Concept
Defective Condition
Claims against manufacturers and other defendants arising out of the manufacture and distribution of products, and defenses to such claims
Defective Condition
Section 402A(1) of the Restatement (Second) of Torts reaches only the seller of a product “in a defective condition unreasonably dangerous” to the user or consumer or to the user's property.1 The earlier sections of this chapter identified the seller, the product, the harm, and the timing. This section takes up the element that decides most products cases: whether the product was defective at all. Because § 402A(2)(a) applies the rule even when the seller has exercised all possible care, the plaintiff must locate the fault in the product and not in the seller's conduct.2
Danger alone supplies no defect. Nearly every product is dangerous when put to certain uses, and most of those dangers are facts of physics, chemistry, or biology that no one can reasonably avoid. Users bear those risks and insure against them. The law shifts a loss to the seller only when a product carries excessive risk, either because consumers do not expect it or because the supplier can avoid it more reasonably than they can.3 A manufacturer is not liable merely because its product involved some risk of injury. Otherwise automakers would pay for every motor vehicle injury and knife makers for every cut finger.4
Comment i to § 402A gives “unreasonably dangerous” its meaning. The article must be dangerous beyond what the ordinary consumer, with the ordinary knowledge of the community about its characteristics, would contemplate. Good whiskey is not unreasonably dangerous merely because it makes some people drunk and is especially dangerous to alcoholics, but bad whiskey containing a dangerous amount of fusel oil is.5
Courts disagreed over whether the two words state one requirement or two. In Cronin v. J.B.E. Olson Corp., the California Supreme Court reasoned that making a plaintiff also prove “unreasonably dangerous” “places upon him a significantly increased burden” and leaves the formula with a negligence complexion, so it banished the phrase from California's strict liability law.6 Most courts kept § 402A's wording despite its awkwardness. They came to treat “defective” as shorthand for a legal conclusion, that the product's excessive dangers make the seller liable, and not as a test for reaching it.7 The qualifier was meant to insist that a defect is required and that risky products are not necessarily defective. The Third Restatement now drops it and uses only “defective.”8
A court that calls a product defective needs a standard for reaching that conclusion, and the right standard depends on how the product went wrong. When Dean Prosser drafted § 402A, products liability was young and the defect notion was unitary, a product being either too dangerous or safe enough, because the cases then before the courts mostly involved production flaws.9 As liability spread to designs and to missing warnings, the differences among those claims became too large to ignore. The American Law Institute's Restatement (Third) of Torts: Products Liability, drafted in the 1990s by Professors Twerski and Henderson, responded by sorting defects into three kinds.10 Section 2 provides that a product is defective when, at the time of sale or distribution, it contains a manufacturing defect, is defective in design, or is defective because of inadequate instructions or warnings, and it then says what each means:11
- the product contains a manufacturing defect when it departs from its intended design even though all possible care was exercised in preparing and marketing it;
- it is defective in design when the foreseeable risks of harm could have been reduced or avoided by a reasonable alternative design, and omitting that design renders the product not reasonably safe; and
- it is defective because of inadequate instructions or warnings when the foreseeable risks of harm could have been reduced or avoided by reasonable instructions or warnings, and omitting them renders the product not reasonably safe.
These three categories, together with a separate claim for misrepresentation, now organize the field and have displaced the older labels of negligence, warranty, and strict liability as the bases of recovery.12 Misrepresentation under § 402B reaches a seller who misstates a material fact about a chattel's character or quality, even though the statement was neither fraudulent nor negligent and the consumer never contracted with the seller.13 Because the three defect categories carry different liability standards, the first step in any products problem is to identify which one the plaintiff alleges.14 Manufacturing defects come first below. Liability for them is strict in the full sense, and the standards for the other two are easier to follow once that baseline is clear.15
Manufacturing Defects
A product has a manufacturing defect when the particular unit departs from its intended design, even though all possible care was exercised in preparing and marketing it.16 Section 402A(2)(a) points the same way, since it applies the strict rule although the seller exercised all possible care.17 The inquiry compares the unit that injured the plaintiff with what the manufacturer meant to make. It does not ask how the flaw arose or whether anyone could have found it before sale. That is why this is the one category in which liability is strict in the full sense: the manufacturer is liable simply for selling a product that turns out to be defective, without proof of any fault.18
Suppose the steel in a medical instrument carried an undetectable flaw when the manufacturer bought it. The manufacturer is liable even so, because the basis of liability is selling the defective product and not any faulty conduct that led to the defect. Courts commonly say that the gravamen of the claim is the condition of the product, not the conduct of the defendant.19 In Keeler v. Richards Manufacturing Co., a surgical screw broke several months after it was inserted in the plaintiff's broken hip. The plaintiff's experts identified four irregularities they considered manufacturing defects, among them internal threads longer than the blueprint allowed and a failure to meet the specified 35% ductility standard.20
Keeler illustrates how such a claim is proved. The plaintiff compares the product with the design specifications, and the comparison may require expert testimony.21 A blueprint sets tolerances, upper and lower limits for each measurement, and a unit passes when every measurement falls within them. One wrench specification, for example, called for a ring hardness of 48 to 52 on the Rockwell C scale, while the accident wrench's ring measured between 45 and 51 in various places.22
When the manufacturer's design says nothing about the failure, ordinary expectations fill the gap. A tire design may not specify how long the tire should last, yet no owner expects a tire to fail with fewer than 2,000 miles on its treads. A tire that fails at that mileage has a manufacturing defect, not because the design promised more, but because the ordinary consumer reasonably expected at least that.23
Several reasons support strict liability for a flawed unit. Buyers pay the standard price for a product that looks like every other and expect its safety to match, so a hidden production flaw defeats an expectation that both sides built into the sale.24 Negligence is often too difficult to prove, and strict liability spares the plaintiff from reconstructing what went wrong in the factory.25 The manufacturer is better placed than the victim to find cost-effective ways of reducing production flaws and to spread the cost of those it is not cost-effective to eliminate.26 It also sets its own flaw rate through the level of investment it chooses to make in production and quality control, so the law may fairly charge it with the flaws that remain.27 Courts kept strict liability for manufacturing defects, unlike other defects, perhaps because the flaw affects a single unit and not the whole product line.28
The rule has deep roots. After MacPherson allowed negligence claims against manufacturers, courts extended warranty claims to defective food, cosmetics, and similar products, and Henningsen v. Bloomfield Motors then let a non-purchaser sue the manufacturer of a defective car in warranty despite the privity and disclaimer defenses.29 The principle underneath is not contractual. Delivering an article is wrong when it lacks the qualities the manufacturer represented, the lack could not readily be detected by the consumer, and the article is therefore unsafe for its ordinary purposes.30
Whether a claim is a manufacturing claim or a design claim turns on whether the dangerous feature appears in every unit. A manufacturing defect occurs in only a small part of a mass-produced line. A design defect occurs in the entire line, as when a car model's gas tank sits where a rear-end collision can rupture it, because every car leaves the line designed that way.31 For a manufacturing defect, the departure of the product from the manufacturer's own design supplies a ready standard. For a design defect that standard is never breached, because every unit conforms to it.32
The label matters because it fixes the plaintiff's burden. In a design case she must carry the expensive burden of proving the design defective under risk-utility analysis or a consumer expectations standard. In a manufacturing case she must show only that the product did not conform to its specifications and that the resulting defect led to her injury.33 The burdens differ because the standards differ. Liability for manufacturing defects is truly strict, while liability for design and warning defects rests largely on negligence principles, despite the strict-sounding language of “defectiveness” in which the definitions are cast.34
A manufacturer cannot answer the claim by pointing out that no production process is perfect. If a predictable flaw rate counted as part of the intended design, every flawed unit would match the design, and liability would again depend on proving negligent quality control, which strict liability exists to avoid. The rule applies “even though all possible care was exercised in the preparation and marketing of the product.”35 In one case a buyer was bitten by a copperhead snake when he reached into a bag of potting soil. The defendant moved to dismiss because the buyer had not alleged how the manufacturing process went wrong, and the court rejected the argument for the simple reason that negligence is not required in a manufacturing defect case.36
Evidence of good quality control seems logically irrelevant, because the issue is the product's defectiveness and not the manufacturer's conduct. Courts nevertheless admit it, especially in food cases involving foreign objects, when it tends to show that the defect arose after the product left the manufacturer's control.37
Strict liability does not excuse the plaintiff from proving the defect. She must show that the product was defective, that it contained the defect when it left the defendant's control, and that the defect proximately caused her harm. When a production flaw causes an accident, she can usually prove all of this by direct evidence, such as an expert's testimony that the product contained an identifiable flaw, a deviation from design specifications, that made it fail.38 Disputes in these cases usually concern causation: whether the product contained a flaw, whether the manufacturer is responsible for it, and whether it caused the harm. Juries resolve those questions, and appellate courts usually leave their findings undisturbed.39 Where a fire might have started in the stove instead of the refrigerator, plaintiffs needed competent evidence excluding the stove to survive summary judgment, and they offered three expert opinions that the fire began in the refrigerator.40 When the relevant part was destroyed in the accident, the plaintiff may rely on circumstantial evidence, by analogy to res ipsa loquitur.41
Food cases show how the departure-from-design standard works when specifications are silent. A recipe may call for chicken and say nothing about bones, so the design cannot say whether a bone in an enchilada is a flaw. Most courts, and the Third Restatement, treat a harm-causing ingredient as a defect if a reasonable consumer would not expect the food to contain it, whether the ingredient is foreign, like a piece of glass in a serving of slaw, or natural, like a sharp one-inch chicken bone in a chicken enchilada.42 A few courts still allow strict liability only for foreign objects and leave natural ingredients to a negligence theory.43 Mix v. Ingersoll Candy Co. illustrates the older approach, under which a bone natural to the meat served was not a foreign substance. Schafer v. JLC Food Systems illustrates the modern one: a restaurant patron felt sharp pain after biting into a pumpkin muffin, and the court applied the reasonable-expectations test. The two cases differ in the question each asked. Mix asked whether the object was foreign or natural, and Schafer asked what the consumer could reasonably expect. Food is the most basic product people buy, and both king and pauper may die by food or drink gone bad.44
A manufacturing claim can also outlast claims that fail for other reasons. In Bergstresser v. Bristol-Myers Squibb Co., claims that Abilify was defectively designed and inadequately warned were preempted because the FDA had approved the drug's design, warning label, and leaflet. The court noted, however, that liability for manufacturing defect is not preempted, as it almost never is. The plaintiff's remaining hope was evidence that the pills he took contained, for instance, more active ingredient than they should have.45 A later section takes up preemption itself.
Consumer Expectations Test
The comments to § 402A turn the statute's words into a working test. Comment g applies the rule where the product, at the time it leaves the seller's hands, is in a condition the ultimate consumer does not contemplate. Comment i adds that the article must be “dangerous to an extent beyond that which would be contemplated by the ordinary consumer who purchases it, with the ordinary knowledge common to the community as to its characteristics.”47 Read together, they yield the consumer expectations test: a product is defective when it is more dangerous than the ordinary buyer, armed with the community's general knowledge of that kind of product, would expect. Courts relied on this test as the dominant measure of defect in the 1960s and 1970s, and many still use it alone or combined with risk-utility.48 This subsection examines where the test works and where it breaks down.
In Greeno v. Clark Equipment Co., a federal court applying Indiana law explained that a product is unreasonably dangerous when it is more dangerous than the ordinary consumer would contemplate. An axe is not unreasonably dangerous, because users contemplate its obvious dangers. A farm combine with a weak lid over the auger is, because that danger lies beyond the contemplation of ordinary users.49 The axe cuts deeply for the very reason a buyer wants one, while the weak lid is a concealed condition that the buyer has no chance to see or weigh. The inquiry looks to the product and to what the user contemplates, and the manufacturer's care drops out, just as § 402A(2)(a) provides.50
The test takes this shape because strict tort liability grew out of warranty, and warranty law protects the reasonable expectations that a product's appearance and its seller's representations create in buyers. When Dean Prosser looked for support for the new tort, the warranty cases were nearly the only authority available, so it was natural to define defect in warranty terms.51 The tort rule does not rest on contract. It rests on the principle that delivering an article is wrong when it lacks the qualities the manufacturer represented and the consumer could not readily detect the lack.52 California's decision in Barker put the same heritage this way: a product's presence on the market carries an implicit representation that it will safely do the jobs for which it was built.53 Section 402A itself grew from cases involving products that could not minimally perform an ordinary function, such as contaminated food and exploding soda bottles.54 The standard is objective. It asks what the ordinary consumer would expect and not what this plaintiff happened to assume, because a manufacturer's design decisions must set one level of safety for consumers as a class.55
The test works best for hidden flaws, where buyer and seller share an understanding of what was sold. A buyer of a car expects the airbag to perform the basic functions that airbags in other cars ordinarily perform, and an airbag that fails to deploy frustrates that expectation without requiring the plaintiff to explain how airbags work.56 Likewise, a grocer who displays cartons of milk invites the buyer to expect milk and nothing worse, and both sides understand that the seller is to supply pure milk and not merely to take care.57 Food remains the test's most durable home. The Third Restatement keeps it there by treating a harm-causing ingredient as a defect if a reasonable consumer would not expect the food product to contain it.58
The test works poorly when the danger is visible, because a consumer who can see a hazard has contemplated it. It therefore almost always defeats a claim about an obvious design hazard, however probable or severe the harm and however cheap the remedy.59 In Vincer v. Esther Williams All-Aluminum Swimming Pool Co., the plaintiff argued that an above-ground pool was defective because the manufacturer had not added a self-closing, self-latching gate at the top of the ladder to keep small children out. The court affirmed dismissal because the safety expectations of an average consumer were plainly met, since such a person could not help but comprehend the danger from the obvious absence of a gate.60 In Halliday v. Sturm, Ruger & Co., a three-year-old was killed playing with a handgun, and the plaintiff argued that simple design changes would have prevented the death. The Maryland court applied the consumer expectations test and held that the gun functioned exactly as expected, so it was not defective even if a simple redesign would have prevented the accident.61 In each case the buyer received the product he paid for. The risk then rests on the buyer, the user, or a third-party victim who had no say in the purchase.62 That result runs against one of the principal justifications for strict liability, the need for safety incentives.63
The test also errs in the other direction, so it is not simply a plaintiff-friendly standard.
Because buyers held no particular expectation about the switch, only a test that weighs the cost of moving it against the risk can reach the claim.64 The reverse also occurs. A product may fall short of consumer expectations even though it cannot be redesigned to be safer, and under the expectations test the plaintiff need show only that the product was less safe than a reasonable consumer would expect.65 That feature explains why plaintiffs often favor the test. It is intuitive, it leaves the jury room for common sense, and it can spare the plaintiff the expense of an expert on alternative designs.66
A product whose danger is inseparable from what it is shows the test's limit from a third angle.
Comment i draws the same boundary: good whiskey is not unreasonably dangerous merely because it intoxicates, but bad whiskey containing a dangerous amount of fusel oil is.67 A risk-utility analysis reaches the same result for vodka, since it cannot be redesigned to be completely benign and still be vodka, so there is no reasonable alternative design.68
The test must also say whose expectations count when the person hurt did not choose the product. In Bellotte v. Zayre Corp., the court looked to the parents' expectations in a claim by a five-year-old burned when his cotton pajama top ignited while he played with matches. In Calles v. Scripto-Tokai Corp., the Illinois Supreme Court reasoned that the ordinary consumer of a utility lighter is an adult, so safety expectations are judged from the adult's perspective and not from that of the young child who used the lighter to start a fatal fire.
The test's vagueness is its deepest problem. The meaning is ambiguous, and one critic wrote that “the test can be utilized to explain most any result that a court or jury chooses to reach,” since the reasonable purchaser in one sense does not “expect” an unknown hazard and in another contemplates the “possibility” of unknown side effects.69 Courts that keep the test therefore tend to confine it to simple products. California, for example, now reserves the consumer expectations prong for cases in which the everyday experience of the product's users permits a conclusion that the design violated minimum safety assumptions.70
The test's decline as the principal design standard followed from these weaknesses. The risk-utility test gradually supplanted consumer expectations as the principal liability test, because cost-benefit analysis seemed a more orderly way to judge the safety of designs and warnings.71 The Third Restatement now treats consumer expectations as one factor among several in the design balance.72 Many courts have not followed that lead and continue to use the expectations test instead of, or as an alternative to, risk-utility.73 Barker v. Lull Engineering Co. gave the alternative its classic form. A jury may be instructed that a design is defective if the plaintiff proves that the product failed to perform as safely as an ordinary consumer would expect when used in an intended or reasonably foreseeable manner, or if the plaintiff proves that the design proximately caused injury and the defendant fails to prove that the design's benefits outweigh its risks.74 At the opposite pole, Branham v. Ford Motor Co. held the risk-utility test, with its requirement of a feasible alternative design, the exclusive test in a design case. The court agreed that the expectations test fits well in manufacturing defect cases but found it ill-suited to design.75 The analysis therefore begins by identifying which formulation the jurisdiction uses.
The third step does the most work in practice, because an obvious danger usually ends the inquiry and sends the claim toward a different standard. The next subsection takes up the risk-utility test, which asks the question this one leaves open: whether a safer design was worth its cost.
Risk-Utility Test
No design is risk-free, so courts needed a way to separate acceptable designs from those carrying avoidable danger.77 The consumer expectations test could not do the job alone. Buyers hold no expectations about most design choices, and an obvious danger passes the test however cheaply it could have been removed. Over the decades courts substituted some form of cost-benefit standard, called risk-utility or risk-benefit, which asks whether the safety benefit of removing a danger was worth its cost.78 The shift carried a price. Because risk-utility analysis circled back to negligence, it hastened the decline of truly “strict” liability under § 402A.79 Section 2(b) of the Restatement (Third) writes the balance into a rule: a product is defective in design when the foreseeable risks of harm could have been reduced or avoided by a reasonable alternative design, and omitting that design renders the product not reasonably safe.80
Judge Learned Hand's negligence formula, restated in cost-benefit terms, supplies the arithmetic. A defect is suggested if the safety benefits of an untaken precaution foreseeably exceed the resulting costs, including any diminished usefulness or diminished safety.81 The precaution required rises with the risk. A failing steering system or a punch press ram that may descend on an operator's hand demands the utmost precautions, while a risk of scratches or fabric stains demands very little.82 The formula puts consumers' safety on the same scale as the manufacturer's interest in cost and profit, so a manufacturer that already weighs usefulness and price in choosing a design can fairly be asked to weigh foreseeable injuries too.83
The balance favors the plaintiff when a cheap precaution would have prevented a serious injury. In Helicoid Gage Division of American Chain & Cable Co. v. Howell, a pressure gauge burst and threw a piece of its lens into the plaintiff's eye, and shatterproof glass would have added only about a dollar to each gauge without reducing its usefulness. In Dorsey v. Yoder Co., the defendant moved for judgment notwithstanding a verdict for the plaintiff. The court observed that a guard would not eliminate the machine's usefulness, that a $200 to $500 guard on an $8,000 machine was not unreasonable, and that the potential harm was great. All of these facts, it said, were questions for the jury.84
The same arithmetic governs an industrial press that delivers 60 tons of force. If the manufacturer can cheaply build the press to activate only when the operator pushes two separate buttons placed away from the press bed, a jury could readily find the single-switch design defective, because a large reduction in risk was available at small cost. When the numbers run the other way, with a low risk and a costly precaution, the design is not defective and the defendant is not liable under either negligence or strict liability.85
The weighing is done one precaution at a time. The question courts actually try is whether the marginal cost of changing the design in some particular way, such as adding a guard or relocating a switch, was worth the foreseeable safety benefit. This narrow “micro-balance” describes the test as lawyers and judges actually use it, whereas early formulations that weighed a product's total risks against its total utility did not.86 Naming the precaution also gives the manufacturer a concrete target to answer. That matters because a ruling that a design is defective condemns the entire product line.87
Courts give juries lists of considerations to structure the balance. In Barker, the California Supreme Court said a jury may consider, among other relevant factors:
- the gravity of the danger posed by the challenged design;
- the likelihood that the danger will occur;
- the mechanical feasibility of a safer alternative design;
- the financial cost of an improved design; and
- the adverse consequences to the product and to the consumer that an alternative design would produce.88
A longer list, the Wade factors, adds the product's usefulness and desirability, the user's ability to avoid the danger by careful use, the user's likely awareness of the danger, and the feasibility of spreading the loss through price or insurance. Modern opinions often recite the list and then decide the case on the cost and benefit of the specific alternative the plaintiff proposes, which is where the real contest lies.89 Apart from loss spreading, these are the factors negligence law routinely weighs.90
The Restatement (Third) goes beyond listing factors and, with limited exceptions, requires the plaintiff to prove a reasonable alternative design. The court then compares the accident product with the product as the plaintiff's proposed feature would have changed it.91 The requirement rests on the commonsense notion that liability for harm caused by product designs should attach only when the harm is reasonably preventable. The court weighs foreseeable risks of harm, consumer expectations, usefulness, costs, aesthetics, longevity, responsibility for maintenance, marketability, and the other advantages and disadvantages of both designs.92 The premise is that safety has a limit. Society gains nothing from a car governed to 20 miles per hour, any more than it gains from a product that is too risky, and it benefits most when product safety reaches its optimal amount.93
The vest facts show why the alternative must be reasonable. A plaintiff who cannot point to an economically feasible modification has no design claim, however serious the injury.
Proving the balance does not always require a quantified study or an expert. The plaintiff must give the jury evidence from which it can reasonably estimate the design's risks and the utility lost if the design were made safer, but jurors can often do this from familiar products and familiar risks.94 Sharp, propeller-like blades protruding from a hubcap endanger anyone who touches the spinning blades and offer almost nothing beyond appearance, so the plaintiff needs no special testimony about the degree of risk or the lack of utility.95 When the claim involves a technical product and the plaintiff must show a feasible alternative design or causation, expert testimony is practically required, and a ruling excluding the expert as unreliable can end the case.96
Because the test weighs an untaken precaution against its safety benefit, it resembles a negligence analysis aimed at the product and not at the maker's conduct. The jury asks whether the manufacturer made a reasonable trade-off between risk and the expense of preventing it, an assessment very close to the reasonable-person standard of negligence law.97 Most courts also require that the risk have been foreseeable when the product was marketed, or at least treat the absence of foreseeability as a defense.98 A later section asks how strict products liability really is once that resemblance is taken seriously.
Risk-utility also reaches cases that the expectations test cannot. A machine with an exposed switch may satisfy buyers, who can see the switch and give it no thought, yet fail the balance if the switch could easily have been relocated to prevent inadvertent start-ups.99 The change in test alters the treatment of obvious dangers too. A few courts hold that an open and obvious danger bars a design claim outright, but most treat obviousness as one factor in the balance. In Perkins v. Wilkinson Sword, Inc., children died in a fire they started while playing with a cigarette lighter, and the Ohio Supreme Court treated the lighter's obvious danger as one factor in the risk-utility analysis and not as a determinative one. In Timpte Industries, Inc. v. Gish, the Texas Supreme Court said that the obviousness of the danger from a trailer's top rail is not itself a bar to liability but an important consideration that may even be decisive in a particular case. The user's ability to see and avoid a danger enters the balance as one more fact about risk and utility, which is why the Wade list includes it.
The same balance explains cases in which no defect exists because the danger cannot be removed without destroying the product. A plaintiff who attacks a whole category of widely distributed product, such as alcoholic beverages, firearms, or above-ground swimming pools, must still identify some reasonable way of designing the danger out. That is impossible when the inherent risk is unavoidable by hypothesis.100 In Adamo v. Brown & Williamson Tobacco Corp., the plaintiff proposed a cigarette with lower tar and nicotine, and the claim failed because the plaintiff did not show that the proposed cigarette would serve the same function or utility as the one sold. Courts have been markedly unreceptive to invitations to outlaw whole product categories and have left that choice to legislatures and regulators.101 A narrow exception remains for a product so dangerous and so useless that no rational person, fully aware of the facts, would choose to use it or let children use it. Toy guns that shoot hard pellets at high velocity, exploding novelty cigars, lawn darts, and clothing made of highly flammable fabrics are examples, and courts should normally rule on such claims as a matter of law.102 O'Brien v. Muskin Corp. shows the road most courts have declined to take. A plaintiff who dove into a shallow above-ground pool argued that if the vinyl liner was the only technically feasible one, the whole category of such pools was defective, and the New Jersey Supreme Court let a jury decide whether the pools' risks and benefits were worth the candle.103 New Jersey's legislature then superseded that result by statute, providing that a manufacturer is not liable if no practical and technically feasible alternative design would have prevented the harm without substantially impairing the product's intended function.104
A small group of states changes who carries the balance. Under Barker, once the plaintiff shows that the design proximately caused her injury, the burden shifts to the defendant to prove that the design's benefits outweigh its inherent risk.105 Most courts that have considered the shift have rejected it, as has the Third Restatement, so in the usual jurisdiction the plaintiff carries the burden on every part of the balance.106 Some jurisdictions combine this test with consumer expectations, as the previous subsection described, and others reject the combination. In Walker v. Ford Motor Co., the Colorado Supreme Court recognized risk-utility as the exclusive test and held it error to include consumer expectations in the jury instruction.
Warning Defects
Comment j to § 402A recognizes that a seller may have to give directions or a warning on the container to keep a product from being unreasonably dangerous.108 The Restatement (Third) makes the point a category of defect. A product is defective because of inadequate instructions or warnings when the foreseeable risks of harm it poses could have been reduced or avoided by reasonable instructions or warnings, and their omission renders it not reasonably safe.109 The category stands alone. A manufacturer is liable for a manufacturing defect however clear the warnings or perfect the design, and for a warning defect however perfect the manufacture or impeccable the design.110 One treatise calls this branch the “informational defect.” It contrasts it with manufacturing defects, which it understands as strict liability for a seller's misrepresentation of its product.111
Warning law addresses the risks that remain after design, those that no reasonable design removes. A warning can serve one or both of two functions: it can instruct users how to use the product safely, and it can tell consumers about unavoidable risks so that they can decide whether to take them.112 Sellers will not reliably supply that information unprompted, because disclosing a risk alerts buyers to a cost they had not counted and decreases demand, which gives the seller an incentive to withhold it.113 The law therefore compels the disclosure. In return, an adequate warning generally relieves the seller of liability for the risk it describes, because once the warning is given the purchaser, not the seller, is the superior bearer of the risk of injury.114
Adequacy is the test of liability in every warning claim, and it turns on reasonableness. To be adequate, a warning must provide information about a product's material hidden dangers and how to avoid them, in a manner and with an intensity justified by the risks and calculated to reach and be understood by those likely to need it. Courts look to its content and comprehensibility, its intensity, and the characteristics of the expected users. They weigh the severity of the danger, the likelihood that the warning will catch the attention of foreseeable users, and the cost of improving it.115 Content must convey the nature of the danger and not merely issue an order. In Post v. American Cleaning Equipment Corp., the court doubted that a sign reading “Keep off the Grass” could apprise a reasonable person that the grass was infested with deadly snakes. A reasonable person might risk the penalty of walking on the grass, but not if he knew the real consequences.116 Content, placement, and audience can fail together. A notice on a medical needle's wrapper that the “shelf life of needle is three years” says too little about the danger of broken needle tips. The wrapper is ripped off and discarded to get at the needle, and the notice may be aimed at the physician and not the patient.117
The duty attaches to risks that are foreseeable, hidden, and material, and comment j shows the limits through allergies. A seller may assume that people with common allergies, such as to eggs or strawberries, know of them and need no warning. But when a substantial number of people are allergic to an ingredient whose danger is not generally known, or which consumers would not expect to find in the product, and the seller knows or should know of the ingredient and the danger, a warning is required.118 Materiality is a judgment of reasonableness. Because warnings are easy to provide, they will usually be warranted when the ordinary consumer would want to know of the risk, and a serious reaction suffered by 3 percent of users is the kind of risk a manufacturer probably should warn about.119 A risk that most consumers already know may still call for a warning. If a risk is virtually certain to kill those who are unaware of it, the ordinary consumer would want the warning even if 99 percent of consumers know of the risk, because the expected benefit of avoiding a one percent chance of dying exceeds the small cost of reading the disclosure.120
A manufacturer that neglects to supply reasonable warnings and instructions about foreseeable, latent, and material hazards is subject to strict liability in tort for the resulting physical harm.121 Because adequacy turns on reasonableness, and the Third Restatement grounds warning defects explicitly in negligence principles, the claim behaves much like negligence even when it is pleaded as strict liability.122 A seller that instead misstates a product's character or quality on its label faces separate liability for misrepresentation under § 402B, even without fraud or negligence.123
Strict liability reaches every seller engaged in the business of selling the product, so the duty to warn does not stop at the manufacturer.124 A hardware store that resells a hedge trimmer in its sealed box is not negligent, yet it sells a product whose packaging omitted a needed warning. It therefore answers to the injured buyer and is left to seek indemnity from the manufacturer.125 Indemnity is only as good as the manufacturer's solvency. In Marcon v. Kmart Corp., Kmart was held liable for some $8 million for failing to warn of a risk from a sled it had sold, and the manufacturer was bankrupt.126
A seller must also take account of how people will foreseeably misuse its product. A person stretching to reach the top of a window may be tempted to gain height by resting a foot on a stepladder's paint shelf. If the shelf will not bear her weight, a warning printed on it can avert the injury, and her own carelessness in trusting the shelf is a matter for comparative fault and does not erase the duty.127 The duty has an outer edge. Manufacturers cannot warn about everything, because excessive warnings undercut the impact of each one, and they cannot anticipate every preposterous manner in which people will misuse their products.128 Between the paint shelf and the preposterous lies a wide middle ground, and a prudent judge will usually treat the foreseeability of misuse as a question of fact for the jury.129
The most important limit on warnings is that a warning cannot cure a design that should have been made safer. The last sentence of comment j seems to say otherwise. It provides that where a warning is given the seller may reasonably assume it will be read and heeded, and that a product bearing a warning, safe if the warning is followed, is not defective.130 Read literally, that sentence would let a fan maker print “Metal blades may cut off fingers” on the base in place of a protective cage. A great majority of courts reject that reading, because each type of defect creates its own obligation and satisfying one does not satisfy the others.131 The narrow truth in comment j is that sellers of inherently dangerous products, such as food, alcohol, tobacco, and drugs, have more limited obligations. Their danger cannot be designed out by hypothesis, so beyond avoiding impurities and misrepresentation they must warn of the unavoidable, latent dangers and instruct on safe use.132 The Third Restatement states that where a safer design can reasonably be implemented, “adoption of the safer design is required over a warning that leaves a significant residuum of such risks.” One court put it more vividly: a warning “is not a Band-Aid to cover a gaping wound.”133 A stamping machine that carries a large red label telling operators to keep their hands out of the stamping area, but has no guard, shows the point. A guard would slow production somewhat, but under a risk-utility analysis the lost speed is likely outweighed by the increased safety. It is almost inevitable that an operator's hands will end up in the stamping area now and then, through an accident, the temptation to adjust the work, or sheer inadvertence, so the better argument is that the machine is defective in design and the label does not answer that claim.134
A warning must also reach someone in a position to use it. Upstream suppliers of bulk materials may sometimes rely on downstream buyers to perform the warning function if there is a reasonable basis to believe the buyers will pass warnings along, but a supplier that knows more about a hazard than its purchasers must share that knowledge when there is a reasonable way to do so.135 The bulk-supplier doctrine arises because a supplier of liquids, sand, or gas delivered by tank car or truck usually has no practical way to put a label in front of the workers who will eventually handle the product.136 The supplier's reliance on the intermediary must itself be reasonable, which requires consideration of the purpose for which the product is used, the magnitude of the risk, the burden of warning end users directly, and the reliability of the intermediary as a conduit.137
The learned-intermediary rule is the intermediary doctrine that exams test most often. The manufacturer of a prescription drug or medical device ordinarily satisfies its duty to warn by giving the prescribing physician all the foreseeable risks and therapeutic benefits material to the prescribing decision.138 The rationale is strong. Only medical professionals have the knowledge, training, and judgment to match a particular drug, with its distinctive benefits and dangers, to a particular patient, and to monitor the results. Tort law in turn requires the prescribing doctor to inform the patient of the drug's benefits and dangers.139 Section 6(d) of the Third Restatement states the rule together with its limit. A drug or device is not reasonably safe if reasonable warnings were not given to the health-care providers in a position to reduce the risk, or to the patient when the manufacturer knows or has reason to know that providers will not be in such a position.140
Even an inadequate warning produces liability only if the missing information caused the injury, and warning cases make that proof awkward because the plaintiff must say what would have happened under a different label. A user who never read the label cannot show that a better label would have changed anything, so the omission is not a but-for cause of the harm.141 Many courts respond with a heeding presumption. Because comment j lets a manufacturer that warns adequately assume the warning will be read and heeded, many courts reason that the same provision must work in reverse and give the plaintiff a presumption when the manufacturer gives no warning. They reason that manufacturers should not escape their safety obligation because accident victims cannot prove what they would have done.142 The presumption shifts responsibility for proving causation but does not eliminate the requirement, so a claim still fails where a warning probably would have done no good.143 Defendants rebut it with evidence about the plaintiff. In Kirkbride v. Terex USA, LLC, the plaintiff had not read the warning in the manual and had expressed a negative view of manuals. In Sharpe v. Bestop, Inc., a plaintiff who claimed he would have buckled up had a warning said so admitted that he had ignored the vehicle maker's visor warning to wear a seat belt.
Causation has a second limit in the scope of the risk. The injury must fall within the class of harm the warning was meant to prevent. A worker who was not warned that asbestos can cause cancer, and who instead loses a hand in a machine accident, can show that the missing warning was a factual cause, because a proper warning would have put him elsewhere. It was not a proximate cause, because the injury lies outside the risk the warning addressed. A user who acts out of momentary inadvertence and not ignorance may likewise show that a warning would have made no difference.144
A warning given after the sale raises a separate question. Where a product was already defective at the time of sale within the meaning of the Restatement, liability can be established without reference to any post-sale duty to warn, and a seller that discovers the defect after the sale generally cannot absolve itself by issuing a post-sale warning.145
The last step is where many warning claims fail, because a plaintiff who proves an inadequate label has still not proved that a better one would have prevented the injury.