competence to engage in policy analysis, and various other factors — is an invitation for them to make stuff up.

Chapter 10 takes up three examples: psychotherapist liability for emotional harm caused to a patient’s family member; the definition of design defect in products liability law; and the treatment of those who repost defamatory statements on the internet. While Sharkey supposes that the way to resolve these cases is to ask and answer the cheapest cost avoider question, that is not how the courts have approached them. Indeed, in all three settings, there is often evidence that defendants have engaged in conduct that foreseeably caused serious harm to plaintiffs. In each, moreover, the defendant was relatively well placed to avoid the harm. Thus, from the perspective of cheapest cost avoider analysis, it looks like we have what we need for liability to attach. But that is not how the courts have seen things. Indeed, in each, the aforementioned “features” — causation, foreseeability, preventability — are exactly what create the problem, not what solve it, for there are significant arguments against liability notwithstanding that the defendants are in a position to avoid the harms in question.

Start with the therapist cases. Typically in these cases a young woman undertakes a course of psychotherapy involving unconventional methods such as recovered-memory therapy, and she comes to believe that her emotional problems relate to having been sexually abused by her father when she was a child and having since repressed memories of the abuse. When the allegations of abuse emerge, law enforcement officials investigate, and the father’s life to some degree unravels. If the investigation is dropped for lack of sufficient evidence, the father might sue the therapist, arguing that the therapist was negligent to use the controversial methods that she did, and that this negligence foreseeably caused him serious emotional, financial, and reputational losses. In response, lawyers for therapists have argued that a therapist’s legal duty of care runs only to her patient, not to relatives of the patient, and hence the father’s claim fails as a matter of law.

State appellate courts have come out on both sides of this issue, with the majority holding that there is no duty of care running from therapist to father.46×46. Goldberg & Zipursky, supra note 4, at 295 n.4 (cataloging cases). On first blush — assuming that the reality is that the father was innocent of the alleged abuse — the cheapest cost avoider would seem to be the therapist. But there are also concerns about the impact on future therapists and their patients, the uncertainty of the science, and the alleged conflicts of interest for therapists. How, then, should judges proceed? We argue that they will do better by taking the law seriously, including the courts’ longstanding recognition of fiduciary aspects of the therapist-patient relationship and tort law’s handling, in other settings, of third-party claims against professionals. The answer we defend is that the therapist’s legal duty of loyalty to her patient is best interpreted as undercutting the recognition of a duty of care to avoid causing emotional harm to a patient’s parent, and that therapists thus ordinarily should win these cases as a matter of law.47×47. Id. at 297–302.

As noted, Chapter 10 also addresses liability for defamation via internet postings, and in particular the pressing question of how properly to interpret section 230(c) of the Communications Decency Act of 199648×48. 47 U.S.C. § 230(c); see also Goldberg & Zipursky, supra note 4, at 326–27 (reproducing § 230(c)). (CDA). The issue is whether an individual or website operator/platform (ISP) that actively circulates a defamatory message can avoid liability by showing that someone else posted it first. Almost all courts to address the question say “yes,” but we argue that they have misinterpreted what section 230(c)(1) means when it says that no ISP or user “shall be treated as the publisher or speaker of any information provided by another information content provider.”49×49. § 230(c)(1). Law professors and plaintiffs’ lawyers around the nation have plausibly claimed that online platforms such as Google and Facebook are the cheapest cost avoiders for defamatory statements appearing through their platforms, and on that basis have argued that such platforms should face legal liability.50×50. See, e.g., Danielle Keats Citron & Mary Anne Franks, The Internet as a Speech Machine and Other Myths Confounding Section 230 Reform, 2020 U. Chi. Legal F. 45, 52. Defense lawyers have replied that Congress made the choice to immunize these actors from liability in order to enhance the growth of the internet.51×51. Cf. id. at 51. Drawing from the above-quoted language of the statute, courts have agreed with the defense lawyers.52×52. See Goldberg & Zipursky, supra note 4, at 331–33.

Lost in this high-level theoretical debate is a legally grounded understanding of the statute that permits a more satisfactory and nuanced account of ISP liability. The CDA was indeed aimed at protecting ISPs from liability merely for providing a platform for others’ defamatory statements. Largely in keeping with common law rules, Congress recognized that acting as a conduit for defamatory statements, even a knowing conduit, is (like the post office delivering a defamatory letter) too far removed from actively defaming someone to support the imposition of liability. By contrast, a defendant or platform that recirculates and repeats a defamatory statement and sends it all around the internet — this is an entirely different matter, on which Congress had nothing to say. In other words, under a proper reading of section 230(c), active republication is actionable even if acting as a passive conduit is not.

Cheapest cost avoider analysis gives no purchase on this active/passive distinction, which may help explain why some “sophisticated” courts have purported to find it unintelligible.53×53. See id. at 330–33 (discussing Batzel v. Smith, 333 F.3d 1018 (9th Cir. 2003)). By contrast, we argue that it is perfectly cogent, so long as one interprets the CDA with an appreciation of how the common law of libel actually works. With some modest exceptions, the law has long distinguished republication of another’s defamatory statement (generally a basis for liability) and mere failure to remove such a statement (generally not a basis for liability). In sum, if one reads section 230(c) against its common law backdrop, one comes to see rather quickly that it does not grant blanket “immunity” to ISPs but instead subjects them to liability in cases in which they post, rather than fail to remove, defamatory content.54×54. Id. at 330–39.

The third “applied” topic in Chapter 10 — products liability law — is the area that Sharkey most emphasizes. Yet she essentially ignores our extended treatment of one of the most basic and consequential questions of products liability doctrine — namely, what counts in the eyes of the law as a design defect.

One might presume that, for run-of-the-mill product-related injuries, the cheapest cost avoider is going to be the manufacturer (perhaps regardless of whether the product is defective, and perhaps even regardless of whether the product can be proven to have caused the victim’s injury).55×55. Goldberg & Zipursky, supra note 22, at 378 (noting this feature of Calabresian cheapest cost avoider analysis). If so, courts have never adopted such a simplistic approach. Instead, for decades, they have wrestled with the design defect question, usually framing it as posing a choice between two tests: the consumer expectations test and the risk-utility test.

In Recognizing Wrongs, we argue that a products liability plaintiff should be permitted to have jurors instructed on either test.56×56. Goldberg & Zipursky, supra note 4, at 312. We further argue that jurisdictions should reject the controversial approach adopted in the products liability provisions of the Third Restatement of Torts because it incoherently combines highly pro-defendant features of each.57×57. Id. at 305–06, 314–19 (discussing Restatement (Third) of Torts: Products Liability § 2 (Am. L. Inst. 1998)). While these arguments are anchored in the theoretical work of the first 263 pages of the book, they are explained through a discussion of the California Supreme Court’s pioneering products liability decisions and are illustrated through the examination of a 2015 Texas Supreme Court decision.

Justice Traynor founded strict products liability law in opinions written in two canonical cases: Escola v. Coca Cola Bottling Co.58×58. 150 P.2d 436 (Cal. 1944); accord id. at 440–44 (Traynor, J., concurring). and Greenman v. Yuba Power Products, Inc.59×59. 377 P.2d 897 (Cal. 1963). Greenman held that the wrong of injuring someone through the marketing of a product that fails to live up to its implied warranty of safety belongs in the domain of tort law and therefore should not be hamstrung by commercial warranty law’s technicalities.60×60. Id. at 901. Meanwhile, Justice Traynor’s concurring opinion in Escola maintained that marketing a defective and dangerous product (and thereby injuring someone) is a failure of the seller to act in a manner in which it is obligated to act, such that liability can attach even if the plaintiff is unable to prove a failure to use reasonable care.61×61. Escola, 150 P.2d at 440–41 (Traynor, J., concurring). In sum, whereas Escola casts products liability law as negligence law with a ramped-up duty not to injure people through placing dangerous products in the stream of commerce, Greenman casts products liability law as warranty law without the formalities.

The battle that later emerged between the risk-utility and consumer-expectations tests has essentially been a battle between these two understandings of strict products liability: either it is turbo-charged negligence (in which case risk-utility makes sense), or it is turbo-charged warranty (in which case consumer expectations makes sense). We argue that this framing poses a needless choice. Each is cogently described as a wrong, each embodies values that are in the law of torts more broadly, each can be rendered judicially manageable, and each rings true to understandings of accountability for wrongful injury appropriate to the twenty-first century. Hence either should, in principle, be available.

Nor is this conclusion inconsequential. Focusing on Genie Industries, Inc. v. Matak,62×62. 462 S.W.3d 1 (Tex. 2015). a recent case involving a worker who fell to his death while using the defendant’s aerial lift,63×63. Id. at 3, 6. we criticize the Texas Supreme Court’s reliance on the particular version of the risk-utility test adopted in the Third Restatement.64×64. Restatement (Third) of Torts: Products Liability § 2 (Am. L. Inst. 1998). Applying that test, the court begrudgingly accepted that there was evidence of a superior product design that would have prevented decedent’s death,65×65. See Matak, 462 S.W.3d at 7–9. but it nonetheless ruled that so few accidents had been reported in connection with the lift that the jury was not entitled to find that the failure to adopt the better design rendered it “unreasonably dangerous.”66×66. Id. at 12. The problem with Matak is not just that it seems hell-bent on laying down a doctrinal gauntlet that few plaintiffs will be able to run. It is that, in seeing tort law instrumentally, as providing a set of doctrinal tools to be deployed so as to achieve a sound policy result (apparently, less liability), the court adopted a restrictive test for design defect that is incoherent and unprincipled. The requirement that a products liability plaintiff must, through the provision of expert testimony, “build a better mousetrap” traces back to the negligence side of products liability. Meanwhile, the “unreasonably dangerous” requirement comes from products liability’s warranty side, and specifically its focus on the attributes of the product that consumers can reasonably expect. Thus, rather than issuing a decision true to the distinct warranty and negligence origins and principles of products liability law, the Texas court, following the Restatement, deployed a test for design defect that is the worst of both worlds, an incoherence with obvious practical implications.

The foregoing recitation of our treatment of products liability law permits us to return, full circle, to Sharkey’s starting point — the U.S. Supreme Court’s recent DeVries decision. As we noted at the outset, she presents it as compelling evidence that, when our most enlightened judges face stubborn tort issues, they reach for cheapest cost avoider analysis.67×67. As Sharkey acknowledges, her effort to portray DeVries as a testament to the ability of cheapest cost avoider analysis to resolve hard cases runs into the awkward fact that it is deployed both by the majority and dissenting opinions. See Sharkey, supra note 1, at 1424. The failure of such analysis to reach a determinate result, she assures us, is no reason to doubt the principle’s unique capacity for guidance. It’s just that judges and scholars need a little more time (another fifty years?) to get the hang of it. See id. at 1454. In fact, DeVries makes a better case for the approach to adjudication that we articulate.

It is true that — amid a perfunctory effort to explain why equipment manufacturers owe a duty to warn users of injury risks posed not by the equipment itself, but by component parts that are required to be later integrated into the equipment to render it operable — Justice Kavanaugh’s majority opinion flies the cheapest cost avoider flag. But his invocation is entirely conclusory, relying on nothing more than the unsupported and crucially qualified assertion that the “product manufacturer will often be in a better position than the parts manufacturer to warn of the danger from the integrated product.”68×68. Air & Liquid Sys. Corp. v. DeVries, 139 S. Ct. 986, 994 (2019). Justice Kavanaugh quietly acknowledges the significance of his “often” hedge in an accompanying footnote suggesting that, in some cases, the component manufacturer might well be the cheapest cost avoider. Id. at 994 n.2.

For its part, Justice Gorsuch’s dissent, to borrow poker terminology, sees Justice Kavanaugh’s Calabresi and raises him a Shavell, citing both scholars for the proposition that judges must be cautious about imposing liability on an actor who seems well positioned to take precautions to avoid a harm, because they might thereby disincentivize efficient precaution-taking by other well-positioned actors.69×69. Id. at 997 (Gorsuch, J., dissenting). Here it is important to note that Justice Gorsuch offers this observation only after explaining why, in his view, the majority’s rule departs from applicable precedents.70×70. See id. His main point is not to criticize the majority for having done its sums wrong, thereby misidentifying the cheapest cost avoider. Instead, it is that there is sufficient indeterminacy in the application of economic analysis to a problem of the sort presented by DeVries that the case offers no compelling reason to depart from what Justice Gorsuch took to be the traditional common law rule.71×71. See id. at 998–99. Because her eyes light up at the mere mention of cheapest cost avoider analysis, Sharkey fails to see that Justice Gorsuch is not so much embracing it as using it to “parry” Justice Kavanaugh’s “thrust.”72×72. See Karl N. Llewellyn, Remarks on the Theory of Appellate Decision and the Rules or Canons About How Statutes Are to Be Construed, 3 Vand. L. Rev. 395, 401 (1950) (using the terms “parry” and “thrust” to convey the indeterminacy inherent in the application of canons of statutory construction).

Neither Justice Kavanaugh nor Justice Gorsuch applies economic analysis of law in a manner that explains or justifies the result in DeVries, for neither does any of the work that would need to be done to show who is likely to be the cheapest cost avoider across various scenarios involving equipment that requires add-ons before being used. Instead, as Ivy League–trained lawyers who move in the fanciest legal circles, both use the language of cheapest cost avoider to signal their sophistication.73×73. See DeVries, 139 S. Ct. at 994; id. at 997 (Gorsuch, J., dissenting). Most other courts do not feel compelled to wrap their tort law opinions in economic garb — a fortunate thing given that it lacks substance, determinacy, and fidelity to the concepts underlying our law.

We will go one step further. Although Sharkey offers DeVries as an emblem of what is wrong with our account, we will insist that it actually shows what is right about it. The defendants’ main argument against liability — sometimes referred to as the “bare-metal defense”74×74. Id. at 992 (majority opinion). — strikes a familiar chord in products liability law. How can a plaintiff have a viable claim based on an injury caused by a product if it was not the defendant’s product that actually caused the injury? Isn’t what makes products liability “strict” that the tortiousness lies in the product, not the manufacturer’s conduct? This line of thinking underlies what Justice Gorsuch correctly depicted as the common law’s default support for a no-liability ruling.75×75. See id. at 997 (Gorsuch, J., dissenting). So far as it goes, this reasoning provides the basis for a strong argument against the recognition of a strict products liability failure-to-warn claim in a case such as DeVries.

The trouble for the dissent is that the plaintiffs in DeVries were not relying upon strict products liability law but instead on a claim of negligence, and specifically negligent failure to warn.76×76. Id. at 992 (majority opinion). The focus of this claim was on what the manufacturers said (about the need for insulation when operating their machines) and what they didn’t say (by failing to warn users of possible asbestos exposure). In effect, the plaintiffs were comparing the manufacturers to a physician who recommends that a patient follow a certain diet in conjunction with taking certain prescribed medications, while overlooking the possibility that the patient has an allergy to some of the foods contained in the diet. The fact that the physician has not herself manufactured or sold the foods in question is neither here nor there because the circumstances create an affirmative duty to warn against the risks of an allergic reaction, and it is the failure to warn that is being treated as a cause of such a reaction. This is why, on the DeVries plaintiffs’ negligence theory, it is not decisive that it was another company’s product that actually caused their injuries.

The core issue in DeVries is thus whether tort law as applied to products — which has, drawing on its warranty side, been allowed to expand significantly in plaintiffs’ favor by linking liability to a product’s dangerousness irrespective of whether the seller behaved carefully — should also permit plaintiffs to invoke an expanded, pro-plaintiff version of negligence by building on affirmative duty principles contained in negligence law. Our answer, as explained in the discussion of design defect above, is that both the warranty and negligence sides of products liability law should remain available, so long as there is not incoherent and unsustainable double-dipping. Thus, the interesting and difficult question in DeVries is whether there was sufficient reason for the Court to adopt a relatively robust version of failure-to-warn liability in negligence law as applied to products.

To be fair to Justices Kavanaugh and Gorsuch, both display some appreciation of the issue just described. In particular, both seem to recognize that, while there is a basis in negligence law for a holding in favor of the DeVries plaintiffs, it would require an extension of negligence doctrine.77×77. See id. at 994; id. at 996 (Gorsuch, J., dissenting). Justice Kavanaugh plausibly defends such an extension by: (a) limiting the affirmative duty to cases in which the manufacturer requires the incorporation or use of the dangerous component part; and (b) noting maritime law’s “longstanding solicitude for sailors.”78×78. Id. at 995 (majority opinion). Justice Gorsuch, by contrast, chides the majority for failing to acknowledge the potential breadth of its holding and to provide a principle that would limit it.79×79. See id. at 996–97 (Gorsuch, J., dissenting). Neither identifies and analyzes the issue head-on, however. Justice Gorsuch does not acknowledge the distinctiveness of the negligence claim here and the prima facie plausibility of an affirmative duty of care owed by commercial sellers to certain product users. Justice Kavanaugh does not adequately articulate a principled basis for the Court’s rule, thereby leaving his analysis unduly vulnerable to the dissent’s worry that the imposition of liability on these defendants is an ad hoc response to the plaintiffs’ inability to recover from the “real” culprits, namely, asbestos suppliers (who are insolvent) and the U.S. Navy (which enjoys broad immunity).80×80. See id. at 1000. This opacity is unfortunate. Lawyers are entitled to candor and clarity from this nation’s highest court, and fiddling around with the verbiage of cheapest cost avoider delivers neither.81×81. As Sharkey notes, state high courts are currently wrestling with an issue in some ways analogous to the issue in DeVries. Sharkey, supra note 1, at 1446–48. The state law issue concerns whether a person injured by a generic drug that lacks adequate warnings can sue the manufacturer of the brand-name version of the drug on the ground that federal law requires generic drug manufacturers to provide exactly those warnings that the brand-name manufacturer provides. Plaintiffs have argued that, because generic drug manufacturers are obligated to reproduce verbatim the warnings accompanying the brand-name drug, a failure to warn by the manufacturer of the brand-name drug constitutes a breach of a duty owed to users of generic counterparts. Compare T.H. v. Novartis Pharm. Corp., 407 P.3d 18, 39 (Cal. 2017) (recognizing such a duty), with McNair v. Johnson & Johnson, 818 S.E.2d 852, 867 (W. Va. 2018) (declining to recognize such a duty).

* * *

We certainly had hoped and continue to hope that those who read our book’s treatment of the therapist cases, CDA section 230(c), and products liability law will find it interesting and informative. This said, we never for a moment have supposed that all will be convinced. After all, each raises complex issues of law, principle, and policy.

Contestability is one thing, however. Vacuity is another. And it is something approaching the latter, harsher judgment that Sharkey renders in her Review. As we have explained in this Response, her criticism is grounded in a conceptual framework so narrow, and so dogmatically wielded, as to provide no reason for anyone other than fellow true believers to accept her critique. It is equally grounded in a failure to engage our book’s aspirations, as well as its analyses of the sorts of “modern” topics about which we are supposed to have nothing to say.

One might hope for a day when academic debates about tort law are not cast, simplistically, as posing a choice between “realistic” economic theories, on the one hand, and edifying-but-useless moral theories, on the other. Alas, it seems we are not there yet. Among our fondest — but now we fear extravagant — hopes for Recognizing Wrongs is that it can help point the way toward better conversations.


* Carter Professor of General Jurisprudence and Deputy Dean, Harvard Law School.
** Professor of Law and James H. Quinn Chair in Legal Ethics, Fordham University School of Law.