A Revisionist History of Products Liability

Increasingly courts, including the Supreme Court, rely on ossified versions of the common law to decide cases. This Article demonstrates the risks of this use of the common law. The main contribution of the Article is to demonstrate that the traditional narrative about early products law—that manufacturers were not liable for injuries caused by their products because the doctrine of privity granted producers immunity from suit by the ultimate consumers of their goods—is incorrect. Instead, the doctrinal rule was negligence liability for producers of injurious goods across the United States in the nineteenth century. Courts routinely ignored or rejected privity arguments, and contract was not their paradigm for understanding a producer’s relationship with users of its products. This analysis has implications for how we view the development of the common law today. And it serves as a warning not to rely on potted histories from casebooks in determining what the common law was in the past.


How like a ruin overgrown

With flowers that hide the rents of time,

Stands now the Past that I have known;

Castles in Spain, not built of stone

But of white summer cloud, and blown

Into this little mist of rhyme!

—Henry Wadsworth Longfellow1Henry Wadsworth Longfellow, Castles in Spain, 39 Atlantic 601, 603 (1877), https://cdn.theatlantic.com/media/archives/1877/05/39-235/132122040.pdf [perma.cc/M89S-M8LG].

This Article tells the story of an error that made its way into treatises and casebooks, and became a part of how we understand the history of tort law in America. And it is a warning. Increasingly, courts rely on ideas about what the common law was to decide questions about what the law is today.2See Anita S. Krishnakumar, The Common Law as Statutory Backdrop, 136 Harv. L. Rev. 608, 613 (2022) (demonstrating that the Roberts Court has used the common law to interpret statutes and describing disagreement among the justices as to the content of the common law despite the perception that it consists of “well-settled default rules” (internal quotation marks omitted)); Alexandra D. Lahav, Why Justice Gorsuch Was Wrong About Causation in Comcast, 23 Green Bag 2d 205, 205–06 (2020) (demonstrating that Justice Gorsuch erred in stating the test for causation in the nineteenth century in deciding Comcast Corp. v. Nat’l Ass’n of Afr. Am.-Owned Media, 140 S. Ct. 1009, 1016 (2020)); see also Johnson v. Nat’l Patient Acct. Servs. Sols., L.L.C., 975 F.3d 1244, 1252 (11th Cir. 2020) (relying on precedents from 1882 and 1885 to outlaw the common practice of permitting class representatives to receive an incentive payment under Federal Rule of Civil Procedure 23). When they make mistakes about history, those errors have far-reaching effects.

The traditional narrative of the development of products liability is that the doctrine of privity required parties to have entered into a contract of sale for the consumer to sue the producer for injuries resulting from their product. Privity in this narrative was a “citadel”3Ultramares Corp. v. Touche, 174 N.E. 441, 445 (N.Y. 1931) (“The assault upon the citadel of privity . . . proceed[s] . . . apace.”); William L. Prosser, The Assault Upon the Citadel (Strict Liability to the Consumer), 69 Yale L.J. 1099, 1100 (1960). that was dismantled by Judge Benjamin Cardozo in MacPherson v. Buick Motor Co.,4MacPherson v. Buick Motor Co., 111 N.E. 1050 (N.Y. 1916). a case that was the genesis of products liability law as we know it. Before MacPherson, the story goes, the doctrine of privity largely prevented individuals from suing the manufacturers whose products injured them, except in a narrow set of circumstances.5See G. Edward White, Law in American History, Volume II: From Reconstruction Through the 1920s, at 258, 265 (2016) (stating that “most persons injured in automobile accidents were unable to bring claims against the manufacturers of the vehicles” due to the “so-called ‘privity’ rule” and that the MacPherson decision “signaled that tort law would be the principal arena for personal injury suits by users or consumers injured by defective products” going forward); see also Catherine M. Sharkey, The Remains of the Citadel (Economic Loss Rule in Products Cases), 100 Minn. L. Rev. 1845, 1845, 1851 & n.29 (2016) (noting that “[c]hief” among the “dramatic changes” in tort law was “the fall of the ‘citadel’ of privity”); James R. Hackney, Jr., The Intellectual Origins of American Strict Products Liability: A Case Study in American Pragmatic Instrumentalism, 39 Am. J. Legal Hist. 443, 496 (1995) (“The judicial origins of strict products liability in America can be squarely placed in the MacPherson opinion.”); Robert L. Rabin, The Historical Development of the Fault Principle: A Reinterpretation, 15 Ga. L. Rev. 925, 937 (1981) (“[T]he fault principle gradually overcame limitations based on contract.” (citing MacPherson, 111 N.E. 1050)). Judge Cardozo revolutionized products liability by moving it from contract to tort, magically turning an exception into a rule. As the opinion famously states, in an excerpt that probably appears in every torts casebook in the United States:

We have put aside the notion that the duty to safeguard life and limb, when the consequences of negligence may be foreseen, grows out of contract and nothing else. We have put the source of the obligation where it ought to be. We have put its source in the law.6MacPherson, 111 N.E. at 1053.

While the traditional narrative recognizes exceptions to the privity rule for inherently dangerous products like poisons, it insists that these exceptions were peripheral and that privity was the central rule.7See Prosser, supra note 3, at 1104; Seely v. White Motor Co., 403 P.2d 145, 149 (Cal. 1965) (“An important early step in the development of the law of products liability was the recognition of a manufacturer’s liability in negligence to an ultimate consumer without privity of contract.” (citing MacPherson, 111 N.E. 1050)).

This Article demonstrates that since the advent of mass-market products, the general rule was that manufacturers owed a duty in tort not to injure consumers carelessly, regardless of how the consumers came to use the product. Neither the limitations of contract and private ordering, nor the idea of damnum absque injuria8Damnum Absque Injuria, Black’s Law Dictionary (11th ed. 2019) (“Loss or harm that is incurred from something other than a wrongful act and occasions no legal remedy.”). (a no-liability rule), governed the relationship between the mass producer and end user of common products such as medicine, food, and clothing. Privity was a peripheral doctrine, applied mostly in the context of services rather than products, albeit sometimes (but not always) used to limit injured workers’ legal claims.9See infra Section II.B. The historical narrative of the transition from privity/no liability to negligence is what the poet Longfellow called a castle in Spain, and the “citadel” of privity was but a white summer cloud.

There is a longstanding scholarly debate about whether freedom of contract or regulation dominated the nineteenth century.10See, e.g., Gary T. Schwartz, Tort Law and the Economy in Nineteenth-Century America: A Reinterpretation, 90 Yale L.J. 1717, 1717–19 (1981) (critiquing Morton Horwitz, The Transformation of American Law 1780–1860, at 67–108 (1977)). This Article adds to that debate by demonstrating that regulation (via common law liability) dominated in the area of products liability.11See, e.g., William J. Novak, The People’s Welfare: Law and Regulation in Nineteenth-Century America 1–18 (1996) (explaining that the nineteenth century was a period of substantial local regulation). A thorough investigation of reported state cases between 1850 and 1916 reveals that state appellate courts consistently held that plaintiffs could sue the manufacturers and sellers of products that injured them, whether those plaintiffs purchased the products directly or indirectly. Arguments in favor of privity were certainly made, and a small minority of courts even applied that doctrine.12See, e.g., Wells v. Cook, 16 Ohio St. 67, 67–68, 71 (1865) (finding that a third-party purchaser of sheep, who found out they were diseased after purchase, cannot sue the original seller). Still, in the appellate reports of products cases invoking privity, cases applying the doctrine are substantially outnumbered by cases in which the courts rejected it.13See, e.g., Moon v. N. Pac. R.R. Co., 48 N.W. 679, 679–80 (Minn. 1891) (finding a rail car manufacturer can be liable to a brakeman for a defective railcar it sold to a railroad). Privity was a peripheral doctrine favored by defendants responding to injurious product suits. It was rarely a bar to recovery, especially in cases involving mass-marketed products that caused injuries outside the employment context.

Notably, many (but not all) of the cases in which privity was rejected involved women and children as plaintiffs and products that were used in the home, such as medicine, food, and clothing.14See, e.g., Thomas v. Winchester, 6 N.Y. 397 (1852) (woman nearly killed by medicine); Gould v. Slater Woolen Co., 17 N.E. 531, 531 (Mass. 1888) (woman harmed by clothing dye); Tomlinson v. Armour & Co., 70 A. 314, 316 (N.J. 1908) (woman harmed by canned meat); Olds Motor Works v. Shaffer, 140 S.W. 1047, 1051 (Ky. 1911) (woman injured by defective automobile). It is possible that courts thought that women and children deserved special protection. It may also be the case that recognition that these parties could not contract drove the difference between the decisions described in this Article and those in the employment context where the “free labor” ideology may have been most powerful. This case is hard to make. There is no evidence of special solicitude toward women and children in the opinions, and the treatment of men in many of the consumer cases discussed is the same as that of women.15By contrast, in her work on accident cases during this period, Margo Schlanger finds differences in the opinions that correspond to perceived gender differences. See Margo Schlanger, Injured Women Before Common Law Courts, 1860–1930, 21 Harv. Women’s L.J. 79, 86 (1998); see also John Fabian Witt, From Loss of Services to Loss of Support: The Wrongful Death Statutes, the Origins of Modern Tort Law, and the Making of the Nineteenth-Century Family, 25 Law & Soc. Inquiry 717, 721 (2000) (demonstrating the relationship between wrongful death statutes and gender roles). There is evidence outside the opinions, however, that tort scholarship and tort law have not valued women, and that the subjects of these suits, mostly products for the home, often involved either women or products associated with femininity.16For a discussion on the devaluation of women in tort, see Martha Chamallas, The Architecture of Bias: Deep Structures in Tort Law, 146 U. Pa. L. Rev. 463 (1998). We shall see some of this devaluation in the scholars’ treatment of the subject,17See infra notes 37, 39, 49, 54 and accompanying text. which may explain the longevity of the myth of privity.

The Article proceeds as follows. Part I describes the traditional story. It demonstrates that many of the most highly regarded scholars in the American firmament, from William Prosser, the preeminent torts scholar of his generation, to G. Edward White, one of the leading historians of tort law, have adopted the view that the history of products liability is one of progress from no liability to negligence to strict liability (and then, perhaps back again to negligence) for harmful products. MacPherson plays a crucial role in these histories.18See White, supra note 5, at 258–66; Prosser, supra note 3, at 1100–02. In the traditional narrative, MacPherson is presented as the turning point, when privity was set aside in favor of negligence for dangerous products, and the private ordering of contract was displaced by the public-oriented duty of tort.

Part II retells the story from the beginning, with the failure of privity to take hold in the American legal imagination in the 1850s. It demonstrates that the weight of the case law was always in favor of negligence, described as a “public duty” by courts. Defendants tried, again and again, to argue that privity governed their cases and that they should win on that account. They repeatedly failed. This was true for very dangerous products such as the explosive naphtha and the poison belladonna, but it was also true for products that were only dangerous due to defects, like a folding bed or canned goods. This Part endeavors to show how the dominant liability rule was mischaracterized as an exception to a privity requirement that in fact did not exist. The mischaracterization produced a category error, disguising the availability of redress for defective products.

Part III considers the lessons this corrected history offers modern students of the law and jurists. First, the narrative that the common law of torts evolved from no liability to strict liability is incorrect. There was always liability for dangerous products even though some sectors (manufacturing) pushed defenses (like privity) that never took root. From its inception, the law governing dangerous products was crafted by judges to address, in a flexible way, the social problems that came before them. Second, the lesson for other areas of the law is that American legal history has not been definitively told but bears reinvestigation and retelling. Claims about historical causes of action or defenses should be carefully reviewed, as the accepted wisdom is not always correct, even if it appears to be textbook law. This is especially important today, when courts increasingly refer to ossified versions of the common law in deciding cases.

I. The Traditional Story

The traditional story of products liability law is that its genesis is in 1916 with the New York Court of Appeals decision in MacPherson v. Buick Motor Co.19MacPherson v. Buick Motor Co., 111 N.E. 1050 (N.Y. 1916). Eminent scholars and jurists such as California Supreme Court Justice Roger Traynor all told more or less the same story: products liability came to be when privity of contract fell to tort in 1916.20For the assertion that products liability originates with MacPherson, see, for example, White, supra note 5, at 258–65, describing MacPherson as the genesis of products liability law; Stephen D. Sugarman, A Century of Change in Personal Injury Law, 88 Calif. L. Rev. 2403, 2408 (2000); Hackney, supra note 5, at 496–98; and Rabin, supra note 5, at 937–38, describing MacPherson as an example of the trend to impose liability for defective products. See also Seely v. White Motor Co., 403 P.2d 145, 148 (Cal. 1965). But see John C. P. Goldberg & Benjamin C. Zipursky, The Myths of MacPherson, 9 J. Tort L. 91, 106 (2016) (arguing that MacPherson expressed the existing rule that privity was not a barrier to claims for injurious products). Prior to that time, privity of contract prevented third-party purchasers from suing for injuries caused by dangerous products. Judges imported privity doctrine to the United States through an 1842 English case called Winterbottom v. Wright,21See Winterbottom v. Wright (1842) 152 Eng. Rep. 402. which, the story went, was widely adopted by American courts.

Law students are more or less taught the doctrine in this way.22See, e.g., John C.P. Goldberg, Anthony J. Sebok & Benjamin C. Zipursky, Tort Law: Responsibilities and Redress 60 (4th ed. 2016) (“The New York Court of Appeals’ struggle with the privity rule in negligence was largely resolved in the MacPherson decision . . . . Consider also the new rule of duty his opinion articulates.”); Mark A. Franklin, Robert L. Rabin & Michael D. Green, Tort Law and Alternatives: Cases and Materials 561–62 (10th ed. 2017) (pointing out, after its presentation of the MacPherson opinion, that earlier courts had understood manufacturers to owe a duty only to those with whom they contracted and that the “doctrine of MacPherson came to be accepted generally throughout the United States”). Indeed, until doing the research for this Article, this was the narrative I taught my torts students.23I herewith apologize to them all. Only one article, by John C.P. Goldberg and Benjamin Zipursky, recognizes the truth about privity.24See Goldberg & Zipursky’s discussion:

Cardozo’s MacPherson opinion is thus easily and rightly taken at face value. It says, with powerful precedential support, that New York common law in 1916 was most cogently interpreted as requiring the manufacturer of a product to exercise due care towards persons beyond those in privity, if it is the kind of product that would seriously endanger life and limb if defective.

Supra note 20, at 106. Their article is based on a reading of MacPherson itself, rather than on a survey of the relevant cases predating it, and it is mostly concerned with how it has influenced scholars’ views on the relationship between strict liability and negligence. See id. at 98–99. What this Article adds to the conversation is a historical doctrinal analysis that supports the proposition that negligence was at the center of products liability from the beginning, proving finally and without a doubt that there was no regime of no liability for product defects in the United States in the nineteenth century. The Article also explains the social and economic context for these legal developments, demonstrating how doctrine grows within that context rather than being independent of it.
That article focuses on how MacPherson was received. This one, by contrast, explains how the law evolved in the nineteenth century, providing a revisionist history of the period that departs from that offered by eminent doctrinal scholars, such as William Prosser, and legal historians, such as Morton Horwitz, Robert Rabin, G. Edward White, and John Witt.25See sources cited supra note 5. But see Novak, supra note 11, at 1–18 (demonstrating that the nineteenth century was a period of substantial local regulation). Friedman disputes how effective this regulation was but recognizes that privity was not necessarily the rule. See Lawrence M. Friedman, A History of American Law 344 (3d ed. 2005). We begin with the doctrinalists because they first constructed the narrative.

A. The Doctrinalists

The architects of the privity just-so story were the most famous early twentieth-century tort scholars, although not all their names are top of mind today. For example, in 1925, Lester Feezer argued in favor of a general negligence rule along the lines of MacPherson. He understood privity to have been a rule eroded over time through numerous exceptions.26Lester W. Feezer, Tort Liability of Manufacturers and Vendors, 10 Minn. L. Rev. 1, 8 (1925) (stating that “[e]xception after exception has become as well recognized as the general rule and some of the more recent cases seem to forecast that day when the old rule will be abandoned altogether”). Similarly, in a 1929 article, the scholar Francis Bohlen, who was also the reporter for the American Law Institute’s Restatement (First) of the Law of Torts, explained that “American cases prior to 1903 had uniformly approved the views expressed by Lord Abinger in Winterbottom v. Wright as applicable to determine the normal liability of a manufacturer-vendor.”27Francis H. Bohlen, Liability of Manufacturers to Persons Other Than Their Immediate Vendees, 22 Law. & Banker & Cent. L.J. 291, 300 n.20 (1929); see also Restatement (First) of Torts (Am. L. Inst. 1934) (listing Bohlen as the reporter in the front matter). That is to say, they had uniformly accepted the rule that a manufacturer was only liable to those with whom it had a direct contractual relationship and the cause of action arose out of that contract. Bohlen then lists two “exceptions” that swallow the rule: (1) liability for products that are known to be dangerous but mislabeled and (2) liability for products that are not ordinarily dangerous but designed in such a way as to be injurious.28Bohlen, supra note 27, at 300 n.20. To Bohlen, these exceptions made little sense, which was why it was so important that then-Judge Cardozo in his opinion in MacPherson “cut through all the superficialities and absurdities” and “reject[ed] once and for all the idea that only certain classes of articles can be ‘imminently’ dangerous.”29Id. at 305–06. That opinion’s influence, he said, increased the speed at which courts rejected limits on liability for manufacturers.30See id. at 305–07.

Judge Cardozo himself played some role in this narrative. He invoked it in a 1931 case, stating that “[t]he assault upon the citadel of privity is proceeding in these days apace.”31Ultramares Corp. v. Touche, 174 N.E. 441, 445 (N.Y. 1931). The metaphor that privity was a fortress being attacked by, one presumes, the armies of true justice, was so evocative that it was used in numerous articles in the middle of the twentieth century. Fleming James, one of the foremost authorities on tort law in the 1950s, described the story in similar terms:

Half a century ago judges were reluctant to apply this general test to the man who made a chattel for someone else. Generally there was a contract or agreement, and the duty of care was long thought to be limited to the parties thereto, even though it was the most likely thing in the world that someone beyond the circle of privity would be hurt by any dangerous defects in the thing supplied. This older restrictive doctrine was well adapted to protect the manufacturer from burdens on his activity, but it did so at the expense of the victims of his mistakes. The citadel of privity has crumbled, and today the ordinary tests of duty, negligence and liability are applied widely to the man who supplies a chattel for the use of another.32Fleming James Jr., Products Liability, 34 Tex. L. Rev. 44, 44 (1955) (internal citations omitted); see G. Edward White, Tort Law in America: An Intellectual History 89 (1980) (describing Fleming James).

William Lloyd Prosser, reporter of the Restatement (Second) of Torts and a scholar whose influence cannot be overstated, called his famous article The Assault Upon the Citadel (Strict Liability to the Consumer).33Prosser, supra note 3, at 1099–100. For a discussion of Prosser’s role in American tort law and his narrative about strict liability in particular, see Kenneth S. Abraham, Prosser’s The Fall of the Citadel, 100 Minn. L. Rev. 1823, 1824–25 (2016). Abraham argues that Prosser did not foresee, or address, the problems of design defect that would become much more important in products liability law. Id. at 1825. Prosser implied that Winterbottom v. Wright was immediately adopted in the United States. Thereafter, he wrote,

[t]he courts began by the usual process of developing exceptions to the “general rule” of nonliability to persons not in privity. The most important of these was that the seller of a chattel owed to any one who might be expected to use it a duty of reasonable care to make it safe, provided that the chattel was “inherently” or “imminently” dangerous. In 1916 there came the phenomenon of the improvident Scot who squandered his gold upon a Buick, and so left his name forever imprinted upon the law of products liability. Cardozo, wielding a mighty axe, burst over the ramparts, and buried the general rule under the exception.34Prosser, supra note 3, at 1100.

Food had always been subject to strict liability for direct sales according to Prosser.35Id. at 1103–04. “The extension of the strict liability to third persons with whom the seller had made no contract” in other contexts “came after the turn of the century.”36Id. at 1104. This was after “a prolonged and violent national agitation over defective food, which at times almost reached a pitch of hysteria.”37Id. at 1104–05 (footnote omitted). Prosser then argued that in the 1950s strict liability had been extended from food to other products.38Id. at 1111–12. To pause on Prosser’s language here, it is notable that many of the foremost activists in this area were women and that the accusation of hysteria is a fairly traditional way of diminishing genuine concerns.39To his credit, Prosser focuses most of his critique on the muckraking journalists, especially Upton Sinclair. See id. at 1104–06. For a discussion on the issue of women’s exclusion from torts, see Chamallas, supra note 16, at 464. Margo Schlanger describes the exclusion of women from the scholarship on torts during the midcentury period in greater detail. See Schlanger, supra note 15. The concerns, contrary to the implication, were very real. Children died from poisoned milk by the thousands in the late 1800s, for example.40 Deborah Blum, The Poison Squad 2 (2018) (describing allegations that “thousands of children were killed in New York City every year by dirty (bacteria[-]laden) and deliberately tainted milk”); Richard A. Meckel, Save the Babies: American Public Health Reform and the Prevention of Infant Mortality 1850–1929, at 66 (1990) (describing food adulteration concerns in the 1870s).

In sum, twentieth-century doctrinal legal scholars such as Bohlen, James, and Prosser told a story about how privity had been the law of the land and was gradually eroded, and the narrative stuck.

B. The Historians

Historians, by contrast, have not been much concerned with what legal regime regulated products in the nineteenth century. Their interest in products liability has been largely focused on the twentieth century, which partially explains the attention to MacPherson. Decided in 1916, and by a Great Judge at that, it provided a convenient opening for the new century’s battles over fault versus strict liability. Thus, for example, White began his history of products liability with MacPherson in the second volume of his encyclopedic history of American law, although the volume itself governs the period from Reconstruction through 1920.41 White, supra note 5, at 258–65. Similarly, Morton Horwitz, in his influential history of American law which began with 1870, only discussed MacPherson.42See Morton J. Horwitz, The Transformation of American Law 1870–1960, at 62 (1992). In a very short passage, Horwitz characterized MacPherson as “[c]autiously embracing one of the most radical and controversial opinions of a late-nineteenth-century English judge, who suggested that everybody owed a duty to the entire world not to be negligent.”43Id. Echoing Prosser and Judge Cardozo himself, Horwitz explained that Cardozo “directly attacked the citadel of privity.”44Id.

By contrast, Lawrence Friedman, in his sweeping history of American law, minimized the role of law altogether during this period with respect to products, although he recognized that privity was not a barrier to suit.45Accord Meckel, supra note 40, at 67 (describing lax regulation of milk adulterants). He mentioned criminal penalties for adulterating food and stated that “anybody could sue a seller whose goods were shoddy, or whose food products made you sick.”46. Friedman, supra note 25, at 344. But they didn’t, because “who would or could go to court over a single can of rotten peas?”47Id. Although the sparse historical studies of filings show few suits, these tend to be based on small samples that cannot reliably tell the whole story. Enough people did sue that there are numerous appellate reports of such suits.48The most important of these are Lawrence M. Friedman & Robert V. Percival, A Tale of Two Courts: Litigation in Alameda and San Benito Counties, 10 Law & Soc’y Rev. 267, 281–82 tbls.3 & 4 (1976) (a sample of suits from Alameda and San Benito counties in 1890 contained almost no products liability suits); and Robert A. Silverman, Law and Urban Growth: Civil Litigation in the Boston Trial Courts 1880–1900, at 106 (1981) (a sample of cases from Boston included almost no products liability suits). I don’t think Friedman quite appreciated how dangerous a can of peas could be. In 1885, eight-year-old Mary Martin died because she ate half a canned pickle preserved with copper.49See Poison in Pickles, N.Y. Times, July 3, 1885, at 4, https://timesmachine.nytimes.com/timesmachine/1885/07/03/103026599.html?pageNumber=4 [perma.cc/HP2K-WPYS] (describing a young child poisoned by copper in pickles and calling on prosecutors to bring criminal charges).

Like other historians, Friedman was more concerned with the development of the law of torts around industrial accidents.50See Friedman, supra note 25, at 350–51 (beginning his chapter on torts by discussing industrial accidents). The same was true of John Witt’s wonderful book, The Accidental Republic.51See John Fabian Witt, The Accidental Republic 4370 (2004). The traditional story of labor-related injuries from accidents was of a move from strict liability to fault and even no liability in some cases.52 Friedman, supra note 25, at 356. The historical consensus has been that law was harsh when it came to workers in the newly industrializing United States. Friedman, for example, wrote that the “general contours of nineteenth-century tort law” were as follows: “The thrust of the rules, taken as a whole, came close to the position that businesses, enterprises, and corporations should be flatly immune from actions for personal injury.”53Id.

None of the historians so far mentioned have written much about the laws governing defective products in their histories of the nineteenth century, and to the extent that they did, they insisted that law was essentially absent or very weak. They viewed tort history through the lens of labor rather than the consumer or the home. This was also true of scholars focused on products liability more directly, such as Sally Clarke—who focused on the twentieth century and particularly automobiles.54See Sally H. Clarke, Unmanageable Risks: MacPherson v. Buick and the Emergence of a Mass Consumer Market, 23 Law & Hist. Rev. 1, 1–3 (2005) (tracing the history of liability for defective automobiles).

This should not be surprising. The growth of manufacturing and mass-market goods were twentieth-century phenomena, as was the idea that consumer spending was the primary driver of the economy, which in turn drew increased attention to the legal rules governing injurious products.55See Lizabeth Cohen, A Consumers’ Republic: The Politics of Mass Consumption in Postwar America 9–10, 22–23, 31 (2003) (describing the economic, social, and political developments around mass consumption in the United States after 1945). It would make sense to place the genesis of products liability with the first mass-marketed machine—the car—at the start of the twentieth century, and in a decision rendered by one of the twentieth century’s greatest judges.

In fact, this narrative is in error. Goods that legal historians do not seem to care as much about, such as medicine, canned food, soap, and clothing, were the focus of litigation during this early period, and the standard applied was negligence.

II. Retelling the Story

This Part retells the story of products liability law. It begins with a little background to help the reader understand the massive changes to American economic life during the period. Then it discusses the source of the privity requirement, the English decision Winterbottom v. Wright.56See Winterbottom v. Wright (1842) 152 Eng. Rep. 402, 404. Winterbottom had a much lighter hold on the American legal imagination than modern tort scholars realize. A close reading of the cases shows that Winterbottom failed to thrive. Rather, the story should be retold, focusing on a more influential decision with far-reaching implications: Thomas v. Winchester.57. Thomas v. Winchester, 6 N.Y. 397 (1852). Several cases illustrate the application of Thomas to products, including medicines, clothing, and canned goods. Finally, we arrive at the case that most people say begins the story: MacPherson. The common thread is that case after case rejected privity. Privity was an argument defendants made, to be sure. We know this because it is raised and discussed in a number of opinions.58See, e.g., id. It just wasn’t very successful. More successful in defeating liability were arguments about foreseeability and causation, that is, arguments that are squarely in the domain of torts and should be familiar to modern readers.

A. Background

It is difficult to comprehend the massive social and economic upheaval of the period between 1850 and 1910. In 1840, only 10% of the American population lived in communities with over 2,500 residents.59 Jonathan Hughes & Louis P. Cain, American Economic History 354 tbl.17.6 (8th ed. 2011). By 1910, 46% lived in such communities. 60Id. Many of the people living in these cities were foreign born. Twenty-four million people immigrated to the United States between 1840 and 1914.61Id. at 323. They settled mostly in cities. By 1890, about half of the urban population (53%) was foreign born.62Id. at 329. Why cities? People settled where wages were highest.63See id. at 333. Unlike the denizens of the United States before the Civil War, these urban dwellers needed to purchase food, clothing, and furnishings. And, especially, Americans loved medicines.64See James Harvey Young, The Toadstool Millionaires: A Social History of Patent Medicines in America Before Federal Regulation 147, 150 (1961); J. Worth Estes, The Pharmacology of Nineteenth-Century Patent Medicines, 30 Pharmacy Hist. 3 (1988) (describing the immensely profitable work of a patent medicine inventor, even where the medicine was unproven); see, e.g., Davis v. Guarnieri, 15 N.E. 350 (Ohio 1887) (products liability suit relating to mislabeled medicine brought by an Italian immigrant).

Unlike the farm workers who had dominated a much more rural United States in the earlier part of the nineteenth century, urban dwellers could not be expected to make products to fill their basic needs. They bought medicine, factory-made clothing, and canned food.65See Harold U. Faulkner, The Decline of Laissez Faire, 1897–1917, in 7 The Economic History of the United States 3, 141, 145–46 (Henry David et al. eds., 1951). And people were able to buy more than ever before. Average people could buy twice as much in 1914 as they could in 1860.66 Hughes & Cain, supra note 59, at 349.

The latter half of the nineteenth century was also a period of remarkable technological development that made products more widely available farther afield. Refrigeration and meatpacking were developed in the 1870s.67Id. at 347. Meat and other foods could travel farther, especially with the new network of railroads able to bring goods across the country.68See id. at 285. By 1910 there was more than ten times as much railroad track as there had been in 1860.69Id. In 1860, there were about 30,000 miles of railroad track in the United States; by 1910, that number had risen to a little over 350,000.70Id. Armour & Co., an Illinois company, sold canned meat that was ultimately purchased in New Jersey,71See Tomlinson v. Armour & Co., 70 A. 314, 316 (N.J. 1908). and soap in Wisconsin.72See Hasbrouck v. Armour & Co., 121 N.W. 157, 158 (Wis. 1909). Technological developments allowed for the increase in both domestic sale and export of manufactured food. The exports of manufactured food grew from $39 million in 1860 to $293 million in 1914.73 Hughes & Cain, supra note 59, at 351. Exports of finished goods grew by a factor of twenty, increasing from $36 million in 1860 to $725 million in 1914.74Id.

The number of other mass-market goods that could be sold in interstate commerce grew. Standardized menswear was developed as automatic power looms and better sewing machines aided in creating a market for mass-produced clothing.75Id. at 347. This clothing traveled across the country. A resident of Michigan could buy a coat with a fur collar made in New York.76Gerkin v. Brown & Sehler Co., 143 N.W. 48, 48–49 (Mich. 1913). Machines also traveled long distances, especially starting at the beginning of the twentieth century. MacPherson itself involved an automobile made in Michigan and sold to a New York purchaser.77MacPherson v. Buick Motor Co., 111 N.E. 1050, 1051 (N.Y. 1916).

Courts deciding products cases in this fast-changing environment developed with the times. When it came to mass-market products, particularly what economists call “credence goods”78“A good is a credence good if the consumer cannot readily determine its quality by inspection or even use, so that he has to take its quality ‘on faith.’ ” Richard A. Posner, An Economic Approach to the Law of Evidence, 51 Stan. L. Rev. 1477, 1489 (1999); see also Lee v. Carter-Reed Co., 4 A.3d 561, 579 (N.J. 2010) (using a similar definition in a drug products liability opinion).—the type of product that the consumer could not evaluate the quality of by herself—courts found liability for foreseeable injurious defects.

B. The Case of the Collapsing Carriage: Winterbottom v. Wright

The case most commonly relied on when explaining why products liability actions were largely nonexistent in nineteenth-century America is the English case of Winterbottom v. Wright.79Winterbottom v. Wright (1842) 152 Eng. Rep. 402, 404. The most important thing to know is that it was an employee case.80See id. at 403. A carriage driver named Winterbottom was hired to deliver mail by the postmaster general.81See id. The postmaster provided the carriage, and Winterbottom used that carriage to deliver the mail.82See id. But something was wrong with the carriage; it collapsed and Winterbottom was injured.83Id. He tried to sue the carriage supplier directly, alleging that the carriage was defective.84See id. The Court of Exchequer ruled in 1842 that Winterbottom could not sue the carriage supplier because he had no direct relationship with it.85See id. at 404–05. Only the postmaster, who had contracted for the defective carriage, had grounds to sue the carriagemaker for its shoddy product. This rule was called the rule of “privity.” It meant that only those who had contracted directly with a manufacturer or seller could sue for injuries caused by the product.86There is a question as to why Winterbottom was decided the way it was, and whether it did not mean to create a broad privity doctrine at all but rather was a result of procedural limitations. See Vernon Palmer, Why Privity Entered Tort—An Historical Reexamination of Winterbottom v. Wright, 27 Am. J. Legal Hist. 85, 85 (1983). Palmer argues that the privity doctrine was initially meant to address the problem of concurrence, that is, the problem of bringing both a contract and tort action arising out of the same events, which was a problem in the writ system as it could lead to contradictory decisions or double counting. Id. Only later, he argues, did privity become an independent limit on tort doctrine. Id. This Article does not delve into the true meaning of Winterbottom, only how it was used by courts in the United States. This was because the only duty owed was a contractually created duty running from vendor to purchaser, that is, between the carriage supplier and the postmaster. It was, the court explained, a necessary rule to avoid unending liability.87Winterbottom, 152 Eng. Rep. at 404–05 (“Unless we confine the operation of such contracts as this to the parties who entered into them, the most absurd and outrageous consequences, to which I can see no limit, would ensue.”).

Winterbottom was influential among treatise writers and scholars in particular.88See, e.g., C.G. Addison, Wrongs and Their Remedies, Being a Treatise on the Law of Torts 706 (1860); Francis Wharton, A Treatise on the Law of Negligence 374 (2d ed. 1878). Its holding was interpreted by some to mean that the law of injuries and accidents was governed by contracts alone—not by a separate body of law for accidental injury.89See, e.g., Recent Case, Torts—Contracts—Duty to Third Parties., 13 Harv. L. Rev. 226, 226 (1900) (referencing Gorrell v. Greensboro Water Supply Co., 32 S.E. 720 (N. C. 1899)). But see Note, The True Liability of a Manufacturer for Latent Defects., 6 Harv. L. Rev. 261 (1893) (critiquing the privity rule). Yet there were many who did not see Winterbottom this way. Consider, for example, a 1904 torts casebook which stated that the general rule for injurious products was tort liability, not privity, and did not include an independent discussion of Winterbottom.90 Francis R.Y. Radcliffe & J.C. Miles, Cases Illustrating the Principles of the Law of Torts 408 (1904). This casebook was criticized in the Harvard Law Review for its statement of the rule, but it demonstrates that some scholars recognized what courts were really doing.91Book Note, Cases Illustrating the Principles of the Law of Torts., 18 Harv. L. Rev. 159, 159–60 (1904).

The first case in the United States to cite Winterbottom was decided that same year, 1842. It articulated what came to be known as the fellow-servant rule, which held that an employer is not liable to an employee for an injury caused by another employee.92Farwell v. Bos. & Worcester R.R. Corp., 45 Mass. (4 Met.) 49, 59 (1842) (“[I]t is the ordinary case of one sustaining an injury in the course of his own employment, in which he must bear the loss himself, or seek his remedy, if he have any, against the actual wrong-doer.”). That case, Farwell v. Boston & Worcester Rail Road Corp.,93Id. was decided by one of the most famous jurists of the period, Justice Lemuel Shaw of the Massachusetts Supreme Judicial Court.94See generally Leonard W. Levy, The Law of the Commonwealth and Chief Justice Shaw (1957). It had nothing to do with an injurious product. Rather, it involved a switch operator employed by a railroad who failed to switch tracks, harming the engineer running the train cars.95The case explains that:

[T]he plaintiff, then being in the employment of the defendants, as such engine-man, and running the passenger train, ran his engine off at a switch on the road, which had been left in a wrong condition, (as alleged by the plaintiff, and, for the purposes of this trial, admitted by the defendants,) by one Whitcomb, another servant of the defendants, who had been long in their employment, as a switch-man or tender.

Farwell, 45 Mass. (4 Met.) at 49.
The same is true for many of the subsequent cases citing Winterbottom.

Prior to 1916, 110 state court cases include a reference to Winterbottom v. Wright by name either in the case report or the reporter’s summary of counsel’s arguments96This statement is based on a Westlaw search for the case name (Winterbottom /2 Wright) among all cases in the state courts database prior to 1916. Of those, twenty-five involve what might be described as products, although they include livestock and faulty construction and therefore do not concern mass-market products that we associate with products liability law today. Note how small a number that is: Winterbottom was simply not being cited in every products case, which supports the view that it was not a strong precedent on this question.

Twelve of the twenty-five products cases citing Winterbottom do not involve injury in the course of employment. Of those twelve, four squarely uphold privity: a suit against a contractor for a falling awning,97McCrorey v. Thomas, 63 S.E. 1011, 1013 (Va. 1909). a broken courthouse elevator,98Simons v. Gregory, 85 S.W. 751, 753 (Ky. 1905). steel supports in a building,99Galbraith v. Ill. Steel Co., 133 F. 485, 489 (7th Cir. 1904). and a passenger injured by a defective railroad axel.100Hegeman v. W. R.R. Corp., 13 N.Y. 9, 27 (1855). One last case involving the sale of sickly sheep was largely decided based on a proximate cause analysis but nevertheless stated that a third-party purchaser cannot sue the original seller.101Wells v. Cook, 16 Ohio St. 61 (1865). Seven more of the nonemployee cases reject privity, including cases involving erroneously labeled medicine102See Thomas v. Winchester, 6 N.Y. 397, 397 (1852); Fisher v. Golladay, 38 Mo. App. 531, 542 (Ct. App. 1889); Peters v. Johnson, 41 S.E. 190, 193 (W. Va. 1902). and defective furniture,103See Lewis v. Terry, 43 P. 398, 398 (Cal. 1896). carriages,104See Glenn v. Winters, 40 N.Y.S. 659, 662 (App. Div. 1896). boilers,105See Pittsfield Cottonwear Mfg. Co. v. Pittsfield Shoe Co., 53 A. 807, 811 (N.H. 1902). and soap.106See Hasbrouck v. Armour & Co., 121 N.W. 157, 161 (Wis. 1909). Note how most of these cases involve mass-produced products, whereas the cases embracing privity involve building construction or specialized, installed machinery.107See, e.g., Losee v. Clute, 51 N.Y. 494, 496 (1873) (finding that privity bars suit by a third-party landowner against a boiler manufacturer when the boiler exploded). Notably, Losee did not cite Winterbottom. This well-known case involved a custom product used by a manufacturer rather than a mass-marketed product. One can conclude from this that the mass-product cases prior to MacPherson explicitly considering the Winterbottom precedent rejected privity.

Of the thirteen cases involving injury from products in the course of employment, seven were allowed to proceed even though they would have been barred by privity.108See Keegan v. W. R.R. Corp., 8 N.Y. 175, 175 (1853) (railroad employee injured by defective boiler may sue manufacturer); Moon v. N. Pac. R.R. Co., 48 N.W. 679, 681 (Minn. 1891) (railroad employee may sue manufacturer for defective break); Empire Laundry Mach. Co. v. Brady, 60 Ill. App. 379, 384 (App. Ct. 1895), aff’d, 45 N.E. 486 (Ill. 1896) (defective laundry machine); Huset v. J.I. Case Threshing Mach. Co., 120 F. 865, 872–73 (8th Cir. 1903) (worker injured by defective threshing machine); Kahner v. Otis Elevator Co., 89 N.Y.S. 185, 189 (App. Div. 1904), aff’d, 76 N.E. 1097 (N.Y. 1905) (employee injured by defective elevator); Leas v. Cont’l Fruit Express, 99 S.W. 859, 861 (Tex. App. 1907) (railroad brakeman can sue manufacturer of defective handhold); Davidson v. Montgomery Ward & Co., 171 Ill. App. 355, 367 (App. Ct. 1912) (suit for defective balance wheel on saw). Six were rejected on privity grounds.109See Finan v. Valvoline Oil Co., 100 N.Y.S. 1087, 1089 (Sup. Ct. 1906); Kuelling v. Roderick Lean Mfg. Co., 84 N.Y.S. 622, 629 (App. Div. 1903); Mo., Kan. & Tex. Ry. Co. v. Merrill, 70 P. 358, 361–63 (Kan. 1902); McCaffrey v. Mossberg & Granville Mfg. Co., 50 A. 651, 653 (R.I. 1901); Ziemann v. Kieckhefer Elevator Mfg. Co., 63 N.W. 1021, 1022–23 (Wis. 1895); Hoosier Stone Co. v. Louisville & Chi. Ry. Co., 31 N.E. 365, 367 (Ind. 1892). Thus, even cases involving injury by a product in the course of the employment relationship were just as likely to go forward as to be dismissed for lack of privity.

Overall, in the sample of state cases that invoked Winterbottom, which is the subset of cases most likely to impose privity, courts were more likely than not to reject privity and proceed with a negligence analysis. In all cases of mass-produced products outside the employment relationship, courts rejected privity. Cases involving products did not have to cite Winterbottom, and not all did.110See text accompanying notes 113, 114.

Notably, many of the other state cases invoking Winterbottom—which is to say the bulk of the 110 cases in the sample—involved services such as title searches, will drafting, and telegram delivery.111The strangest of the cases upholding privity in the context of contracts for services was Dennis v. Larkin, 19 Iowa 434 (1865), in which a man who was injured in the Civil War sued a man who had contracted to enlist in place of another and instead fled to Canada. The court held that the man who ran off, Larkin, could only be sued by the man who had contracted with him to avoid service, not by the person called up in his stead who was subsequently injured. Id. at 437; see also State v. Harris, 89 Ind. 363, 364 (1883) (holding that a public officer is not liable to a mortgagee for failure to collect taxes because there was no privity between the officer and the mortgagee); Dundee Mortg. & Tr. Inv. Co. v. Hughes, 20 F. 39, 43 (C.C.D. Or. 1884) (holding that a lawyer cannot be sued by mortgage company for error in title search because there was no privity of contract between the lawyer and the company); Dustin v. Radford, 23 N.W. 715, 716–17 (Mich. 1885) (client cannot sue his lawyer’s agent for failure to collect funds because of a lack of privity between client and agent); W. Union Tel. Co. v. Schriver, 141 F. 538, 543 (8th Cir. 1905) (finding that a telegraph company owes no duty to recipient, only to sender). These consistently (albeit not uniformly) held that a third party has no cause of action against the service provider.112But see Anniston Cordage Co. v. W. Union Tel. Co., 49 So. 770, 770 (Ala. 1909) (finding that a telegraph company can be liable to the addressee of telegram even though there is no privity between them).

The reasonable conclusion from this analysis is that privity was the rule for services, but not for products. Even in the employment context, privity as a bar to suit was at best a controversial doctrine.

C. The Case of the Mixed-Up Medicines: Thomas v. Winchester

The most important and widely cited case involving injurious products in the second half of the nineteenth century was Thomas v. Winchester,113Thomas v. Winchester, 6 N.Y. 397 (1852). involving the sale of a poison as a medicine. Winterbottom was cited 110 times prior to 1916 while Thomas was cited 194 times in the reports during the same period.114These results are based on a search in the Westlaw state decision database including the full case name (Thomas /2 Winchester) before 1916. None of these cases indicate the race of the litigant, which indicates that they were likely all white. See Jennifer B. Wriggins, Damages in Tort Litigation: Thoughts on Race and Remedies, 1865–2007, 27 Rev. Litig. 37, 44 (2007) (discussing how, in the early period, courts only noted race of litigants when the litigant was a person of color). Thomas was a much more powerful decision in the American legal imagination, yet it receives relatively little attention. A more accurate narrative of the history of products liability would center Thomas and move Winterbottom to the periphery.115As, indeed, some did. See Book Note, supra note 91, at 160.

It was the early spring of 1849 and Mary Anne Thomas was sick.116See Transcript of Record at 4, Thomas v. Winchester, 6 N.Y. 397 (on file with author). The facts in the following paragraphs are taken from the trial transcript of Thomas v. Winchester. She lived in Cazenovia, a pretty town on a lake, close to Syracuse, New York and about 250 miles north of Manhattan.117See id. at 1. Her brother, George, was visiting the Thomas family that day in March and he went to their local doctor to get something to help Mary Anne.118Id. Dr. Adams prescribed dandelion extract and sent George to Dr. Alvin Foord, also a medical man and the local druggist, to pick it up.119Id. at 7, 15. At the time, dandelion was used for stomach ailments and liver problems.120See id. at 7. You can still buy it today from natural food stores and purveyors of alternative medicines.121See Dandelion, Mount Sinai, www.mountsinai.org/health-library/herb/dandelion [perma.cc/E47X-NR23].

George went to the druggists to buy an ounce of dandelion.122Transcript of Record, supra note 116, at 7. Dr. Foord was out that day, but his fourteen-year-old assistant, Charles Bates, was in the shop.123Id. at 7–8. Bates went to a jar of dandelion extract that was already open to get the medicine, but it was nearly empty.124Id. at 7. George watched as Bates took a white, paper-wrapped pot labeled “Extract of Dandelion, prepared by A. Gilbert, 108 John street New York” off the shelf, opened it, and put one ounce in a small bottle.125See id. at 8. After paying for the medicine, George took the bottle back to Mary Anne’s house.126Id. at 7.

That night around eight, Mary Anne took half an ounce of the extract as prescribed by Dr. Adams.127Id. at 7, 15. By eleven she was in convulsions, her eyes rolling back in her head, chills running through her.128See id. at 7. Her husband Samuel was distraught.129See id. He called to George, who had stayed over that night.130Id. Both men thought she was dying.131Id. They went to fetch Dr. Adams, who rushed over.132See id. at 7, 15. His ministrations caused her to throw up at least part of the medicine she had been given, but she was not well.133Id. at 7. Brother, husband, and doctor kept watch over Mary Anne all night, believing she could die at any moment.134See id. She survived, but wasn’t the same woman.135See id. For weeks she could not get out of bed on her own, couldn’t sleep, and was in pain all the time.136Id.

The day after the incident, Samuel Thomas, Mary Anne’s husband, went to Foord’s shop to ask about the dandelion extract. 137Id. at 8. Something must have been in it to cause Mary Anne’s near-death experience, he thought. Foord’s assistant, Bates, was there again that day. 138Id. He showed Foord the jar marked “Extract of Dandelion” and Foord tasted it. 139Id. Unlike extract of dandelion, it burned his tongue and his mouth felt parched.140Id. He gave Bates a taste, and both men quickly realized this was not extract of dandelion at all, but belladonna. 141Id.

Belladonna, or deadly nightshade as the English called it, was known to be poisonous since antiquity.142See, e.g., Belladonna, Crimson Sage Medicinal Plant Nursery, https://crimsonsage.com/product/belladonna [perma.cc/8B28-37L2]. Belladonna was the berry allegedly used to murder the Emperor Augustus.143The Powerful Solanaceae: Belladonna, U.S. Forest Serv., https://www.fs.usda.gov/wildflowers/ethnobotany/Mind_and_Spirit/belladonna.shtml [perma.cc/LL2H-LF2V]. Augustus’s death must have been horrible because belladonna interrupts the body’s regulation of the unconscious activities that keep us alive, especially our breathing and heart rate.144See Belladonna, Encyc. Britannica, https://www.britannica.com/plant/belladonna [perma.cc/G5N4-R2WW]; Belladonna, Encyc. Britannica, https://www.britannica.com/plant/belladonna [perma.cc/G5N4-R2WW]; Scopolamine, Encyc. Britannica, https://www.brittanica.com/science/scopolamine [perma.cc/VCT5-H7CV]; Atropine, Encyc. Britannica, https://www.britannica.com/science/atropine [perma.cc/CM8M-FVR2]. At a high enough dose, belladonna can cause loss of balance, hallucinations, crippling headache, convulsions, the collapse of the nervous system, and even death.145The Powerful Solanaceae, supra note 143; see also Mohamed Adnane Berdai, Smael Labib, Khadija Chetouani & Mustapha Harandou, Atropa Belladonna Intoxication: A Case Report, 11 Pan Afr. Med. J. 72 (2012). The Swedish botanist Carl Linnaeus named it Atropa Belladonna because during the Renaissance period in Italy women used the berry in small doses to dilate their pupils for a more beautiful appearance.146The Powerful Solanaceae, supra note 143.

The poison makes your mouth painfully dry.147Berdai, Labib, Chetouani & Harandou, supra note 145, at 72. Indeed, this is what Mary Anne’s husband and brother reported after tasting a tiny bit of the medicine that night; Mary Anne complained that it burned her throat; and Dr. Foord and his assistant reported the same when they tasted it.148See Transcript of Record, supra note 116, at 8. Belladonna was the reason Mary Anne Thomas was in so much pain that night and in the weeks after. She was lucky to be alive, but she suffered greatly. The record reveals that in June she had a miscarriage, which she likely attributed to the poisoning.149See id. at 17.

Mary Anne sued the packager of the medicine.150Id. at 1–2. It turns out that wasn’t A. Gilbert, whose name was on the bottle, but instead a man named Hosea Winchester.151Id. at 20. Winchester had a business in New York City packaging and selling vegetable medicines.152Id. at 21. Gilbert had had his own business, which went under, and he ended up working for Winchester and used some of his old labels on Winchester’s jars. 153Id. at 20–22. Waste not, want not. It was Winchester, the proprietor, who was responsible for mixing up the medicines.

Medicines were among the first real mass-market products in the United States. They were sold via the mail, across state lines, or to communities far away.154See Young, supra note 64, at 105. It is quite surprising when you think about how far the medicine that Mary Thomas ended up taking had to travel to get to her—250 miles was practically a world away in 1850.

Vegetable medicines like dandelion extract were all the rage in the United States during the early- and mid-1800s.155See id. People became wary of various physical medical interventions, and probably for good reason if one considers doctors’ hygiene habits (not great) and available technologies (worse).156Among the most popular cures was “heroic medicine,” which involved bleeding or purging patients. Id. at 36–38. In a report on public health in Massachusetts, published in 1850, Lemuel Shattuck wrote: “Any one, male or female, learned or ignorant, an honest man or a knave, can assume the name of physician, and ‘practice’ upon any one, to cure or to kill, as either may happen, without accountability. It’s a free country!”157Id. at 56. People wanted a pill or a tincture or an elixir to cure what ailed them.

Some of the medicines that were marketed to patients were one-ingredient medicines, like dandelion extract. Others were special combinations created by entrepreneurs and called “patent medicines” because they enjoyed legal protections, mostly in trademark, which allowed the maker to prevent others from copying his recipe and giving a kind of government imprimatur to the mixture.158See id. at 147–48; see also Paul Starr, The Social Transformation of American Medicine: The Rise of a Sovereign Profession and the Making of a Vast Industry 127 (1982). Enterprising people without any expertise in medicine made a killing, both literally in terms of human life and figuratively in terms of profit, selling patent medicines that combined various undisclosed ingredients. They advertised their wares under names like “Swaim’s Panacea” or “Radam’s Microbe Killer.”159See Young, supra note 64, at 137, 148–49. The latter had a picture of a man, bearing a spiked club, attacking a skeleton. Its creator, William Radam, wrote: “The instruments of the surgeon are . . . the means of destroying more lives in our hospitals and colleges than are the weapons of all our desperadoes and lawbreakers.”160Id. at 148–49, 151–52. Presumably, his microbe killer offered a real cure.

Radam had the distinction of being the first person to advertise a patent medicine based on a germ theory of disease in the late nineteenth century, although the theory had been around a while.161Id. at 147–48. This may be attributable to the fact that he immigrated to the United States from Prussia, and was familiar with European discoveries.162Id. at 144. His gloss on that theory seems laughable to modern readers; he advertised that all microbes were alike, and could be killed in the body by large quantities of his nostrum, like a pesticide kills insects dangerous to plants.163Id. at 147–48. It worked, as a way to sell product in any event. Radam started his American career as a gardener in Texas.164Id. at 144. By 1890 he had a mansion on Fifth Avenue, funded by the Microbe Killer, which was sold by the gallon across the country.165Id. at 150.

But what was in the Microbe Killer? Nobody knew. Radam claimed to harness electricity (he called it lightening) to create the potion.166Id. at 148. Around 1890, one doctor and chemist, R.G. Eccles, who worked at Long Island Hospital in Brooklyn, New York, decided to find out what was in the stuff.167Id. at 153. He tested some of it in his lab and found it contained water, a bit of red wine, and sulfuric and hydrochloric acids.168Id. at 148, 153–55.

It is not clear that patent medications were worse than some of the interventions prescribed by doctors, especially bloodletting. They were often laxatives of various kinds, or, like the Microbe Killer, very diluted dangerous chemicals. But they were untested and unregulated, and sometimes, as in Mary Anne’s case, they could be truly deadly.

There are cases in the reports similar to Thomas v. Winchester involving poisons sold instead of what passed at the time for medication. Two cases decided in the 1890s involved the same scenario: a patient seeking dandelion was given belladonna.169See Smith v. Hays, 23 Ill. App. 244, 249 (App. Ct. 1887) (upholding an award of 0 for injuries caused by erroneous provision of belladonna instead of dandelion); cf. Gwynn v. Duffield, 15 N.W. 594, 595 (Iowa 1883) (finding no liability when the patient was orally told medicine was dandelion even though he saw the label that said belladonna). Other cases involved different medicinal mix-ups: replacing tincture of rhubarb with laudanum,170See Norton v. Sewall, 106 Mass. 143, 144 (1870). the accidental confusion of vitriol (also called sulphate of zinc) with harmless Epsom salt,171See Brown v. Marshall, 11 N.W. 392, 392 (Mich. 1882); see also Peters v. Johnson, 41 S.E. 190, 191 (W. Va. 1902) (reversing a defense verdict on grounds that privity did not prevent liability for drug seller who mixed up saltpeter with Epsom salts and stating that a person who sells dangerous drugs “does so at his peril”). and a fatal confusion between oil of sweet almond and oil of bitter almond that killed a young immigrant mother.172See Davis v. Guarnieri, 15 N.E. 350, 352, 360–61 (Ohio 1887). Still other lawsuits involved patent medicines that included poisons in their recipe, such as cheap but poisonous wood alcohol replacing grain alcohol,173. See Darks v. Scudders-Gale Grocer Co., 130 S.W. 430, 431 (Mo. Ct. App. 1910). or iodide of potash in a patent medicine called “Blood Balm” that was supposed to purify the blood.174See Blood Balm Co. v. Cooper, 10 S.E. 118, 119 (Ga. 1889).

Medicines were dangerous to human life into the twentieth century. They were often inaccurately labeled and contained hidden poisons. Patients took these medicines and found themselves suffering debilitating illnesses. When these injured patients sued the manufacturers, they often won, at least on appeal.

Importantly, the medicine did not have to be poisonous or result in death or near-death experiences to lead to liability. In Blood Balm Co. v. Cooper,175Id. the case involving iodide of potash, the Supreme Court of Georgia held that any patent medicine that harmed a patient could be the basis for a tort suit against the manufacturer, whether purchased directly or indirectly.176Id. As the court explained: “It was a wrong on the part of the proprietor to extend to the public generally an invitation to take the medicine in quantities sufficient to injure and damage persons who might take it.”177Id.

The idea behind these court rulings was that a manufacturer had an obligation to warn when the medicine was dangerous, and to be careful in packaging it so that mistakes like the one that harmed Mary Anne Thomas didn’t occur. The rationale was that when it came to products like patent medicine, where buyers could not judge the safety of the product, the manufacturer owed buyers a duty to sell a safe product. The contents of these patent medicines were secret, and patients had no way of determining what was in them. They therefore relied on the good word of the manufacturer. “There is no way for a person who uses the medicine to ascertain what its contents are, ordinarily, and in this case the contents were only ascertained after an analysis made by a chemist,” explained the Supreme Court of Georgia, “which would be very inconvenient and expensive to the public; nor would it be the duty of a person using the medicine to ascertain what poisonous drugs it may contain.”178. Id.

The United Kingdom adopted a similar approach in 1869, more than fifteen years after Thomas v. Winchester was decided.179George v. Skivington (1869) 21 LT 495 (Exch.). Joseph George and his wife, Emma, were on holiday, maybe even their honeymoon, and they bought a hair tonic for Emma at Skivington’s chemist shop.180See David Ibbetson, George v. Skivington (1869), in Landmark Cases in the Law of Tort 69–92 (Charles Mitchell & Paul Mitchell eds., 2010). The tonic was toxic, apparently, because after Emma used it, large chunks of her hair fell out and she developed a rash and burns on her scalp.181See id. The injuries were bad enough that she sued the chemist. Under the privity rule of Winterbottom, Mrs. George shouldn’t have been able to sue because she had not been a party to the original purchase agreement.182Winterbottom v. Wright (1842) 152 Eng. Rep. 402.

Why was that so? Today, a married woman has the same rights as anyone else. But in 1869 that was not the case. A married woman could not enter into contracts. Instead, under the legal doctrine of “coverture” she was under the legal authority of her husband.183See Coverture, Black’s Law Dictionary (11th ed. 2019). Only Mr. George could enter into a contract with Skivington, the chemist, to buy the tonic. This meant that Mrs. George had no direct legal relationship with the chemist. She was, as far as the law was concerned, in a similar position to Mr. Winterbottom: a third party to a deal between two other people.

But where Mr. Winterbottom lost, Mrs. George won. The appellate court held that she could sue the apothecary for her injuries.184See Skivington, 21 LT at 495. Why was that? The court may have seen the injustice of the marriage relationship preventing a woman who was injured from recovering for what seem to have been some pretty significant damages. But that is not what the court said. It did not create a special marriage exception to the privity rule or give women a new right to enter into contracts and sue on their own behalf. Perhaps the court could have held that Mrs. George, being a person subsumed into the legal identity of her husband, did not need to be a party to the contract for that reason. But it did not.

Rather, the court said that because Skivington had made the hair tonic from a recipe known only to him, which was (apparently) dangerous, and because he knew that Mrs. George would use it on her head, he was responsible for the injury.185See id. The chemist owed a duty that the article sold be “compound[ed] . . . with due care and ordinary skill . . . [and] that the use of it by the person buying it, or for whom it is bought, shall not be productive of personal injury.”186See id.

George v. Skivington has been described as an anomaly by later scholars.187See Ibbetson, supra note 180, at 69. But in the United States, the citation patterns do not bear this out. While Skivington is not cited nearly as often as Thomas (or Winterbottom), it still appeared in both case reports and treatises, and it was consistent with the holdings of state cases.188Skivington is cited nineteen times in the state reports available on Westlaw prior to 1916. (The search term used was “Skivington”). Winterbottom is cited 110 times during the same period in the same database (using the search term “Winterbottom”). Thomas, the counterpoint to Winterbottom, is cited 196 times during this period in the same database (using Westlaw’s Keycite function). It is important to remember that Winterbottom is cited in many cases involving personal services rather than products. See, e.g., Buckley v. Gray, 42 P. 900 (Cal. 1895) (attorney not liable to third party). What Thomas, Skivington, and MacPherson all have in common is that they were all cases involving indirect purchasers injured by consumer goods in which the consumer was permitted to sue the producer.

D. The Case of the Noxious Coat: Gerkin v. Brown & Sehler

The negligence rule for products did not only apply to medicines. Courts granted relief in cases where judges thought the product was injurious, and concluded that the defendant was at fault for being careless with respect to other products, such as, for example, clothing.189See Gerkin v. Brown & Sehler Co., 143 N.W. 48, 52 (Mich. 1913). Then, as now, people spent a significant part of their budget on cloth and clothing.190See Laura F. Edwards, Only the Clothes on Her Back: Clothing and the Hidden History of Power in the Nineteenth-Century United States 59–60 (2022) (describing sales and purchases of clothing in the nineteenth-century United States). The technologies that produced clothing had been radically transformed by industrialization in the late nineteenth century in ways that were unfamiliar and untested.191See id. at 79–101 (discussing mechanization of cloth production between the colonial period and 1900). Like food and medicine, the dangers posed by clothing affected bodies directly. Many layers of clothing touched a wearer’s skin. Merchants used new chemical discoveries to take a product that might not have been of a high enough quality to sell and transform it into a valuable thing. In the case of cloth and furs, this process often involved dyes.192Linda Jean Thorsen, The Merchants and the Dyers: The Rise of a Dyeing Service Industry in Massachusetts and New York 1800–1850, 15 Textile Soc’y Am. Symp. Proc. 499 (2016).

In early December of 1910, Henry Gerkin bought a coat with a muskrat fur collar, dyed black.193Dave Golowenski, Price of Muskrat Pelts Soars, Columbus Dispatch (Jan. 15, 2012, 10:31 AM), https://www.dispatch.com/story/sports/outdoors/2012/01/15/price-muskrat-pelts-soars/23615816007 [perma.cc/2B3S-ZSKY]. (Muskrat fur is sometimes dyed to resemble more expensive mink.)194See Gerkin, 143 N.W. at 48. Soon after, Mr. Gerkin started to suffer a terrible inflammation on his face and neck. Malignant sores developed on his skin.195See id. at 51. His eyes swelled. His face and hands became inflamed and discolored. His condition got so bad as “to incapacitate him from work and confine him to his house in misery.”196Id. He went to his local doctor but couldn’t get any relief. Finally, in mid-March, he traveled to Detroit to see a skin specialist. The doctor suggested the cause of his suffering was the coat.197Id. Mr. Gerkin stopped wearing it, and, gradually, his symptoms eased.198Id.

Mr. Gerkin bought the coat from a store called Young Brothers in Howell, Michigan.199Id. at 48. They, in turn, got the coat from a clothing wholesaler named Brown & Sehler.200See id. The wholesaler got the fur collars from a New York furrier.201Id. at 49. It turns out, John Sehler, the president of the wholesale company, knew that sometimes the dyed coat collars caused an allergic reaction.202See id. at 52. Not all the time, but about one in a hundred times, he testified, the coat would cause a rash or worse. He tried to find out why this was but had no luck. So, he developed a scheme. He wouldn’t disclose the fact that sometimes dyed fur collars caused rashes.203See id. Instead, knowing that “some people simply could not wear dyed furs,” Sehler explained, he

adopted the plan, which the traveling salesman was told to observe, to have the coat sent in and change the collar on it for one that was not dyed when a complaint was made, giving the purchaser no preliminary warning, and leaving it for him to make the discovery by the slow and dangerous process of painful experience.204Id.

When Mr. Gerkin sued the Brown & Sehler Company for his injuries, he faced some problems. Brown & Sehler blamed the New York furrier, whom Mr. Gerkin couldn’t find, and in any event, Mr. Gerkin would probably have had to sue them in New York, far from his home, whereas he could sue Brown & Sehler in Michigan.205Id. at 49–50. The court allowed him to sue Brown & Sehler without joining the New York furrier. The court record reveals no argument that the fact that Mr. Gerkin hadn’t bought the coat directly from the wholesaler, but rather through the intermediary of Young Brothers, was an impediment to suit.

At trial, Mr. Gerkin introduced evidence that John Sehler knew that the collar might cause rashes and didn’t disclose this fact. He also put the skin specialist from Detroit on the stand, asking him: “How do you account then, Doctor, for the fact that some people are able to wear dyed furs without any difficulty and others not?”206Id. at 52. “We account for it in two ways,” the doctor said:

One is a possibility of there having been something in the manufacture of that particular coat or the dyeing of that particular coat which has not been—I don’t know whether it is a different dye used or something else that has not removed the irritant. . . . [T]he other one is the manner in which the collar is worn. For instance, a man turns down his collar, didn’t turn it up on his face, might wear a fur coat for a long, long while, never have it . . .207Id.

The doctor did not state which chemical might have caused Henry Gerkin’s problem, or if a particular mechanism was to blame for Mr. Gerkin’s inflammation and other issues. Nor did he know why Mr. Gerkin had the problem and others didn’t, although his theory seems like a good one.208Id. Nor was John Sehler sure why some people were sensitive to the chemicals in dyed fur collars and others weren’t. But there had been warnings that the dye caused problems. Mr. Sehler knew it sometimes did.209Id. Had he warned the buyers, he might have been able to get the case dismissed. As it was, a jury got to decide.210Id. at 54.

Contrast this with an earlier case, decided in 1888, where the plaintiff was unable to prove that the defendant clothier had any knowledge of the risks of rash. Frances Gould of Boston had bought some woolen cloth from the Slater Woolen Company, and the dye used in the cloth gave her a terrible rash.211Gould v. Slater Woolen Co., 17 N.E. 531, 531 (Mass. 1888). More precisely, the problem was the fixative or “mordant” that was used to make sure that the dye adhered to the cloth and didn’t rub off. Finding a good fixative was a real challenge in the late 1800s.212For a sense of the debates, see Charles Harrington, Letter to the Editor, Chrome Considered as a Poison., 10 Science 104–05 (1887), discussing the risks of using chrome in dyes.

It turned out that the Slater Woolen Company used a perfectly ordinary dye—the most common dye used at the time, according to the court record—and there hadn’t been any reports of that dye causing harm, at least not by the time Mrs. Gould sued.213Transcript of Record, Plaintiff’s Exceptions at 1–5, Gould v. Slater Woolen Co., 17 N.E. 531 (Mass. 1888) (No. 1963) (on file with author). Indeed, it had been used for many years prior without incident to dye cotton stockings black. As the case worked its way to the courts, there were some reports of injuries caused by the dye. But the problem was that when Mrs. Gould had filed her lawsuit, there hadn’t been any such reports, at least none that had been documented.214See id. So there wasn’t any reason, the court said when it finally decided the case in 1888, that the company would know to take more care.215See id. This was the case even though Mrs. Gould was able to show that she was injured by the cloths she bought.216Id.

The privity defense would not have been raised in Gould v. Slater Woolen Co. because Mrs. Gould had bought the cloth directly from the manufacturer, as many people did in small cities in the nineteenth century. They purchased products, from food to clothing to medicine to musical instruments, from local producers, artisans, and apothecaries.217 Hughes & Cain, supra note 59, at 349–51 (describing the expansion of the U.S. market in goods). Even so, the judges were not concerned with what the contract promised but rather with what duties a clothier owed consumers in the new economy. Gould, in other words, was treated as a tort case. The fact that there had been a purchase, and therefore a purchase agreement, played no role in the decision. It was hardly mentioned at all in the opinion, other than to say that Mrs. Gould bought the items from Slater Woolen Company.

What we learn from these cases is that contract did not dominate judicial thinking during the period of growth in consumer goods. Rather, judges hearing tort cases relating to injurious products were concerned with three main issues: what the customer could be expected to know (or the obviousness of the danger), the defendant’s conduct (or blameworthiness), and what the science showed. Slater Woolen Company was not blameworthy. It did not know the risk posed by the dye, and it seems the court thought the company could not have known the risk at the time of injury. Brown & Sehler were blameworthy. They did know that the dyes they used caused terrible rashes, and the court was satisfied that the causal link had been proven. None of this had anything to do with agreements, warranties, or other contract-adjacent doctrines, but rather with duties outside of agreements that people in a society owe one another. That is the essence of tort.

E. The Case of the Pokey Soap: Hasbrouck v. Armour & Co.

The major barrier for plaintiffs was not privity, or the idea that manufacturers owed no duty to consumers, but rather the scope of the duty manufacturers owed, especially when it came to surprising or unexpected defects. Products liability in the nineteenth century, in other words, was not about relationship rules but conduct rules. This brings us to the case of the needle buried deep in a bar of soap.

After the Civil War, demand for inexpensive soap increased among consumers.218Judith Ridner, The Dirty History of Soap, The Conversation (May 12, 2020, 1:35 PM), https://theconversation.com/the-dirty-history-of-soap-136434 [perma.cc/RJ45-YPDN]; Kimblery L. Kostka & David D. McKay, Chemistry Keeps Us Clean. Chemists Clean Up: A History and Exploration of the Craft of Soapmaking—How Soap Came to Be Common in America, 79 J. Chem. Educ. 1172, 1172–73 (2002). As people moved into cities, they were unable to make soap out of rendered fat or lard as they would have done on farms, and increasingly needed to buy their soap. Culturally, there was also an increased awareness of the importance of hygiene after the war, which spurred soap sales and the growth of companies to fill this new consumer need.219Ridner, supra note 218.

One of the companies making soap in the late nineteenth century was Armour & Co., a meatpacking company founded in Chicago that processed “all the parts of the animal—‘everything but the squeal’—making such products as glue, lard, gelatin, and fertilizer.”220Mark R. Wilson, Stephen R. Porter & Janice L. Reiff, Armour & Co., Elec. Encyc. of Chi. (2005), http://www.encyclopedia.chicagohistory.org/pages/2554.html [perma.cc/MK2Z-ZZHW]. Among these products was soap, Armour & Co. Toilet Soap No. 175, to be precise. The company held “out to the public that this soap would supply every need for all toilet purposes, and guarantee the purity and harmlessness thereof, and that the soap is free and clear from all harmful ingredients or foreign substances which might injure persons using the same in the ordinary manner.”221Hasbrouck v. Armour & Co., 121 N.W. 157, 159 (Wis. 1909).

Somehow, one of the many bars of soap Armour & Co. sold to the S. Heymann Company in Oshkosh, Wisconsin, which thereafter sold the soap to the general public, contained a needle buried deep into the bar. F.M. Hasbrouck bought the offending bar of soap. “While properly using the soap so purchased for toilet purposes, the plaintiff was injured by this needle in the soap entering the palm of his right hand and producing the most serious consequences, including paralysis and disability.”222Id. Neither the manufacturer, Armour & Co., nor the merchant, Heymann, knew of the existence of this needle. The question for the Wisconsin Supreme Court, where the case of Hasbrouck v. Armour & Co. eventually landed, was whether either Armour & Co. or Heymann could be liable for the injury to the plaintiff’s hand.223See id. at 161.

Today, this type of injury would be considered a manufacturing defect and would be subject to strict liability.224See Restatement (Third) of Torts: Prods. Liab. § 2 (Am. L. Inst. 1998) (“A product: (a) contains a manufacturing defect when the product departs from its intended design even though all possible care was exercised in the preparation and marketing of the product . . . .”); Douglas A. Kysar, The Expectations of Consumers, 103 Colum. L. Rev. 1700, 1709 (2003) (stating that courts have little difficulty finding liability in manufacturing defect cases); Mark A. Geistfeld, Strict Products Liability 2.0: The Triumph of Judicial Reasoning Over Mainstream Tort Theory, 14 J. Tort L. 403, 417 (2021) (discussing the varying approaches of the Restatements to the issue of manufacturing defects). A needle manifestly does not belong in soap, and there would be liability for such a dangerous condition of the soap even if the manufacturer took proper care. This was not the rule that the Wisconsin court applied. But it didn’t apply privity either.

Armour & Co., predictably, argued that the suit was barred by privity. The Wisconsin Supreme Court rejected this argument. The court began in a way that seemed good for manufacturers: manufacturers owe no general duty to the public, it said.225Hasbrouck, 121 N.W at 160. If one stopped reading at that point, one might think that the rule was damnum absque injuria, that is, a no-liability rule.226Robert Rabin argues that the rule was no liability. See Rabin, supra note 5, at 937–38. As we shall see, there was no liability in Hasbrouck, but not because the rule was a no-liability rule. But immediately after asserting that there was no general duty, the court went on to say:

The manufacturer or dealer who puts out, sells, and delivers, without notice to others of its dangerous qualities, an article which invites a certain use, and which article is not inherently dangerous, but which by reason of negligent construction he knows to be imminently dangerous to life or limb, or is manifestly and apparently dangerous when used as it is intended to be used, is liable to any person who suffers an injury therefrom, which injury might have been reasonably anticipated.227Hasbrouck, 121 N.W. at 160.

In other words, a manufacturer who was careless and thereby produced a dangerous product owed “a general duty toward the public to whom the wares are offered.”228Id. This duty ran to secondary purchasers such as Hasbrouck.

Yet despite recognizing this general duty, the court held that Armour & Co. could not be held liable. The problem was not that Armour & Co. owed no duty but rather that a needle in soap was such a “remote possibility, an extraordinary occurrence,” that “serious injury resulting from such act to persons using the soap for toilet purposes is an unusual and remote consequence of the careless dropping of such needle into the mixture.”229Id. at 161. The defect was so unexpected that the company could not have been required to guard against it.

In the case of the needle in the soap, the difference between negligence and strict liability had real bite. Although the language used by the court speaks of proximate causation, the link between duty and proximate cause is also evident. Because a needle falling into soap was a “remote possibility,” there was no duty to guard against this unlikely event; allowing a needle to fall into the soap was not negligence.230As the court quoted: “Negligence in the law is not mere carelessness, but is careless conduct under such circumstances that an ordinarily prudent person would anticipate some injury to another as a reasonable and probable result thereof.” Id. (quoting Johanson v. Webster Mfg. Co., 120 N.W. 832 (Wis. 1909)) (internal quotation marks omitted). It was just a terrible, unforeseeable accident.

The most important takeaway from a case like Hasbrouck is that a manufacturer had a “public” duty to all consumers of its product not to be careless.231The idea that this is a public duty comes from the court. By this the court meant, I think, that the duty was imposed by law rather than by private ordering through contractual agreement. This should be distinguished, probably, from laws enforced only by public entities, such as the criminal laws or regulatory regimes. Where the accident could not have been anticipated, one the court would not expect a soap manufacturer to guard against, there would be no duty and no proximate cause. But the implication was that in cases of true carelessness (or worse) there would be a duty. Thus, an injury from a mixed-up medicine was expected and created an obligation; an injury from a stray needle in a bar of soap was unimaginably rare and created none.

This distinction also explains the different outcomes in Gould and Gerkin, the two clothing cases discussed earlier.232. Gould v. Slater Woolen Co., 17 N.E. 531 (Mass. 1888); Gerkin v. Brown & Sehler Co., 143 N.W. 48 (Mich. 1913). The clothier in Gould had no knowledge that its mordant could cause a debilitating rash, and therefore could not be expected to protect or warn consumers against this risk.233Gould, 17 N.E. at 531–32. By contrast, the wholesaler in Gerkin not only should have known, but in fact did know, that the fur collars on its coats sometimes caused rashes yet took no care to prevent or warn of this risk.234Gerkin, 143 N.W. at 51–52. Therefore, it was found liable. It seems likely, however, that even if it did not in fact know but should have known (if this was the type of thing one would expect the manufacturer to guard against), it would have been held liable.

F. The Case of the Decayed Can: Tomlinson v. Armour & Co.

Yet another case involving Armour & Co., this time concerning canned meat, demonstrates that the general rule applied by courts was that a predictable or expected risk, such as food poisoning from canned ham, could be best avoided by the manufacturer of the can. This raised an additional issue that came to be important in the nineteenth century: the problem that many manufactured products could not be competently evaluated by the consumer. Neither the consumer nor the grocer was able to determine the quality of the canned meat until the can was opened, and by then, it was too late. Courts recognized this fact and ruled in favor of consumers.

Armour & Co.’s primary business was meat. Among other things, the company canned meat in Illinois for sale all over the United States. Canned meat in the late nineteenth and early twentieth centuries was a scandal. Upton Sinclair’s novel The Jungle (based on the realities of the meatpacking plants in Chicago) was published serially in 1905 and as a book in 1906.235See Upton Sinclair, The Jungle (1906); Blum, supra note 40, at 120 (stating that Sinclair’s book was serialized in the socialist magazine Appeal to Reason in 1905). It caused a national outpouring of anger; citizens wrote letters of outrage to President Theodore Roosevelt, demanding to know how he “planned to fix the problem of the country’s disgusting food supply.”236 Blum, supra note 40, at 143.

Roosevelt appointed two independent investigators, and their reports of the sanitary conditions in the meat packing plants were appalling.237See id. at 145–47. The authors of the reports wrote that they “saw meat shoveled from filthy wooden floors, piled on tables rarely washed, pushed from room to room in rotten box cards, in all of which processes it was in the way of gathering dirt, splinters, floor filth and expectoration of tuberculous and other diseased workers.”238Id. at 147. A superintendent at the plant told investigators that the meat would be cooked, so these filthy conditions weren’t a problem.239Id.

The publication of The Jungle was not the first time that meat made headlines across the nation. In 1899, a court of inquiry—called the “Beef Court”—investigated the harm caused by the canned beef served to soldiers in the Spanish-American War.240Id. at 53–59. The meat smelled like a dead body, soldiers testified, and when cooked tasted of boric acid.241Id. at 56. The poet Carl Sandburg recalled the odor of the meat as “more pungent than ever reaches the nostrils from a properly embalmed cadaver.”242Id. at 59. Soldiers were sickened. Even future President Theodore Roosevelt testified to what he had seen in the First U.S. Voluntary Cavalry Regiment (known as the “Rough Riders”). He described men who “ate the meat and vomited.”243Id. at 55–56. The food provided to the soldiers, he said, was “utterly unsafe and utterly unfit” and the result was that soldiers nearly starved.244Id.

This was the same stuff that was being marketed to consumers.245See id. at 58. That was part of the problem, it turned out. The Beef Court did not confirm the view that embalmed beef caused sickness among the troops, instead attributing illness to the hot climate in Cuba, in part because the provisions were no “better or worse than any other” product on the grocery store shelves.246See id. at 59. This conclusion does not speak well for the products on the shelves. There were many dangers lurking in canned and preserved foods, including bacteria; formaldehyde; boric acid; and heavy metals, such as lead, which were used to solder the cans.247Id. at 56.

A few years after the publication of The Jungle, Sara Tomlinson sued Armour & Co. for food poisoning resulting from a canned ham that made her sick.248“[T]he plaintiff after purchasing said can of ham, and without fault or negligence on her part, ate a piece of ham taken from said can, and in consequence thereof became poisoned and sick with ptomaine poison.” See Tomlinson v. Armour & Co., 70 A. 314, 316 (N.J. 1908), overruled in part by Henningsen v. Bloomfield Motors, Inc., 161 A.2d 69 (N.J. 1960). She had purchased the can of ham from a local grocer, but sued Armour & Co. for damages.249See id. The highest court in New Jersey, then called the Court for Errors and Appeals,250See Organizational Chart for Pre 1948 Court System, N.J. Dep’t of State (2011), https://www.nj.gov/state/archives/catcourtstructure.html [perma.cc/EF8D-XPVW]. held that Ms. Tomlinson could sue Armour & Co. for the poisoned meat and that the company owed a duty to consumers to sell healthful meat.251Tomlinson, 70 A. at 319. The court rejected contractual reasoning, accepting without deciding that there was no implied warranty on the canned meat.252See id. at 316. It placed liability squarely in the tort frame, holding that the company owed a duty to consumers not to sell tainted meat.253See id. at 319.

The reason for this ruling was that canned goods are credence goods: their quality cannot be ascertained prior to purchase, indeed prior to eating.254Posner, supra note 78, at 1489. “Canned goods are, at the present day, in such common use,” explained the court, “that we may judicially recognize that the contents are sealed up, not open to the inspection or test, either of the retailer, or of the customer, until they are opened for use; and not then susceptible to practical test, except the test of eating.”255Tomlinson, 70 A. at 317. The New Jersey high court rejected the maxim of buyer beware, writing that “the fundamental condition upon which the common-law doctrine of caveat emptor is based—that the buyer should ‘look out for himself’—is conspicuously absent; for he has no opportunity to look out for himself.”256Id.

But the court did not stop there. It added to this the argument that it was a societal duty to take care of others’ health and wellbeing: “To assert, therefore, that one living in a state of society, organized, as ours is, according to the principles of the common law, need not be careful that his acts do not endanger the life or impair the health of his neighbor seems to offend against the fundamentals.”257Id. The consumer, the court asserted, “has a right to insist that the manufacturer shall at least exercise care that they are [fit to be eaten], and are not unwholesome and poisonous.”258Id. These rationales were notably public and societal—the issue was not what was promised in some bargained-for exchange but rather that liberal ideal that one was free to engage in conduct so long as it did not harm another. Such a harm would not lie where it fell.259This is a reference to Oliver Wendell Holmes, Jr.’s statement that “[t]he general principle of our law is that loss from accident must lie where it falls.” Oliver Wendell Holmes, Jr., The Common Law 94 (Boston, Little, Brown & Co., 1923) (1881). It seems to me this is a better description of what Holmes thought the law ought to be than what it was.

The New Jersey Court of Errors and Appeals was not an outlier. Sellers of rotten food were responsible for the injuries that their food caused in a number of late nineteenth-century cases, whether or not that food was canned.260See Craft v. Parker, Webb & Co., 55 N.W. 812, 813 (Mich. 1893) (holding that a seller of meat is liable for injuries caused by food poisoning); Bishop v. Weber, 1 N.E. 154, 154 (Mass. 1885) (same). For example, Anna Wiedeman sued her butcher, Henry Keller, for selling meat that gave her family food poisoning. In 1897, the Supreme Court of Illinois held that the butcher owed his customer food that was safe. The court explained that the butcher was in a better position than the customer to know whether the meat was safe to eat: “[I]t is much safer to hold the vendor liable than it would be to compel the purchaser to assume the risk.”261Wiedeman v. Keller, 49 N.E. 210, 210–11 (Ill. 1897).

The Illinois court arguably went further than the one in New Jersey, recognizing liability not only in situations where the customer could not inspect the product, as in the case of canned goods, but with respect to any meat that was defective.262Id. at 210. It based its views on an argument that would become popular in the late twentieth century: that the best cost avoider should bear the risk of loss.263See Guido Calabresi, The Cost of Accidents: A Legal and Economic Analysis 175 (1970) (describing the best cost avoider as “that reduction or alteration in activities which gives us the optimal primary accident cost avoidance”) (emphasis omitted). The butcher and the canner are in a better position than the customer to take precautions to prevent losses due to food poisoning, and therefore they should bear the cost of injury to incentivize them to take greater care.264See Wiedeman, 49 N.E. at 211.

There are other cases, not found in the case reports but in journalistic accounts, that were decided by juries—meaning that if the plaintiff lost it was not a doctrinal or legal impediment that stood in her way but a factual or cultural one. For example, in 1888, a woman named Theodora Kayler of New York sued the large provision firm of Thurber & Co., claiming that a can of tomatoes that she bought was rendered poisonous by the use of “muriate of zinc” or zinc chloride on the top of the can.265See Henry A. Riley, Medico-Legal Cases, 54 Med. & Surg. Rep. 15 (1888) (on file with author). She sought $50,000 in damages,266Id. a substantial sum in those days.267To get a sense of the amount, ,000 in 1913 (the earliest available date from the Bureau of Labor Statistics) was approximately ,490,000 in 2022 dollars. See CPI Inflation Calculator, U.S. Bureau of Lab. Stat., https://www.bls.gov/data/inflation_calculator.htm [perma.cc/HMV8-PFHZ]. The case went to the jury, which found for the defendant. According to the news reports, there were no other reports of incidents from the cans and thus there would have been no way for Thurber & Co., or Shade & Johnson, the Maryland firm that canned the tomatoes, to know of the alleged risk.268See Riley, supra note 265, at 454.

Some might wish to distinguish the case of the pokey soap from the case of the decayed can on the grounds that poisoned food was subject to strict liability, whereas products were not.269See Prosser, supra note 3, at 1104. Prosser wrote that “[t]he early American decisions thought that ‘warranty’ was the name for it, and imposed strict liability upon the seller of food, in favor of his purchaser, as ‘a principle, not only salutary, but necessary to the preservation of health and life.’ Id. But he said that these cases only involved direct sales, not secondary sales. And he claimed that “[t]he extension of the strict liability to third persons with whom the seller had made no contract came after the turn of the century.” Id. Leaving out of the discussion for the moment what the standard of liability was (strict or fault), it is not true that courts did not recognize third-party causes of action against purveyors of defective food before 1906. See Bishop v. Weber, 1 N.E. 154, 154 (Mass. 1885) (finding a seller of food liable to third parties for injury). But after a careful read of the courts’ arguments, the key to the different outcomes lies not in any categorical threshold that is the result of a relationship rule but in the foreseeability analysis.270Hasbrouck v. Armour & Co., 121 N.W. 157, 161 (Wis. 1909). The same is true of Kayler, the canned tomato case; Gerkin, the irritating collar case; and Gould, the woolen sock case.271See Gould v. Slater Woolen Co., 17 N.E. 531 (Mass. 1888); Gerkin v. Brown & Sehler Co., 143 N.W. 48 (Mich. 1913); Riley, supra note 265, at 454. In all of these cases, the outcome turned on what indicia the producer had of a potential defect as a proxy for foreseeability, an idea that harkens all the way back to Thomas v. Winchester, where the court explained: “[n]othing but mischief like that which actually happened could have been expected from sending the poison falsely labeled into the market.”272Thomas v. Winchester, 6 N.Y. 397, 410 (1852).

G. The Case of the Collapsing Car: MacPherson v. Buick Motor Co.

As we have seen, there was no generalized privity rule in the United States that limited a producer’s liability to injured consumers. Judge Cardozo correctly observed that products liability suits that failed were not rejected on privity grounds but rather on the grounds of proximate cause.273MacPherson v. Buick Motor Co., 111 N.E. 1050, 1053 (N.Y. 1916). Cardozo’s analysis in MacPherson is consistent with everything described above except for one thing: its decision to place privity and contract at the center and liability and tort at the periphery of the law governing injurious products.

The opinion framed the issue as follows: whether a manufacturer owed “a duty of care and vigilance to any one but the immediate purchaser.”274See id. at 1051. It then segued into a doctrinal history. The first salvo was the 1852 case of Thomas v. Winchester,275See Thomas, 6 N.Y. at 404. in which a woman injured by the sale of a poison negligently labeled as a medicine was able to recover from the original seller, a manufacturer of herbal medicines, even though the medicine was purchased from a local druggist. Cardozo wrote that the doctrine was initially one of “narrow construction.”276MacPherson, 111 N.E. at 1052. Only cases in which the product “put human life in imminent danger” would lie in tort.277Id. at 1051 (quoting Thomas, 6 N.Y. at 397) (internal quotation marks omitted). Otherwise, contract governed and persons not party to the contract of sale could not recover.

But soon, Cardozo explained, the doctrine expanded beyond poisons. He highlighted several key cases, such as Devlin v. Smith,278Devlin v. Smith, 89 N.Y. 470 (1882). in which a builder was found liable for an improperly constructed scaffold to the painters injured using it, even though they did not commission the structure.279MacPherson, 111 N.E. at 1052 (discussing Devlin, 89 N.Y. 470). “Building it for [the employees’] use,” wrote Cardozo, “he owed them a duty, irrespective of his contract with their master, to build it with care.”280Id. A similar case was Statler v. Ray Manufacturing Co.,281See Statler v. Ray Mfg. Co., 88 N.E. 1063 (N.Y. 1909). in which the court held the manufacturer of a commercial coffee urn liable for injury to a restaurant employee.282See MacPherson, 111 N.E. at 1052. Other cases included a servant injured by an exploding bottle of soda (reminding one of Escola v. Coca Cola)283See Torgesen v. Schultz, 84 N.E. 956, 957 (N.Y. 1908). Escola involved liability for the inexplicable explosion of a bottle of Coca-Cola and is famous for Justice Traynor’s concurrence arguing that a manufacturing defect should be subject to strict liability. Escola v. Coca Cola Bottling Co. of Fresno, 150 P.2d 436, 437, 440 (Cal. 1944). and a child who was working in a building and injured by defective maintenance of the elevator.284See Kahner v. Otis Elevator Co., 89 N.Y.S. 185, 187 (App. Div. 1904), aff’d, 76 N.E. 1097 (N.Y. 1905).

One thing these cases all had in common was that they rejected the privity doctrine. The other was that, except for Thomas,285See Thomas v. Winchester, 6 N.Y. 397 (1852). they all involved employees injured on the job: painters injured by scaffolding, servants injured by exploding bottles, waitresses injured by exploding coffee urns, and workers injured by elevators. This was the circumstance where the doctrine of privity was strongest. The appellate court in Kahner v. Otis Elevator Co., the case about the child injured by an elevator, explained this strong version of privity doctrine (before rejecting it):

[T]he ordinary rule of law would be that, in the absence of contractual relations or of privity between the manufacturer and a stranger, there is no liability for injuries either to person or property by reason of defects that may exist in machinery or in mechanical contrivances or appliances. That rule is established in other jurisdictions as well as by the courts of this state.286Kahner, 89 N.Y.S. at 188.

The Kahner court rejected the privity defense on the grounds that the elevator was “made dangerous by the treatment it received from the [defendant’s] servants in the performance of its duty and obligation to make it safe for the use of those authorized to use it.”287Id.

The application of privity to workers’ injuries seems to have been an available rule288See supra notes 106–107 and accompanying text. (although I have found many exceptions such as Kahner), and its use is consistent with what historians such as John Witt have shown to be a judicial hostility to working men and women during the Industrial Revolution.289See Witt, supra note 51, at 43–70. But MacPherson was not about workers’ injuries. Donald MacPherson was a consumer, and the case was about an allegedly defective car that he bought from a dealership.290White explains how the underlying facts of the case are not what they appear from reading the opinion. There was apparently strong evidence that MacPherson was driving too fast. For that backstory, see White, supra note 5, at 258–65.

In fact, few courts rejected products liability suits on privity grounds outside of the employment context.291See MacPherson v. Buick Motor Co., 111 N.E. 1050, 1052 (N.Y. 1916) (characterizing the “trend of judicial thought” as rejecting privity). Judge Cardozo was right to notice that in many cases where the plaintiff-victim lost, the loss was based on proximate or intervening causes, not privity.292See id. at 1054 (describing contrary cases and stating that “[s]ome of them, at first sight inconsistent with our conclusion, may be reconciled upon the ground that the negligence was too remote, and that another cause had intervened”). He did not go far enough, however, as he still conceded that early cases provided only for a narrow duty of the manufacturer to the end user.293See id. at 1052 (“These early cases suggest a narrow construction of the rule. Later cases, however, evince a more liberal spirit.”). And he conceded that there was a time when the dominant idea in products liability was the “notion that the duty to safeguard life and limb, when the consequences of negligence may be foreseen, grows out of contract and nothing else.”294Id. at 1053.

Scholars who believe there was a strong privity doctrine that applied to workers and consumers alike largely follow the reasoning of Chief Judge Willard Bartlett’s dissent in MacPherson. He asserted, “I do not see how we can uphold the judgment in the present case without overruling what has been so often said by this court and other courts of like authority.”295Id. at 1056 (Bartlett, C.J., dissenting). The dissent and scholarly commentary imply that Cardozo was pulling a fast one, reading exceptions as the rule.296See, e.g., Prosser, supra note 3, at 1100. Even Cardozo himself seemed to concede that he was moving a peripheral doctrine to the center, making the exception into the rule, when he stated that the courts had “put aside the notion” that contract not tort governed products liability.297See MacPherson, 111 N.E. at 1053. But as we have seen, the dissent was wrong about this. Chief Judge Bartlett may have wanted privity to be the rule, but wanting does not make it so.

Another car case, decided a few years before MacPherson, proves the point. The case involved a young woman, Lucy Shaffer, who was riding in a “rumble seat”—a seat that was supposed to be affixed to the back of the car’s frame. The seat was not well-affixed to the car and fell off, throwing her and injuring her permanently and, according to the court, quite badly. In 1911, the Kentucky Court of Appeals decided Olds Motor Works v. Shaffer,298See Olds Motor Works v. Shaffer, 140 S.W. 1047 (Ky. 1911). holding that she could sue Olds directly for the defect that caused her injury.299Id. at 1051. Automobiles, the court stated,

are in general use throughout the country, and are employed as means of transportation by great numbers of people, and the liability of their occupants to injury from defects in material or construction is so great as to put upon manufacturers the duty of exercising a high degree of care in their construction, and equipping them in such a manner as will make them, when used with proper care, reasonably safe.300Id.

The court found a general “inclination” based on “sound principles” to hold “manufacturers of articles, intended for public use” liable to third parties where those products were likely “if defectively constructed, to inflict harm.”301Id.

In sum, privity wasn’t the rule when it came to automobiles any more than it was for other products. Cardozo was more correct than he is usually given credit for being, even if he comes off as somewhat less magical.

There were two exceptions to this general approach to automobiles decided almost simultaneously with MacPherson. In one case, decided by the Second Circuit the year before MacPherson and also involving defective wooden wheels, the court stated that “one who manufactures articles dangerous only if defectively made, or installed, . . . is not liable to third parties for injuries caused by them, except in case of willful injury or fraud.”302Cadillac Motor Car Co. v. Johnson, 221 F. 801, 803 (2d Cir. 1915). One of the interesting aspects of this case is that the court indicates that it thought the manufacturer was careful, so that even under a negligence theory it should not have been liable. See id. at 804 (“The practice of the manufacturers of automobiles and of the trade as to the examination of wheels, while not controlling, was certainly relevant.”). If privity were such a well-established rule, there would be no need to consider any alternative theories. Indeed, one wonders, if it was such a bright line and easy rule to follow, why the trial court did not apply it and dismiss the case. The answer is that it wasn’t the rule. In a second case, Ford Motor Co. v. Livesay, issued only a few months after Cardozo’s opinion, the Supreme Court of Oklahoma held that privity prevented a man injured by a broken spoke on the wooden wheel of his Ford car from suing Ford directly.303Ford Motor Co. v. Livesay, 160 P. 901, 902 (Okla. 1916). Livesay cited largely irrelevant cases for the proposition that privity was the rule. For example, in Vincent v. Crandall & Godley Co., 115 N.Y.S. 600, 602 (App. Div. 1909), a chauffeur left his car while he went into a store. Several boys came and started the vehicle, crashing it. The court held that the boys were an intervening cause, and that a car can be left on the street unlike inherently dangerous items like dynamite. Other irrelevant cases include Steffen v. McNaughton, 124 N.W. 1016 (Wis. 1910) (finding that the owner of a car was not responsible for a chauffeur hurting a pedestrian in a car accident); Cunningham v. Castle, 111 N.Y.S. 1057 (App. Div. 1908) (same); and Jones v. Hoge, 92 P. 433, 434 (Wash. 1907) (same). These were minority decisions, and to some extent they demonstrate that privity as a concept was most successful in the area of car manufacturing around the time of MacPherson, rather than a longstanding rule that was abrogated by that case. There are no state car-defect cases before 1916 that prevented plaintiff recovery due to privity, nor (as we have seen) many other product cases in the state courts that prevented recovery on those grounds.304I have not studied the prevalence of privity among the federal courts during this period, which notably was before Erie was decided such that the federal courts could have applied their own tort law. See Erie R.R. Co. v. Tompkins, 304 U.S. 64 (1938). If there was a split between state and federal courts on this question, that would be a worthwhile project for another article.

H. Category Errors: Inherently Dangerous and Dangerous as Designed

The canonical view is that courts were able to depart from the privity rule through three exceptions.305Prosser, supra note 3, at 1100. This Section explains that these exceptions were not uniformly described as such in the case law, and that the salient theme was liability for consumer goods (as opposed to services, for example). While it is true that some judges characterized liability as the exception and privity as the rule during the latter half of the nineteenth century, many others did not. Often judges only addressed privity at all because they were responding to a defense argument that privity barred recovery.306See, e.g., Dow v. Kan. Pac. Ry. Co., 8 Kan. 642 (1871) (showing that only counsel mentioned Winterbottom, which was not raised by the court at all); Thomas v. Winchester, 6 N.Y. 397 (1852) (rejecting privity arguments raised by defense counsel). In Thomas v. Winchester, the appellate record reflects that the defendant made the privity argument, which the court rejected. See Points for Appellant at 1–4, Thomas, 6 N.Y. 397 (on file with author). The so-called “exceptions” to privity in fact constituted a set for rules for injurious consumer goods, an emerging problem in the second half of the nineteenth century that had simply not been an issue in a world where most Americans lived on farms.

The first rule was that there could be liability for imminently dangerous goods that were “intended to preserve, destroy, or affect human life.”307See Huset v. J.I. Case Threshing Mach. Co., 120 F. 865, 870 (8th Cir. 1903). Under this category fell goods such as the explosive naphtha sold as lamp oil,308See Wellington v. Downer Kerosene Oil Co., 104 Mass. 64, 64 (1870). or the poison belladonna sold as harmless dandelion extract.309See Thomas, 6 N.Y. 397. These might be described as mislabeling cases. The second rule concerned owners who invited others to use the dangerous product, such as the owner of a building who constructed a defective scaffold and was sued by the construction workers.310See Huset, 120 F. at 870–71 (citing Coughtry v. Globe Woolen Co., 56 N.Y. 124 (1874)). The duty to construct a safe scaffold for the workers existed independent of the contract, a New York court held in 1874.311See Coughtry v. Globe Woolen Co., 56 N.Y. 124 (1874). The third rule was that anyone who sold a product he knew or should have known was dangerous was liable to anyone injured by that product.312See Huset, 120 F. at 871. While sometimes limited by a fraud-like requirement that the defendant knew of the danger and hid it, or a requirement of foreseeability, this rule created substantial tort liability independent of contractual relations and governed anyone injured by the product.

There were prominent cases in which these three rules were described as exceptions to privity, as they were in MacPherson itself. That claim has been taken at face value despite many other cases, including those described above, which did not treat these rules as exceptions to privity at all but rather as stand-alone justifications for liability. Among the cases that have been relied on for the list of exceptions is Huset v. J.I. Case Threshing Machine Co., which involved a worker who was injured by a machine on the job, and which featured prominently in MacPherson.313Id.; see MacPherson v. Buick Motor Co., 111 N.E. 1050, 1053–54 (N.Y. 1916). Unlike the cases above involving consumers, Huset involved a fact pattern more similar to Winterbottom and raised the same employment issues as that case.

The case involved a man whose leg was crushed by a threshing machine he himself did not buy directly from the manufacturer.314Huset, 120 F. at 866. A threshing machine removes the seeds and husks from the grain stalks.315See Cook v. Massey, 220 P. 1088, 1091 (Idaho 1923) (describing a threshing machine as defined under Idaho law). The machine was made of an iron-and-steel cylinder set with rows of steel teeth and two-inch spikes, and a frame set in front of and under the cylinder with similar teeth so that the two sets of teeth would pass between one another as the cylinder revolved at a high rate of speed, threshing grain in the process.316Huset, 120 F. at 865. The machine included a covering for the cylinder to prevent a person using the machine from walking into the moving teeth.317Id. It was expected, given the design, that a person could walk on this cover.318Id. at 866.

The problem with the threshing machine that Mr. Huset was using was that the cover was too thin, so that if a man stepped on it, the covering would bend and expose him to the moving teeth.319Id. at 865. And that is indeed what happened. Mr. Huset walked on the covering, it collapsed, and his right foot was crushed up to the knee. His leg had to be amputated above the knee joint.320Id. at 866.

The company argued that, as Mr. Huset had not purchased the machine from them directly, it owed him no responsibility.321Id. That simple argument was unavailing, although if privity were the law, it should have been dispositive. This was not a case governed by contract, the court explained, but by negligence:

The case falls fairly within the third exception. It portrays a negligence imminently dangerous to the lives and limbs of those who should use the machine, a machine imminently dangerous to the lives and limbs of all who should undertake to operate it, a concealment of this dangerous condition, a knowledge of the defendant when it was shipped and supplied to the employer of the plaintiff that the rig was imminently dangerous to all who should use it for the purpose for which it was made and sold, and consequent damage to the plaintiff.322Id. at 872.

As with the coat in Gerkin, the key issue for the court in Huset was the question of which dangers were obvious in a machine and which were nonobvious, and whether the company hid the nonobvious dangers from the consumer or end user. If the company was careless in its construction of the machine, and it hid the defect so that a discerning consumer could not protect himself, it would be liable.

The court in Huset stated unequivocally that the general rule is privity, and that the threshing company was liable as a result of an exception to that rule.323See id. at 870 (referring to the privity rule and stating: “It is, perhaps, more remarkable that the current of decisions throughout all the courts of England and the United States should be so uniform and conclusive in support of this rule, and that there should, in the multitude of opinions, be but one or two in conflict with it, than it is that such sporadic cases should be found. They are insufficient in themselves, or in the reasoning they contain, to overthrow or shake the established rule which prevails throughout the English speaking nations”). It described the three exceptions to that rule (1) for products “imminently dangerous to the life or health of mankind,”324Id. (2) for those who are “invited” to use an owner’s “defective appliance,”325Id. at 870–71. and (3) for “one who sells or delivers an article which he knows to be imminently dangerous to life or limb to another without notice of its qualities” such that the seller or deliverer is “liable to any person who suffers an injury . . . which might have been reasonably anticipated.”326Id. at 871. The last is not a privity exception at all but a negligence rule applicable to defective products.

Other early twentieth-century courts predating Huset also adopted the narrative of expanding liability from inherently dangerous to merely injurious products. For example, in Lebourdais v. Vitrified Wheel Co., a case involving a defective emery wheel,327See Lebourdais v. Vitrified Wheel Co., 80 N.E. 482, 483 (Mass. 1907). An emery wheel is a wheel covered in emery powder and used for grinding or polishing. Lumatz v. Am. Car & Foundry Co., 273 S.W. 1089, 1090 (Mo. Ct. App. 1925) (describing factory uses of emery wheel to grind tools). the Massachusetts Supreme Judicial Court started out its opinion by delineating the categories for which liability was possible despite privity: products “intrinsically dangerous to life or property,”328Lebourdais, 80 N.E. at 483. or “where a manufacturer or vendor knowingly sells for general use without disclosing the existence of the defect, a machine, mechanical instrumentality, or other article, which because of its defective construction or condition when put out causes injury.”329Id. These rules, the court explained, arose from “the implied duty which he owes to the community to refrain from the commission of acts of negligence whereby injury follows to its members in person or property.”330Id. That is, a public duty not to sell defective products. What the court said with one side of its mouth (privity limits liability with few exceptions), it disclaimed with the other (there is a general duty not to sell defective products that injure people).

One way to frame the turn of the century is as a transitional moment, where courts were uncertain as to which approach—contract or tort—governed until developments in the economy made the cases increasingly common and forced the courts to commit to one understanding. They repeated defense claims that privity was a rule, but then disclaimed it at the same time because the law was in fact unsettled as to whether a privity defense would lie.331I thank Aneil Kovvali for this useful framing of the argument. But I think that the weight of the evidence favors an emerging consensus rejecting privity in tandem with the rise of mass-market products.

It was not just Winchester’s dangerous medicine, Sehler’s toxic coat, or Case Company’s destructive thresher. There were cases from numerous high courts that recognized a duty to the buyer whenever a product had a hidden defect, either because the plaintiff could not know of the defect (as with medicines) or because the defendant actively hid it (as with the coat and the thresher). In Minnesota, the high court held that a painter named Shubert could sue the ladder manufacturer when he was severely injured falling off a defective ladder, even though his employer had purchased the ladder from a local hardware store.332See Schubert v. J.R. Clark Co., 51 N.W. 1103, 1104 (Minn. 1892). In California, the court allowed a tenant, Mr. Lewis, to sue the manufacturer of a folding bed that nearly crushed him, even though the purchasers were his landlords.333See Lewis v. Terry, 43 P. 398, 398 (Cal. 1896). The Supreme Court of Illinois found that a company that manufactured wringing machines for laundry was liable for injuries caused by defects in the machine.334See Empire Laundry Mach. Co. v. Brady, 45 N.E. 486, 486, 488 (Ill. 1896). And in Massachusetts, the Supreme Judicial Court held that a partygoer could sue the caterer of a fancy dress ball for food poisoning.335See Bishop v. Weber, 1 N.E. 154, 154–55 (Mass. 1885). The idea that privity dominated tort would argue in favor of seeing Tomlinson (poisoned food) and Thomas (poisoned medicine) as similar on the basis of the type of product at issue, that is, ingested products that could poison people. But food that has gone bad is not inherently dangerous or poisonous. It has been made so by lack of care in the sale. By contrast, the medicine sold to Mrs. Thomas’s brother was in fact a poison.336This was the point made in a 1904 Harvard Law Review Note. See Note, Tort Liability of a Vendor of Chattels to Others Than His Vendee., 17 Harv. L. Rev. 274, 274 (1904). Furthermore, looking at the other cases discussed, those involving everything from coats to folding beds to wringing machines for washing clothes, the pattern that emerges is one of a solid rule of negligence not dependent on contract, and frequently involving the assertion of what courts called a “public” duty, owed to a community of consumers, as compared with the private duties created by bilateral contracts. Consider the language in Gerkin, the case of the noxious coat collar: “the duty the dealer owes to the public generally, which includes all whom it may concern, to give notice of any concealed dangers in the commodity in which he traffics, and to exercise a reasonable precaution for the protection of others commensurate with the peril involved.”337Gerkin v. Brown & Sehler Co., 143 N.W. 48, 53 (Mich. 1913).

Only some courts, like Huset, described the rule that a manufacturer can be liable to anyone for a hidden defect in a product as an exception to privity. Others merely asked whether the injury was foreseeable.338Hasbrouck v. Armour & Co., 121 N.W. 157 (Wis. 1909). In any event, characterizing the negligence rule as an exception did not make it so, as case after case demonstrated. What was likely going on was a confusion about which rules applied to workers, which to services, and which to consumer goods.

This is not to say that privity never governed. Often the decision bottomed on whether the court thought that the manufacturer owed a public duty, in which case tort governed, or a narrower duty arising out of contract.339See, for example, the Young v. Bransford case:

The right of action is confined to those who stand in the relation of contracting parties, or to cases where the injury is caused by the disregard or neglect of some obligation or duty which the party causing it owes to the party injured, or some public duty has been broken by the former.

80 Tenn. 232, 245 (1883).
Cases rejecting a public duty often involved limitations of liability to employees, supporting the widely held scholarly view that the courts were hostile to workers in the nineteenth and early twentieth centuries.340See Witt, supra note 51, at 52, 55. By contrast, in nonemployee, third-party purchaser situations, and even some cases involving workers, a “public” duty, which is to say a duty not created by the private ordering regimes of contract but by tort law, often governed.

As noted at the outset, and explained in Huset, there is a theory that the exceptions to privity were based on categories of products arranged in order of danger: from poisons and explosives (most dangerous to human life) to poorly designed but ordinary products (where no liability should lie unless there was a breach of contract).341See Huset v. J.I. Case Threshing Mach. Co., 120 F. 865, 870–71 (8th Cir. 1903). If one takes MacPherson seriously, one might believe that a tort duty slowly expanded to include rotten food, tools, and household products. Was there something special in these early years about products that were ingested such as canned meat and medicine that explains the decisions in cases like Tomlinson v. Armour & Co. and Thomas v. Winchester? After all, the court found no liability in Hasbrouck v. Armour & Co., the case of the pokey soap, where the product was not ingested but was also dangerous.

As product after product hit the consumer market—at first medicine and food, soon mass-produced clothing, and eventually machines—the rule of negligence encompassed those products. That these exceptions eradicated the rule was observed by commentators, although they erred in thinking it was ever the rule to begin with. A better way of characterizing the doctrine should have said that privity was an exception to the regular negligence rule. A Harvard Law Review note dated 1904 opined on the one hand that it was a settled rule “that a person not privy to the contract cannot recover for an injury caused him by the seller’s negligence, unless the article sold is one dangerous to human life.”342Note, supra note 336, at 274. Yet, the note pointed out that in its application, the rule hardly made sense: “it is difficult to see why a fly-wheel or a steam-boiler is intrinsically safer than a scaffolding, or why any of them is not as dangerous as food the negligent preparation of which rendered a seller liable at the suit of the vendee’s guest.”343Id.

The evolution was not categorical, that is, it was not an evolution that expanded exceptions from one type of product to another, from more dangerous products to less dangerous ones until MacPherson blew the door off of privity. Instead, products liability law was the product of social and economic changes from its inception. Medicines were sold to unsuspecting consumers, and tort offered a means for redress. Canned food came into widespread use, was poisonous, and tort offered a means for redress.344See supra Section II.F. The same was true for other common goods such as clothing, and ultimately machines such as cars or wringers or what have you.345See supra Section II.H.

Courts conceived the issue to be whether the private ordering regime of contract or the public duty regime of tort would govern. The Supreme Court of California explained in 1896: “One who, knowing an article to be defectively constructed, represents it to be safe, and sells it to a person who has no knowledge of the defect, is liable in damages to one who, without fault on his part, was injured while lawfully using the same.”346Lewis v. Terry, 43 P. 398, 398 (Cal. 1896) (holding that a manufacturer owes a duty to anyone injured by their product); see also Schubert v. J.R. Clark Co., 51 N.W. 1103 (Minn. 1892) (same); Wellington v. Downer Kerosene Oil Co., 104 Mass. 64, 67–68 (1870) (same). How much the manufacturer had to know about the defect, and to what extent the basis for liability was a species of fraud rather than mere negligence, is an important question in cases like this. Courts disagreed. But note what was not at play: the idea that the duty is limited to a contractual relationship.

Courts repeatedly referred to products liability as a public duty. By this they likely meant a duty owed by manufacturers to the community of consumers and users. It was a duty that could reach beyond the buyer-seller relationship to anyone who was injured by a product, as distinct from a contractual duty between parties to a private agreement. These courts expressed no concern that liability needed to be limited, implicitly rejecting arguments such as that put forth in Winterbottom that “[u]nless we confine the operation of such contracts as this to the parties who entered into them, the most absurd and outrageous consequences, to which I can see no limit, would ensue.”347Winterbottom v. Wright (1842) 152 Eng. Rep. 402, 405. Instead, they held that a manufacturer has a “public duty . . . to furnish safe appliances.”348Leas v. Cont’l Fruit Express, 99 S.W. 859, 862 (Tex. Civ. App. 1907). There are contrary cases, such as, for example, The Mary Stewart, in which the district court judge explained in dicta that a ship owner owed no public duty to provide a stable rope, but rather that duty ran only to those with whom he had contracted. 10 F. 137, 139 (E.D. Va. 1881). The court contrasted this situation with that of a defective mast or scaffold, which would expose any invitee to injury. Id. Because only employees were reliant on the rope, no public duty was owed. See id.

The Minnesota Supreme Court, for example, in the case of a brakeman injured by a defective handle, wrote:

One may owe two distinct duties, in respect to the same thing,—one of a special character to one person, growing out of special relations to him; and another of a general character, to those who would necessarily be exposed to risk and danger from the negligent discharge of such duty.349Moon v. N. Pac. R.R. Co., 48 N.W. 679, 680 (Minn. 1891).

What was the scope of this public duty? First, the duty ran to anyone who might foreseeably use and therefore be injured by using the product. It was not quite a duty to all the world but it was a broad duty that encompassed many potential victims. For example, holding that a railcar manufacturer was liable for a defective break handle that injured an employee, the Texas appellate court in Leas explained:

When it assumed the duty and responsibility of furnishing the cars, it assumed the duty, towards all who might, in furtherance of the business of those to whom the cars were furnished, be engaged with the cars, not only to use ordinary care to construct the cars so as to make them safe, but from time to time, as the cars might be furnished to others, to use ordinary care to see that they are in a safe and suitable condition for use. The cars were not only to be properly constructed but to be kept in that condition. 350Leas, 99 S.W. at 861.

That is, whether the injured party was a direct purchaser or an accidental third-party victim, the manufacturer owed a duty of care. Second, the injury had to be something that seemed to the court the predictable outcome of carelessness. That is, the rationale was not that every injury should be compensated to spread the costs of industry, for example. Those ideas developed later.351For a discussion of the development of these ideas, see generally Witt, supra note 51, at 71–102, describing the rise of cooperative insurance in the twentieth century. Thus, claims that struck judges as only remote possibilities which they did not think manufacturers should guard against were not subject to a public duty but rather to a no-liability rule.

Finally, courts struggled with the scope of these duties when it came to the then-extant analogue of public utilities that injured citizens by their alleged negligence.352For example, in one case involving an injury caused by a fallen lamppost, the court struggled with whether to hold liable the gas company that had contracted with the city to put up lampposts. See Lampert v. Laclede Gas-Light Co., 14 Mo. App. 376, 383–84 (Ct. App. 1883). On the one hand, the court explained that:

The rule which requires privity of contract in order to support an action does not always apply where the duty which is undertaken by the contract is a public duty, to be performed by the obligor in the contract for the benefit of the public generally, or for the benefit of particular members of the public distributively.

Id. at 383–84. “On the other hand, it does not follow that because a public duty has been broken, any person who is thereby damaged may have a right of action against the party who had undertaken the duty.” Id. at 384. The court ultimately concluded that the plaintiff injured by the fallen lamppost had a cause of action, but evidence that the company was not negligent required dismissal of the action:

Our opinion is that the contract declared upon, as recited in the petition, raises, on the part of the defendant, a public duty to be performed for the benefit of the inhabitants of St. Louis distributively, and that, for the negligent non-performance of this duty, an action will lie, either by the city of St. Louis suing upon the contract, or by any individual specially damaged thereby, proceeding as for the non-performance of a public duty.

Id. at 393.
Cases both limited and expanded the public duty under those circumstances, but they were treated somewhat differently than products liability cases in which the courts usually found a duty of the manufacturer to the consuming public.353See, e.g., Lewis v. Terry, 43 P. 398, 398 (Cal. 1896).

III. Implications

What are the implications of the fact that privity was never a barrier to suit when it came to mass-market products? First, it is important to correct the record. Casebooks and historians should distinguish between powerful arguments made by defendants and arguments that won the day. More broadly, modesty is the best approach to the historical development of the common law.

Recently, judges have harkened back to common law rules in analyzing statutes and treated these rules as set in stone by early case law.354See Comcast Corp. v. Nat’l Ass’n of Afr. Am.-Owned Media, 140 S. Ct. 1009, 1016 (2020) (stating that, with few exceptions, the common law required but-for causation in 1866 and applying this standard to an 1866 statute). This trend is most evident and debated in public law, but is also present in private law. Yet, as this analysis demonstrates, a historical narrative that legal scholars accepted for a century can be wrong.355See Lahav, supra note 2, at 206 (demonstrating that but-for causation was not the common law tort rule in the middle of the nineteenth century, although most scholars and judges think that it has always been the rule). The same can be true of mistaken narratives in constitutional law. See, e.g., Julian Davis Mortenson & Nicholas Bagley, Delegation at the Founding, 121 Colum. L. Rev. 277, 280 (2021) (demonstrating that there was no prohibition on congressional delegation at the founding, contra the claims of some originalist scholars); Bernadette Meyler, Common Law Confrontations, 37 Law & Hist. Rev. 763, 770 (2019) (“[O]riginalist emphasis on a singular common law rather than multiple common laws in England and America has conduced to a distorted vision of what the Constitution guarantees.”). Judges and scholars would do well to take heed of the poet’s warning:

History has many cunning passages, contrived corridors

And issues, deceives with whispering ambitions,

Guides us by vanities.356T.S. Eliot, Gerontion, Poets.Org, https://poets.org/poem/gerontion [perma.cc/9M5E-5ERV]. Readers should know that this poem contains antisemitic lines and that there is credible evidence beyond its content that T.S. Eliot was antisemitic. See Louis Menand, Eliot and the Jews, N.Y. Rev. Books (June 6, 1996), https://www.nybooks.com/articles/1996/06/06/eliot-and-the-jews [perma.cc/K89X-SBYM].

To be clear, this does not mean that history and tradition have no place in common law analysis but rather that one ought to be wary of potted histories and convenient narratives.

This Part suggests two lessons that can be drawn from the analysis set forth in this Article. First, the teleological narrative of the fall of the citadel of privity is mistaken. Second, the common law evolves in response to social needs and social forces.

A. The Teleological Narrative Is Mistaken

The traditional story is not about privity but the assault on the citadel of privity. To have a successful assault, one must first have a citadel. Thus, it was necessary to invent the citadel in order to dismantle it. Prosser, for example, argued that privity had been rejected as one step toward a strict liability regime.357See Prosser, supra note 3, at 1100–03; see also Goldberg & Zipursky, supra note 20, at 106. It was more convenient for the narrative arc to move from contract or no liability to negligence to strict liability, crafting a teleological story of an ever forward march towards improving the law. The match between the progressive agenda and the fall of privity may be the reason this story had such longevity.

Or this narrative may have survived so well because it involved products and people that were of less interest to scholars. Scholars discounted the harm people suffered due to defective medicine, food, and clothing.358See supra notes 34, 43 and accompanying text. Perhaps this is because these products are associated with the home and women rather than the workplace and working men. The opinions from the period, however, do not discount these injuries.

The narrative that privity was powerful in the nineteenth century and was rejected in the beginning of the twentieth century is also consistent with histories of the progressive era that describe the late nineteenth century as a period of laissez-faire, governed largely by private ordering. Stories that fit with preconceived notions may have greater staying power. But more recent scholarship has demonstrated that local law was thick with regulations.359See, e.g., Novak, supra note 11, at 1–16 (demonstrating in great detail the amount of local regulation that characterized the United States before 1900, focusing mostly on the antebellum period). This Article contributes to that line of historical work, showing that state tort law protected people’s ability to sue for products that injured them, even if they didn’t sue very often.

B. The Common Law Evolves

Products liability doctrine was, and remains, an area of struggle between injured consumers who sought recompense and manufacturers who sought immunity.360Compare A. Mitchell Polinsky & Steven Shavell, The Uneasy Case for Product Liability, 123 Harv. L. Rev. 1437, 1438 (2010) (arguing, based on formal economic models, against most product liability law), with John C.P. Goldberg & Benjamin C. Zipursky, The Easy Case for Products Liability Law: A Response to Professors Polinsky and Shavell, 123 Harv. L. Rev. 1919 (2010). See also Sharkey, supra note 5. For the most part, nineteenth-century courts recognized what they called a “public duty” on the part of manufacturers to consumers as a class, but this result was not preordained. The best way to understand products liability law is as a response to social and economic trends.

The relationship between law and historical context is illustrated by the 1896 case of the murphy bed that injured the boarder who sued not his landlord but the manufacturer.361See Lewis v. Terry, 43 P. 398 (Cal. 1896). In 1850, his case would not have been prevented by the doctrine of privity but rather his injury would have been exceedingly unlikely to have occurred at all because of the structure of American life. There were few people who lived and worked in cities who would have needed to rent a room. Few whose landlords would have furnished that room with items made by anyone other than themselves, much less a murphy bed, intended to fit more strangers into smaller spaces, as compared with familial arrangements where everyone slept together in one bed or one room. Indeed, William Murphy did not patent his eponymous invention until 1911, although he was experimenting with it in the late 1800s, which is likely how Mr. Lewis’s landlords in San Francisco had one in 1895.362Joseph Caputo, Curator Finds Murphy Bed’s Place in American History, Smithsonian Mag. (Apr. 28, 2009), https://www.smithsonianmag.com/smithsonian-institution/curator-finds-murphy-beds-place-in-american-history-48350873 [perma.cc/8AJP-95L7]. The occasion for that lawsuit could not have arisen before the 1890s. With urbanization, a new workforce, mass immigration, and similar developments, his suit was first a possibility and then an eventuality.

The earliest cases demonstrate that courts believed producers had a broad public duty not to sell harmful products, a duty which ran to all end users of those products. The doctrinal growth of products liability law was related to significant changes, wrought by the Industrial Revolution, in what products were available that could injure people, who was buying them, and social changes in terms of what people were buying. Courts had always recognized that manufacturers had a duty in tort. What expanded was not the duty but rather innovation. In other words, the number of products to which that duty could apply increased exponentially. As people bought more mass-produced and mass-marketed products, they were injured by these products. A minority of them sued, and defendants litigated, leading to the many opinions discussed in this Article.

In 1852, when Thomas v. Winchester was decided, there were very few mass-marketed products people could buy. Farmers and farm workers did not buy a threshing machine or a saw from a catalogue. Many machines that later came to harm people simply had not been invented or were not available for sale. Those that were sold were made by locals. The most widely available products for sale by mail would have been medicines.363See Young, supra note 64, at 150. A report of the American Medical Association sounding the alarm was published in 1912, but by then it had been a problem for years. Am. Med. Ass’n, Nostrums and Quackery 37 (2d ed. 1912). One of the most interesting takeaways from the report is the extent to which people complained of bad medicines, but there was no recognition in the report of any products liability lawsuits. (Notably, most of the largest instances of multidistrict litigation today involve products liability cases, and many of these involve pharmaceuticals or medical devices.)364See Zachary D. Clopton, MDL as Category, 105 Cornell L. Rev. 1297, 1316 (2020). As more products were brought to market that could injure—from coats to threshers to wringing machines—cases were decided, opinions were written, and a products liability doctrine developed.

Over the nineteenth century, courts built a set of rules for mass-produced products as those products emerged as part of a growing national economy: first medicines in the 1850s and 1860s,365See Thomas v. Winchester, 6 N.Y. 397, 397 (1852); Blood Balm Co. v. Cooper, 10 S.E. 118, 119 (Ga. 1889); Darks v. Scudders-Gale Grocer Co., 130 S.W. 430, 435 (Mo. Ct. App. 1910). then canned meat,366See Tomlinson v. Armour & Co., 70 A. 314, 316 (N.J. 1908). clothing,367See Gould v. Slater Woolen Co., 17 N.E. 531 (Mass. 1888); Gerkin v. Brown & Sehler Co., 143 N.W. 48 (Mich. 1913). furnishings,368See Lewis v. Terry, 43 P. 398, 398 (Cal. 1896) (murphy bed). and finally machines.369See Moon v. N. Pac. R.R. Co., 48 N.W. 679, 681 (Minn. 1891) (railcar manufacturer liable to brakeman for defective brakes); Empire Laundry Mach. Co. v. Brady, 60 Ill. App. 379, 383 (App. Ct. 1895), aff’d, 45 N.E. 486 (Ill. 1896) (automated wringer machine); Huset v. J.I. Case Threshing Mach. Co., 120 F. 865, 872 (8th Cir. 1903) (thresher); Leas v. Cont’l Fruit Express, 99 S.W. 859, 861 (Tex. Civ. App. 1907) (defective handhold on railcar brake); Davidson v. Montgomery Ward & Co., 171 Ill. App. 355, 367 (App. Ct. 1912) (defective sawing machine). When these products were produced locally and sold by the producer, the local producer/seller was liable. As mass markets developed and producers and sellers were separated by chains of commerce, liability followed.

The dominant idea in the opinions was that manufacturers or producers owed the consuming public a duty to be careful, but if they were careful, they would not have to pay for injuries that their products caused. This is why when Armour & Co. was sued for the needle in the soap, which severely injured the unsuspecting user, it was freed from liability. The likelihood of a needle falling into soap was so miniscule that there could be no expectation that the defendant take more care. But when Gerkin sued the fur seller, he did recover because the injury had happened before and the seller knew allergic reactions to the cheap and dangerous dye it used for furs were a real possibility.

These rules seem to have had some effect. There is historical evidence that they spurred some manufacturers to test their products. Sally Clarke demonstrates in her history of automobile litigation, which centers on MacPherson, that, in the decade before that case was decided, car companies were testing their products.370See Clarke, supra note 54, at 14–15 (“Although companies in a few industries, notably railroads, had established research laboratories in the late nineteenth century, most firms in other industries maintained small testing labs to assess materials on their arrival to a plant.”). In 1907, the industry magazine Horseless Age called for more testing, and companies were implementing the suggestion.371“In the auto industry, Horseless Age noted in 1907 both the need for and lack of adequate research facilities. Within a few years, however, public and private entities had begun to establish research institutions.” Id. Clarke writes that many of the cases involving defective cars seem to have sounded in contract, with the buyers (often society men) wanting their money back rather than compensation for injury. See id. at 19.

Liability of producers, even distant ones, was an outgrowth of the separation between producers and direct sellers that occurred in the second half of the nineteenth century. There was no occasion for a defendant to invoke privity in the early days when producers sold their own goods. But it made perfect sense to try to make inroads with this argument for a national manufacturing company like Armour & Co. once it came into being. As this Article has demonstrated, those arguments largely failed. Courts during this period routinely held that manufacturers owed a general duty to the consuming public.

It is tempting to think that because the common law recognized causes of action for injurious products as the method of manufacture and distribution changed, similar liability today for new technologies and inventions is somehow a natural and necessary requirement of our common law tradition. But this would be the wrong lesson to take from this analysis. The common law’s evolution will continue to be responsive to social context as well as intellectual trends.

The common law is by its nature an evolving body of law—to freeze it at some historical point is to render it something else. Flexibility, combined with an eye towards consistency, was and remains the greatness of the American common law tradition.372I think Lemuel Shaw’s explanation of the common law’s flexibility is the best: “[T]he common law consists of a few broad and comprehensive principles, founded on reason, natural justice, and enlightened public policy, modified and adapted to the circumstances of all the particular cases which fall within it.” Nor. Plains Co. v. Bos. & Me. R.R., 67 Mass. (1 Gray) 263, 267 (1854); see also Bernadette Meyler, Towards a Common Law Originalism, 59 Stan. L. Rev. 551, 567, 585–86 (2006) (describing the concept of the common law as disunified and contested at the founding and describing the understanding of the common law during that period as being at once immemorial and flexible). When legal analysis relies on English or U.S. historical sources without understanding the extent to which common law doctrines were contested in U.S. legal history, it fails to understand its own subject in a fundamental way. Judges deciding cases in the common law tradition apply their understanding of general principles to present social circumstances. The benefit of flexibility is that if they were wrong in the past, or even if they are wrong today, they can correct that error going forward.373An example of this is the debate over whether Amazon is a seller, subject to products liability law for defective products. Although the courts are interpreting statutes in this area, they draw on common law understandings. Compare Amazon.com, Inc. v. McMillan, 625 S.W.3d 101, 112 (Tex. 2021) (holding that Amazon is not a seller and therefore not subject to products liability law in Texas), with Loomis v. Amazon.com L.L.C., 277 Cal. Rptr. 3d 769 (App. Ct. 2021) (holding that Amazon can be held liable for the sale of defective products because of the link through a virtual chain of distribution that is sufficient to give rise to liability).


Tort doctrine has always evolved to address new social problems—whether dangerous medicines in 1850, risky machines in 1910, or opioids, talc, and earplugs today. The through-line of tort history from the beginning of mass-marketed products has been to impose a duty on manufacturers that ran to all users of their products. Nineteenth-century American courts agreed that manufacturers who sold goods far and wide in the emerging industrial economy had an obligation to consumers to produce safe products. Judges consistently rejected Lord Abinger’s fearful warning that there must be contractual limits on personal injury suits because otherwise “the most absurd and outrageous consequences, to which I can see no limit, would ensue.”374Winterbottom v. Wright (1842) 152 Eng. Rep. 402, 405. This was because these judges did not think privity was good social policy and did not think it was required by either reason or natural justice. When it comes to products liability, the citadel of privity was but a castle in the air. It is time to correct the casebooks and treatises.

*Professor of Law, Cornell Law School. Many thanks to John Goldberg and Adam Zim- merman for their careful read and comments that improved the piece immeasurably, and to Deborah Dinner, Valerie Hans, Luke Herrine, Lawrence Friedman, Greg Keatin g, Aneil Kovvali, Jeff Rachlinski, David Simon, Tim Waters, John Witt, and participants in the Cornell and IU Maurer faculty workshops for their comments that helped improve the piece. I benefitted greatly from conversations with Nora Freeman Engstrom, Law rence Friedman, Bill Novak, and Robert Rabin in developing this project. I am indebted to Adam Mackie and N ina Scholtz, librarians at the University of Connecticut and Cornell, whose research help was invaluable to this project, and to Alison Draikiwicz fo r research assistance. I am also grateful to the deans of both Cornell Law and the University of Connecticut, as well as the Radcliffe Fellowship, for financial support