class actions, disabled rights, copyright, attorneys general, online speech, law schools, obesity, New York, mortgages, legal blogs, safety, CPSC, pharmaceuticals, patent trolls, ADA filing mills, international human rights, humor, hate speech, illegal drugs, immigration law, cellphones, international law, real estate, bar associations, Environmental Protection Agency, First Amendment, insurance fraud, slip and fall, smoking bans, emergency medicine, regulation and its reform, dramshop statutes, hotels, web accessibility, United Nations, Alien Tort Claims Act, lobbyists, pools, school discipline, Voting Rights Act, legal services programs

A Short History of Product Liability Law
By Richard Epstein, 11-10-2005

Scope of Product Liability Laws The topic of product liability law arises so often in the media today that some brief explanation of its scope, origins, and impact might prove useful. This short essay is intended to be a primer for those who appreciate the importance of the topic, but would like some guidance on how to think about it more rigorously.

The first order of business is to realize that not all cases that involve products are products liability cases. Let driver A sue driver B in an automobile accident, and there is a suit that involves products, but there is not a product liability suit. To count as a product liability suit, it is necessary that the defendant selected for suit not be in possession of the car at the time that the accident took place. All products liability cases take place at one level removed, and thereby raise the question of whether the seller (i.e., defined in this context as manufacturer, distributor, retailer, or even component part manufacturer) is responsible for an injury caused by its product once it enters into the stream of commerce.

To many people, the initial challenge is why that "remote" party should be held responsible at all, given that each of two drivers could be held responsible for hitting the other or blocking a right of way. Stated otherwise, the initial center of gravity in any lawsuit should always be those parties who have the greatest ability to prevent the accident. In general the parties in possession of dangerous instruments fill that bill. As a matter of principle, finding product liability should be an uphill battle. How then did the law evolve historically so that product cases have become so common?

Common Law Origins The earliest common law view was that the plaintiff could never win this uphill battle. The operative doctrine went under the somewhat mysterious name of the "privity limitation," which meant that any upstream party in the chain of distribution could only be sued by the person to whom he sold and delivered the product in question. This rule owes its origin to the notorious 1842 English case of Winterbottom v. Wright. There the Court held that the party who poorly repaired a coach owned by the Postmaster General could not be sued at all (because of sovereign immunity) when the coach had been let out to a contractor whose driver was injured when the coach broke in ordinary use. The driver was only in privity with his employer. The employer could not be sued because he had done nothing wrong. The repairman, who did something wrong, could not be sued because of the absence of any direct connection.

That outcome was troublesome for this reason: everyone conceded that the defendant had done a poor job in fixing the coach, which was in defective condition when it left his hands. It was also clear that neither the Postmaster General nor the employer had done anything to weaken the wheel further. Why then should a defendant that makes a product with a latent defect, such that the product fails while in its original condition, escape from liability when the injured party used it in its intended fashion? Note here that the three key prerequisites adopted later for liability work together to weed out weak cases. The requirement that there be a latent defect means that the driver has no reasonable opportunity to find out about the condition and alter his conduct accordingly. The requirement that the product be still in its original condition means that the two subsequent parties in possession of it did nothing to make the situation worse. The proper conduct of the plaintiff neatly negates his responsibility for bringing about the loss.

This criticism of the Winterbottom decision set the roadmap for the legal developments of the next 100 years. The first crack in the "citadel" of privity came in the 1852 New York case of Thomas v. Winchester, which involved an instructive variation from the Winterbottom facts. Here a druggist by mistake prepared some deadly belladonna, which he sold to a customer as harmless extract of dandelion. Had the customer been injured by taking the deadly potion, the privity condition would have been satisfied and he could have recovered for any harms suffered from consuming the inherently dangerous product. But the victim in this case was his wife, one person removed. The Court brushed aside this privity objection, and allowed the action for inherently dangerous substances, which were after a fashion distinguishable from the defect in Winterbottom, which came from defective craftsmanship, not a mislabeled poison.

That intermediate step allowed recovery for the plaintiff in Thomas, but it did not overturn the basic privity rule of Winterbottom. The next legal step adopted a compromise position for those products that were imminently dangerous because defectively made. If the plaintiff could show that the defendant knew of the defect when the product was put on the market, then recovery was allowed on a theory tantamount to fraud. But ordinary negligence did not support recovery in this second class of cases.

The great 1916 decision of Judge Cardozo in MacPherson v. Buick is often hailed as the breakthrough case in modern product liability law because of its rejection of the privity doctrine. But in fact it was nothing of the sort. As depicted by Cardozo, a wooden wheel on an expensive Buick just spontaneously broke while the car was traveling along a public road at eight miles per hour. Cardozo allowed the suit to go forward by rejecting the distinction between inherently dangerous products (such as poisons or guns) and products imminently dangerous because defectively made (such as weak wheels or farm equipment). Rather, he said sensibly that both kinds of defect should be treated alike and subject to a uniform negligence standard; i.e., the plaintiff had to show that the defect was attributable to the negligence of the defendant in preparing the wheel, or in this case in its negligent failure to inspect the wheel which had been defectively made by a tire manufacturer.

In the majestic recounting of Cardozo, the case still fit into the paradigm of strong product liability cases. The defect was latent, the car was in its original condition when the accident occurred, and the plaintiff had done nothing wrong. But it is just here that the extended chain of custody in product liability cases opens these cases to serious abuses on matters of evidence. Professor James Henderson of Cornell did a close check on the factual record on MacPherson, only to come up with a frightenly different account. The car had been used to haul cement for some time and the wheel broke when the plaintiff tried to turn it while driving at 30 miles per hour through four inches of gravel. One strong, if belated, reason for the Winterbottom rule is that it cuts out these weak cases without having to bear the expense of the trial. The basic rule of thumb is to beware of the downstream user.

The short-term impact of MacPherson was quite minor, because the litigated cases fell within the original paradigms, with many involving people eating food, purchased by others, that contained ground glass or similar impurities. So far so good. The next great step in liability took place in Escola v. Coca-Cola Bottling Co. (1944), where Judge Traynor, in an opinion that repeatedly invoked the "public interest", insisted that actions for defective products should be governed by a strict liability standard, so that negligence in fabrication or design need not be proved at all. But he insisted that liability only attached for latent defects that caused harm in their original condition to individuals who made "normal and proper" use of the product. The standard was sensible, but the evidentiary risks were now greater than ever. Many bottle cases brought against Coca-Cola and its bottlers at the time were fraudulent, as claimants were instructed by their lawyers to say that they handled the bottle correctly even though some of the bottles had been stored in hot places or opened by knocking them on the side of a counter. Downstream mischief again. But that said, liability still remained so infrequent that common insurance practice of the period did not ask for separate premiums to cover product liability insurance.

The Modern Period The true confusion in this area came only in the period after Escola with two further developments. The first of these was that the original limitations on product liability actions were eliminated one by one. The second was the rejection of all contractual defenses to liability.

Doctrinal Changes On the first issue, the key event here was not the well-known 1965 Second Restatement of Torts, which endorsed the strict liability rule of Escola. Rather it came in the years afterwards when all the limitations in the initial formulation were cast to one side. Here are some of the shifts.

First, there was no longer a requirement that the defect in question be latent. Obvious conditions could count as product defects. At this point, information is no longer necessarily asymmetrical, and the sensible approach is to hold the downstream user responsible for using the product with knowledge of the risk that it created. In some instances, as with machine tools, there might be a responsible downstream party such as an employer obligated to pay workers' compensation to an injured employee, but by the 1970s that low-level remedy was thought to be insufficient. The older cases were unceremoniously overruled.

The scope of liability expanded still further as courts ruled that more and more obvious conditions did not bar liability, among them the common knowledge that automobiles could not withstand certain kinds of collisions. Now liability was allowed on a general crashworthiness theory, even when the design in question met all regulatory standards. Liability became a question of a general cost/benefit analysis, where juries could decide what alternative designs were appropriate and what not. But judgments made in hindsight are always subject to bias that overrates the risks that eventually come to pass. And juries were never asked to decide whether the inclusion of one safety feature required adjustments elsewhere. Inconsistent verdicts could arise in separate cases imposing requirements that could not be simultaneously satisfied, so that no jury verdict could provide any guidance of what to do in the next case.

Similar developments occurred in duty-to-warn cases, which were the successors to the early mislabeling cases. Now the requirement was to disclose adequately all risks, which is hard to do in the abstract with vaccines and drugs that have odd constellations of advantages and disadvantages. But here too the possibility of relying on, say, FDA standards was systematically rebuffed so that once again the question of what counts as a product defect was not answered, though it was certain that defects covered far more than the old cases of contamination or mislabeling. To this day there is no clear safe-harbor on warning issues for dangerous drugs.

The second condition also received a beating, as newer cases held that liability could be imposed on the original manufacturer even if an intermediate party, such as an employer, altered equipment in ways that made it more dangerous than before. The rule was never that the modification was irrelevant. Rather, it was that the jury could decide whether the liability could be imposed on the ground that these alterations were foreseeable or easily preventable.

Finally, the rules on the conduct of the product user were altered in ways that expanded the risk of liability. The older rule that required normal and proper or intended use as a precondition for product liability gave way to rules that held that "foreseeable misuse," even if deliberate, did not necessarily sever liability, although in most states it could be used to reduce the potential verdict by some unspecified amount. But reckless driving does not bar a product liability action against an automobile manufacturer, and misuse of drugs may well have the same consequence.

Contractual Invalidity We now come to the second fatal decision. Once it became clear that product (and medical malpractice) liability were on the rise, manufacturers did not wish to hide behind the broken wall of privity. Rather they sought to enter, often through dealers, into direct contractual relations with their customers to limit liability and restore the principle that downstream control was preferred. These contractual limitations were not perfectly drafted, but the drafters were never given the chance to improve on their intial offerings. Rather, starting with Henningsen v. Bloomfield Motors (1960), courts consistently held that all efforts to waive or disclaim liability for defective products were against public policy, and that rule is ironclad today. The upshot is that the mistakes that courts make cannot be subject to private correction. Any change must instead proceed through legislation where the new interests that have grown up around the expanded product liability rules (e.g., plaintiff lawyers) have been exceedingly adept at preserving the status quo.

The Joint Payoff The combination of poor liability rules and weak enforceability of contracts proved to be of enormous force. Matters reached their peak in the next generation of cases, which involved toxic substances that did not cause injury all at once, but were ingested or inhaled over long periods. These cumulative trauma cases are much harder to prove, and even though many of the earlier asbestos cases did arise out of cases of knowing concealment of dangers not known to the other side, as the years went on, the factual predicate of the plaintiff's cases became weaker and weaker, yet the dollars obtained in liability became ever greater. Some sign of the difficulty is found in the simple observation that exposure to asbestos went down sharply for all sorts of reasons in the post-1970 period, yet the number and severity of suits continues to rise apace. The recent revelations of serious irregularities in the silicosis cases points to the same conclusion. There is an irony here. The issue of fraudulent or exaggerated claims that received little attention in MacPherson and Escola may well be the central problem in the modern product liability area. The great challenge is to find ways to cut out these cases without destroying the limited but useful function that product liability law can serve.

A Short Recipe for Reform Remember this formula: Product liability only works well for latent defects that cause harm in their original conditions in products that were used in their normal and proper way. We are a long way from that gold standard today.




Published by the Manhattan Institute

The Manhattan Insitute's Center for Legal Policy.