Concept
Defenses to Strict Liability
Claims against manufacturers and other defendants arising out of the manufacture and distribution of products, and defenses to such claims
Defenses to Strict Liability
Under Section 402A(1) of the Restatement (Second) of Torts, a seller is liable for physical harm caused by a product sold in a defective condition unreasonably dangerous to the user or consumer, if the seller is in the business of selling such a product and the product is expected to reach the user, and does reach the user, without substantial change.1 Subsection (2) closes off two answers the seller might otherwise offer. The rule applies even though the seller exercised all possible care in preparing and selling the product, and even though the user never bought from or contracted with the seller.2 A seller who cannot win on care or privity, and cannot knock out an element, still has arguments left. This section collects them.
Not every argument called a defense is one. Comparative fault and assumption of the risk are true defenses: the defendant asks the court to bar or reduce recovery even though the plaintiff has proved the claim.3 The state-of-the-art argument, the argument that a danger was open and obvious, and misuse are also called defenses, but they might better be characterized as challenges to the plaintiff's ability to make a prima facie case of product defect.4 The distinction has consequences. A defendant who defeats an element defeats the whole claim, while a true defense may bar the claim, reduce it, or do nothing, depending on the state.
Misuse shows how one fact can do several jobs. An unusual use can bear on whether the product was defective at all, and on whether any defect was a proximate cause of the harm. Neither role is a defense in itself, yet misuse can also operate as a defense in its own right.5 The first role follows from the theory of liability, which treats delivery of an article as wrongful when it is not safe for the purposes for which the consumer would ordinarily use it.6 One form of misuse gets special treatment. A user's failure to heed reasonable warnings or follow instructions on safe use ordinarily should bar recovery, provided that the danger of ignoring them is evident, the disregard is a substantial cause of the harm, and there is no simple means or apparent reason for the manufacturer to design the danger out of the product. Many courts, and a few statutes (Ariz., N.C., N.J.), bar recovery for injuries from ignoring warnings and instructions.7
Consider a plaintiff hurt when a power saw kicks back while she uses it for a task its maker never contemplated. The seller's lawyer has three distinct arguments. The saw was not defective for the purposes for which consumers ordinarily use it. The unusual use, not any defect, caused the injury. And, where misuse is a defense, the use itself defeats recovery. The first two deny an element of the plaintiff's case. Only the third concedes that case and asks to avoid it.
Tort law is state law, and the states differ on each defense that follows. A plaintiff's negligence may bar a products recovery entirely in one jurisdiction, reduce it in others, and have no effect at all in others.8 Assumption of the risk shows the drift. A good many states have adopted comparative negligence and already treat assumption of risk as a form of negligence in negligence cases. They now treat a plaintiff's conscious choice to encounter the risk posed by a defective product as a form of comparative fault in strict liability claims as well.9 Because identical facts can produce different results under different rules, each argument should be run through the same short sequence.
The sections that follow track that sorting. Conduct comes first: comparative fault, assumption of the risk, and misuse and modification. Arguments about the product and its danger come next: open and obvious danger, state of the art, and unavoidably unsafe products. Time bars, preemption, and compliance end the chapter. Comparative fault opens the series because it also supplies the framework many states now use for assumption of the risk.
Comparative Fault
Strict liability under § 402A ignores the seller's conduct: the seller is liable although it has exercised all possible care in preparing and selling the product.10 The question here is whether the rule also ignores the plaintiff's conduct. Under comment n it largely did, because the plaintiff's negligence was no defense unless it amounted to conscious assumption of the risk, the subject of the next subsection.11
The drafters added § 402A in 1965, when contributory negligence was the general rule in negligence cases and fully barred a negligent plaintiff from recovery. They concluded that contributory negligence should not bar recovery on a strict products liability claim.12 In their view strict liability was a different type of liability, one that placed the risk of injury from defective products on the seller even when the plaintiff's own negligence also caused the injury, and early cases under § 402A followed comment n.13
The results could be dramatic. In Kimco Development Corp. v. Michael D's Carpets, a store ordered carpet backing and piled it to the ceiling, where it ignited from the heat of the lights. The jury found the store 80 percent at fault for the fire, but it also found that the seller had failed to warn that the product was flammable. The Pennsylvania court held that the store could recover its fire loss in full, because the store's negligence was no defense to its strict liability claim for failure to warn.14
The rule also shields the plaintiff who merely fails to find a defect, and that shield has a reason beyond history. Strict liability rests on the special responsibility for public safety that a seller undertakes by entering the business of supplying products, and on the forced reliance of buyers upon that undertaking.15 As one formulation puts it, the wrong lies in delivering an article that lacks the qualities the manufacturer represented, when the consumer could not readily detect the lack.16 A consumer who relied on the product and never looked for a hidden flaw has done little more than rely, which is the very reliance on which the rule is built.
Comparative fault changed the question. Once most courts and legislatures abandoned the all-or-nothing effect of contributory negligence in the 1970s and 1980s, the issue was no longer whether a careless plaintiff should lose everything but whether she should lose a share.17 Merging the two doctrines drew three objections. Comparing a plaintiff's negligence with a defendant's strict liability mixes apples and oranges. A reduction in damages weakens the manufacturer's incentive to make safe products. And jurors cannot assess, measure, or compare such different things.18
In Daly v. General Motors Corp., the California Supreme Court merged comparative fault with strict products liability and rejected each objection. It concluded that the goals of strict liability would not be frustrated by comparative principles. The plaintiff remains relieved of proving that the manufacturer was negligent, the defendant's liability for a defective product remains strict, and the plaintiff's recovery is reduced only to the extent that his own lack of reasonable care contributed to the injury. The cost of compensating the victim, though proportionately reduced, stays on the manufacturer and is spread through it, and the court saw no reason of policy why the share of damages flowing from the plaintiff's own fault should be borne by others.19 Courts that followed answered the deterrence objection the same way: manufacturers remain responsible for all damages attributable to defects and are shielded only from the portion of the loss attributable to the plaintiff's own failure to act properly. They answered the apples-and-oranges objection by observing that factfinders have no trouble apportioning responsibility between negligent plaintiffs and manufacturers subject to strict liability.20 Daly found the practical objection unpersuasive for the same reason, since jurors can fairly apportion liability.21
Daly also exposed an anomaly in the old rule. A plaintiff who sued in negligence could have even a negligent assumption of the risk merged into comparative principles, so that it no longer defeated recovery, but identical conduct in a strict liability case still barred recovery completely. Strict liability, which was developed to free injured consumers from the constraints of negligence and warranty, thus put the consumer in a worse position than a simple negligence claim would.22 The anomaly mattered because the two theories often travel together: a design found unreasonably dangerous under the risk-utility test will usually support a finding that the manufacturer was negligent as well.23
Not everyone accepted the merger. A critic predicted that every defendant charged with marketing a defective product would assert that the injured plaintiff had done something, anything, that could be deemed contributory negligence: driving 56 miles an hour instead of 54, failing to discover a latent defect, or yielding to the salesman's persuasion. Pleading the defense, he said, would become boilerplate.24 Some courts reject the merger outright. Bowling v. Heil Co. held that comparative negligence or comparative fault has no application to a products case based on strict liability in tort. Strict liability, the Ohio court reasoned, focuses on the product rather than on the conduct of its manufacturer or seller, so it "does not seek to apportion a loss among all persons who have caused or contributed to it. Rather, it seeks to spread the loss among all users of the product."25
Where comparative fault applies, the mechanics are familiar. In most (though not all) jurisdictions today, the jury decides whether the product was defective and also whether the plaintiff was negligent. If it finds both, it assigns percentages of fault to the seller for selling the defective product and to the plaintiff for her negligence, and the judgment follows those percentages.26 One formulation of the apportionment standard directs the factfinder to consider the nature of each person's risk-creating conduct, including any awareness or indifference with respect to the risks and any intent with respect to the harm, and the strength of the causal connection between that conduct and the harm.27 Weighing of this kind is not foreign to strict liability itself. For abnormally dangerous activities, the Second Restatement replaced the First Restatement's fixed elements with factors, and strict liability becomes more likely when, for example, the activity creates a high risk.28
Statutes can also settle the vocabulary problem. Some states call their statutes comparative fault schemes but define fault broadly enough to include the sale of a defective product. Arkansas's statute reaches "any act, omission, conduct, risk assumed, breach of warranty, or breach of any legal duty which is a proximate cause of any damages sustained by any party," and the phrase "breach of any legal duty" surely includes strict liability for selling a defective product. Maine's statute applies, in addition to negligence claims, to a "breach of statutory duty or other act or omission that gives rise to a liability in tort or would, apart from this section, give rise to the defense of contributory negligence," which would also include a strict products liability claim.29
The hypothetical tracks the two regimes as the sources describe them. Under § 402A, Ortney would recover her full damages despite her negligence in failing to have the brakes checked, and under apportionment an 80 percent share of fault on the seller yields an 80 percent recovery for her.30 Adopting comparative fault therefore often works against the products plaintiff. In negligence it rescued plaintiffs from a complete bar, but in strict products liability, comment n had already spared the plaintiff any reduction for ordinary negligence, so the new rule adds a reduction that was not there before.
One question survives the merger: does a plaintiff's careless failure to discover a defect now join other fault in the apportionment? The original reason for ignoring it, the all-or-nothing bar, is gone, and a number of states say simply that all fault must be compared. Yet almost every court that has considered this precise issue has shielded consumers from responsibility for failing to watch for product defects, holding that a plaintiff's careless failure to discover a defect lies outside damages apportionment.31 The shield is narrow. It covers the precise matter of failing to discover a defect, and it leaves other negligence to the jury's ordinary assignment of percentages.32
A plaintiff who ignores a warning is treated more harshly than one who merely misses a defect. Failing to respect reasonable warnings of danger or to follow instructions on safe use is a special form of misuse that ordinarily should bar recovery, provided that the danger of ignoring them is evident, the disregard is a substantial cause of the harm, and there is no simple means or apparent reason for the manufacturer to design the danger out of the product. Many courts and a few statutes (Arizona, North Carolina, New Jersey) bar recovery on this ground.33 The Misuse and Modification subsection returns to the point.
Crashworthiness claims raise a last wrinkle. There the manufacturer is liable only for the exacerbated injuries caused by a defective design, and the question is whether that liability should be reduced further by a percentage matching the plaintiff's negligence in causing the collision. Case law is divided. The logically sounder view holds that the enhanced-injury concept already apportions damages on a comparative basis, since the defendant answers only for the increased injury and not for the injury of the crash itself, so no further reduction follows.34
Assumption of the Risk
Section 402A makes a seller liable for physical harm caused by a defective, unreasonably dangerous product, and the rule applies although the seller has exercised all possible care.37 The seller's diligence is no defense, so the question becomes which plaintiff conduct is. Comment n answers in two halves. The first excuses a plaintiff whose only fault was failing to discover the defect or to guard against its possible existence; under comment n, that carelessness leaves the plaintiff's recovery intact.38 The second half reaches the user who has found the defect: "If the user or consumer discovers the defect and is aware of the danger, and nevertheless proceeds unreasonably to make use of the product and is injured by it, he is barred from recovery."39
That conduct is assumption of the risk: a person's voluntary decision to face a danger he understands. Under § 402A it is a complete bar, so a plaintiff who proves every element of the claim still recovers nothing. The defendant must plead and prove it, because it is an affirmative defense.40
The defense rests on consent. A user who understands a danger and freely chooses to run it accepts responsibility for the result, and the seller who created the danger is excused. Because a person cannot consent to what he does not know or to what is forced upon him, the encounter must be informed and voluntary.41 Comment n adds a third requirement: the plaintiff's decision to proceed must have been unreasonable.42 Knowledge, voluntariness, and unreasonableness together make up the defense, and warranty law draws a similar line. In Hayes v. Ariens Co., a Massachusetts decision listing the defenses that can defeat an established breach of warranty, the court repeated that a user who unreasonably proceeds with a product he knows to be defective and dangerous "relinquishes the protection of the law."43
Knowledge is the element on which most disputes turn. The plaintiff must have appreciated the particular danger that hurt him, and a vague sense that a product could be dangerous if handled carelessly does not qualify, because a person who does not know how a product can injure him cannot weigh the likelihood and seriousness of the harm or choose to accept it.44 In Hadar v. AVCO Corp. (Pa. Super. Ct. 2005), the plaintiff knew the general risk of getting a hand caught in a corn picker's rollers, yet the court found a question of fact whether he understood the specific risk that the rollers could pull in a three-foot corn stalk so quickly that he could not release it before his hand followed.45 Appreciation does not require technical understanding. Most courts hold that a plaintiff need not grasp the mechanical, chemical, or biological mechanism of the harm, so a user who knew a gas grill might explode if he failed to light and vent it properly appreciated the risk even without understanding where unlit gas would pool.46 Because the inquiry concerns what this plaintiff actually understood, it is ordinarily a jury question, and the jury need not credit the plaintiff's account of his own state of mind but may test it against what people of similar age and experience ordinarily understand about such a hazard.47
The two drivers differ in what each understood, and in a state that follows comment n that difference decides the case. Carelessness about a danger a person only sensed is ordinary negligence, which comment n ignores; a choice made with a clear picture of a particular danger is assumption of the risk, which comment n treats as fatal. A fine difference in the plaintiff's understanding therefore separates full recovery from none.48
Warnings show the same logic from another side. A plaintiff who never learned of a hidden risk cannot have assumed it, whereas a plaintiff who already knew and understood the risk may be barred from claiming that a warning about it was missing.49 A plaintiff who disregards a reasonable warning meets a related bar. Many courts and a few statutes treat the disregard as a special form of misuse that defeats recovery when the danger of ignoring the warning is evident and the disregard substantially causes the harm, a subject the next subsection takes up.50
The second element asks whether the plaintiff chose freely, and the test is whether a reasonable alternative existed. A person who faces two bad options acts under a kind of duress that defeats consent; a person who can avoid the danger by a reasonable course and knowingly takes the dangerous one cannot claim compulsion.51 The Second Restatement treats acceptance of a risk as involuntary when the defendant's wrong left the plaintiff no reasonable way to avert harm or to exercise a right the defendant had no power to take away.52 Employment is the standard application. An employee who knows a machine is unguarded must run it or quit, changing jobs usually costs a great deal, and the manufacturer has no right to take the job away, so the decision to keep working is typically reasonable and, for that very reason, not voluntary.53 A number of courts give rescuers similar protection.54
Courts disagree about how far the workplace point reaches. In Rhoads v. Service Machine Co., a press operator's arm and hand slipped into a large punch press when she lost her balance while activating it, and the court upheld the jury's rejection of the defense despite her knowledge that the machine was unguarded, calling the voluntariness of such an assigned worker's assumption of risk "illusory."55 The Ohio Supreme Court went further in Cremeans v. Willmar Henderson Mfg. Co. and abolished the defense in the workplace setting, reasoning that workers are often effectively trapped into dangerous tasks.56 Other courts treat workplace encounters like any other. They reason that any compulsion comes from the employer, a third party, and that workers sometimes make deliberate choices to face risks they fully understand, as the operator in Karim v. Tanabe Mach., Ltd. did when he grabbed a control box near a moving chain and sprocket that caught his hand.57
Comment n's third requirement departs from the classic defense, which required only that the plaintiff appreciate the risk and choose it willingly.58 The added test is objective: would a prudent person looking after his own safety have chosen to encounter that risk in those circumstances? The requirement seldom changes a result. A plaintiff who had a reasonable alternative and still took the dangerous course has usually acted unreasonably, and a plaintiff who had no reasonable alternative was never voluntary to begin with, so few reasonable assumptions of risk remain.59 Comment n itself calls the defense a form of contributory negligence, a description that prepared the way for what comparative fault later did to it.60
Comparative fault undermined the complete bar. Where a state has adopted comparative fault, a plaintiff who negligently assumes a risk in a negligence suit loses only a share of the damages, yet under § 402A the same conduct wiped out a strict liability claim. The California Supreme Court in Daly v. General Motors Corp. called that result an anomaly: strict liability, built to free consumers from the limits of negligence and warranty, left the plaintiff worse off than a negligence claim would and rewarded the careful choice of a pleading theory.61 The court added a point about incentives. A complete bar tells a manufacturer that it profits when its product is so defective that an injured user must have known of the defect, and that inverts the safety incentive the liability exists to create.62 Daly therefore abolished the separate defense of assumption of risk to the extent that it is a form of contributory negligence and sent the plaintiff's choice to the jury as one more share of fault.63 The Third Restatement takes the same position, as do statutes in Iowa and Utah, the first applying comparative negligence to "unreasonable assumption of risk" and the second defining fault to include assumption of the risk.64
The merger costs little, because a plaintiff who is truly at fault in running a risk is already reached by the doctrines of no duty, comparative negligence, and sole proximate cause.65 Some conduct once labeled assumption of risk turns out, on analysis, to mean that the product was not defective or that the plaintiff's own act was a superseding cause.66 Where the defense survives only as apportionment, a deliberate choice may draw a larger percentage than vague carelessness would, but the plaintiff still recovers the remainder.67
Some jurisdictions keep the complete bar. Pennsylvania's Ferraro v. Ford Motor Co. applied it to a driver who kept driving a truck knowing that its wheels would lock on left turns and that its accelerator pedal tended to dislodge, and Reott v. Asia Trend, Inc. later explained that voluntary assumption of the risk requires proof that the user knew of a defect and voluntarily and unreasonably proceeded to use the product, whereas misuse is use that was unforeseeable or outrageous and does not depend on knowledge of any defect.68 Colorado, as applied by the Tenth Circuit in Tafoya v. Sears Roebuck & Co., and Mississippi, by statute, also treat conscious assumption as a complete defense.69 In such a state the plaintiff's conduct sorts into two kinds: a plaintiff who should have recognized the danger but did not, and so was only negligent, has damages reduced, while a plaintiff who recognized the danger and went ahead recovers nothing.70 Justice Mosk's dissent in Daly shows why some courts hold to the bar. He would have kept it as a total defense for a consumer who elects to use a patently defective product when other alternatives are available, and he predicted that affirmative defenses alleging plaintiff's fault would become boilerplate in every products case.71 The defense also has a practical use, since it can support summary judgment when the facts are clear that a plaintiff knowingly and unreasonably took a calculated risk.72
Knowing a specific danger differs from being able to see one, and that difference keeps assumption of risk from collapsing into the open-and-obvious rule taken up in a later subsection. If awareness of an obvious risk sufficed, the defense would bar every claim about a patent danger. Most courts reject that patent-danger rule, because a patent danger is not the same as a patent defect and a consumer does not take responsibility for a risk simply by knowing of it and using the product.73 Comment n also covers only what courts call secondary assumption of risk, in which the seller has already breached its duty by distributing a defective product and the plaintiff then chooses to use it anyway. Primary assumption of risk, which treats the inherent risks of an activity such as skiing as accepted by those who take part, rarely appears under that name in product cases; courts reach the same result by rejecting claims that rest on a whole category of product being dangerous.74
Misuse and Modification
Section 402A(1) of the Restatement (Second) of Torts makes a seller liable for a product “in a defective condition unreasonably dangerous” to the user, and subsection (1)(b) adds a condition that looks past the sale: the product must be expected to reach the user, and must actually reach the user, “without substantial change in the condition in which it is sold.”78 Those words make what happens to a product after the sale part of the seller’s liability. A later alteration changes the product the seller answers for, and an unusual use changes what the seller was bound to expect.
The tort roots of the doctrine explain why use matters. Liability for a defective article does not rest on contract.79 The wrong lies in delivering an article that lacks the qualities the manufacturer represented it as having, when the consumer could not readily detect the lack.80 And the measure of the wrong is whether the article is safe for the purposes for which the consumer would ordinarily use it.81 That measure makes the plaintiff’s actual use of the product, and any change someone has made to it, relevant, and foreseeability turns the measure into a working test. In MacPherson v. Buick Motor Co., 111 N.E. 1050 (N.Y. 1916), lack of privity was no defense if it was foreseeable that the product, negligently made, would injure a class of persons that included the plaintiff, which is essentially the test for negligence itself.82
Cases speak of a “misuse defense,” but misuse can affect a strict products claim in three distinct ways.83 An unforeseeable use can defeat the plaintiff’s proof of defect. A product that is defective for foreseeable risks can injure the plaintiff only through an unforeseeable misuse, which breaks causation. And a foreseeable misuse can leave the claim intact while counting against the plaintiff as fault. Foreseeability determines which applies, and only the last is an affirmative defense; the first two limit the plaintiff’s prima facie case.84
The first role removes the plaintiff’s proof of defect. The tort duty covers only foreseeable risks: a manufacturer cannot weigh, when it makes a safety decision, a risk it could not foresee, and the law gains nothing by holding it liable for harms outside the risks that made its conduct tortious.85 A use so unusual that no reasonable seller would have anticipated it therefore cannot make the product defective, however dangerous the product proves to be in that use.86 Section 402A and the Third Restatement draw the same line. Under Section 402A, a product safe for “normal handling” is not defective, and a seller is shielded from strict liability for injuries from “mishandling,” “over-consumption,” “excessive use,” and a failure to read and heed an adequate warning; the Third Restatement confines both design and warning defects to “the foreseeable risks of harm posed by the product” and bars recovery when a product is put to an unforeseeable use.87
In Venezia v. Miller Brewing Co., the plaintiff threw an empty beer bottle against a telephone pole, the bottle shattered, and he was hurt. The court held the brewer not liable, because beer bottles are not designed to survive that treatment.88 Such a use is “so unusual that the average consumer could not reasonably expect the product to be designed and manufactured to withstand it,” in the words of an Oregon decision, Findlay v. Copeland Lumber Co.89 The implied warranty of merchantability draws the same line. Its guarantee of fitness for ordinary use includes a guarantee that the product will withstand foreseeable misuse incident to its intended use, but the Venezia court refused to stretch the warranty to the deliberate smashing of a bottle in a manner unrelated to any normal or intended use.90
Foreseeability is elastic, however, and courts do not draw the line with the confidence the bottle case suggests. The decisions on foreseeability of misuse are “all over the map,” and a prudent judge will almost always treat the question as one of fact for the jury.91 A prudent lawyer hesitates to predict whether a jury will find that a seller should foresee a teenager scenting a candle by pouring cologne on it below the flame, or a buyer sitting in a chair too frail for her weight.92 Trivino v. Jamesway Corp. shows how readily the question reaches a jury. The trial court granted the manufacturer summary judgment, but the appellate court reversed, holding that it was a jury question whether the plaintiff’s use of the puffs was an unforeseeable misuse.93 It held the same of the manufacturer’s alternative argument that the danger of the puffs igniting was obvious.94
A user who pries open a paint can with a paring knife shows how a manufacturer should argue. Calling the user negligent is the weaker move, because in some jurisdictions a plaintiff’s negligence is no defense at all and in most it is only a partial defense under comparative negligence rules.95 The stronger move is to argue that the knife was never intended for that purpose. To recover, the user must prove that the knife was defective when used for the purpose for which it was intended, or at least for one the manufacturer should have anticipated. If the use was an unforeseeable misuse, which it probably was, the knife is not defective, even though it was unfit for that use.96
The second role works through causation and applies even when the product is defective. A product may be defective with respect to a foreseeable risk of misuse while the plaintiff’s injury comes from a different misuse that no one could foresee; the defect then did not proximately cause the injury.97 The seller answers for harms within the risks that made its conduct tortious, not for a freakish one.98 Daniell v. Ford Motor Co. illustrates the limit. The plaintiff locked herself in a car trunk to attempt suicide, changed her mind, and could not escape because the trunk had no inside release. The court assumed that the car may have been defective for lack of a release but held Ford not liable for so unforeseeable an occurrence, because the bizarre misuse barred recovery.99 The case differs from the beer-bottle case in where the claim fails: the bottle plaintiff could not prove a defect for his use at all, while the trunk plaintiff might prove a defect and still lose because it did not cause her injury.
The third role arises when the misuse is foreseeable. Foreseeable misuse leaves the plaintiff’s claim intact and puts her conduct on the scale as fault, which the defendant raises as an affirmative defense.100 The effect depends on the jurisdiction. Some ignore a plaintiff’s negligence in strict products cases, so that stepping on a shelf would not affect her claim at all.101 A clear majority now treat negligent misuse as comparative fault: the jury assigns the plaintiff a percentage for misusing the shelf as a step, and her recovery shrinks by that percentage.102
Foreseeable misuse cannot bar recovery outright without emptying the duty it triggers. Suppose tires are defectively designed to explode when a car is driven five miles an hour over the speed limit. Speeding is negligent per se, but if that negligence barred recovery, the manufacturer would never be liable for the exploding tire. To avoid that result, the Restatement (Second) refused to recognize this form of contributory negligence.103 Comparative fault answers a different concern, which is cost. A damages award raises the price of the product for all consumers, including those who never misuse it, so a major policy reason courts give for comparative responsibility is that making the victim’s negligence irrelevant is unfair, because it makes careful users bear the costs created by careless ones.104
Not everyone welcomed the change. One critic of admitting plaintiff negligence into strict liability predicted that every defendant would assert that the plaintiff did something, anything, that conceivably could be called contributory negligence: driving 56 miles an hour instead of 54, failing to discover a latent defect, or yielding to the salesman’s persuasion.105 Pleading such defenses, the critic said, would become boilerplate.106
Knowing misuse sits closest to assumption of the risk, covered in the previous subsection. A plaintiff who runs a table saw without its guard, knowing it should be there, has made the deliberate choice that assumption of the risk describes, and the choice is plaintiff misconduct that reduces or bars recovery.107 Ignoring a warning is a special form of misuse. It ordinarily should bar recovery when the danger of ignoring the warning or instruction is evident, the disregard is a substantial cause of the harm, and the manufacturer had no simple means or apparent reason to design the danger out of the product.108 Many courts, and statutes in Arizona, North Carolina, and New Jersey, bar recovery for injuries from ignoring warnings and instructions.109 The bar assumes a full and fair warning. If the disregarded warning was itself inadequate, the user’s failure to follow it is foreseeable and excusable, and misuse that an adequate warning would likely have prevented is no defense. Because warnings may be disregarded, modern courts also widely hold that manufacturers have an independent duty to design away dangers when a reasonable way exists.110
A third party’s alteration of the product sorts the same way, and the without-substantial-change requirement of Section 402A(1)(b) anchors it. Suppose an employer strips the safety devices from a machine after buying it. Whether that change relieves the manufacturer depends on foreseeability, which courts use as the test for whether an intervening act supersedes: almost every case turns principally or exclusively on that single question.111 A change the manufacturer should have foreseen does not break the chain, and the defect and causation questions stay open; a change it could not have foreseen does, and the product did not reach the user in the condition in which it was sold.
The difference between the two presses is foreseeability, and foreseeability tracks what a manufacturer can do. It can design against a change it should expect; no precaution reaches a change it had no reason to anticipate, and liability for the unforeseeable would not induce conduct anyone could perform.112
The three roles end in different places, so the label an answer gives the facts decides the result it reaches. An answer that calls a foreseeable use of a shelf a complete defense ends the case too early, and an answer that treats a suicide attempt in a trunk as a mere reduction of damages fails to stop a claim that proximate cause defeats.
Open and Obvious Danger
Suppose a cook slices a finger on an Accu-Cut knife and sues Accu-Cut for failing to warn that knives cut. The claim fails, because everyone already understands that danger.115 Yet “open and obvious” is a loosely used phrase. The same visible danger can end one kind of claim, count as a single factor in another, and carry no weight in a third.116 Section 402A(1) makes a seller liable for physical harm from a product sold “in a defective condition unreasonably dangerous to the user or consumer,” and that opening sentence says nothing about obviousness.117 Obviousness therefore matters only through the questions whether the product was defective and whether it needed a warning, and each kind of defect claim weighs it differently.
In a warning claim, obviousness can be a complete answer. A warning exists to tell users what they do not already know, so a manufacturer has no duty to warn of dangers people already understand. When the danger is both obvious and impossible to eliminate from the product, as a knife’s edge is, obviousness defeats the plaintiff’s prima facie case for failure to warn.118 The reverse holds for a hidden danger. A user who does not understand the risk may recover if the manufacturer failed to warn of it.119
Design claims once met the opposite rule. Before the 1970s, victims of accidents caused by obvious design dangers were broadly barred from recovery under the “patent-danger doctrine.”120 The classic case is Campo v. Scofield. A man working on his son’s farm was feeding onions into an “onion topping” machine when his hands were caught in its revolving steel rollers. He sued the manufacturers for negligently failing to equip the machine with a guard or stopping device.121 The New York Court of Appeals held that a manufacturer satisfies the law if the machine has no latent defect and creates no peril unknown to the user. “We have not yet reached the state where a manufacturer is under the duty of making a machine accident proof or foolproof,” the court said, and the manufacturer “is under no duty to guard against injury from a patent peril or from a source manifestly dangerous.”122
The rule’s simple statement of individual responsibility for self-protection struck a responsive chord. After Campo, many courts adopted it or reaffirmed it, as Bartkewich v. Billinger did, and by the 1950s and early 1960s it was one of the most firmly entrenched doctrines in products liability. Even two decades after Campo, the Maryland Court of Appeals aligned itself with it.123 The rule’s logic is easy to trace to older formulations of manufacturer liability, which were framed around qualities whose absence “could not be readily detected by the consumer.” A danger the consumer could see sat outside that frame.124 Fleming James derided the no-duty rule as “a vestigial carryover from pre-MacPherson days when deceit was needed for recovery.”125
Courts abandoned the rule because of what it rewarded. A manufacturer whose machine carried a plainly exposed hazard was safer from suit than one whose hazard was hidden, so the more obvious the misdesign, the better the defense. In Palmer v. Massey-Ferguson, Inc., a Washington appellate court called the rule “somewhat anomalous.” The maker of an obviously defective product “ought not to escape because the product was obviously a bad one,” the court wrote, and the law “ought to discourage misdesign rather than encouraging it in its obvious form.”126
The rule also assumed that seeing a risk means accepting it. Consumers are often ignorant of latent dangers, and even for apparent risks they may not fully understand the magnitude of the risk or that it can be eliminated at reasonable cost. Lacking that information, they do not adequately value safety improvements, which gives manufacturers an incentive to forgo reasonable investments.127 A buyer can see that a blade is exposed without any idea that a guard would cost almost nothing.
Pike v. Frank G. Hough Co. supplied the working alternative. If the product is a carrot-topping machine with exposed moving parts, or an electric clothes wringer dangerous to the operator’s limbs, and a guard or safety release would be feasible, whether reasonable care demanded the precaution is a jury question even though its absence is obvious. The court reversed a nonsuit and concluded that “the obviousness of peril is relevant to the manufacturer’s defenses, not to the issue of duty.”128
Micallef v. Miehle Co. made the same move in New York. The court held that a manufacturer must design its products with reasonable care to avoid unreasonable risks of harm to persons exposed to foreseeable dangers, based on “a balancing of the likelihood of harm, and the gravity of harm if it happens, against the burden of the precaution which would be effective to avoid the harm.”129 Obviousness often reduces the likelihood of accidents, so it enters that calculus, and it may also bear on contributory negligence and assumption of risk. But “the patent-danger doctrine should not, in and of itself, prevent a plaintiff from establishing his case.”130 The doctrine has since lost its remaining strongholds, Georgia and Illinois. Only Missouri keeps it, and only for negligent design claims.131
The modern placement follows from the mechanism. A visible hazard lowers the probability of an accident, and that belongs on the risk side of the scale. It does not change the burden of the precaution on the other side. An inexpensive guard against a grave injury can therefore still tip the balance toward the plaintiff, however plain the danger.132
Two limits keep that statement from being universal. The first is the consumer expectations test. In a jurisdiction that measures design defect by that test alone, obviousness tends to be fatal. In a pool case, the plaintiff claimed that the design was defective because the manufacturer omitted a self-closing, self-latching gate at the top of the ladder to keep small children out. The trial court dismissed, and the appellate court affirmed. It reasoned that an average consumer’s safety expectations were plainly met, because such a person could not help but comprehend the danger from the obvious absence of a gate.133 Risk-utility analysis asks something else: whether the foreseeable risks could have been reduced or avoided by a reasonable alternative design whose omission renders the product not reasonably safe. The visibility of the missing gate does not answer that question.134
The second limit is the inseparable danger. A knife must be sharp to do its job, so the risk of cuts is not the result of a design flaw. It is not an unreasonable risk, and it cannot be eliminated without undermining the knife’s utility. The knife is therefore not defectively designed, despite the danger of cuts.135 That failure belongs to the design claim on its own terms, and obviousness is what defeats the companion warning claim. It is no accident that obviousness enters design and warning claims as an ingredient of reasonable care. The Third Restatement grounds design and warning defects explicitly in negligence.136
The hazard is visible in both the knife and the snow blower, so visibility cannot be what separates the two outcomes. What separates them is whether the danger can be removed at reasonable cost without destroying the product’s function. If the snow blower’s manufacturer could answer that its design was dangerous but obviously so, the duty to design reasonably safe products would weaken as designs grew more blatantly dangerous, and the more egregious the design, the less liability the manufacturer would risk. When a danger can be eliminated at reasonable cost, it should be, whether it is obvious or not.137
Obviousness cannot excuse a manufacturing defect at all. A unit that departs from the maker’s own design is defective although the seller exercised all possible care,138 and the Third Restatement continues to impose liability for manufacturing defects in terms that are truly strict.139 Whether a plaintiff noticed a visible flaw and used the product anyway is a question for the conduct defenses discussed earlier in this section, such as contributory negligence and assumption of risk, and the defect element does not reach it.140
State of the Art
Section 402A makes a seller liable for physical harm caused by a product sold in a defective condition unreasonably dangerous to the user or consumer,145 and subsection (2)(a) applies that liability even though the seller has exercised all possible care in preparing and selling the product.146 The clause seems to close the door on the argument that a seller could not have known of a danger. Yet that argument sits near the center of modern products litigation, and a drug maker shows how it arises. The maker tests a drug reasonably, has no reason to expect an allergic reaction, and markets it. After widespread use, a small percentage of patients turn out to react. The injured patient sues in strict liability, and the maker answers that, given the state of the art (the state of scientific knowledge about the drug’s risks), it had no reason to anticipate the injury.147 In negligence that answer would win. Under strict liability, whether it wins is the whole question, and the label “strict” does not settle it. Even for abnormally dangerous activities, the Second Restatement dropped the First Restatement’s fixed elements in favor of factors, such as the level of risk the activity creates, that make strict liability more or less likely.148
The traditional premise of products liability points toward no. Liability does not rest on contract. It rests on the principle that delivering an article is itself wrong when the article lacks the qualities the manufacturer represented, the lack could not readily be detected by the consumer, and the article is therefore unsafe for ordinary use.149 That principle asks about the article and not about the seller’s foresight, so the seller’s ignorance of a danger appears beside the point.150 The inference explains why courts divided so sharply during the formative period of strict products liability, from the 1960s to the 1980s. The dispute was whether a product is defective if it posed dangers that were unknown when it was sold and could not reasonably have been discovered through investigation and testing.151
Some decisions followed that reasoning to its end. Beshada v. Johns-Manville Products Corp., an asbestos case, is one of the few that refused to let a manufacturer rely on the state of the art.152 The Wade-Keeton test likewise pressed truly strict liability on manufacturers for dangers in design, until both of the deans whose names it bore repudiated it in the early 1980s.153 Almost every court has since gone the other way. Decisions such as Feldman v. Lederle Laboratories, Brown v. Superior Court, and Anderson v. Owens-Corning Fiberglas Corp. confine liability to what a seller knew or could have learned. Today almost all courts allow state-of-the-art evidence to defeat strict liability claims based on unknowable risks. Put differently, they refuse to impose a duty to warn of risks no one could know, and so they decline to apply a standard that is truly “strict.”154
The reasons for the retreat are practical. Courts have given thin support to hindsight in warning cases because the goal of the law is to induce conduct that can actually be performed, and liability for failing to warn of risks no one could know does not advance that goal.155 One state high court, revising its implied-warranty law in line with the Third Restatement, held that a defendant is not liable for failing to warn or instruct about risks that were not reasonably foreseeable at the time of sale or could not have been discovered by reasonable testing before marketing. The same court held the manufacturer to the knowledge of an expert in the field and left it under a continuing duty to warn of risks discovered after the sale.156
The defense is not a license to stay ignorant. The seller is charged with knowledge of what reasonable testing would have revealed, so the defense protects only the manufacturer that looked.157 The charge answers an evidentiary gap. A plaintiff who must prove that a risk was reasonably knowable has to establish the parameters of a reasonable research program and then show that it would have found this risk, and the proof is usually extraordinarily demanding.158 Left alone, the gap can immunize the manufacturer that conducts no research and weaken its incentive to test, because manufacturers usually benefit by remaining ignorant.159
Genuinely unknowable risks are also uncommon. Makers of durable products can almost always foresee the kinds of harm their products may cause: a punch press may crush a hand, a defective appliance may start a fire, a car may crush inward in an accident. Claims of unknowable risk are far more common for toxic substances such as drugs, chemicals, asbestos, and tobacco, and even there proper testing and lab analysis make truly unforeseeable dangers very unusual.160
The phrase “state of the art” is itself a source of trouble, because its definitions vary considerably. Some lean toward industry custom, others toward theoretical technological capability, and the best lie somewhere between.161 Some courts and legislatures use the term without defining it, which leaves decisionmakers with little guidance.162 Neither extreme works. Almost all courts properly distinguish state-of-the-art evidence from evidence of custom, since, as Judge Learned Hand observed in a different context, “a whole calling may have unduly lagged in the adoption of new and available devices.”163 Defining the term as the best science and technology in existence goes too far the other way. It would require manufacturers to put new safety theories into practice at once, though the theories may sit in arcane journals, perhaps in another language, years from real-world application.164
Most statutes and decisions settle on a middle ground. Nebraska defines state of the art as “the best technology reasonably available at the time.” Missouri says that, as a defense, the phrase “means that the dangerous nature of the product was not known and could not reasonably be discovered at the time the product was placed into the stream of commerce.” Arizona’s statute speaks of the technical, mechanical, and scientific knowledge of manufacturing, designing, testing, or labeling the same or similar products that was in existence and reasonably feasible for use at the time of manufacture.165 An airbag example shows the difference. A maker who says that no workable airbag could be built with the technology available when the car was sold is making a state-of-the-art argument. A maker who says only that no competitor offered airbags yet is describing custom. The two statements differ in kind: “no manufacturer would have been aware of the need for this, or been able to provide it” is not the same as “no manufacturer was doing it at the time I sold the product.”166
The argument works differently in each kind of defect claim. In a warning claim it goes to foreseeability. Unforeseeable risks, by definition, cannot specifically be warned against, and the trend is that a manufacturer has no duty to warn of risks it could not have discovered by reasonable care and pre-sale investigation.167 The drug maker above wins for that reason. Had reasonable pre-market testing revealed the reaction, the same patient would win, because the seller is charged with what testing would have shown.
In a design claim the argument goes to feasibility. The manufacturer is not liable for omitting a safety feature if, given the technology available when it sold the product, it would not have been aware of the need for the feature or had the technical means to provide it.168 The Third Restatement builds the same idea into its definition. A product is defective in design when the foreseeable risks of harm could have been reduced or avoided by a reasonable alternative design whose omission renders the product not reasonably safe, so the design obligation is limited to foreseeable risks and grounded on reasonableness.169 Unforeseeable risks are rare in design cases except for highly dangerous and unusual uses. Having acknowledged the definitional problem, the Third Restatement sidesteps it by dropping the term “state of the art” altogether, while keeping the underlying proposition, now widely accepted, that manufacturers answer only for what they can foresee and reasonably prevent.170
In a manufacturing defect claim the argument has no place. Courts have kept strict liability for manufacturing defects, perhaps because, unlike a design or warning defect, the flaw affects only one unit and not the entire product line.171 The gist of the state-of-the-art defense is that a manufacturer cannot be faulted if its design is reasonably safe in light of the technology available when it designed and sold the product. Importing that defense into a manufacturing defect case misapplies it.172 Strict liability for production flaws also supplies an incentive that a fault standard cannot. Courts adopted it for manufacturing defects to give sellers an adequate financial incentive to adopt reasonably safe systems of quality control, since a plaintiff can rarely prove what a reasonable system would have caught.173
Who carries the burden depends on how the jurisdiction characterizes the argument. It may be treated as a defense, or it may be treated as part of the plaintiff’s prima facie case, on the view that a manufacturer commits no tort by selling a product with an unknowable risk. If it is part of the prima facie case, the plaintiff must show that the manufacturer should have been aware of or discovered the danger before the sale. The states are mixed on the point.174 Missouri’s statute, for one, speaks of the term “as a defense.”175 The Third Restatement writes foreseeability of the risk and a reasonable alternative design into the definition of defect, so under it those issues fall within the plaintiff’s proof.176
A few courts still resist. Sternhagen v. Dow Co. is among the decisions that decline to let state-of-the-art evidence defeat a strict liability claim, and these outliers are exceptions to a rule that almost all courts now follow.177 The Wade-Keeton test also lingers in places despite its repudiation. Some courts adopted it after its “official” demise, New York for one, and others, Arizona for one, restate it. Oregon’s legislature reversed its judicial adoption, while Tennessee’s legislature affirmatively adopted it.178
Every version of the argument is anchored to the time of sale. A risk discovered later does not make the product defective as sold. A seller that learns afterward of a substantial risk may still owe a post-sale warning, but only if those at risk can be identified and can reasonably be assumed to be unaware of it, a warning can be effectively communicated and acted on, and the risk is great enough to justify the burden of warning.179 Where the product was already defective at the time of sale, liability can be established without reference to any post-sale duty to warn, and a post-sale warning does not generally absolve the seller.180 A later safety improvement is treated similarly. If a new safety device is developed but the old design was not defective when sold, courts generally impose no automatic duty to recall and retrofit.181
Unavoidably Unsafe Products
Section 402A imposes liability for physical harm caused by a product sold “in a defective condition unreasonably dangerous to the user or consumer.”184 Comment k exempts an entire class of products from strict liability: those whose danger no seller can remove.185 The comment describes products that, “in the present state of human knowledge, are quite incapable of being made safe for their intended and ordinary use,” and it calls them especially common among drugs. Its example is the vaccine for the Pasteur treatment of rabies, which not uncommonly leads to very serious and damaging consequences when injected. Because rabies itself invariably leads to a dreadful death, both the marketing and the use of the vaccine are fully justified despite the unavoidable high degree of risk.186
A product of this kind, “properly prepared, and accompanied by proper directions and warning, is not defective, nor is it unreasonably dangerous.”187 Vitanza v. Upjohn Co. quotes the rest of the comment, which extends the point to “drugs, vaccines, and the like,” many of which cannot legally be sold except to physicians or on a physician’s prescription. A seller of such a product, the comment concludes, is not held to strict liability for unfortunate consequences of its use “merely because he has undertaken to supply the public with an apparently useful and desirable product, attended with a known but apparently reasonable risk.”188 The ordinary §402A case looks quite different: a manufacturer, a wholesaler, and a retailer each answer to the injured consumer, and privity is not required.189
This differs from the state-of-the-art defense, which concerns dangers no one could know when the product was sold. Under comment k the risk is known, and the sale is justified by the product’s benefit. Two features of drugs explain why the law treats that sale differently. The first is that the danger usually cannot be designed out. The chemical properties that can cause great harm are usually the same properties that make a drug therapeutic, so a change that avoided an adverse effect would usually reduce or eliminate the drug’s benefit; in general, the “design” of pharmaceuticals cannot be changed.190 Liability for such a risk therefore usually leaves the maker no safer formula to adopt. The second feature is the supply of products the public needs. Strict liability ordinarily gives sellers an incentive to adopt quality-control measures, but the bankruptcy of blood suppliers and other pharmaceutical manufacturers would create social problems different in kind from the bankruptcy of automobile or soda makers, because blood and other pharmaceutical products are necessary for public health and safety. For such makers, strict liability can have a pernicious safety effect.191
The comment sets two conditions, and each does work. “Properly prepared” keeps production flaws outside the exemption. Because the comment says that a properly prepared and warned product of this kind is not defective, such a product could be defective only in design, and most courts and commentators therefore hold that comment k does not apply to manufacturing defects.192 The Third Restatement agrees: makers of prescription drugs and medical devices face true strict liability for manufacturing defects, just as sellers of other products do.193 A manufacturing defect can make a product malfunction, as in an exploding soda bottle, and a contaminated lot of vaccine departs in the same way from what its maker meant to sell.194 Liability in such cases does not rest on contract; it rests on the wrong of delivering an article that, because it lacks qualities the manufacturer represented and the consumer could not readily detect, is unsafe for its ordinary use.195 A later warning does not cure the flaw either, since a seller that learns after the sale that its product was defective when sold generally cannot absolve itself by issuing a post-sale warning.196
“Proper directions and warning” is the second condition, and it matters most once design is off the table. Sellers of inherently dangerous products, in addition to avoiding impurities and misrepresentations of safety, need only warn consumers of the unavoidable, latent dangers the products intrinsically contain; because such a product by hypothesis cannot be designed more safely, adequate warnings of danger and instructions on safe use are the seller’s principal remaining duty.197 The Restatement’s commentary on warnings adds that, for poisonous drugs or drugs unduly dangerous for other reasons, a warning as to use may be required.198 For prescription products, the duty to provide danger information runs to healthcare professionals, not directly to consumers.199 Under the learned intermediary rule that the vast majority of states follow, the maker satisfies the duty by disclosing to the prescribing physician all the foreseeable risks and therapeutic benefits material to the prescribing decision, which lets the physician make an individualized medical judgment about whether the patient should use the product.200 If the maker gave the physician that information, it owes the consumer no warning directly; if it did not, it is liable for failure to warn.201
A problem about a lifesaving drug with a serious side effect tempts a reader to stop at the benefit, but the benefit is only the starting point. The comment protects a product that is unavoidably unsafe, properly prepared, and properly warned, and an exam stem usually varies one of the last two.
The same reasoning reaches products that are not drugs. The Third Restatement holds that entire categories of products commonly understood to be inherently dangerous, such as alcoholic beverages, firearms, and above-ground swimming pools, but not cigarettes, cannot be judicially classified as defective in design.202 Its design test asks whether the foreseeable risks could have been reduced or avoided by a reasonable alternative design, a test it adopted in rejecting the consumer expectations test to rein in what was perceived as excessive liability under Barker.203 A plaintiff cannot meet that test when the product’s risks are, by hypothesis, unavoidable.204 Vodka shows the result. Consumers understand its dangers and accept them as inseparable from the product, so it passes the consumer expectations test, and it passes the risk-utility test because it cannot be redesigned to be completely benign and still be vodka.205 One commentator reads comment j’s closing sentence along the same lines: sellers of a very small category of inherently dangerous products, among them food, alcohol, tobacco, and drugs, have more limited safety obligations than sellers of ordinary products.206
Because comment k exempts a class of products, a court must decide who shows that a particular drug belongs in it. The majority rule treats the comment as an affirmative defense, so the maker bears the burden of showing that the design of its drug or device passes the risk-utility test.207 That allocation runs against the usual one. For other products the plaintiff in the vast majority of jurisdictions must prove that the design fails the test, so a defense meant to benefit manufacturers imposes on pharmaceutical makers an evidentiary burden heavier than the one other makers carry.208
The Third Restatement gives prescription drugs and medical devices their own rule in §6. Courts have tended to treat makers of these products differently because the FDA scrutinizes and regulates the design of prescription drugs and devices and the warnings that accompany them. Manufacturing defects stay under true strict liability, while the design test turns on the judgment of health-care providers: a drug or device is not reasonably safe in design if its foreseeable risks are sufficiently great in relation to its foreseeable therapeutic benefits that reasonable health-care providers, knowing of both, would not prescribe it for any class of patients.209 That standard leaves so small a window for design claims that almost no drug claim could fit through it. Even thalidomide would not qualify, because of its value in treating leprosy, and one commentator defends that result: people with leprosy should not lose a beneficial therapy because some doctors improperly give the drug to child-bearing women.210 A problem on this topic therefore calls for stating which approach the jurisdiction follows before applying it.
Time Bars
A plaintiff can prove a defect, causation, and injury and still lose because the suit came too late. Section 402A(1) of the Restatement (Second) of Torts makes a seller liable for "physical harm thereby caused" by a product in a defective condition unreasonably dangerous to the user,214 but it says nothing about how long the injured person has to sue. State statutes answer that question, and statutes of limitation and repose were among the principal subjects of products liability reform.215 Up to three clocks can therefore run against a plaintiff: a tort limitations period, a warranty limitations period, and a statute of repose. Each starts at a different event, so each must be checked on its own.
The tort clock is the short one. A claim in strict liability or negligence is a tort claim, and tort limitations periods typically run two or three years from discovery.216 Discovery matters because section 402A liability attaches only to physical harm, so a plaintiff has no claim to sue on until the harm occurs, and may not learn of the claim until later still. The period can also be tolled. If a defendant fraudulently prevents the plaintiff from learning of the claim, a court may apply equitable tolling to delay the statute until the plaintiff does, or should, discover it.217 The defendant must have acted affirmatively to hide facts necessary to the plaintiff's claim and thereby prevented the plaintiff from discovering it.218
The warranty clock comes from the sales contract. Article 2 requires an action for breach of a sales contract to be commenced within four years after the cause of action accrues,219 and in the products setting that usually means four years from sale.220 One formulation of the strict liability rule says that it "does not rest upon contractual obligations," but on the principle that delivering an article is wrong when it lacks the qualities the manufacturer represented and the lack could not readily be detected.221 A rule with that tort foundation is not naturally governed by the Code's contract period. Four years from sale is normally longer than tort's two or three years from discovery, which is one reason plaintiffs plead warranty.222 A late injury reverses that advantage. Measured from sale, the warranty clock may expire before the buyer is hurt, while the tort clock, measured from discovery, has not yet started.
The gap between the clocks explains why plaintiffs often plead more than one theory. There is often little point in pleading negligence when strict liability applies, but in some states statutes of repose or limitations or other defenses might defeat one claim and not the other.223
A statute of repose works from a different starting point. Where a state has one, the period typically counts from an event in the product's history, such as its first sale, and ignores when the plaintiff was injured or discovered the injury. A claim can therefore expire before the plaintiff is ever hurt, and a plaintiff who sues promptly and proves a real defect can still be barred. Beginning in the 1970s, a number of states enacted statutory schemes to decrease the size and number of recoveries against manufacturers and sellers, many in response to dramatic increases in products liability insurance rates, and statutes of limitation and repose were among the principal changes.224 The injured plaintiff bears the cost of that finality.
Time bars also differ in method from the merits rules. For abnormally dangerous activities, the Second Restatement dropped fixed elements and told courts to weigh factors, such as whether the activity creates a high risk.225 A time bar involves no weighing. The court compares dates, and a strong case on the merits does not change the arithmetic. Limitations asks how quickly the plaintiff acted once the claim was or should have been known; repose asks only how old the product was. When an exam stresses a proven defect and a plaintiff who sued at once, check the product's age before reaching the merits.
Preemption and Compliance
Section 402A makes a seller liable for physical harm caused by a product sold in a defective condition unreasonably dangerous to the user or consumer, and it applies that rule even though the seller exercised all possible care in preparing and selling the product.230 A manufacturer that built to a government safety standard has taken care that an outside agency can verify, so the obvious question is whether that compliance answers the claim. At common law it does not: compliance is one piece of evidence for the jury to weigh. Federal preemption is the exception, and it operates at a different level, because it decides whether a state claim may go forward at all. This subsection takes the common-law default first and then preemption.
At common law, compliance with a safety statute or regulation is evidence that a product was not defective, and nothing more. A jury normally remains free to find a complying design or warning defective, because governmental safety requirements are generally set at levels minimally acceptable for regulatory purposes.231 Section 4(b) of the Restatement (Third) of Torts: Products Liability states the rule in two halves. Compliance "is properly considered in determining whether the product is defective with respect to the risks sought to be reduced by the statute or regulation," but it "does not preclude as a matter of law a finding of product defect."232 The first half confines the evidence to the risks the rule addresses, and the second leaves the verdict with the jury. Treating compliance as one consideration among several matches how strict liability is built elsewhere: for abnormally dangerous activities, the Second Restatement dropped fixed elements in favor of factors to be considered, such as whether the activity creates a high risk.233
Compliance cannot settle the matter because the common-law claim asks a different question from the regulator's. One formulation of the strict liability rule locates the wrong in delivering an article that is not safe for the purposes for which the consumer would ordinarily use it, because it lacks qualities the manufacturer represented it as having and the consumer could not readily detect their absence.234 Under California's version, a product is defective in design if it fails to perform as safely as an ordinary consumer would expect when used in an intended or reasonably foreseeable manner. That test grows out of warranty and out of the idea in Greenman that putting a product on the market represents that it will safely do the jobs for which it was built.235 Neither formulation asks whether the maker satisfied a regulator, so a regulation's minimum informs the inquiry without replacing it.
Vehicle cases illustrate the point. A car may be defective and unreasonably dangerous because it is insufficiently crashworthy,236 and in Volkswagen of America, Inc. v. Young the court observed that the most significant federal vehicle-safety legislation shows Congress did not view safe motor vehicle design as solely a legislative problem. The National Traffic and Motor Vehicle Safety Act of 1966 says so directly: "Compliance with any Federal motor vehicle safety standard issued under this subchapter does not exempt any person from any liability under common law."237
Compliance can still count for a great deal. In very unusual situations, where no special circumstances suggest a need for greater caution, a court may rule that conformity to a statutory or regulatory safety standard amounts to due care.238 Wilson v. Piper Aircraft Corp. illustrates the exception in a closely regulated field: the Oregon court gave weight to the Federal Aviation Administration's approval of the very design the plaintiffs attacked when it asked whether their evidence of defect could reach the jury. Some states go further than the Restatement and provide that compliance with an applicable statute or regulation raises a rebuttable presumption that the product is not defective.239 Noncompliance cuts the other way, though less sharply than one might expect. The doctrine of "defectiveness per se" is undeveloped, and the cases suggest only that noncompliance with an applicable safety statute or regulation may be considered evidence, though not dispositive, that a product was defective.240
Preemption changes the question from how much compliance weighs to whether the claim may proceed. When federal law displaces state law on a subject, the state tort claim on that subject fails however a jury might have weighed the evidence. Which claims federal law displaces depends on the particular statute, and the answer can differ from claim to claim within a single lawsuit. The examples below show that pattern and do not state a rule for all products.
The drug case Bergstresser shows the claim-by-claim split. The plaintiff sued over the drug Abilify, and the courts held that Pennsylvania law preempted the claims because the Food and Drug Administration had approved the drug's design, warning label, and leaflet. Liability for a manufacturing defect, however, was not preempted, as it almost never is.241 The line follows from what each claim asks. A design or warning claim asks whether the approved design or label should have been different, which is the very judgment the agency made, and under the Third Restatement those claims rest largely on negligence principles.242 A manufacturing defect is a departure from the intended design even though all possible care was exercised.243 Such a defect also affects only one unit and not the whole product line,244 so a manufacturing claim leaves the approved design alone and holds the maker to it. The plaintiff's remaining route was to prove that the pills Bergstresser took contained more active ingredient than they should have. That is a slim reed, but it is one preemption leaves standing.245
The Federal Insecticide, Fungicide and Rodenticide Act raises the same question under a different statute. It authorizes the Environmental Protection Agency to regulate pesticide sale and use and makes it unlawful to use a registered pesticide in a manner inconsistent with its labeling. That "label use" provision limits the ability of injured parties to sue pesticide manufacturers on inadequate-labeling theories, and it has raised a significant question: whether the Act preempts state damage claims for crop injury, and whether its preemption reaches only the subjects EPA reviews when it first approves a label.246 In boating litigation, federal courts held state products liability claims preempted by the Federal Boat Safety Act in Lewis v. Brunswick Corp., and maritime decisions more generally show that courts do not always agree.247 Each result rests on its own statute, so a problem that mentions a federal regulation calls for one question: what does this statute displace? An assumption that federal law always wins does not answer it.
Both makers did the same thing: they followed the government's label rule. The difference lies in whether the rule came from a government whose law displaces the state tort standard, and that fact decides whether a jury ever hears the claim.
Compliance with government specifications can also defeat a claim through the government contractor defense. In Boyle v. United Technologies, the Supreme Court recognized a defense that shields a contractor from liability to third parties for design defects when the government approved reasonably precise specifications, the product conformed to them, and the contractor warned the government of dangers that the contractor knew about and the government did not. The defense has limits. If the government merely ratifies a privately designed product, as when it selects among competing designs, Boyle does not apply, and the designer remains responsible for defects. In failure-to-warn cases, if the government itself discovers unavoidable design risks and warns its service members of them, the manufacturer avoids liability because another gave the warning.248 The first limit separates a design the government chose from one it merely accepted.