Cardozo, J.

The defendant is a manufacturer of automobiles. It sold an automobile to a retail dealer. The retail dealer resold to the plaintiff. While the plaintiff was in the car, it suddenly collapsed. He was *385thrown out and injured. One of the wheels was made of defective wood, and its spokes crumbled into fragments. The wheel was not made by the defendant; it was bought from another manufacturer. There is evidence, however, that its defects could have been discovered by reasonable inspection, and that inspection was omitted. There is no claim that the defendant knew of the defect and willfully concealed it. The case, in other words, is not brought within the rule of Kuelling v. Lean Mfg. Co. (183 N. Y. 78). The charge is one, not of fraud, but of negligence. The question to be determined is whether the defendant owed a duty of care and vigilance to any one but the immediate purchaser.

The foundations of this branch of the law, at least in this state, were laid in Thomas v. Winchester (6 N. Y. 397). A poison was falsely labeled. The sale was made to a druggist, who in turn sold to a customer. The customer recovered damages from the seller who affixed the label. “The defendant’s negligence,” it was said, “put human life in imminent danger.” A poison falsely labeled is likely to injure any one who gets it. Because the danger is to be foreseen, there is a duty to avoid the injury. Cases were cited by way of illustration in which manufacturers were not subject to any duty irrespective of contract. The distinction was said to be that their conduct, though negligent, was not likely to result in injury to any one except the purchaser. We are not required to say whether the chance of injury was always as remote as the distinction assumes. Some of the illustrations might be rejected to-day. The principle of the distinction is for present purposes the important thing.

Thomas v. Winchester became quickly a landmark of the law. In the application of its principle there may at times have been uncertainty or even error. There has never in this state been doubt or disavowal of the principle itself. The chief cases are well known, yet to recall *386some of them will he helpful. Loop v. Litchfield (42 N. Y. 351) is the earliest. It was the case of a defect in a small balance wheel used on a circular saw. The manufacturer pointed out the defect to the buyer, who wished a cheap article and was ready to assume the risk. The risk can hardly have been an imminent one, for the wheel lasted five years before it broke. In the meanwhile the buyer had made a lease of the machinery. It was held that thti manufacturer was not answerable to the lessee. Loop v. Litchfield was followed in Losee v. Clute (51 N. Y. 494), the case of the explosion of a steam boiler. That decision has been criticised (Thompson on Negligence, 233; Shearman & Redfield on Negligence [6th ed.], § 117); but it must be confined to its special facts. It was put upon the ground that the risk of injury was too remote. The buyer in that case had not only accepted the boiler, hut had tested it. The manufacturer knew that his own test was not the final one. The finality of the test has a bearing on the measure of diligence owing to persons other than the purchaser (Beven, Negligence [3d ed.], pp. 50, 51, 54; Wharton, Negligence [2d ed.], § 134).

These early cases suggest a narrow construction of the rule. Later cases, however, evince a more liberal spirit. First in importance is Devlin v. Smith (89 N. Y. 470). The defendant, a contractor, built a scaffold for a painter. The painter’s servants were injured. The contractor was held liable: He knew that the scaffold, if improperly constructed, was a most dangerous trap. He knew that it was to he used by the workmen. He was building it for that very purpose. Building it for their use, he owed them a duty, irrespective of his contract with their master, to build it with care.

From Devlin v. Smith we pass over intermediate cases and turn to the latest case in this court in which Thomas v. Winchester was followed. That case is Statler v. Ray Mfg. Co. (195 N. Y. 478, 480). The defendant *387manufactured a large coffee urn. It was installed in a restaurant. When heated, the urn exploded and injured the plaintiff. W e held that the manufacturer was liable. We said that the urn “was of such a character inherently that, when applied to the purposes for which it was designed, it was liable to become a source of great danger to many people if not carefully and properly constructed.” It may be that Devlin v. Smith and Statler v. Ray Mfg. Co. have extended the rule of Thomas v. Winchester. If so, this court is committed to the extension. The defendant argues that things imminently dangerous to life are poisons, explosives, deadly weapons — things whose normal function it is to injure or destroy. But whatever the rule in Thomas v. Winchester may once have been, it has no longer that restricted meaning. A scaffold (Devlin v. Smith, supra) is not inherently a destructive instrument. It becomes destructive only if imperfectly constructed. A large coffee urn (Statler v. Ray Mfg. Co., supra) may have within itself, if negligently made, the potency of danger, yet no one thinks of it as an implement whose normal function is destruction. What is true of the coffee urn is equally true of bottles of aerated water (Torgeson v. Schultz, 192 N. Y. 156). We have mentioned only cases in this court. But the rule has received a like extension in our courts of intermediate appeal. In Burke v. Ireland (26 App. Div. 487), in an opinion by Cullen, J., it was applied to a builder who constructed a defective building; in Kahner v. Otis Elevator Co. (96 App. Div. 169) to the manufacturer of an elevator; in Davies v. Pelham Hod Elevating Co. (65 Hun, 573; affirmed in this court without opinion, 146 N. Y. 363) to a contractor who furnished a defective rope with knowledge of the purpose for which the rope was to be used. We are not required at this time either to approve or to disapprove the application of the rule that was made in these cases. It is enough that they help to characterize the trend of judicial thought.

*388 Devlin v. Smith was decided in 1882. A year later a very similar case came before the Court of Appeal in England (Heaven v. Pender, L. B. [11 Q. B. D.] 503). We find in the opinion of Brett, M. B., afterwards Lord Esher (p. 510), the same conception of a duty, irrespective of contract, imposed upon the manufacturer by the law itself: “Whenever one person supplies goods, or machinery, or the like, for the purpose of their being used by another person under such circumstances that every one of ordinary sense would, if he thought, recognize at once that unless he used ordinary care and skill with regard to the condition of the thing supplied or the mode of supplying it, there will be danger of injury to the person or property of him for whose use the thing is supplied, and who is to use it, a duty arises to use ordinary care and skill as to the condition or manner of supplying such thing.” He then points out that for a neglect of such ordinary care or skill whereby injury happens, the appropriate remedy is an action for negligence. The right to enforce this liability is not to be confined to the immediate buyer. The right, he says, extends to the persons or class of persons for whose use the thing is supplied. It is enough that the goods “would in all probability be used at once * * * before a reasonable opportunity for discovering any defect which might exist,” and that the thing supplied is of such a nature “that a neglect of ordinary care or skill as to its condition or the manner of supplying it would probably cause danger to the person or property of the person for whose use it was supplied, and who was about to use it.” On the other hand, he would exclude a case “in which the goods are supplied under circumstances in which it would be a chance by whom they would be used or whether they would be used or not; or whether they would be used before there would probably be means of observing any defect,” or where the goods are of such a nature that “a want of care or skill as to their condition or the manner 'of supplying them would not probably *389produce danger of injury to person or property.” What was said by Lord Esher in that case did not command the full assent of his associates. His opinion has been criticised “as requiring every man to take affirmative precautions to protect his neighbors as well as to refrain from injuring them” (Bohlen, Affirmative Obligations in the Law of Torts, 44 Am. Law Reg. [N. S.] 341). It may not be an accurate exposition of the law of England. Perhaps it may need some qualification even in our own state. Like most attempts at comprehensive definition, it may involve errors of inclusion and of exclusion. But its tests and standards, at least in their underlying principles, with whatever qualification may be called for as they are applied to varying conditions, are the tests and standards of our law.

We hold, then, that the principle of Thomas v. Winchester is not limited to poisons, explosives, and things of like nature, to things which in their normal operation are implements of destruction. If the nature of a thing is such that it is reasonably certain to place life and limb in peril when negligently made, it is then a thing of danger. Its nature gives warning of the consequences to be expected. If to the element of danger there is added knowledge that the thing will housed by persons other than the purchaser, and used without new tests, then, irrespective of contract, the manufacturer of this thing of danger is under a duty to make it carefully. That is as far as we are required to go for the decision of this case. There must be knowledge of a danger, not merely possible, but probable. It is possible to use almost anything in a way that will make it dangerous if defective. That is not enough to charge the manufacturer with a duty independent of his contract. Whether a given thing is dangerous may be sometimes a question for the court and sometimes a question for the jury. There must also be knowledge that in the usual course of events the danger will he shared by others than the buyer. Such knowledge may often be *390inferred from the nature of the transaction. But it is possible that even knowledge of the danger and of the use will not always be enough. The proximity or remoteness of the relation is a factor to be considered. We are dealing now with the liability of the manufacturer of the finished product, who puts it on the market to be used without inspection by his customers. If he is negligent, where danger is to be foreseen, a liability will follow. We are not required at this time to say that it is legitimate to go back of the manufacturer of the finished product and hold the manufacturers of the component parts. To make their negligence a cause of imminent danger, an independent cause must often intervene; the manufacturer of the finished product must also fail in his duty of inspection. It may be that in those circumstances the negligence of the earlier members of the series is too remote to constitute, as to the ultimate user, an actionable wrong (Beven on Negligence [3d ed.], 50, 51, 51; Wharton on Negligence [2d ed.], § 134; Leeds v. N. Y. Tel. Co., 178 N. Y. 118; Sweet v. Perkins, 196 N. Y. 482; Hayes v. Hyde Park, 153 Mass. 511, 516). We leave that question open. We shall have to deal with it when it arises. The difficulty which it suggests is not present in this case. There is here no break in the chain of cause and effect. In such circumstances, the presence of a known danger, attendant upon a known use, makes vigilance a duty. We have put aside the notion that the duty to safeguard life and limb, when the consequences of negligence may be foreseen, grows out of contract and nothing else. We have put the source of the obligation where it ought to be. We have put its source in the law.

From this survey of the decisions, there thus emerges a definition of the duty of a manufacturer which enables us to measure this defendant’s liability. Beyond all question, the nature of an automobile gives warning of probable danger if its construction is defective. This *391automobile was designed to go fifty miles an hour. Unless its wheels were sound and strong, injury was almost certain. It was as much a thing of danger as a defective engine for a railroad. The defendant knew the danger. It knew also that the car would be used by persons other than the buyer. This was apparent from its size; there were seats for three persons. It was apparent also from the fact that the buyer was a dealer in cars, who bought to resell. The maker of this car supplied it for the use of purchasers from the dealer just as plainly as the contractor in Devlin v. Smith supplied the scaffold for use by the servants of the owner. The dealer was indeed the one person of whom it might be said with some approach to certainty that by him the car would not be used. Yet the defendant would have us say that he was the one person whom it was under a legal duty to protect. The law does not lead us to so inconsequent a conclusion. Precedents drawn from the days of travel by stage coach do not fit the conditions of travel to-day. The principle that the danger must be imminent does not change, but the things subject to the principle do change. They are whatever the needs of life in a developing civilization require them to be.

In reaching this conclusion, we do not ignore the decisions to the contrary in other jurisdictions. It was held in Cadillac M. C. Co. v. Johnson (221 Fed. Rep. 801) that an automobile is not within the rule of Thomas v. Winchester. There was, however, a vigorous dissent. Opposed to that decision is one of the Court of Appeals of Kentucky (Olds Motor Works v. Shaffer, 145 Ky. 616). The earlier cases are summarized by Judge Sanborn in Huset v. J. I. Case Threshing Machine Co. (120 Fed. Rep. 865). Some of them, at first sight inconsistent with our conclusion, may be reconciled upon the ground that the negligence was too remote, and that another cause had intervened. But even when they cannot be reconciled, the difference is rather in the applica*392tion of the principle than in the principle itself. Judge Sanborn says, for example, that the contractor who builds a bridge, or the manufacturer who builds a car, cannot ordinarily foresee injury to other persons than the owner as the probable result (120 Fed. Rep. 865, at p. 867). We take a different view. We think that injury to others is to be foreseen not merely as a possible, but as an almost inevitable result. (See the trenchant criticism in Bohlen, supra, at p. 351). Indeed, Judge Sanborn concedes that his view is not to be reconciled with our decision in Devlin v. Smith (supra). The doctrine of that decision has now become the settled law of this state, and we have no desire to depart from it.

In England the limits of the rule are still unsettled. Winterbottom v. Wright (10 M. & W. 109) is often cited. The defendant undertook to provide a mail coach to carry the mail bags. The coach broke down from latent defects in its construction. The defendant, however, was not the manufacturer. The court held that he was not liable for injuries to a passenger. The case was decided on a demurrer to the declaration. Lord Esher points out in Heaven v. Pender (supra, at p. 513) that the form of the declaration was subject to criticism. It did not fairly suggest the existence of a duty aside from the special contract which was the plaintiff’s main reliance. (See the criticism of Winterbottom v. Wright, in Bohlen, supra, at pp. 281, 283). At all events, in Heaven v. Pender (supra) the defendant, a dock owner, who put up a staging outside a ship, was held liable to the servants of the shipowner. In Elliott v. Hall (15 Q. B. D. 315) the defendant sent out a defective truck laden with goods which he had sold. The buyer’s servants unloaded it, and were injured because of the defects. It was held that the defendant was under a duty “not to he guilty of negligence with regard to the state and condition of the truck.” There seems to have been a *393return to the doctrine of Winterbottom v. Wright in Earl v. Lubbock (L. B. [1905] 1 K. B. 253). In that case, however, as in the earlier one, the defendant was not the manufacturer. He had merely made a contract to keep the van in repair. A later case (White v. Steadman, L. R. [1913], 3 K. B. 340, 348) emphasizes that element. A livery stable keeper who sent out a vicious horse was held liable not merely to his customer but also to another occupant of the carriage, and Thomas v. Winchester was cited and followed (White v. Steadman, supra, at pp. 348, 349). It was again cited and followed in Dominion Natural Gas Co. v. Collins (L. R. [1909] A. C. 640, 646). From these cases a consistent principle is with difficulty extracted. The English courts, however, agree with ours in holding that one who invites another to make use of an appliance is bound to the exercise of reasonable care (Caledonian Ry. Co. v. Mulholland, L. R. [1898] A. C. 216, 227; Indermaurv. Dames, L. R. [1 C. P.] 274). That at bottom is the underlying principle of Devlin v. Smith. The contractor who builds the scaffold invites the owner’s workmen to use it. The manufacturer who sells the automobile to the retail dealer invites the dealer’s customers to use it. The invitation is addressed in the one case to determinate persons and in the other to an indeterminate class, but in each case it is equally plain, and in each its consequences must be the same.

There is nothing anomalous in a rule which imposes upon A, who has contracted with B, a duty to C and D and others according as he knows or does not know that the subject-matter of the contract is intended for their use. We may find an analogy in the law which measures the liability of landlords. If A leases to B a tumbledown house he is not liable, in the absence of fraud, to B’s guests who enter it and are injured. This is because B is then under the duty to repair it, the lessor has the right to suppose that he will fulfill that duty, and, if he *394omits to do so, his guests must look to him (Bohlen, supra, at p. 276). But if A leases a building to be used by the lessee at once as a place of public entertainment, the rule is different. There injury to persons other than the lessee is to be foreseen, and foresight of the consequences involves the creation of a duty (Junhermann v. Tilyou R. Co., 213 N. Y. 404, and cases there cited).

In this view of the defendant’s liability there is nothing inconsistent with the theory of liability on which the case was tried. It is true that the court told the jury that “an automobile is not an inherently dangerous vehicle.” The meaning, however, is made plain by the context. The meaning is that danger is not to be expected when the vehicle is well constructed. The court left it to the jury to say whether the defendant ought to have foreseen that the car, if negligently constructed, would become “imminently dangerous.” Subtle distinctions are drawn by the defendant between things inherently dangerous and things imminently dangerous, but the case does not turn upon these verbal niceties. If danger was to be expected as reasonably certain, there was a duty of vigilance, and this whether you call the danger inherent or imminent. In varying forms that thought was put before the jury. We do not say that “the court would not have been justified in ruling as a matter of law that the car was a dangerous thing. If there was any error, it was none of which the defendant can complain.

We think the defendant was not absolved from a duty of inspection because it bought the wheels from a reputable manufacturer. It was not merely a dealer in automobiles. It was a manufacturer of automobiles. It was responsible for the finished product. It was not at liberty to put the finished product on the market without subjecting the component parts to ordinary and simple tests (Richmond & Danville R. R. Co. v. Elliott, 149 U. S. 266, 272). Under the charge of the trial judge nothing more was *395required of it. The obligation to inspect must vary with the nature of the thing to be inspected. The more probable the danger, the greater the need of caution. There is little analogy between this case and Carlson v. Phoenix Bridge Co. (132 N. Y. 273), where the defendant bought a tool for a servant’s use. The making of tools was not the business in which the master was engaged. Reliance on the skill of the manufacturer was proper and almost inevitable. But that is not the defendant’s situation. Both by its relation to the work and by the nature of its business, it is charged with a stricter duty.

Other rulings complained of have been considered, but no error has been found in them.

The judgment should be affirmed with costs.

Willard Bartlett, Ch. J. (dissenting).

The plaintiff was injured in consequence of the collapse of a wheel of an automobile manufactured by the defendant corporation which sold it to a firm of automobile dealers in Schenectady, who in turn sold the car to the plaintiff. The wheel was purchased by the Buick Motor Company, ready made, from the Imperial Wheel Company of Flint, Michigan, a reputable manufacturer of automobile wheels which had furnished the defendant with eighty thousand wheels, none of which had proved to be made of defective wood prior to the accident in the present case. The defendant relied upon the wheel manufacturer to make all necessary tests as to the strength of the material therein and made no such tests itself. The present suit is an action for negligence brought by the subvendee of the motor car against the manufacturer as the original vendor. The evidence warranted a finding by the jury that the wheel which collapsed was defective when it left the hands of the defendant. The automobile was being prudently operated at the time of the accident and was moving at a speed of only eight miles an hour. There was *396no allegation or proof of any actual knowledge of the defect on the part of the defendant or any suggestion that any element of fraud or deceit or misrepresentation entered into the sale.

The theory upon which the case was submitted to the jury by the learned judge who presided at the trial was that, although an automobile is not an inherently dangerous vehicle, it may become such if equipped with a weak wheel; and "that if the motor car in question, when it was put upon the market was in itself inherently dangerous by reason of its being equipped with a weak wheel, the 'defendant was chargeable with a knowledge of the defect so far as it might be discovered by a reasom able inspection and the application of reasonable tests. This liability, it was further held, was not limited to the original vendee, but extended to a subvendee like the plaintiff, who was not a party to the original contract of sale.

I think that these rulings, which have been approved by the Appellate Division, extend the liability of the vendor of a manufactured article further than any case which has yet received the sanction of this court. It has heretofore been held in this state that the liability of the vendor of a manufactured article for negligence arising out of the existence of defects therein does not extend to strangers injured in consequence of such defects but is confined to the immediate vendee. The exceptions to this general rule which have thus far been recognized in New York are cases in which the article sold was of such a character that danger to life or limb was involved in the ordinary use thereof; in other words, where the article sold was inherently dangerous. As has already been pointed out, the learned trial judge instructed the jury that an automobile is not an inherently dangerous vehicle.

The late Chief Justice Cooley of Michigan, one of the most learned and accurate of American law writers, *397states the general rule thus: “The general rule is that a contractor, manufacturer, vendor or furnisher of an article is not liable to third parties who have no contractual relations with him for negligence in the construction, manufacture or sale of such article.” (2 Cooley on Torts [3d ed.], I486.)

The leading English authority in support of this rule, to which all the later cases on the same subject refer, is Winterbottom v. Wright (10 Meeson & Welsby, 109), which was an action by the driver of a stage coach against a contractor who had agreed with the postmaster-general to provide and keep the vehicle in repair for the purpose of conveying the royal mail over a prescribed route. The coach broke down and upset, injuring the driver, who sought to recover against the contractor on account of its defective construction. The Court of Exchequer denied him any right of recovery on the ground that there was no privity of contract between the parties, the agreement having been made with the postmaster-general alone. i£If the plaintiff can sue,” said Lord Abinger, the Chief Baron, “every passenger or even any person passing along the road, who was injured by the upsetting of the coach, might bring a similar action. Unless we confine the operation of such contracts as this to the parties who enter into them, the most absurd and outrageous consequences, to which I can see no limit, would ensue.”

The doctrine of that decision was recognized as the law of this state by the leading New York case of Thomas v. Winchester (6 N. Y. 397, 408), which, however, involved an exception to the general rule. There the defendant, who was a dealer in medicines, sold to a druggist a quantity of belladonna, which is a deadly poison, negligently labeled as extract of dandelion. The druggist in good faith used the poison in filling a prescription calling for the harmless dandelion extract and the plaintiff for whom the prescription was put up was poisoned by the *398belladonna. This court held that the original vendor was liable for the injuries suffered by the patient. Chief Judge Ruggles, who delivered the opinion of the court, distinguished between an act of negligence imminently dangerous to the lives of others and one that is not so, saying: “If A. build a wagon and sell it to B., who sells it to C. and C. hires it to D., who in consequence of the gross negligence of A. in building the wagon is overturned and injured, D. cannot recover damages against A., the builder. A.’s obligation to build the wagon faithfully, arises solely out of his contract with B. The public have nothing to do with it. * * * So, for the same reason, if a horse be defectively shod by a smith, and a person hiring the horse from the owner is thrown and injured in consequence of the smith’s negligence in shoeing; the smith is not liable for the injury. ”

In Torgeson v. Schultz (192 N. Y. 156, 159) the defendant was the vendor of bottles of aerated water which were charged under high pressure and likely to explode unless used with precaution when exposed to sudden changes of temperature. The plaintiff, who was a servant of the purchaser, was injured by the explosion of one of these bottles. There was evidence tending to show that it had not been properly tested in order to insure users against such accidents. We held that the defendant corporation was liable notwithstanding the absence of any contract relation between it and the plaintiff “under the doctrine of Thomas v. Winchester (supra), and similar cases based upon the duty of the vendor of an article dangerous in its nature, or likely to become so in the course of the ordinary usage to be contemplated by the vendor, either to exercise due care to warn users of the danger or to take reasonable care to present the article sold from proving dangerous when subjected only to customary usage.” The character of the exception to the general rule limiting liability for negligence to the original parties to the contract of sale, was still more clearly stated by Judge *399Hiscock, writing for the court in Statler v. Ray Manufacturing Co. (195 N. Y. 478, 482), where he said that “in the case of an article of an inherently dangerous nature, a manufacturer may become liable for a negligent construction which, when added to the inherent character of the appliance, makes it imminently dangerous, and causes or contributes to a resulting injury not necessarily incident to the use of such an article if properly constructed, but naturally following from a defective construction.” In that case the injuries were inflicted by the explosion of a battery of steam-driven coffee urns, constituting an appliance liable to become dangerous in the course of ordinary usage.

The case of Devlin v. Smith (89 N. Y. 470) is cited as an authority in conflict with the view that the liability of the manufacturer and vendor extends to third parties only when the article manufactured and sold is inherently dangerous. In that case the builder of a scaffold ninety feet high which was erected for the purpose of enabling painters to stand upon it, was held to be liable to the administratrix of a painter who fell therefrom and was killed, being at the time in the employ of the person for whom the scaffold was built. It is said that the scaffold if properly constructed was not inherently dangerous; and hence that this decision affirms the existence of liability in the case of an article not dangerous in itself but made so only in consequence of negligent construction. Whatever logical force there may be in this view it seems to me clear from the language of Judge Rapallo, who wrote the opinion of the court, that the scaffold was deemed to be an inherently dangerous structure; and that the case was decided as it was because the court entertained that view. Otherwise he would hardly have said, as he did, that the circumstances seemed to bring the case fairly within the principle of Thomas v. Winchester.

I do not see how we can uphold the judgment in the *400present case without overruling what has been so often said by this court and other courts of like authority in reference to the absence of any liability for negligence on the part of the original vendor of an ordinary carriage to any one except his immediate vendee. The absence of such liability was the very point actually decided in the English case of Winterbottom v. Wright (supra), and the illustration quoted from the opinion of Chief Judge Ruggles in Thomas v. Winchester (supra) assumes that the law on the subject was so plain that the statement would be accepted almost as a matter of course. In the case at bar the defective wheel on an automobile moving only eight iniles an hour was not any more dangerous to the occupants of the car than a similarly defective wheel would be to the occupants of a carriage drawn by a horse at the same speed; and yet unless the courts have been all wrong on this question up to the present time there would be no liability to strangers to the original sale in the case of the horse-drawn carriage.

The rule upon which, in my judgment, the determination of this case depends, and the recognized exceptions thereto, were discussed by Circuit Judge Sanborn of the United States Circuit Court of Appeals in the Eighth Circuit, in Huset v. J. I. Case Threshing Machine Co. (120 Fed. Rep. 865) in an opinion which reviews all the leading American and English decisions on the subject up to the time when it was rendered (1903). I have already discussed the leading New York cases, but as to the rest I feel that I can add nothing to the learning of that opinion or the cogency of its reasoning. I have examined the cases to which Judge Sanborn refers, but if I were to discuss them at length I should be forced merely to paraphrase his language, as a study of the authorities he cites has led me to the same conclusion; and the repetition of what has already been so well said would contribute nothing to the advantage of the bench, the bar or the individual litigants whose case is before us.

*401A few cases decided since his opinion was written, however, may he noticed. In Earl v. Lubbock (L. R. 1905 [1 K. B. Div.] 253) the Court of Appeal in 1904 considered and approved the propositions of law laid down by the Court of Exchequer in Winterbottom v. Wright (supra), declaring that the decision in that case, since the year 1842, had stood the test of repeated discussion. The master of the rolls approved the principles laid down by Lord Abinger as based upon sound reasoning; and all the members of the court agreed that his decision was a controlling authority which must he followed. That the Federal courts still adhere to the general rule, as I have stated it, appears by the decision of the Circuit Court of Appeals in the Second Circuit, in March, 1915, in the case of Cadillac Motor Car Co. v. Johnson (221 Fed. Rep. 801). That case, like this, was an action by a subvendee against a manufacturer of automobiles for negligence in failing to discover that one of its wheels was defective, the court holding that such an action could not be maintained. It is true there was a dissenting opinion in that case, but it was based chiefly upon the proposition that rules applicable to stage coaches are archaic when applied to automobiles and that if the law did not afford a remedy to strangers to the contract the law should he changed. It this be true, the change should he effected by the legislature and not by the courts. A perusal of the opinion in that case and in the Huset case will disclose how uniformly the courts throughout this country have adhered to the rule and how consistently they have refused to broaden the scope of the exceptions. I think we should adhere to it in the case at bar and, therefore, I vote for a reversal of this judgment.

Hiscock, Chase and Cuddeback, JJ., concur with Cardozo, J., and Hogan, J., concurs in result; Willard Bartlett, Ch. J., reads dissenting opinion; Pound, J., not voting.

Judgment affirmed.

*456GIBSON, C. J.

Plaintiff, a waitress in a restaurant, was injured when a bottle of Coca Cola broke in her hand. She alleged that defendant company, which had bottled and delivered the alleged defective bottle to her employer, was negligent in selling “bottles containing said beverage which on account of excessive pressure of gas or by reason of some defect in the bottle was dangerous . . . and likely to explode.” This appeal is from a judgment upon a jury verdict in favor of plaintiff.

Defendant’s driver delivered several cases of Coca Cola to the restaurant, placing them on the floor, one on top of the other, under and behind the counter, where they remained at least thirty-six hours. Immediately before the accident, plaintiff picked up the top case and set it upon a near-by ice cream cabinet in front of and about three feet from the refrigerator. She then proceeded to take the bottles from the case with her right hand, one at a time, and put them into the refrigerator. Plaintiff testified that after she had placed three bottles in the refrigerator and had moved the fourth bottle about eighteen inches from the ease “it exploded in my hand.” The bottle broke into two jagged pieces and inflicted a deep five-inch cut, severing blood vessels, nerves and muscles of the thumb and palm of the hand. Plaintiff further testified that when the bottle exploded, “It made a sound similar to an electric light bulb that would have dropped. It made a loud pop.” Plaintiff’s employer testified, “I was about twenty feet from where it actually happened and I heard the explosion.” A fellow employee, on the opposite side of the counter, testified that plaintiff “had the bottle, I should judge, waist high, and I know that it didn’t bang either the case or the door or another bottle . . . when it popped. It sounded just like a fruit jar would blow up. . . .” The witness further testified that the contents of the bottle “flew all over herself and myself and the walls and one thing and another.”

The top portion of the bottle, with the cap, remained in plaintiff’s hand, and the lower portion fell to the floor but did not break. The broken bottle was not produced at the trial, the pieces having been thrown away by an employee of the restaurant shortly after the accident. Plaintiff, however, described the broken pieces, and a diagram of the bottle was made showing the location of the “fracture line” where the bottle broke in two.

*457One of defendant’s drivers, called as a witness by plaintiff, testified that he had seen other bottles of Coca Cola in the past explode and had found broken bottles in the warehouse when he took the cases out, but that he did not know what made them blow up.

Plaintiff then rested her case, having announced to the court that being unable to show any specific acts of negligence she relied completely on the doctrine of res ipsa loquitur.

Defendant contends that the doctrine of res ipsa loquitur does not apply in this case, and that the evidence is insufficient to support the judgment.

Many jurisdictions have applied the doctrine in cases involving exploding bottles of carbonated beverages. (See Payne v. Rome Coca-Cola Bottling Co., 10 Ga.App. 762 [73 S.E. 1087] ; Stolle v. Anheuser-Busch, 307 Mo. 520 [271 S.W. 497, 39 A.L.R. 1001] ; Bradley v. Conway Springs Bottling Co., 154 Kan. 282 [118 P.2d 601] ; Ortego v. Nehi Bottling Works, 199 La. 599 [6 So.2d 677] ; MacPherson v. Canada Dry Ginger Ale, Inc., 129 N.J.L. 365 [29 A.2d 868]; Moeres v. Coca-Cola Bottling Co., 290 Mich. 567 [287 N.W. 922] ; Benkendorfer v. Garrett (Tex. Civ. App.), 143 S.W.2d 1020.) Other courts for varying reasons have refused to apply the doctrine in such cases. (See Gerber v. Faber, 54 Cal.App.2d 674 [129 P.2d 485] ; Loebig’s Guardian v. Coca-Cola Bottling Co., 259 Ky. 124 [81 S.W.2d 910]; Stewart v. Crystal Coca-Cola Bottling Co., 50 Ariz. 60 [68 P.2d 952]; Glaser v. Seitz, 35 Misc. 341 [71 N.Y.S. 942]; Luciano v. Morgan, 267 App. Div. 785 [45 N.Y.S.2d 502]; cf. Berkens v. Denver Coca-Cola Bottling Co., 109 Colo. 140 [122 P.2d 884]; Ruffin v. Coca Cola Bottling Co., 311 Mass. 514 [42 N.E.2d 259]; Slack v. Premier-Pabst Corporation, 40 Del. 97 [5 A.2d 516] ; Wheeler v. Laurel Bottling Works, 111 Miss. 442 [71 So. 743, L.R.A. 1916E 1074] ; Seven-Up Bottling Co. v. Gretes, _Va._[27 S.E.2d 925]; Dail v. Taylor, 151 N.C. 284 [66 S.E. 135, 28 L.R.A.N.S. 949].) It would serve no useful purpose to discuss the reasoning of the foregoing cases in detail, since the problem is whether under the facts shown in the instant case the conditions warranting application of the doctrine have been satisfied.

Res ipsa loquitur does not apply unless (1) defendant had exclusive control of the thing causing the injury and (2) the accident is of such a nature that it ordinarily *458would not occur in the absence of negligence by the defendant. (Honea v. City Dairy, Inc., 22 Cal.2d 614, 616-617 [140 P.2d 369], and authorities there cited; cf. Hinds v. Wheadon, 19 Cal.2d 458, 461 [121 P.2d 724]; Prosser on Torts [1941], 293-301.)

Many authorities state that the happening of the accident does not speak for itself where it took place some time after defendant had relinquished control of the instrumentality causing the injury. Under the more logical view, however, the doctrine may be applied upon the theory that defendant had control at the time of the alleged negligent act, although not at the time of the accident, provided plaintiff first proves that the condition of the instrumentality had not been changed after it left the defendant’s possession. (See cases collected in Honea v. City Dairy, Inc., 22 Cal.2d 614, 617-618 [140 P.2d 369].) As said in Dunn v. Hoffman Beverage Co., 126 N.J.L. 556 [20 A.2d 352, 354], “defendant is not charged with the duty of showing affirmatively that something happened to the bottle after it left its control or management; ... to get to the jury the plaintiff must show that there was due care during that period.” Plaintiff must also prove that she handled the bottle carefully. The reason for this prerequisite is set forth in Prosser on Torts, supra, at page 300, where the author states: “Allied to the condition of exclusive control in the defendant is that of absence of any action on the part of the plaintiff contributing to the accident. Its purpose, of course, is to eliminate the possibility that it was the plaintiff who was responsible. If the boiler of a locomotive explodes while the plaintiff engineer is operating it, the inference of his own negligence is at least as great as that of the defendant, and res ipsa loquitur will not apply until he has accounted for his own conduct.” (See, also, Olson v. Whitthorne (& Swan, 203 Cal. 206, 208-209 [263 P. 518, 58 A.L.R. 129].) It is not necessary, of course, that plaintiff eliminate every remote possibility of injury to the bottle after defendant lost control, and the requirement is satisfied if there is evidence permitting a reasonable inference that it was not accessible to extraneous harmful forces and that it was carefully handled by plaintiff or any third person who may have moved or touched it. (Cf. Prosser, supra, p. 300.) If such evidence is presented, the question becomes one for the trier of fact (see, e. g., *459 MacPherson v. Canada Dry Ginger Ale, Inc., 129 N.J.L. 365 [29 A.2d 868, 869]), and, accordingly, the issue should be submitted to the jury under proper instructions.

In the present case no instructions were requested or given on this phase of the case, although general instructions upon res ipsa loquitur were given. Defendant, however, has made no claim of error with reference thereto on this appeal. Upon an examination of the record, the evidence appears sufficient to support a reasonable inference that the bottle here involved was not damaged by any extraneous force after delivery to the restaurant by defendant. It follows, therefore, that the bottle was in some manner defective at the time defendant relinquished control, because sound and properly prepared bottles of carbonated liquids do not ordinarily explode when carefully handled.

The next question, then, is whether plaintiff may rely upon the doctrine of res ipsa loquitur to supply an inference that defendant’s negligence was responsible for the defective condition of the bottle at the time it was delivered to the restaurant. Under the general rules pertaining to the doctrine, as set forth above, it must appear that bottles of carbonated liquid are not ordinarily defective without negligence by the bottling company. In 1 Shearman and Redfield on Negligence (rev. ed. 1941), page 153, it is stated that: “The doctrine . . . requires evidence which shows at least the probability that a particular accident could not have occurred without legal wrong by the defendant.”

An explosion such as took place here might have been caused by an excessive internal pressure in a sound bottle, by a defect in the glass of a bottle containing a safe pressure, or by a combination of these two possible causes. The question is whether under the evidence there was a probability that defendant was negligent in any of these respects. If so, the doctrine of res ipsa loquitur applies.

The bottle was admittedly charged with gas under pressure, and the charging of the bottle was within the exclusive control of defendant. As it is a matter of common knowledge that an overcharge would not ordinarily result without negligence, it follows under the doctrine of res ipsa loquitur that if the bottle was in fact excessively charged an inference of defendant’s negligence would arise. If *460the explosion resulted from a defective bottle containing a safe pressure, the defendant would be liable if it negligently-failed to discover such flaw. If the defect were visible, an inference of negligence would arise from the failure of defendant to discover it. Where defects are discoverable, it may be assumed that they will not ordinarily escape detection if a reasonable inspection is made, and if such a defect is overlooked an inference arises that a proper inspection was not made. A difficult problem is presented where the defect is unknown and consequently might have been one not discoverable by a reasonable, practicable inspection. In the Honea case we refused to take judicial notice of the technical practices and information available to the bottling industry for finding defects which cannot be seen. In the present case, however, we are supplied with evidence of the standard methods used for testing bottles.

A chemical engineer for the Owens-Illinois Glass Company and its Pacific Coast subsidiary, maker of Coca Cola bottles, explained how glass is manufactured and the methods used in testing and inspecting bottles. He testified that his company is the largest manufacturer of glass containers in the United States, and that it uses the standard methods for testing bottles recommended by the glass containers association. A pressure test is made by taking a sample from each mold every three hours—approximately one out of every 600 bottles—and subjecting the sample to an internal pressure of 450 pounds per square inch, which is sustained for one minute. (The normal pressure in Coca Cola bottles is less than 50 pounds per square inch.) The sample bottles are also subjected to the standard thermal shock test. The witness stated that these tests are “pretty near” infallible.

It thus appears that there is available to the industry a commonly-used method of testing bottles for defects not apparent to the eye, which is almost infallible. Since Coca Cola bottles are subjected to these tests by the manufacturer, it is not likely that they contain defects when delivered to the bottler which are not discoverable by visual inspection. Both new and used bottles are filled and distributed by defendant. The used bottles are not again subjected to the tests referred to above, and it may be inferred that defects not discoverable by visual inspection do not develop in bottles after they are manufactured. Obviously, if such defects do *461occur in used bottles there is a duty upon the bottler to make appropriate tests before they are refilled, and if such tests are not commercially practicable the bottles should not be re-used. This would seem to be particularly true where a charged liquid is placed in the bottle. It follows that a defect which would make the bottle unsound could be discovered by reasonable and practicable tests.

Although it is not clear in this case whether the explosion was caused by an excessive charge or a defect in the glass, there is a sufficient showing that neither cause would ordinarily have been present if due care had been used. Further, defendant had exclusive control over both the charging and inspection of the bottles. Accordingly, all the requirements necessary to entitle plaintiff to rely on the doctrine of res ipsa loquitur to supply an inference of negligence are present.

It is true that defendant presented evidence tending to show that it exercised considerable precaution by carefully regulating and cheeking the pressure in the bottles and by making visual inspections for defects in the glass at several stages during the bottling process. It is well settled, however, that when a defendant produces evidence to rebut the inference of negligence which arises upon application of the doctrine of res ipsa loquitur, it is ordinarily a question of fact for the jury to determine whether the inference has been dispelled. (Druzanich v. Criley, 19 Cal.2d 439, 444 [122 P.2d 53]; Michener v. Hutton, 203 Cal. 604, 610 [265 P. 238, 59 A.L.R. 480].)

The judgment is affirmed.

Shenk, J., Curtis, J., Carter, J., and Sehauer, J., concurred.

TRAYNOR, J.

I concur in the judgment, but I believe the manufacturer’s negligence should no longer be singled out as the basis of a plaintiff’s right to recover in cases like the present one. In my opinion it should now be recognized that a manufacturer incurs an absolute liability when an article that he has placed on the market, knowing that it is to be used, without inspection, proves to have a defect that causes injury to human beings. McPherson v. Buick Motor Co., 217 382 [111 N.E. 1050], Ann.Cas. 1916C 440, L.R.A. 1916F, established the principle, recognized by this court, that irrespective of privity of contract, the manufacturer *462is responsible for an injury caused by such an article to any person who comes in lawful contact with it. (Sheward v. Virtue, 20 Cal.2d 410 [126 P.2d 345]; Kalash v. Los Angeles Ladder Co., 1 Cal.2d 229 [34 P.2d 481].) In these cases the source of the manufacturer’s liability was his negligence in the manufacturing process or in the inspection of component parts supplied by others. Even if there is no negligence, however, public policy demands that responsibility be fixed wherever it will most effectively reduce the hazards to life and health inherent in defective products that reach the market. It is evident that the manufacturer can anticipate some hazards not. Those who suffer injury from defective products are unprepared to meet its consequences. The cost of an injury and the loss of time or health may be an overwhelming misfortune to the person injured, and a needless one, for the risk of injury can be insured by the manufacturer and distributed among the public as a cost of doing business. It is to the public interest to discourage the marketing of products having defects that are a menace to the public. If such products nevertheless find their way into the market it is to the public interest to place the responsibility for whatever injury they may cause upon the manufacturer, who, even if he is not negligent in the manufacture of the product, is responsible for its reaching the market. However intermittently such injuries may occur and however haphazardly they may strike, the risk of their occurrence is a constant risk and a general one. Against such a risk there should be general and constant protection and the manufacturer is best situated to afford such protection.

The injury from a defective product does not become a matter of indifference because the defect arises from causes other than the negligence of the manufacturer, such as negligence of a submanufacturer of a component part whose defects could not be revealed by inspection (see Sheward v. Virtue, 20 Cal.2d 410 [126 P.2d 345]; O’Rourke v. Bay & Night Water Heater Co., Ltd., 31 Cal.App.2d 364 [88 P.2d 191]; Smith v. Peerless Glass Co., 259 N.Y. 292 [181 N.E. 576]), or unknown causes that even by the device of res ipsa loquitur cannot be classified as negligence of the manufacturer. The inference of negligence may be dispelled by an affirmative showing of proper care. If the evidence against the fact in*463ferred is “clear, positive, uncontradicted, and of such a nature that it cannot rationally be disbelieved, the court must instruct the jury that the nonexistence of the fact has bee established as a matter of law.” (Blank v. Coffin, 20 Cal.2 457, 461 [126 P.2d 868].) An injured person, however, is not ordinarily in a position to refute such evidence or identify the cause of the defect, for he can hardly be familiar with the manufacturing process as the manufacturer himself is. In leaving it to the jury to decide whether the inference has been dispelled, regardless of the evidence against it, the negligence rule approaches the rule of strict liability. It is needlessly circuitous to make negligence the basis of recovery and impose what is in reality liability without negligence. If public policy demands that a manufacturer of goods be responsible for their quality regardless of negligence there is no reason not to fix that responsibility openly.

In the case of foodstuffs, the public policy of the state is formulated in a criminal statute. Section 26510 of the Health and Safety Code prohibits the manufacturing, preparing, compounding, packing, selling, offering for sale, or keeping for sale, or advertising within the state, of any adulterated food. Section 26470 declares that food is adulterated when “it has been produced, prepared, packed, or held under insanitary conditions whereby it may have been rendered diseased, unwholesome or injurious to health.” The statute imposes criminal liability not only if the food is adulterated, but if its container, which may be a bottle (§ 26451), has any deleterious substance (§26470 (6)), or renders the product injurious to health. (§26470 (4)). The criminal liability under the statute attaches without proof of fault, so thaf the manufacturer is under the duty of ascertaining whether an article manufactured by him is safe. (People v. Schwartz, 28 Cal.App.2d Supp. 775 [70 P.2d 1017].) Statutes of this kind result in a strict liability of the manufacturer in tort to the member of the public injured. (See eases cited in Prosser, Torts, p. 693, note 69.)

The statute may well be applicable to a bottle whose defects cause it to explode. In any event it is significant that the statute imposes criminal liability without fault, reflecting the public policy of protecting the public from dangerous products placed on the market, irrespective of negligence in their manufacture. While the Legislature imposes criminal lia*464bility only with regard to food products and their containers, there are many other sources of danger. It is to the public interest to prevent injury to the public from any defective goods by the imposition of civil liability generally.

The retailer, even though not equipped to test a product, is under an absolute liability to his customer, for the implied warranties of fitness for proposed use and merchantable quality include a warranty of safety of the product. (Goetten Owl Drug Co., 6 Cal.2d 683 [59 P.2d 142] ; Mix v. Ingersoll Candy Co., 6 Cal.2d 674 [59 P.2d 144] ; Gindraux v. Maurice Mercantile Co., 4 Cal.2d 206 [47 P.2d 708]; Jensen v. Berris, Cal.App.2d 537 [88 P.2d 220]; Ryan v. Progressive Grocery Stores, 255 N.Y. 388 [175 N.E. 105; 74 A.L.R. 339] ; Race v. Krum, 222 N.Y. 410 [118 N.E. 853, L.R.A. 1918F 1172].) This warranty is not necessarily a contractual one (Chamberlain Co. v. Allis-Chalmers etc. Co., 51 Cal.App.2d 520, 524 [125 P.2d 113]; see 1 Williston on Sales, 2d ed., §§ 197-201), for public policy requires that the buyer be insured at the seller’s expense against injury. (Race v. Krum, supra; Ryan v. Progressive Grocery Stores, supra; Chapman v. Roggenkamp, 182 Ill.App. 117, 121; Ward v. Great Atlantic & Pacific Tea Co., 231 Mass. 90, 94 [120 N.E. 225, 5 A.L.R. 242]; see Prosser, The Implied Warranty of Merchantable Quality, 27 Minn.L.Rev. 117, 124; Brown, The Liability of Retail Dealers For Defective Food Products, 23 Minn.L.Rev. 585.) The courts recognize, however, that the retailer cannot bear the burden of this warranty, and allow him to recoup any losses by means of the warranty of safety attending the wholesaler’s or manufacturer’s sale to him. (Ward v. Great Atlantic & Pacific Tea Co., supra; see Waite, Retail Responsibility and Judicial Law Making, 34 Mich.L.Rev. 494, 509.) Such a procedure, however, is needlessly circuitous and engenders wasteful litigation. Much would be gained if the injured person could base his action directly on the manufacturer’s warranty.

The liability of the manufacturer to an immediate buyer injured by a defective product follows without proof of negligence from the implied warranty of safety attending the sale. Ordinarily, however, the immediate buyer is a dealer who does not intend to use the product himself, and if the warranty of safety is to serve the purpose of protecting health and safety it must give rights to others than the dealer. In the words *465of Judge Cardozo in the McPherson ease: “The dealer was indeed the one person of whom it might be said with some approach to certainty that by him the car would not be used.” Yet, the defendant would have us say that he was the one person whom it was under a legal duty to protect. The law does not lead us to so inconsequent a solution.” While the defendant’s negligence in the McPherson case made it unnecesssary for the court to base liability on warranty, Judge Cardozo’s reasoning recognized the injured person as the real party in interest and effectively disposed of the theory that the liability of the manufacturer incurred by his warranty should apply only to the immediate purchaser. It thus paves the way for a standard of liability that would make the manufacturer guarantee the safety of his product even when there is no negligence.

This court and many others have extended protection according to such a standard to consumers of food products, taking the view that the right of a consumer injured by unwholesome food does not depend “upon the intricacies law of sales” and that the warranty of the manufacturer to the consumer in absence of privity of contract rests on public policy. (Klein v. Duchess Sandwich Co., Ltd., 14 Cal.2d 282 [93 P.2d 799]; Ketterer v. Armour & Co., 200 F. 321, 322, 323 [160 C.C.A. 111, L.R.A. 1918D 798]; Decker & Sons v. Capps, 139 Tex. 609 [164 S.W.2d 828, 142 A.L.R. 1479]; see Perkins, Unwholesome Food As A Source of Liability, 5 Iowa L.Bull. 6, 86.) Dangers to life and health inhere in other consumers’ goods that are defective and there is no reason to differentiate them from the dangers of defective food products. (See Bohlen, Studies in Torts, Basis of Affirmative Obligations, American Cases Upon The Liability of Manufacturers and Vendors of Personal Property, 109, 135; Llewellyn, On Warranty of Quality and Society, 36 Col.L.Rev. 699, 704, note 14; Prosser, Torts, p. 692.)

In the food products cases the courts have resorted to various fictions to rationalize the extension of the manufacturer’s warranty to the consumer: that a warranty runs with the chattel; that the cause of action of the dealer is assigned to the consumer; that the consumer is a third party beneficiary of the manufacturer’s contract with the dealer. They have also held the manufacturer liable on a mere fiction of negli*466gence: “Practically he must know it [the product] is fit, or bear the consequences if it proves destructive.” (Parks v. C. C. Yost Pie Co., 93 Kan. 334 [144 P. 202, L.R.A. 1915C 179]; see Jeanblane, Manufacturer’s Liability to Persons Other Than Their Immediate Vendees, 24 Va.L.Rev. 134.) Such fictions are not necessary to fix the manufacturer’s liability under a warranty if the warranty is severed from the contract of sale between the dealer and the consumer and based on the law of torts (Decker & Sons v. Capps, supra; Prosser, Torts, p. 689) as a strict liability. (See Green v. General Petroleum Corp., 205 Cal. 328 [270 P. 952, 60 A.L.R. 475]; McGrath v. Basich Bros. Const. Co., 7 Cal.App.2d 573, 46 P.2d 981]; Prosser, Nuisance Without Fault, 20 Tex.L. Rev., 399, 403; Feezer, Capacity To Bear The Loss As A Factor In The Decision Of Certain Types of Tort Cases, 78 U. of Pa.L.Rev. 805, 79 U. of Pa.L.Rev. 742; Carpenter, The Doctrine of Green v. General Petroleum Corp., 5 So.Cal.L.Rev. 263, 271; Pound, The End of Law As Developed In Legal Rules And Doctrines, 27 Harv.L.Rev. 195, 233.) Warranties are not necessarily rights arising under a contract. An action on a warranty “was, in its origin, a pure action of tort,” and only late in the historical development of warranties was an action in assumpsit allowed. (Ames, The History of Assumpsit, 2 Harv.L.Rev. 1, 8; 4 Williston on Contracts (1936) § 970.) “And it is still generally possible where a distinction of procedure is observed between actions of tort and of contract to frame the declaration for breach of warranty in tort.” (Williston, loc. cit.; see Prosser, Warranty On Merchantable Quality, 27 Minn.L.Rev. 117, 118.) On the basis of the tort character of an action on a warranty, recovery has been allowed for wrongful death as it could not be in an action for breach of contract. (Greco v. S. S. Kresge Co., 277 N.Y. 26 [12 N.E.2d 577, 115 A.L.R. 1020]; see Schlick v. New York Dugan Bros., 175 Mise. 182 [22 N.Y.S.2d 238]; Prosser, op. cit., p. 119.) As the court said in Greco v. S. S. Kresge Co., supra, “Though the action may be brought solely for the breach of the implied warranty, the breach is a wrongful act, a default, and, in its essential nature, a tort.” Even a seller’s express warranty can arise from a noncontractual affirmation inducing a person to purchase the goods. (Chamberlain Co. v. Allis-Chalmers etc. Co., 51 Cal.App.2d 520 [125 P.2d 113].) “As an actual agreement to contract is not essential, the obli*467gation of a seller in such a case is one imposed by law as distinguished from one voluntarily assumed. It may be called an obligation either on a quasi-contract or quasi-tort, because remedies appropriate to contract and also to tort are applicable.” (1 Williston on Sales, 2d ed. §197; see Ballantine, Classification of Obligations, 15 Ill.L.Rev. 310, 325.)

As handicrafts have been replaced by mass production with its great markets and transportation facilities, the close relationship between the producer and consumer of a product has been altered. Manufacturing processes, frequently valuable secrets, are ordinarily either inaccessible to or beyond the ken of the general public. The consumer no longer has means or skill enough to investigate for himself the soundness of a product, even when it is not contained in a sealed package, and his erstwhile vigilance has been lulled by the steady efforts of manufacturers to build up confidence by advertising and marketing devices such as trade-marks. (See Thomas v. Winchester, 6 N.Y. 397 [57 Am.Dec. 455]; Baxter v. Ford Motor Co., 168 Wash; 456 [12 P.2d 409, 15 P.2d 1118, 88 A.L.R. 521]; Crist v. Art Metal Works, 230 App.Div. 114 [243 N.Y.S. 496], affirmed 255 N.Y. 624 [175 N.E. 341]; see also Handler, False and Misleading Advertising, 39 Yale L.J. 22; Rogers, Good Will, Trade-Marks and Unfair Trading (1914) ch. VI, A Study of The Consumer, p. 65 et seq.; Williston, Liability For Honest Misrepresentations As Deceit, Negligence Or Warranty, 42 Harv.L.Rev. 733; 18 Cornell L.Q. 445.) Consumers, no longer approach products warily but accept them on faith, relying on the reputation of the manufacturer or the trade mark. (See Max Factor & Co. v. Kunsman, 5 Cal.2d 446, 463 [55 P.2d 177]; Old Dearborn etc. Co. v. Seagram-Distillers Corp., 299 U.S, 183 [57 S.Ct. 139, 81 L.Ed. 109, 106 A.L.R. 1476]; Schechter, The Rational Basis of Trade Mark Protection, 40 Harv.L.Rev. 813, 818.) Manufacturers have sought to justify that faith by increasingly high standards of inspection and a readiness to make good on defective products by way of replacements and refunds. (See Bogert and Pink, Business Practices Regarding Warranties In The Sale Of Goods, 25 Ill.L.Rev. 400.) The manufacturer’s obligation to the consumer must keep pace with the changing relationship between them; it cannot be escaped because the marketing of a product has become so complicated as to require one or more *468intermediaries. Certainly there is greater reason to impose liability on the manufacturer than on the retailer who is but a conduit of a product that he is not himself able to test. (See Soule, Consumer Protection, 4 Encyclopedia of The Social Sciences, 282; Feezer, Manufacturer’s Liability For Injuries Caused By His Products: Defective Automobiles, 37 Mich.L. Rev. 1; Llewellyn, Cases And Materials on Sales, 340 et seq.)

The manufacturer’s liability should, of course, be defined in terms of the safety of the product in normal and proper use, and should not extend to injuries that cannot be traced to the product as it reached the market.

Appellant’s petition for a rehearing was denied August 3, 1944. Edmonds, J., voted for a rehearing.

DAVIS, Circuit Judge:

These are wrongful death actions, brought under diversity jurisdiction, in which defendant Jeppesen & Company (Jeppesen) appeals from jury verdicts and judgments rendered against it in the United States District Court for the District of Connecticut, including challenges to the trial court’s denial of its motions for a new trial and for judgment notwithstanding the verdict. We affirm.

I

On August 31,1975, Captain Willard Vernon Wahlund, a Braniff International pilot with approximately seven thousand hours of flight experience, departed from an airport at Charleston, West Virginia, on a flight to Danbury, Connecticut. Wahlund was off-duty and was piloting his own Beechcraft Sierra; the plane was equipped with a King 214 receiver for instrument flight purposes and carried ample fuel. The flight was one leg of a trip which originated in Dallas and which was slated to end in Danbury. Wahlund’s father and Erik, Wahlund’s six-year-old son, accompanied him.

Wahlund carried navigational charts, produced by Jeppesen, on board the Beechcraft Sierra. Braniff purchased those charts from Jeppesen in Colorado for Wahlund’s use (as well as comparable charts for Braniff’s other pilots). Jeppesen furnishes its customers generally with three types of charts. Enroute charts display large geographic areas — several states or so in size— and flight paths or airways. Area charts portray geographic areas around major metropolitan areas and their correlative airways. Approach charts (or plates) depict runways as well as the vertical and horizontal coefficients of the approach paths to those runways.

While enroute from Charleston to Dan-bury, • Wahlund decided — for a now unknown reason — to land at the Martinsburg, West Virginia airport.1 Wahlund had the area chart for the Washington area, but he did not have an approach plate for the Martinsburg airport (as Braniff did not service that facility). The Martinsburg airport did not then possess a full instrument landing system; it was equipped with a localizer beam, but not a glidescope beam.2 The *673Jeppesen chart for the Washington area, however, portrayed Martinsburg airport as containing a full instrument landing system; this was done through use of the designation “ILS” on the chart, adjacent to the designation of the Martinsburg airport.

There was substantial evidence from which the jury could also conclude the following: At 3:25 p.m. (EST) on the afternoon of August 81, Wahlund informed air traffic control that he wished to land at the Martinsburg airport and asked for vectors. Wahlund received them and was transferred to Dulles Arrival Radar. At 3:34, Dulles Arrival requested Wahlund to advise them as to what type of landing he desired. Wahlund replied, “[W]e’d like a uh ILS if . ... ” At this point, unknown to Wahlund, his transmission was cut off by another pilot’s transmission. Nonetheless, Dulles Arrival responded, “Roger [i.e., “we have received your full message”] ... expect the uh localizer back course runway two six approach ...,” and transferred control of Wahlund’s plane to Dulles Departure Radar. Dulles Departure advised the pilot that visibility at Martinsburg was three miles, with light rain, fog and wind — adequate ILS conditions for a pilot of Wahlund’s experience. At 3:36, Wahlund asked for a “front course ILS” instead of a “localizer back course” approach. Dulles Departure responded affirmatively and gave Wahlund appropriate vectors; shortly thereafter, Dulles Departure gave Wahlund the option of landing on “runway zero eight,” which he accepted.

At 3:40 p.m., Dulles Departure told Wahlund that he was “cleared for the front course localizer” and not to “descend below two thousand nine hundred feet until crossing Gerrard inbound.” Gerrard is a point in the sky — fixed by the intersection of the localizer beam and a cross beam — which served as the final approach fix for a localizer landing at Martinsburg. Gerrard appears on approach plates for Martinsburg, but not on the Washington area chart.

Wahlund’s immediate response to Dulles Departure’s altitude instructions is now unintelligible. Shortly thereafter, Wahlund reported that he was “established,” or fixed, on the localizer. At 3:42, Dulles Departure advised, “[R]adar service terminated five miles from Gerrard you can contact Martinsburg radio .... ” Four minutes later, Wahlund contacted Martinsburg and reported that he was located at Gerrard. Martinsburg gave Wahlund the weather and runway advisory. This was the last contact with the aircraft. - Shortly thereafter the plane crashed, killing the three occupants.

The National Transportation Safety Board report on the crash noted that the plane wreckage was located at 1400' on the west side of a 1600' ridge near Gerrard. The plane apparently struck the ridge at a normal descent angle in virtually exact alignment with the runway and the localizer beam. All instruments on board were destroyed by a fire which apparently occurred on impact. No evidence of aircraft malfunction or pilot infirmity was discovered.

Following the tragedy, appellee Saloomey, as administratrix of Wahlund’s estate, and appellee Halstead, as administrator of Erik’s, filed separate actions in the district court. Saloomey’s complaint named Jeppesen and the United States as defendants; Halstead’s complaint named Jeppesen, the United States and Wahlund as defendants.3 Both actions, insofar as they named Jeppesen as a defendant, were consolidated for jury trial. The counts against Jeppesen included negligence, breach of warranty, and strict products liability.

After a lengthy trial and two days of deliberation, the jury returned verdicts against Jeppesen on behalf of both plain*674tiffs. The jury also answered 18 separate interrogatories; those responses showed that the jury believed Jeppesen to be liable on all theories of liability urged by the plaintiffs — negligence, breach of implied and express warranties, and strict products liability. In a separate damages trial, the jury awarded Wahlund’s estate $1,500,000 and Erik’s estate $5,000. The court then denied Jeppesen’s motions for a new trial and for judgment notwithstanding the verdict. This appeal followed.

II

Jeppesen’s first argument is that the district court erred in applying the substantive law of Colorado. See Halstead v. United States, 535 F.Supp. 782 (D.Conn.1982) (Halstead II). Although we consider the question close, we cannot agree with appellant that the substantive law of West Virginia should govern.

Federal jurisdiction is based here on diversity of citizenship. See 28 U.S.C. § 1332(a)(1) (1976). We must therefore employ the choice-of-law rule of the forum state, Connecticut, to ascertain the controlling substantive law. Day & Zimmermann, Inc. v. Challoner, 423 U.S. 3, 4, 96 S.Ct. 167, 168, 46 L.Ed.2d 3 (1975); Klaxon Co. v. Stentor Electric Manufacturing Co., 313 U.S. 487, 496, 61 S.Ct. 1020, 1021, 85 L.Ed. 1477 (1941). Connecticut courts traditionally have held — in automobile tort cases— that the rule of lex loci delicti applies. See, e.g., Gibson v. Fullin, 172 Conn. 407, 411, 374 A.2d 1061, 1064 (1977); Menczer v. Menczer, 160 Conn. 563, 564, 280 A.2d 875, 876 (1971) (per curiam); Landers v. Landers, 153 Conn. 303, 304, 216 A.2d 183, 184 (1966); Chasse v. Albert, 147 Conn. 680, 683, 166 A.2d 148, 150 (1960). Jeppesen relies on that rule.

However, the Supreme Court of Connecticut has never had occasion to apply the rule of lex loci delicti to a wrongful death action arising from an aviation accident. We think that, in the absence of a direct ruling by the Connecticut courts on aviation accidents, the choice-of-law rule a Connecticut court would use in these circumstances is uncertain, and a federal diversity court has the duty to ascertain what rule the Connecticut court would establish. See West v. American Telephone & Telegraph Co., 311 U.S. 223, 237, 61 S.Ct. 179, 183, 85 L.Ed. 139 (1940); Bailey Employment System, Inc. v. Hahn, 655 F.2d 473, 477 (2d Cir.1981);4 see also Meredith v. Winter Haven, 320 U.S. 228, 237, 64 S.Ct. 7, 12, 88 L.Ed. 9 (1943).

In our view, the district court should not be faulted in predicting, on the facts here, that a Connecticut court would choose to follow the “most significant relationship” test embodied in Restatement (Second) of Conflict of Laws § 145 (1971). In Gibson v. Fullin, supra, the Connecticut Supreme Court indicated that its endorsement of the lex loci delicti rule was made in “motor vehicle eases.” 172 Conn. at 411, 374 A.2d at 1064. The court recognized the judicial trend toward adoption of the “most significant relationship” test for selecting substantive law, but noted that Gibson presented “no compelling reason to abandon the traditional role.” Id.

The current case presents very good reasons to replace the rule of lex loci delicti by the Restatement (Second) approach,5 for *675aviation accidents.6 In contrast to automotive travel, aviation accidents — especially those occurring in interstate air travel— more frequently pose situations in which the place of actual injury is wholly fortuitous and unimportant. See In re Air Crash Disaster Near Chicago, 644 F.2d 594, 615 (7th Cir.), cert. denied, 454 U.S. 878, 102 5. Ct. 358, 70 L.Ed.2d 187 (1981); Kilberg v. Northeast Airlines, Inc., 9 N.Y.2d 34, 39, 211 N.Y.S.2d 133, 135, 172 N.E.2d 526, 527 (1961). That is true in this case. Appellees’ decedents were domiciled in Connecticut, and appellant is a Colorado corporation engaged there in map-making; they would not normally expect the relationship between them to be governed by West Virginia law. The happenstance of an aviation accident in West Virginia as a result of appellant’s map-making in Colorado should not alter the normal expectations of the parties. See Restatement (Second) of Conflict of Laws § 145 comment e (1971); id. § 6.

Other considerations support that conclusion. Connecticut, through its legislature and courts, has indicated strong interest in fully compensating its domiciliaries when they file meritorious wrongful death and strict products liability actions.7 Those interests would not be served by the application of West Virginia law; a statute in force in West Virginia at the time of the accident limited wrongful death recoveries to an amount far below that permitted by Connecticut or Colorado. The then applicable West Virginia statute limited wrongful death recoveries to $10,000 for the death itself, $100,000 for monetary losses suffered by decedent’s dependents, and expenses. See West Virginia Code § 55-7-6 (1965), cited in Halstead II, supra, 535 F.Supp. at 785 n. 1. This limit has since been repealed.8 In addition, West Virginia did not recognize actions in strict products liability at that time, although Connecticut and Colorado did. See Halstead II, supra, 535 F.Supp. at 786. But West Virginia has no interest in having its limitations enforced by a Connecticut court to protect a foreign corporation such as appellant, which does little or no business in West Virginia, against non-West Virginia plaintiffs seeking remuneration for their losses. Cf. De-Fourneaux v. Sturm, Ruger & Co., 503 F.Supp. 2, 4 (D.Conn.), aff’d, 639 F.2d 768 (2d Cir.1980) (mem.), cert. denied, 451 U.S. 908, 101 S.Ct. 1975, 68 L.Ed.2d 295 (1981).

Moreover, the Restatement (Second) “most significant relationship” approach is in accord, for interstate aviation cases, with the policy expressed by the Supreme Court of Connecticut in Gibson. There, the court emphasized that choice-of-law rules “should not only be simple and easy to determine and apply, but should also lead to predictable and desirable results.” Gibson, supra, 172 Conn. at 412, 374 A.2d at 1064. Invocation of the lex loci delicti rule in aviation-generated wrongful death actions often produces unpredictable and undesirable results; the locale of injury may well have no connection to other relevant factors such as the parties’ domiciles or residences, their places of incorporation, their principal *676places of business, or the location of the wrongful conduct.

The Restatement (Second) approach recognizes this inherent unpredictability and ameliorates it by including the place of injury in a choice-of-law analysis which also examines the relationship and other contacts between the parties. See Restatement (Second) of Conflict of Laws § 145(2) (1971). The principles underlying this analysis harmonize with the concerns voiced by the court in Gibson; they include certainty and predictability of result as well as ease in determination and application of substantive law. Compare Gibson, supra, 172 Conn. at 412, 374 A.2d at 1064, with Restatement (Second) of Conflict of Laws § 6(2) (1971). Moreover, the district judge’s interpretation of Connecticut choice-of-law precedent is entitled to special deference because of his experience and familiarity with Connecticut law. See Bernhardt v. Polygraphic Co. of America, 350 U.S. 198, 204, 76 S.Ct. 273, 276, 100 L.Ed. 199 (1956).

Ill

We likewise accept the district court’s ruling, under the significant-factors test, that Colorado substantive law governs, rather than West Virginia’s. See Halstead II, supra, 535 F.Supp. at 788-89. Colorado possesses the most significant relationship to the parties and the accident. Ticking off the contacts enunciated in § 145 of the Restatement (Second), see supra note 4, we see that Jeppesen is a Colorado corporation; Colorado is also Jeppesen’s principal place of business. In addition, Braniff purchased from Jeppesen in Colorado the navigational charts that were to be used by Wahlund and other pilots; appellant cannot contend that the application of Colorado substantive law to actions involving alleged defects in those charts was unforeseeable, unpredictable, or fortuitous.

As noted in Parts I and II, supra, West Virginia’s only significant relation or contact with the parties and issues was the circumstance of the crash itself. This fact, standing alone, is insufficient to trigger the application of West Virginia substantive law to the measure of recovery and putative strict products liability.

Though both decedents were domiciled in Connecticut, that is not enough to justify the application of Connecticut law, given the strong impact of the Colorado contacts. We agree with the trial court, see Halstead II, supra, 535 F.Supp. at 789, that Connecticut’s interest in amply compensating its residents who file wrongful death actions is adequately served in this instance by Colorado law.

IV

Jeppesen also says that the trial court erroneously determined that the navigational charts are products, rather than services. If the charts should be characterized as the provision of services, then appellees’ claims in strict products liability could not stand under Colorado law. See Halstead II, supra, 535 F.Supp. at 789.

Colorado recognizes actions in strict products liability as defined by Restatement (Second) of Torts § 402A (1965). See Hiigel v. General Motors Corp., 190 Colo. 57, 63, 544 P.2d 983, 987 (1975) (en banc); Bradford v. Bendix-Westinghouse Automotive Air Brake Co., 33 Colo.App. 99, 106, 517 P.2d 406, 411 (1973). But the Colorado courts have never decided whether navigational charts are products for the purposes of § 402A. The Ninth Circuit has assumed, without discussion, that appellant’s portrayal of Federal Aviation Administration flight data on its charts is a “product” for strict liability purposes. See Aetna Casualty & Surety Co. v. Jeppesen & Co., 642 F.2d 339, 342-43 (9th Cir.1981).

We believe that the trial court did not err in classifying appellant’s charts as products. The charts, as produced by Jeppesen and supplied to Wahlund by Braniff, reached Wahlund without any individual tailoring or substantial change in contents — they were simply mass-produced. The comments to § 402A, supra, envision strict liability against sellers of such items in these circumstances. By publishing and *677selling the charts, Jeppesen undertook a special responsibility, as seller, to insure that consumers will not be injured by the use of the charts; Jeppesen is entitled — and encouraged — to treat the burden of accidental injury as a cost of production to be covered by liability insurance. See Restatement (Second) of Torts § 402A comment c (1965). This special responsibility lies upon Jeppesen in its role as designer, seller and manufacturer. See Pust v. Union Supply Co., 38 Colo.App. 435, 439, 561 P.2d 355, 359 (1976), aff’d, 196 Colo. 162, 583 P.2d 276 (1978) (en banc); Restatement (Second) of Torts § 402A comment f (1965).

Appellant’s position that its navigational charts provide no more than a service ignores the mass-production aspect of the charts. Though a “product” may not include mere provision of architectural design plans or any similar form of data supplied under individually-tailored service arrangements, see Gibson v. Sonstrom, 2 Conn.L. Trib. No. 103, at 3 (Super.Ct. Hartford Cty. 1976), the mass production and marketing of these charts requires Jeppesen to bear the costs of accidents that are proximately caused by defects in the charts. See Halstead II, supra, 535 F.Supp. at 791; K-Mart Corp. v. Midcon Realty Group, 489 F.Supp. 813, 816-19 & 818 n. 7 (D.Conn.1980); Restatement (Second) of Torts § 402A comments c, f (1965).

V

Jeppesen also levies an assault on the jury’s findings of negligence and proximate causation — urging us to overturn the district court’s refusal to grant appellant’s motions for judgment notwithstanding the verdict and for a new trial.

We take up the former contention first. In reviewing that argument, we must scan the evidence in the light most favorable to appellees, and draw all reasonable inferences from the evidence on their behalf. See Simblest v. Maynard, 427 F.2d 1, 4 (2d Cir.1970). In other words, the motion for judgment notwithstanding the verdict should have been granted only if there was but one conclusion reasonable people could have reached — in appellant’s favor. Id.; see Billiar v. Minnestoa Mining & Manufacturing Co., 623 F.2d 240, 246 (2d Cir.1980); Bigelow v. Agway, Inc., 506 F.2d 551, 554 (2d Cir.1974). Neither the trial court nor we can assess the credibility of the witnesses or weigh the evidence if a rational jury could hold for appellees. Schulz v. Pennsylvania Railroad, 350 U.S. 523, 526, 76 S.Ct. 608, 610, 100 L.Ed. 668 (1956); Simblest v. Maynard, supra, 427 F.2d at 4. On that strict principle we cannot say that the district court erred in denying appellant judgment notwithstanding the verdict.

First, there was certainly adequate evidence to sustain the jury’s specific findings that Jeppesen’s area chart was defective in designating Martinsburg as having a full instrument landing system — through the designation ILS — and that Jeppesen was negligent in the manufacture or inspection of that chart. See supra Part I and infra note 10. Those facts are not seriously disputed. The jury also found that Willard Wahlund was negligent in the operation of the aircraft and the conduct of the flight.9 The critical question, then, is whether there was sufficient evidence to support the jury’s additional findings that Wahlund’s negligence was not the proximate cause of the crash but that appellant’s negligence and defective chart were. The trial judge noted, in his findings in appellees’ actions against the United States, that Wahlund was negligent and that the air traffic controllers’ negligence was not a proximate cause of the crash. See supra note 3. Nonetheless, the court held that the jury verdicts in this action against Jeppesen had to stand. That ruling was not improper merely because the judge would apparently have reached a different conclusion on appellant’s causal responsibility for the crash. See Sentilles v. Inter-Carribean Corp., 361 *678U.S. 107, 110, 80 S.Ct. 173, 175, 4 L.Ed.2d 142 (1959).

We are constrained to agree with the court below in its evaluation of the jury’s verdict. Initially, it was reasonable for the jury to conclude that Wahlund would not have attempted to land at Martinsburg but for the erroneous “ILS” designation on the Jeppesen-produced Washington area chart.10 Wahlund told the air traffic controllers during two different radio exchanges that he desired a full, precision ILS approach, see supra Part I; the chart designation could well have led him to believe that one was available. When Wahlund first requested an ILS landing, his transmission was interrupted, yet the controllers responded “Roger,” thereby apparently leading Wahlund to believe that his entire transmission was received and an ILS approach was feasible. Moments later, Wahlund received a second affirmative response when he requested a “front course ILS” approach instead of the proffered “localizer back course” landing. At no time did the controllers specifically inform him that the glidescope component for an ILS was unavailable at Martinsburg. The jury could conclude that Wahlund’s actions were consistent with those of a pilot who believed that he was engaged in a proper ILS approach procedure.

The jury also possessed enough evidence to conclude that, once Wahlund was established on the localizer, he would not have known that glidescope instrumentation to go along with the localizer was unavailable. As we noted earlier (see supra note 1), Wahlund had previously informed air traffie control that his King 214 receiver was not performing properly, and some trial testimony indicated that the receiver had a history of occasional malfunctions. The jury could have determined that the warning flag on the receiver’s glidescope needle was inoperative during the descent. Alternatively, the jury could have believed that Wahlund attributed the presence of the warning flag on the glidescope needle during the fatal approach to a weak signal; because Wahlund was fixed on the localizer, he could have assumed that the flag would disappear as he approached the source of the glidescope beam. Though there was conflicting evidence on these points, there was affirmative evidence from which the jury could have reached these conclusions.11

Contrary to appellant's position, we think, too, that the jury could have found that the aircraft was not in fact five miles from Gerrard intersection when Dulles Departure terminated radar service and contact with Wahlund. Three different witnesses for appellees testified that, in their opinion, Wahlund was further west of Gerrard at that time. The testimony of the controller, Gackowski, who gave Wahlund the location, indicated that radar reception in that area was not always accurate.12 The jury could have inferred from this testimony that Wahlund’s actual position, as reported by the controller, was incorrect (but that Wahlund did not know that fact). Nor was it unreasonable for the jury to find that Wahlund, incorrectly (but not negligently) believing he was five miles from Gerrard, could later report (incorrectly) to *679Martinsburg radio that he was passing Gerrard — having timed and monitored the assumed five-mile distance. Conversely, the jury could find, as it specifically did, that the pilot did not violate the Federal Air Regulations in continuing on the course he believed he had been given. As shown by the difference between the trial judge’s findings (in the action against the United States) and the jury’s findings (in the suit against Jeppesen), there could obviously be a parting of the ways on the ultimate issue of proximate causation,13 but we cannot overturn the district judge’s own conclusion that the difference was a rational one.

We hold, finally, that the district court did not abuse its discretion in declining to grant Jeppesen’s motion for a new trial. See Bevevino v. Saydjari, 574 F.2d 676, 683-84 (2d Cir.1978); Diapulse Corp. v. Birtcher Corp., 362 F.2d 736, 744 (2d Cir.), cert. dismissed, 385 U.S. 801, 87 S.Ct. 9, 17 L.Ed.2d 48 (1966). In his denial of the new trial, the judge carefully considered the jury verdicts and his own position. Appellees were not compelled to demonstrate proximate causation with either absolute or mathematical certainty. Nor was the court required to order a new trial merely because of its disagreement with the fact finding of the jury, see Bevevino, supra, 574 F.2d at 685. Our review of the evidence leads us to conclude that the court below was not wrong in holding that the jury’s verdict was not “seriously erroneous.” See id. at 684-85.14

Affirmed.

CARD AMONE, Circuit Judge,

dissenting:

I .agree with all parts of the majority opinion except section II. In section II my colleagues conclude that the district court properly held that the applicable Connecticut choice of law rule in this case was the “most significant relationship” test, rather than lex loci delicti. Because in my view Connecticut’s long-standing choice of law rule for tort actions, lex loci delicti, should have been used, I respectfully dissent.

Less than two years ago this Court held that “Connecticut’s choice of law is governed by traditional principles, and, in the case of tort actions, the rule of lex loci delicti still obtains.” Bailey Employment System, Inc. v. Hahn, 655 F.2d 473, 476 (2d Cir.1981). No post-Bailey development in Connecticut law supports the majority’s^ deviation from our holding in Bailey. Nor can the majority opinion point to any Connecticut state court decision in which the “most significant relationship” test was the appropriate choice of law rule in a tort case governed by Connecticut law.

Instead, the majority argues that the district court was free to predict that the “most significant relationship” test should apply because the Supreme Court of Connecticut has never had occasion to apply lex loci delicti to a tort action specifically arising from an aviation accident. While it is true that the Connecticut Supreme Court has not addressed the choice of law issue in a tort case resulting from an aviation mishap, as opposed to some other type of transportation accident, nothing in the language-of the Connecticut decisions applying lex loci delicti in accident cases warrants the strained distinction now relied upon, see, e.g., Gibson v. Fullin, 172 Conn. 407, 374 A.2d 1061, 1064 (1977); Menczer v. Menczer, 160 Conn. 563, 280 A.2d 875, 876 (1971); Bissonnette v. Bissonnette, 145 Conn. 733, 142 A.2d 527, 528 (1958).

There is no real uncertainty as to what choice-of-law rule Connecticut’s Supreme Court would apply in this case. Only six years ago in Gibson v. Fullin, the Connecticut Supreme Court unanimously refused to deviate from its traditional rule of lex loci delicti, see 374 A.2d at 1064, under circum*680stances at least as compelling as, if not more compelling than, those present here. In Gibson a Connecticut plaintiff brought a tort action against a Connecticut defendant for damages arising out of an automobile accident which had occurred in Florida. Connecticut’s highest court held that the law of Florida, the situs of the accident, was controlling under the doctrine of lex loci delicti. Significantly, this decision was made despite the following facts: all of the relevant parties were residents of the forum state, Connecticut, whose tort law was not applied; the location of the accident was entirely fortuitous; application of Florida’s guest statute totally barred the plaintiff’s claim. By contrast, in the instant case not all of the relevant parties are domiciliaries of the forum state, and application of lex loci delicti would not totally bar plaintiffs’ action. Instead it would simply limit the legal theories upon which plaintiffs could rely and the amount of recoverable damages. Given that the Connecticut Supreme Court expressly declined to abandon lex loci delicti in Gibson, it seems clear that they would not desert it here.

The majority’s further argument — that Connecticut’s inclination to fully compensate tort victims militates in favor of the “most substantial interest” test in this case — overlooks the fact that the purported inclination did not carry the day in Gibson. Additionally, while the majority opinion strongly asserts that an “interest” analysis lends itself to easy and predictable application, a policy consideration voiced in Gibson, see id. at 1064, lex loci delicti would typically be easier to apply than an “interest” analysis in an aircraft accident case.

It is true, of course, that many jurisdictions have shifted to one form or another of “interest” analysis. Nonetheless, Connecticut’s Supreme Court has noted that trend and resolutely declined to follow it. Because the district court was obligated to follow Connecticut’s clear choice of law rule in this diversity action, see Klaxon Co. v. Stentor Electric Manufacturing Co., 313 U.S. 487, 61 S.Ct. 1020, 85 L.Ed. 1477 (1941), it was error to apply an “interest” analysis. Accordingly, I vote to reverse and remand this case to the district court.

SNEED, Circuit Judge:

Plaintiffs are mushroom enthusiasts who became severely ill from picking and eating mushrooms after relying on information in The Encyclopedia of Mushrooms, a book published by the defendant. Plaintiffs sued the publisher and sought damages under various theories. The district court granted summary judgment for the defendant. We affirm.

*1034 I.

FACTS AND PROCEEDINGS BELOW

The Encyclopedia of Mushrooms is a reference guide containing information on the habitat, collection, and cooking of mushrooms. It was written by two British authors and originally published by a British publishing company. Defendant Putnam, an American book publisher, purchased copies of the book from the British publisher and distributed the finished product in the United States. Putnam neither wrote nor edited the book.

Plaintiffs purchased the book to help them collect and eat wild mushrooms. In 1988, plaintiffs went mushroom hunting and relied on the descriptions in the book in determining which mushrooms were safe to eat. After cooking and eating their harvest, plaintiffs became critically ill. Both have required liver transplants.

Plaintiffs allege that the book contained erroneous and misleading information concerning the identification of the most deadly species of mushrooms. In their suit against the book publisher, plaintiffs allege liability based on products liability, breach of warranty, negligence, negligent misrepresentation, and false representations. Defendant moved for summary judgment asserting that plaintiffs’ claims failed as a matter of law because 1) the information contained in a book is not a product for the purposes of strict liability under products liability law; and 2) defendant is not liable under any remaining theories because a publisher does not have a duty to investigate the accuracy of the text it publishes. The district court granted summary judgment for the defendant. Plaintiffs appeal. We affirm. 1

II.

DISCUSSION

A book containing Shakespeare’s sonnets consists of two parts, the material and print therein, and the ideas and expression thereof. The first may be a product, but the second is not. The latter, were Shakespeare alive, would be governed by copyright laws; the laws of libel, to the extent consistent with the First Amendment; and the laws of misrepresentation, negligent misrepresentation, negligence, and mistake. These doctrines applicable to the second part are aimed at the delicate issues that arise with respect to intangibles such as ideas and expression. Products liability law is geared to the tangible world.

A. Products Liability

The language of products liability law reflects its focus on tangible items. In describing the scope of products liability law, the Restatement (Second) of Torts lists examples of items that are covered. 2 All of these are tangible items, such as tires, automobiles, and insecticides. 3 The American Law Institute clearly was concerned with including all physical items but gave no indication that the doctrine should be expanded beyond that area.

The purposes served by products liability law also are focused on the tangible world and do not take into consideration the unique characteristics of ideas and expression. Under products liability law, strict *1035 liability is imposed on the theory that “[t]he costs of damaging events due to defectively dangerous products can best be borne by the enterprisers who make and sell these products.” Prosser & Keeton on The Law of Torts, § 98, at 692-93 (W. Keeton ed. 5th ed. 1984). Strict liability principles have been adopted to further the “cause of accident prevention ... [by] the elimination of the necessity of proving negligence.” Id. at 693. Additionally, because of the difficulty of establishing fault or negligence in products liability cases, strict liability is the appropriate legal theory to hold manufacturers liable for defective products. Id. Thus, the seller is subject to liability “even though he has exercised all possible care in the preparation and sale of the product.” Restatement § 402A comment a. It is not a question of fault but simply a determination of how society wishes to assess certain costs that arise from the creation and distribution of products in a complex technological society in which the consumer thereof is unable to protect himself against certain product defects.

Although there is always some appeal to the involuntary spreading of costs of injuries in any area, the costs in any comprehensive cost/benefit analysis would be quite different were strict liability concepts applied to words and ideas. We place a high priority on the unfettered exchange of ideas. We accept the risk that words and ideas have wings we cannot clip and which carry them we know not where. The threat of liability without fault (financial responsibility for our words and ideas in the absence of fault or a special undertaking or responsibility) could seriously inhibit those who wish to share thoughts and theories. As a New York court commented, with the specter of strict liability, “[w]ould any author wish to be exposed ... for writing on a topic which might result in physical injury? e.g. How to cut trees; How to keep bees?” Walter v. Bauer, 109 Misc.2d 189, 191, 439 N.Y.S.2d 821, 823 (Sup.Ct.1981) (student injured doing science project described in textbook; court held that the book was not a product for purposes of products liability law), aff'd in part & rev’d in part on other grounds, 88 A.D.2d 787, 451 N.Y.S.2d 533 (1982). One might add: “Would anyone undertake to guide by ideas expressed in words either a discrete group, a nation, or humanity in general?”

Strict liability principles even when applied to products are not without their costs. Innovation may be inhibited. We tolerate these losses. They are much less disturbing than the prospect that we might be deprived of the latest ideas and theories.

Plaintiffs suggest, however, that our fears would be groundless were strict liability rules applied only to books that give instruction on how to accomplish a physical activity and that are intended to be used as part of an activity that is inherently dangerous. We find such a limitation illusory. Ideas are often intimately linked with proposed action, and it would be difficult to draw such a bright line. While “How To” books are a special genre, we decline to attempt to draw a line that puts “How To Live A Good Life” books beyond the reach of strict liability while leaving “How To Exercise Properly” books within its reach.

Plaintiffs’ argument is stronger when they assert that The Encyclopedia of Mushrooms should be analogized to aeronautical charts. Several jurisdictions have held that charts which graphically depict geographic features or instrument approach information for airplanes are “products” for the purpose of products liability law. See Brocklesby v. United States, 767 F.2d 1288, 1294-95 (9th Cir.1985) (applying Restatement for the purpose of California law), cert. denied, 474 U.S. 1101, 106 S.Ct. 882, 88 L.Ed.2d 918 (1986); Saloomey v. Jeppesen & Co., 707 F.2d 671, 676-77 (2d Cir.1983) (applying Restatement for the purpose of Colorado Law); Aetna Casualty & Surety Co. v. Jeppesen & Co., 642 F.2d 339, 342-43 (9th Cir.1981) (applying Nevada law); Fluor Corp. v. Jeppesen & Co., 170 Cal.App.3d 468, 475, 216 Cal.Rptr. 68, 71 (1985) (applying California law). Plaintiffs suggest that The Encyclopedia of Mushrooms can be compared to aeronautical charts because both items contain representations of natural features and both are intended to be used while engag *1036 ing in a hazardous activity. We are not persuaded.

Aeronautical charts are highly technical tools. They are graphic depictions of technical, mechanical data. The best analogy to an aeronautical chart is a compass. Both may be used to guide an individual who is engaged in an activity requiring certain knowledge of natural features. Computer software that fails to yield the result for which it was designed may be another. In contrast, The Encyclopedia of Mushrooms is like a book on how to use a compass or an aeronautical chart. The chart itself is like a physical “product” while the “How to Use” book is pure thought and expression. 4

Given these considerations, we decline to expand products liability law to embrace tóle ideas and expression in a book. 5 We know of no court that has chosen the path to which the plaintiffs point. 6

B. The Remaining Theories

As discussed above, plaintiffs must look to the doctrines of copyright, libel, misrepresentation, negligent misrepresentation, negligence, and mistake to form the basis of a claim against the defendant publisher. Unless it is assumed that the publisher is a guarantor of the accuracy of an author’s statements of fact, plaintiffs have made no case under any of these theories other than possibly negligence. Guided by the First Amendment and the values embodied therein, we decline to extend liability under this theory to the ideas and expression contained in a book.

*1037 In order for negligence to be actionable, there must be a legal duty to exercise due care. 6 B. Witkin, Summary of California Law, Torts § 732 (9th ed. 1988). The plaintiffs urge this court that the publisher had a duty to investigate the accuracy of The Encyclopedia of Mushrooms’ contents. We conclude that the defendants have no duty to investigate the accuracy of the contents of the books it publishes. A publisher may of course assume such a burden, 7 but there is nothing inherent in the role of publisher or the surrounding legal doctrines to suggest that such a duty should be imposed on publishers. Indeed the cases uniformly refuse to impose such a duty. 8 Were we tempted to create this duty, the gentle tug of the First Amendment and the values embodied therein would remind us of the social costs. 9

Finally, plaintiffs ask us to find that a publisher should be required to give a warning 1) that the information in the book is not complete and that the consumer may not fully rely on it or 2) that this publisher has not investigated the text and cannot guarantee its accuracy. With respect to the first, a publisher would not know what warnings, if any, were required without engaging in a detailed analysis of the factual contents of the book. This would force the publisher to do exactly what we have said he has no duty to do — that is, independently investigate the accuracy of the text. We will not introduce a duty we *1038 have just rejected by renaming it a “mere” warning label. With respect to the second, such a warning is unnecessary given that no publisher has a duty as a guarantor.

For the reasons outlined above, the decision of the district court is AFFIRMED.

OPINION OF THE COURT

Read, J.

The United States Court of Appeals for the Second Circuit has asked us whether a company that sold one of its used machines (itself purchased used) to a different company can be held strictly liable for a workplace accident involving that machine, which occurred 16 years later. In light of our precedents and the policy considerations underlying strict products liability, we answer the certified question in the negative.

I.

On March 9, 2002, plaintiff Mario Miguel Jaramillo seriously injured his right hand when he caught it between two rollers of an industrial Flexo Folder Gluer machine (FFG) that he was operating while working for his employer, Universal Glenwood Packaging Products Corporation, a manufacturer of corrugated containers, at its plant in Yonkers, New York. Jaramillo had worked for Glenwood for about five years and had operated the FFG uneventfully many times before.

In 1986, Glenwood purchased the FFG as a used machine from defendant Weyerhaeuser Company, an international forest products company with its principal place of business in the state of Washington. As part of its business, Weyerhaeuser operates numerous plants where cardboard boxes are fabricated from corrugated cardboard sheets; FFGs are used to make these boxes.

Weyerhaeuser’s Investment Recovery Business (IRB), the division through which the company generally disposes of its obsolete or unneeded equipment, marketed the FFG sold to Glenwood. The IRB distributes quarterly catalogs of items for sale, advertises in trade journals, telemarkets, and conducts market research on potential buyers and dealers of used equipment. The United States District Court for the Southern District of New York found that there was no genuine dispute of fact that the FFG was sold to Glenwood in “as-is, where-is” condition, and the Second Circuit agreed (Jaramillo v Weyerhaeuser Co., 2007 WL 194011, *1 n 7, 2007 US Dist LEXIS 7413, *5, n 7 [SD NY, Jan. 24, 2007], on appeal 536 F3d 140, 144 n 1 [2d Cir 2008]).

The IRB grossed somewhere between $7.5 million and $8.5 million in 1986, the year Glenwood bought the FFG involved in *185 Jaramillo’s accident. This accounted for approximately .15% of Weyerhaeuser’s net sales of about $5.65 billion for that year. There is no evidence in the record to show Weyerhaeuser’s sales of used FFGs before 1986, but Jaramillo contends that Weyerhaeuser sold more than 60 FFGs from 1986 to 2006.

Older FFGs such as the one sold by Weyerhaeuser to Glen-wood have “open architecture,” or open spaces between operating sections that a worker can enter while the machine is running, which is what Jaramillo did when he injured his right hand. By contrast, “closed architecture” machines, as the name implies, do not permit entry. Open architecture machines may be retrofitted with a safety device to shut them down automatically if an open space is accessed; for example, an interlocking device or a safety mat, which causes the machine to stop when stepped on, may be installed. Weyerhaeuser has furnished some of its open architecture FFGs with interlocking devices, or has asked the manufacturer to do this. Further, the Second Circuit noted

“evidence that Weyerhaeuser owns patents related to technology used in FFGs, and that the company maintains relationships with FFG manufacturers. [Weyerhaeuser] has occasionally made recommendations to manufacturers about how to improve FFG design, including with regard to safety features. When it has detected safety issues with an FFG, the company has also sometimes suggested that the manufacturer install a new safety mechanism in the machine in question” (536 F3d at 143).

The FFG involved in Jaramillo’s accident was originally manufactured in about 1964 by S&S Manufacturing, a Brooklyn-based company that went bankrupt around 1986; it was sold new to the General Foods company. Weyerhaeuser purchased the FFG from General Foods in 1971 for roughly $36,500, and placed the machine in its Lynchburg, Virginia box plant. Weyerhaeuser added new parts to the FFG over the years and replaced worn ones, at a total cost of about $282,000, but did not change the safety mechanisms installed by S&S in the original manufacture. The machine was disassembled in Virginia by a Glenwood employee, and transported to Yonkers, where it was reassembled at the Glenwood plant.

Jaramillo filed a complaint in Supreme Court against Weyerhaeuser, seeking damages for his personal injuries. Jaramillo’s *186 claim against Weyerhaeuser sounded principally in strict products liability. He argued that by 1986 open architecture FFGs were defective if not equipped with a safety device to shut off operation in the event a person or a foreign object, such as a human extremity, was sensed in the machine’s open spaces, and that Weyerhaeuser did not add any such device to the FFG that it sold to Glenwood that year.

On March 7, 2003, Jaramillo’s suit was removed to federal court. On May 22, 2006, Weyerhaeuser sought summary judgment on the basis that it was a casual seller of FFGs and therefore was not strictly liable under New York law for any defects. On June 20, 2006, Jaramillo cross-moved for summary judgment on the issue of whether Weyerhaeuser was an ordinary seller of FFGs, potentially subject to strict products liability.

On January 24, 2007, the District Court Judge granted Weyerhaeuser’s motion for summary judgment, denied Jaramillo’s cross motion, and dismissed the complaint. The Judge noted that “[t]he New York Court of Appeals has held that strict products liability is imposed on manufacturers and sellers who engage in product sales in the ordinary course of their business” (Jaramillo, 2007 WL 194011 at *3, 2007 US Dist LEXIS 7413 at *10, citing Sprung v MTR Ravensburg, 99 NY2d 468, 473 [2003]), but has limited “the duty of a casual or occasional seller . . . [to] warn[ing] the person to whom the product is supplied of known defects that are not obvious or readily discernible” (2007 WL 194011 at *3, 2007 US Dist LEXIS 7413 at *11-12, quoting Sukljian v Ross & Son Co., 69 NY2d 89, 97 [1986] [internal quotation marks omitted]). “This is so because unlike the manufacturer or the seller of a product in the normal course of business, the occasional seller is not part of the regular commercial network for that product” (2007 WL 194011 at *3, 2007 US Dist LEXIS 7413 at *12, quoting Sukljian, 69 NY2d at 97 [internal quotation marks omitted; citations omitted by District Court]).

Acknowledging that there was no controlling authority from our Court requiring a ruling in his favor, Jaramillo “urge[d]” the District Court Judge “to adopt dicta in Sukljian and find that under these facts defendant [was] a regular seller of the machine” (2007 WL 194011 at *4, 2007 US Dist LEXIS 7413 at *13). The Judge “decline[d] plaintiff’s invitation” for three reasons: “First, . . . the facts . . . [were] not. . . meaningfully distinguishable from those that the Sukljian court found supportive of a casual seller finding. Second, a policy analysis *187 supported] a casual seller holding. Third, on the facts of this case, there [was] no cogent reason to impose strict liability” (id.).

Jaramillo appealed the District Court’s order. In an opinion and order issued August 1, 2008, the Second Circuit discussed the current state of strict products liability law in New York at length. Citing Sukljian, the court noted that New York courts make a distinction between “ ‘ordinary’ or ‘regular’ sellers of a product—those who sell the product on a regular basis, through the ordinary course of business,” who are “strictly liable for injuries caused by any manufacturing, design, or warning defect”; and “ ‘casual’ or ‘occasional’ sellers—those who sell the product in sporadic transactions that are incidental to their businesses,” whose “liability . . . extends only insofar as it fails ‘to warn the person to whom the product is supplied of known defects that are not obvious or readily discernible’ ” (Jaramillo, 536 F3d at 145). The court observed that “whether the doctrine of strict products liability applies to regular sellers of used goods is an open question in New York” (id., citing Stiles v Batavia Atomic Horseshoes, 81 NY2d 950, 951 [1993]).

The Second Circuit recited the “two principal reasons” that we have given “for imposing strict products liability on ordinary sellers”:

“The first is that those who sell products in the normal course of business, by reason of their continuing relationships with manufacturers, are most often in a position to exert pressure for the improved safety of products and can recover increased costs within their commercial dealings, or through contribution or indemnification in litigation . . . The second is that those who market products as a regular part of their businesses may be said to have assumed a special responsibility to the public, which has come to expect them to stand behind their goods” (Jaramillo, 536 F3d at 145-146, quoting Sukljian, 69 NY2d at 95 [internal quotation marks omitted]).

By way of contrast, the occasional seller “has neither the opportunity, nor the incentive, nor the protection of the manufacturer or seller who puts that product into the stream of commerce as a normal part of its business, and the public consumer does not have the same expectation when it buys from such a seller” (id. at 146, quoting Sukljian, 69 NY2d at 96 [internal quotation marks omitted]).

*188 Commenting that “[w]hile the reasons behind the ordinary seller rule and occasional seller exception [were] clear, where to draw the line between ordinary and occasional sellers [was] not,” the court next examined “[t]he principal case on which Jaramillo relie[d] to argue that Weyerhaeuser [was] an ordinary seller of used FFGs”—the decision by the United States Court of Appeals for the Fifth Circuit in Galindo v Precision Am. Corp. (754 F2d 1212 [5th Cir 1985]) (Jaramillo, 536 F3d at 146). In Galindo, the Fifth Circuit suggested that a business selling its obsolete assets could be held strictly liable for defects in those products under Texas law where certain hypothetical circumstances existed: specifically, a policy of replacing the equipment at all of its facilities at regular intervals; a division or department devoted to selling the retired equipment; widespread advertising; and substantial revenues from this activity (see Galindo, 754 F2d at 1219). The Second Circuit also reviewed our decision in Sukljian, where we discussed the Galindo hypothetical (see Sukljian, 69 NY2d at 96-97).

Concluding that

“resolution of the summary judgment motions turn[ed] principally on . . . whether the New York courts would recognize the Galindo hypothetical as a case where strict liability should be imposed, and whether the facts of this case [were] sufficiently similar to that hypothetical such that Weyerhaeuser should be deemed an ordinary seller of used FFGs,” the Second Circuit certified the following question to us: “Construing the evidence in the light most favorable to Jaramillo, is Weyerhaeuser Company a ‘regular seller’ of used Flexo Folder Gluers such that it can be held strictly liable under New York law?” (Jaramillo, 536 F3d at 147, 149.)

II.

We have long held that “ [m] anufacturers of defective products may be held strictly liable for injury caused by their products” (Sukljian, 69 NY2d at 94, citing Voss v Black & Decker Mfg. Co., 59 NY2d 102, 106 [1983] [additional citations omitted]). “Imposition of this onerous liability rests largely on considerations of public policy” (Sukljian, 69 NY2d at 94-95). Unlike manufacturer strict liability, however, which applies broadly, “not every seller is subject to strict liability” (id. at 95 [emphasis added]). As we explained in Sukljian-.

*189 “The policy considerations that have been advanced to justify the imposition of strict liability on manufacturers and sellers in the normal course of business obviously lack applicability in the case of a party who is not engaged in the sale of the product in issue as a regular part of its business. (See, Restatement [Second] of Torts § 402 A, comment f, concluding: ‘This Section is also not intended to apply to sales of the stock of merchants out of the usual course of business, such as execution sales, bankruptcy sales, bulk sales, and the like.’) The casual or occasional seller of a product does not undertake the special responsibility for public safety assumed by those in the business of regularly supplying those products, nor is there the corollary element of forced reliance on that undertaking by purchasers of such goods. As a practical matter, the occasional seller has neither the opportunity, nor the incentive, nor the protection of the manufacturer or seller who puts that product into the stream of commerce as a normal part of its business, and the public consumer does not have the same expectation when it buys from such a seller. We recognized, for example, in Gobhai v KLM Royal Dutch Airlines (57 NY2d 839 [1982], affg 85 AD2d 566 [1st Dept 1981]) . . . that where distribution of an allegedly defective product is incidental to defendant’s regular business the principles of strict products liability have no relevance” (69 NY2d at 95-96).

In Sukljian, an employee of Ardex Corporation injured his hand while cleaning the rollers of a high-speed, three-roll grinding mill manufactured and sold by Charles Ross & Son Company Inc. to General Electric Company (GE) in January 1962. In June 1973, more than 11 years after its purchase of the mill, GE decided the machine was no longer needed and included it as part of a June 1973 sale of surplus property where several hundred lots were offered. There was evidence that GE held two or three sales of surplus equipment a year, but no evidence that the company derived any profit therefrom. The terms of the June 1973 sale were “As Is, Where Is,” and it was unclear how well-publicized the sale was—the only evidence of publicity was a snippet from an internal GE memorandum, which stated in its entirety “will be taken to advertise the sale through the various medias available, such as newspapers, trade magazines, *190 trade papers, etc.” (Sukljian, 69 NY2d at 93.) The mill was sold by GE for $35 and subsequently resold multiple times until it ended up in the hands of Ardex, where the accident victim worked. The accident occurred in 1978, five years after the mill was sold as surplus by GE; the victim’s father brought suit against GE on a theory of strict products liability.

Considering these facts, we specifically held that a “corporation that sold a machine previously used in its own production as surplus property was not liable to remote purchasers ... in strict products liability . . . for injuries allegedly resulting from a defect in the machine, and should have been awarded summary judgment dismissing the complaint” (id. at 92). Despite the facts that GE regularly held sales of surplus equipment, that these sales were quite extensive (although not especially lucrative), and that they were apparently publicized (although the amount of publicity was in dispute), we flatly rejected the “assertion] that there is a question of fact whether [GE] was in reality in the regular business of a second-hand equipment dealer” (id. at 96), an assertion explicitly relying on Galindo.

In Galindo itself, the court was faced with a strict products liability claim against the Georgia-Pacific Corporation, which operated more than 240 sawmills and plants and regularly sold its used sawmill equipment to dealers. The plaintiff in Galindo was severely injured while operating a sawmill trimmer that Georgia-Pacific sold used to someone who sold it to a company that, in turn, sold it to the plaintiff’s employer. The Fifth Circuit “d[id] not believe that it necessarily follow[ed]” under Texas law “that getting rid of the depreciated equipment used for production of the goods and services that constitute the business of the seller can never constitute a business in which the seller is engaged” (Galindo, 754 F2d at 1219 [internal quotation marks omitted]), and hypothesized a particular situation in which the policies underlying strict products liability might exist in this context. In particular, the Fifth Circuit

“[s]uppose[d], for example, that Georgia-Pacific [had] a policy of purchasing new equipment at all of its sawmills every five years. Suppose further that Georgia-Pacific [had] a division or department devoted to selling the used equipment, that the department advertise[d] the availability of equipment on a widespread basis, and that it realize[d] substantial revenues from this activity. Clearly, on *191 such facts, Georgia-Pacific would not be exempted from [strict products] liability simply because the equipment [was] used” (id.).

In Sukljian, we distinguished Galindo on the facts. We also explicitly declined to recognize its reasoning, stating that

“[e]ven if we were to recognize that liability might be imposed in a case such as hypothesized in Galindo . . . [t]here is no basis in the undisputed facts before us to support an inference that [GE], in incidentally disposing of the mill at a surplus equipment sale in the manner it did, undertook any special responsibility to the public for product safety, or was perceived by the public as having done so, or was better able to spread any loss caused by such defective products, making it more equitable to impose such loss on the seller. While we recognize that more than the single or isolated sale of a jam jar, a pound of sugar or an automobile described in the Restatement (Restatement [Second] of Torts § 402 A, comment f) is at issue here, still the principle is the same whether an individual, or an airline, or a huge corporation disposing of its surplus equipment in an occasional sale is involved: there is no basis in public policy for the imposition of strict liability” (Sukljian, 69 NY2d at 96-97).

Seven years later in Stiles, we reaffirmed Sukljian. In Stiles, the plaintiff was injured while operating a punch press owned by his employer, which had purchased the press in used condition from Batavia. Batavia bought five punch presses at auction, kept two and resold the other three, including the one sold to the plaintiffs employer (see Stiles v Batavia Atomic Horseshoes, 174 AD2d 287, 289 [4th Dept 1992]).

The jury awarded damages to the plaintiff, “concluding] that Batavia was a regular seller of used goods and thus, under the court’s instructions, accountable to plaintiff under the theory of strict products liability” (Stiles, 81 NY2d at 951). We found the trial court’s affirmed finding of fact that defendant “engaged in sales of equipment as a ‘regular part of its business’ ” to be unsupported by the record (id.), despite evidence, cited in the Appellate Division’s decision, that “during the two years Batavia was in business, Batavia bought and resold used industrial machinery on at least two other occasions, one before and the other after the auction at which Batavia bought the punch press *192 in question,” and “that on each occasion Batavia realized a substantial profit on the sale of the used machinery” (Stiles, 174 AD2d at 290).

We noted, however, that “Batavia’s business was the manufacture and sale of horseshoes,” and that on the three occasions when it had engaged in the purchase and resale of industrial equipment, “it purchased [the] used equipment in lots intending to keep much of the machinery for its own purposes”; indeed, “[t]he press sold to [the plaintiffs employer] was never unloaded at Batavia, but instead [was] sold directly off the truck that had transported it from an equipment auction and was in the possession of Batavia at its facility for approximately one hour before [the plaintiffs employer] took delivery of it” (Stiles, 81 NY2d at 951). Because we concluded that “this evidence fail[ed] as a matter of law to support a finding that Batavia was a regular seller of used goods,” we did “not reach the question reserved in Sukljian and submitted by the parties to [the] appeal, i.e., whether the doctrine of strict products liability applies to regular sellers of used goods” (id.).

This case is controlled by Sukljian and the policy considerations underlying our holding in that case. The “onerous” burden of strict liability is only imposed on “certain sellers” because of “continuing relationships with manufacturers” and a “special responsibility to the public, which has come to expect [these sellers] to stand behind their goods” (Sukljian, 69 NY2d at 94-95). The second of these policy goals is clearly absent here: buyers of Weyerhaeuser’s used (third-hand, in fact) equipment at irregularly-scheduled “as is, where is” surplus sales cannot reasonably “expect [Weyerhaeuser] to stand behind [someone else’s] goods.” As to the first policy goal, while Weyerhaeuser may have had a closer relationship with FFG manufacturers than a customer would have with a supplier of run-of-the-mine equipment not unique to its particular industry, this relationship was still general in nature and even more attenuated with respect to the FFGs that Weyerhaeuser sold as surplus. Indeed, the FFG involved in Jaramillo’s accident was actually bought used by Weyerhaeuser from a third party rather than new from an FFG manufacturer. Simply put, there is no reason to believe that imposing strict liability on Weyerhaeuser’s sales of its scrap, used FFGs would create any measurable “pressure for the improved safety of products” on FFG manufacturers. Indeed, the most likely effect would be exactly what the District Court predicted: Weyerhaeuser would stop *193 selling its used machinery, thus depriving small businesses of the ability to purchase otherwise unaffordable equipment.

While there may be some imaginable future case in which the facts justify imposition of strict products liability on a seller of used goods, this case is not—just as Sukljian and Stiles were not—that case. In sum, ‘‘[c]onstruing the evidence in the light most favorable to Jaramillo,” Weyerhaeuser Company is not a “ ‘regular seller’ of used Flexo Folder Gluers such that it can be held strictly liable under New York law” (Jaramillo, 536 F3d at 149).

Accordingly, the certified question should be answered in the negative.

Judges Ciparick, Graffeo, Smith, Pigott and Jones concur; Chief Judge Lippman taking no part.

Following certification of a question by the United States Court of Appeals for the Second Circuit and acceptance of the question by this Court pursuant to section 500.27 of the Rules of Practice of the Court of Appeals (22 NYCRR 500.27), and after hearing argument by counsel for the parties and consideration of the briefs and the record submitted, certified question answered in the negative.

*598 Opinion

HUFFMAN, J.

In this opinion we discuss the two-pronged test for a strict liability manufacturing defect as applied to the production of latex gloves. (Barker v. Lull Engineering Co. (1978) 20 Cal.3d 413, 432 [143 Cal.Rptr. 225, 573 P.2d 443, 96 A.L.R.3d 1] (Barker).) This products liability action filed by plaintiff Christine McGinnis was the first to go to trial in a group of cases in coordinated proceedings involving allegations against various defendants who manufactured or distributed latex gloves, used by the plaintiffs at their work, that contained natural or artificial substances that were alleged to be causative factors in the development of the plaintiffs’ serious latex allergies. McGinnis’s individual action against Baxter Healthcare Corporation (Baxter) went to jury trial on her manufacturing defect strict liability theory, as well as negligence and failure to warn. A defense verdict was obtained on failure to warn and negligence, but the jury found a manufacturing defect had been proven and awarded McGinnis compensatory damages.

However, in posttrial proceedings, Baxter moved for judgment notwithstanding the verdict (JNOV) and for new trial on the manufacturing defect issue, and both motions were granted by the trial court. (Code Civ. Proc., §§ 629, 657.) McGinnis appeals the ensuing judgment, arguing the trial court applied an incorrect legal standard in ruling on the motions, such that under a proper analysis, there was substantial evidence that Baxter’s latex gloves were defective, either as a defective product that differed from the manufacturer’s intended result or that differed “from other ostensibly identical units of the same product line.” (Barker, supra, 20 Cal.3d at p. 429.)

McGinnis also makes a last-ditch claim that the court erred in refusing to instruct the jury on her design defect theory, based on a consumer expectations approach. (See Morson v. Superior Court (2001) 90 Cal.App.4th 775, 784 [109 Cal.Rptr.2d 343] (Morson), a prior mandamus proceeding in these consolidated cases, originating in this McGinnis case, in which this court rejected the application of that same consumer expectations theory to latex gloves in the health care context.)

Our examination of the record and the applicable legal principles persuades us that the trial court correctly granted JNOV under the two-part test for a manufacturing defect and that no new trial on that theory is warranted in light of that determination. Nor was there any reversible instructional error regarding design defect. We need not reach the alternative grounds on which the trial court denied Baxter’s new trial motion (lack of causation evidence and inconsistent verdicts). We affirm the judgment in favor of Baxter.

*599 Factual and Procedural Background

A

Pleadings and Background

On review of this order granting JNOV, we state the facts in the light most favorable to the jury’s verdict. (Hansen v. Sunnyside Products, Inc. (1997) 55 Cal.App.4th 1497, 1510 [65 Cal.Rptr.2d 266].) As this court outlined the background facts in Morson, supra, 90 Cal.App.4th 775, the plaintiffs in these coordinated cases pursue a theory of product liability that the latex gloves supplied to them caused a serious, disabling, and potentially life-threatening allergy to all forms of natural rubber latex (referred to as NRL) to develop, even though they did not have this condition prior to their extensive use of latex gloves. They accordingly claim improperly designed and manufactured NRL gloves caused this allergy by allowing excessive levels of allergenic agents, latex proteins, to remain present on the surface of the gloves during manufacture. It is not disputed that such agents may be greatly reduced or eliminated through washing and chlorinating procedures in the design and manufacture of these gloves. The issue is whether, as plaintiff complains here in the context of her manufacturing defect claim, Baxter “took too long” to make that its standard practice, in light of its knowledge and research.

McGinnis (sometimes referred to as Plaintiff) was employed as a respiratory technician by various hospitals and care facilities for a number of years between 1982 and 1996, and used thousands of pairs of Baxter NRL gloves during her career. The brands she used over 93 percent of the time, Flexam powdered exam gloves and Triflex powdered surgical gloves, were manufactured at Baxter plants in the United States and Malaysia. In addition, Baxter purchased gloves from other manufacturers to sell under its brand names (“buy-in gloves”). Baxter began to receive reports around 1988-1989 that some health care workers commonly exposed to NRL products were developing severe latex allergies. It began a research and development effort in its glove production and purchases around that time, as we later describe.

Both through her own use of NRL products and the use by others around her, McGinnis became sensitized to that substance to the point of developing a serious type I latex allergy, which caused her in 1995 to experience symptoms going beyond mild symptoms of itching and skin irritation, to a life-threatening anaphylactic reaction (respiratory distress, hives and other symptoms). She was forced to leave health care work, has undergone emergency medical treatment for such reactions, and must carry medication to treat them at all times, as her allergy is a lifelong condition.

*600 McGinnis sued Baxter and other defendants (who were no longer involved in the case by the time of trial and this appeal) on various products liability and negligence theories. The matter went to jury trial on strict liability theories of manufacturing defect and failure to warn of a defective product, and well as negligence through manufacture and failure to warn.

B

Jury Trial: Evidence

Extensive testimony and documentary evidence was presented at trial about the manufacturing process of NRL gloves. The critical qualities provided by rubber gloves to the health care profession include barrier protection and tactile sensitivity. The market for gloves grew tenfold from 1983 to 1990 after the Food and Drug Administration (FDA) recommended and then in 1987 adopted universal precautions for health care workers to prevent the spread of AIDS and hepatitis, requiring expanded use of gloves and other barrier protection equipment. By 1990, Baxter was manufacturing and distributing approximately four billion gloves per year, which represented approximately half of the American medical glove market. Most of these gloves were made of NRL.

The multistep manufacturing process begins with the tapping of rubber trees and centrifuging and mixing of raw rubber, the preparation of glove molds to be positioned on a continuous conveyor line, the dipping of the mold in coagulant and rubber compounds, the leaching in water of the molds, curing, rinsing, powdering, chlorination and sterilization, and packaging of the gloves. (Morson, supra, 90 Cal.App.4th at pp. 780-781.) Plaintiff presented evidence that additional washing and chlorination of the gloves would reduce allergenic protein levels, while Baxter presented evidence that these steps might lead to defects in barrier protection such as pinholes, tearing, or a change in texture. (Id. at p. 782.) Baxter continued its defense: “ ‘Additionally, each of these steps may differ from line to line, glove to glove, and plant.to plant depending on line speed, temperature, plant configuration and other conditions.’ ” (Id. at p. 781.) Accordingly, “ ‘[ejach of the steps must be performed with an acute awareness of barrier protection issues, with an eye toward ensuring that the function of this life-saving medical device will not be compromised.’ ” (Id. at p. 782.)

Baxter’s witnesses testified about the company’s efforts to reduce the protein levels in their products, in response to consumer complaints received as part of the process of federal regulatory monitoring of the production of gloves. From 1989 through 1991, Baxter sold over 15 billion gloves, and *601 received a total of 10 user complaints describing type I severe allergic reactions to NRL gloves. Dr. Wava Truscott, Baxter’s manager of laboratories for the applicable division, was assigned to head the investigation of the glove protein allergy problem, beginning in 1988. The task included assessing many different production Unes and dealing with many different aspects of production, and it appears that Truscott’s efforts were at first an uphill battle. She was concerned that during these years, Baxter was perceived as dragging its corporate feet on the matter. By 1991, her research had proposed a target or threshold level of protein for gloves, and she worked with product engineer Charles Gosnell to test her research. Complaints had been received about both high- and low-protein-level gloves.

In response to the problem, by 1992, Baxter had implemented protein reduction techniques on all its production lines. By 1994, Baxter was requiring all its gloves to undergo a postcure rinse to reduce surface protein levels. Later, a postcure protein leach procedure was added. Baxter also made educational efforts about NRL allergies for health care workers, presenting traveling seminars as part of its marketing efforts. Around 1992-1994, it was advertising its gloves as “The Right Choice,” due to Baxter’s research and improvements in its manufacturing processes regarding allergens.

Plaintiff presented evidence from a manufacturing expert, Dr. Broutman, about the methods for and the feasibility of reducing protein levels, and his opinion that Baxter did not make desirable changes as quickly as possible. Extensive protein level evidence was offered by Plaintiff of Baxter’s research and manufacturing changes for different glove types at different production facilities. For example, documents were introduced about tests run at different Baxter plants in the early 1990’s, its internal memoranda among scientific and manufacturing staff to propose methods to reduce protein levels, including data on different gloves tested, and its dealings with the makers of its buy-in gloves on the subject.

With respect to the buy-in line of gloves that Baxter bought from others and distributed, mainly after 1992, Baxter was not requiring the other manufacturers to use the techniques it was developing to reduce protein levels in the gloves during the time that McGinnis was using them. The buy-in gloves were labeled the same as Baxter-produced gloves, so that a user could not tell whether a particular pair was manufactured by Baxter or another company. Beginning in 1994, Baxter tested every lot of these gloves for protein levels.

Another line of evidence presented had to do with the sufficiency of the warnings provided by Baxter on its products regarding potential allergic *602 reactions to NRL. Baxter began to label its latex gloves for latex content in 1992. In March 1993, the FDA announced its plans to issue regulations requiring manufacturers to state the latex content of medical devices such as gloves. Although the FDA was conducting research in the area of latex allergies, it did not require warning labels on that subject until 1996. Until 1998, due to concerns about the success of the AIDS, etc., universal precautions requirement of glove usage, the FDA prohibited glove manufacturers from making comparisons about protein levels of their product. Also until 1998, there were no FDA protein level requirements or standards, such as had been established for strength and barrier protection qualities of the gloves. Specifically, these 1998 FDA protein level requirements or standards were implemented only for a low-protein line of gloves.

Further, extensive testimony from various health care professionals was presented regarding the genesis of allergic conditions in the human body, specifically with reference to NRL allergies. (Morson, supra, 90 Cal.App.4th at p. 782.) McGinnis presented testimony about her health condition and her claimed damages.

C

Jury Trial: Instructions and Argument

At the outset of trial, the jury was preinstructed on the elements of claims for both manufacturing defects and design defects. However, by the conclusion of the evidence, McGinnis’s claims no longer included a strict liability design defect cause of action, due in part to the ruling that we reviewed in the prior mandamus proceeding, Morson, supra, 90 Cal.App.4th at page 784, that precluded her from presenting a design defect theory based on consumer expectations. McGinnis did not present any alternative design defect theory under a risk-benefit analysis. (Id. at p. 785.) In accordance with that ruling, the trial court declined to instruct the jury on the consumer expectations test for a design defect. McGinnis contends this amounted to a nonsuit on design defect, while Baxter argues she simply did not pursue this claim further.

Thus, at trial, McGinnis went to the jury on her manufacturing defect claim, failure to warn of a defect and/or through negligence, and negligent manufacture and/or design. She stated during the in limine motion hearings that it was never her theory that the gloves were defective simply because they were made of latex as opposed to some other material, but rather that there were manufacturing defect problems, as well as failure to warn problems.

On the manufacturing defect claim, the jury was instructed in the language of BAJI No. 9.00.3 as follows:

*603 “The essential elements of a claim based upon an alleged manufacturing defect are:

“1. The defendant, Baxter Healthcare Corporation, was the manufacturer and supplier of a product, namely natural rubber latex gloves. flO 2. The product possessed a defect in its manufacture. ftD 3. The defect in manufacture existed when the product left the defendant’s possession. flO 4. The defect in manufacture was a cause of injury to the plaintiff, and flQ Plaintiff’s injury resulted from the use of the product that was reasonably foreseeable to the defendant. flQ A defect in the manufacture of a product exists if the product differs from the manufacturer’s intended result, or if the product differs from apparently identical products from the same manufacturer.”

The jury also received instructions about the failure to warn and negligence theories.

Plaintiff’s counsel presented closing argument that focused upon the instruction about the manufacturing defect claim, BAJI No. 9.00.3: “A defect exists if the product differs from the intended result.” He argued that the Baxter witnesses testified they had the intent, starting in 1990, to produce a low-protein glove, but that although “their intentions were good, their execution was bad. And that creates a defect. They didn’t execute their intent.” Also, Plaintiff’s counsel argued that the product could also be defective under the test “if the product differs from apparently identical products from the same manufacturer.” He compared the gloves manufactured in Malaysia by Baxter, the buy-in gloves, and the United States-made gloves, and argued that these apparently similar products were actually different, so the test was satisfied.

In contrast, Baxter argued that the protein level evidence offered by Plaintiff had not been placed in context with any applicable government requirements, and that at the time Dr. Truscott was investigating the problem, complaints had been received about both high-protein and low-protein gloves, which made analysis at that point inconclusive. Before and after 1992, Baxter was constantly tinkering with the system to get the best protein testing system in place. This protein testing system had to be implemented while keeping production up, due to the health care profession’s need for universal precautions equipment. Baxter’s position was that its personnel were at the top of the heap in the production field, and although they were not perfect, they acted reasonably.

*604 D

Jury Verdict

The jury returned a verdict finding that a manufacturing defect had been proven and awarded McGinnis net compensatory damages of $886,921.20. The jury also found Baxter had been negligent but there had been no causation of her injuries through negligence. A comparative-fault finding was made assessing 70 percent of the negligence to Baxter, 15 percent to McGinnis, and 15 percent to her previous hospital employer (not a party to the action). The jury also rejected McGinnis’s claim that a warning defect was present.

E

Baxter’s Motions for JNOV and New Trial

Following trial, Baxter filed its motions for JNOV and new trial. It argued that with respect to the manufacturing defect finding, no substantial evidence supported the verdict. It also argued the causation findings were inconsistent and lacked support in the evidence, and excessive damages had been awarded.

After briefing and argument, the trial court granted the Baxter motion for JNOV on the single cause of action on which McGinnis had prevailed, manufacturing defect under a strict products liability theory. The court noted that no cause of action for design defect was submitted to the jury. The court found that the jury’s verdict on manufacturing defect was not supported by the evidence, stating its reasoning as follows: “A manufacturing defect occurs when a product deviated from its intended design. Plaintiff had the burden of demonstrating there was a flaw in the manufacturing process whereby the injury-inflicting product or products deviated from the manufacturer’s design or specification and thus was manufactured differently from the prototype. [1Q No evidence was introduced at trial that any of the gloves involved in the contributing [sic] to Plaintiff’s injuries, in any manner deviated from the design for said gloves. Plaintiff’s evidence tended to show that the gloves were defective in their design, since the design did not require the elimination or substantial reduction of protein that collected on the glove surface during its manufacture. [^] The issue of design defect was not submitted to the jury for its determination, although a substantial portion of the evidence adduced at trial related to that issue.”

In the alternative, the trial court granted the motion for new trial on the same grounds, with respect to the manufacturing defect theory. Although the *605 order does not reflect the details of the new trial ruling, the transcript of the hearing shows that the trial court denied the new trial motion insofar as it argued insufficient evidence of causation and/or inconsistent special verdicts on causation. The excessive damages claim was also rejected.

McGinnis appealed the posttrial orders and the judgment, and her appeals were consolidated.

Discussion

McGinnis’s appeal raises the same basic concerns about the trial court’s order with respect to both the Baxter motion for JNOV and its motion for new trial. She argues the trial court incorrectly applied both the alternative tests for a manufacturing defect as set forth in Barker, supra, 20 Cal.3d 413. There, the Supreme Court opined (optimistically, in hindsight) that defining the concept of a product defect “raises considerably more difficulties in the design defect context than it does in the manufacturing or production defect context, ffl] In general, a manufacturing or production defect is readily identifiable because a defective product is one that differs from the manufacturer’s intended result or from other ostensibly identical units of the same product line.” (Id. at p. 429.) These concepts form the basis of BAJI No. 9.00.3, defining a manufacturing defect, which was given to the jury here. 1

McGinnis’s initial criticism is that the trial court mistakenly evaluated the evidentiary record only in light of the “intended result test,” by first requiring evidence that the high protein gloves that McGinnis used departed from Baxter’s formal product design, prototype, or specifications, and secondly, by finding there was no such evidence. McGinnis next argues that the trial court failed to recognize that the alternative formulation of the test applied and was satisfied here, because there is substantial evidence that the high-protein gloves she used differed from “apparently identical” gloves that Baxter also produced or distributed, due to the different sources from which Baxter obtained its gloves and its varying production techniques over time. We will discuss these arguments in terms of the two procedural contexts in *606 which the order was made, with attention to the separate standards for each ruling.

I

Judgment Notwithstanding the Verdict

The standards to apply in reviewing a ruling on a motion for JNOV are well established. “ ‘The trial court’s discretion in granting a motion for judgment notwithstanding the verdict is severely limited.’ [Citation.] ‘ “The trial judge’s power to grant a judgment notwithstanding the verdict is identical to his power to grant a directed verdict [citations]. The trial judge cannot reweigh the evidence [citation], or judge the credibility of witnesses. [Citation.] If the evidence is conflicting or if several reasonable inferences may be drawn, the motion for judgment notwithstanding the verdict should be denied. [Citations.] ‘A motion for judgment notwithstanding the verdict of a jury may properly be granted only if it appears from the evidence, viewed in the light most favorable to the party securing the verdict, that there is no substantial evidence to support the verdict. If there is any substantial evidence, or reasonable inferences to be drawn therefrom, in support of the verdict, the motion should be denied.’ [Citation.]” ’ [Citation.] The trial court cannot consider witness credibility. [Citation.]” (Hansen v. Sunnyside Products, Inc., supra, 55 Cal.App.4th at p. 1510.)

When an appellate court reviews an order granting JNOV, it will “ ‘ “resolve any conflict in the evidence and draw all reasonable inferences therefrom in favor of the jury’s verdict.” ’ ” (Hansen v. Sunnyside Products, Inc., supra, 55 Cal.App.4th at p. 1510.)

A

Comparison of Definitions of Product Defects: Manufacturing/Design

As explained by the Supreme Court in Brown v. Superior Court (1988) 44 Cal.3d 1049 [245 Cal.Rptr. 412, 751 P.2d 470] (Brown), under Barker’s strict products liability analysis, there are three types of product defects: “First, there may be a flaw in the manufacturing process, resulting in a product that differs from the manufacturer’s intended result. The archetypal example of such a defect was involved in Escola [v. Coca Cola Bottling Co. of Fresno (1944)] 24 Cal.2d 453 [150 P.2d 436], a Coca Cola bottle that exploded. . . . [f] Second, there are products which are ‘perfectly’ manufactured but are unsafe because of the absence of a safety device, i.e., a defect in design. This was the defect alleged in Barker. It held that a product *607 is defectively designed if it failed to perform as safely as an ordinary consumer would expect when used as intended or reasonably foreseeable, or if, on balance, the risk of danger inherent in the challenged design outweighs the benefits of the design. [Citation.]” (Brown, supra, 44 Cal.3d at p. 1057.) The third type of defect “is a product that is dangerous because it lacks adequate warnings or instructions.” (Ibid.)

As a further illustration of the development of the concept of a defect in the strict liability field, we may turn to Jiminez v. Sears, Roebuck & Co. (1971) 4 Cal.3d 379, 383 [93 Cal.Rptr. 769, 482 P.2d 681, 52 A.L.R.3d 92], where the court reviewed prior case law to say that “a defective product is viewed as one which fails to match the quality of most like products, and the manufacturer is then liable for injuries resulting from deviations from the norm: the lathe did not like other lathes have a proper fastening device, the brakes of the automobile went on unexpectedly, the drive shaft of a new car became disconnected.”

Here, McGinnis’s case relied on the first and third types of product defects listed in Barker, supra, 20 Cal.3d 413, and the jury rejected the third (failure to warn of a defect). Only the first type, manufacturing defect, is squarely presented as an issue in this appeal. (See pt. Ill, post, where we reject her effort to revive a direct design defect theory under a consumer expectations rubric, and where we note she did not effectively pursue the remaining design defect theory, risk-benefit.) Hence, our task is to see if, as McGinnis contends, substantial evidence was presented to support a manufacturing defect theory under either of the Barker formulations. We do this in tandem with evaluating the Baxter argument that McGinnis actually tried this case under a design defect approach, and her appellate argument on the manufacturing defect theory does not fit the facts like a glove.

In Morson, supra, 90 Cal.App.4th 775, this court relied on Dierks v. Mitsubishi Motors Corp. (1989) 208 Cal.App.3d 352, 354-355 [256 Cal.Rptr. 230] as a statement of the difference between a defect in manufacture and a defect in design: “ ‘The latter focuses upon whether the product was designed to perform as safely as an ordinary consumer would expect or whether the risk of danger inherent in the design outweighed the benefits of the design. [Citations.] The former focuses on whether the particular product involved in the accident was manufactured in conformity with the manufacturer’s design. [Citations.]’ ” (Morson, supra, 90 Cal.App.4th at pp. 788-789.) There, we also cited to a treatise writer for the concept that “ ‘the line between design and manufacturing defects is not necessarily a sharp one. This is clear when one considers that the choice of quality control techniques may determine the rate of metallurgical flaws, which ordinarily would be *608 characterized as “manufacturing defects” rather than design defects. One may note, moreover, that those kinds of entrepreneurial decisions are quite analogous to the sorts of choices that are made in selecting one product configuration or another on the basis of cost considerations.’ ” (Id. at p. 789, citing 1 Shapo, The Law of Products Liability (3d ed. 1994) 1 9.01(2), pp. 9-5 to 9-6.)

Also in Morson, supra, 90 Cal.App.4th 775, we noted that in general, the plaintiffs in these coordinated latex glove cases are alleging a theory of injury that “appears to have aspects common to both the design defect and manufacturing defect theories” (id. at p. 789), in that “the materials from which the latex gloves were made contained excessive amounts of latex rubber proteins close to the surface of the gloves, causing the Plaintiffs to become sensitized to them and to develop or to exacerbate an existing allergy.” (Ibid.) We discussed the particular nature of this product as having primarily a protective or barrier function, leading to the choice of latex as an appropriate material, but also noting that the effect of this material and the manufacturing processes used “may well be to create in their users many degrees of allergic reactions. Understanding and assessing responsibility for such allergic reactions is a matter that is driven by the science of the manufacturing and preparation procedures, as well as the medical aspects of an individual’s allergic reactions to various substances.” (Id. at p. 793.)

From these observations, we are led to a further conclusion, that it is useless to focus on a distinction between a raw material of which a product is made (NRL), and the format or construction of the product itself (glove), for purposes of deciding if a strict liability defect is one of design or manufacture. Here, McGinnis stated during the hearings on the in limine motions that it was never her theory that the gloves were defective simply because they were made of latex as opposed to some other material, but rather she claimed there were manufacturing defect problems, as well as failure to warn problems. It is appropriate to view the product as a whole, composed of different components, including latex.

In other factual contexts, similar difficulties have arisen in analyzing the component parts of a whole product for purposes of applying the manufacturing defect test. In Pierce v. Pacific Gas & Electric Co. (1985) 166 Cal.App.3d 68 [212 Cal.Rptr. 283], the appellate court concluded that a commercial supplier of electricity is subject to strict liability in tort for personal injuries caused by delivery of electricity at dangerously high voltage resulting from defects in a transformer. The facts were that due to a transformer malfunction, the defendant utility’s electricity arrived at the plaintiff’s home at nearly 60 times its intended voltage, ultimately causing *609 the plaintiff to suffer bodily injury. (Id. at p. 77.) However, the court noted that the utility defendant (PG&E) was not the manufacturer of the defective transformer, which some other electric company had made, such that the defendant PG&E “never placed the transformer ‘on the market’ or in the stream of commerce. PG&E was, in essence, a consumer rather than a manufacturer of the transformer, and cannot be held strictly liable in tort for the transformer’s defects per se. [Citation.]” (Id. at p. 76.) However, the appellate court concluded that since the product, electricity, was delivered to the plaintiff’s home by way of the defective transformer, and arrived at a harmful level of voltage and injured the plaintiff, it was error to grant a nonsuit in favor of the defendant utility, as a cause of action for strict liability in tort existed and personal injuries were shown. It did not make any difference that the utility could not be held strictly liable in tort for the transformer’s defects “per se,” as the subject product, electricity, harmed plaintiff anyway. (Id. at pp. 77, 84.) However, the court did not directly identify whether a manufacturing or design defect was shown, as to the electricity itself.

One explanation for this blending of theories was alluded to in Artiglio v. Superior Court (1994) 22 Cal.App.4th 1388, 1393 [27 Cal.Rptr.2d 589] (Artiglio), in which this court noted “pre-Brown authorities lumped together the concepts of proper manufacture, lack of design defect and adequate warning.” However, in Artiglio we further stated that in Brown, supra, 44 Cal.3d 1049, which rejected the theory of design defect for pharmaceuticals, the Supreme Court very clearly distinguished among the three concepts of fault, in the prescription drug context: “Liability for defective design could not be premised on strict liability, but would require proof of negligence. [Citation.] Strict liability would continue applicable for manufacturing defects; and liability for failure to warn of known or reasonably knowable risks in the use of the product remains viable ‘under general principles of negligence.’ [Citation.]” (Artiglio, supra, 22 Cal.App.4th at p. 1393.) Artiglio similarly concluded, based on the public policies identified in Brown, “that the entire category of medical implants available only by resort to the services of a physician are immune from design defect strict liability.” (Artiglio, supra, 22 Cal.App.4th at p. 1397.)

Of course, latex gloves, even as used in the health care field as this plaintiff did, are not prescription drugs, nor are they available “only by resort to the services of a physician,” as in Artiglio, supra, 22 Cal.App.4th 1388, and Brown, supra, 44 Cal.3d 1049. Nor was this case pursued or instructed as a design defect case, as in Artiglio and Brown. Nevertheless, many of the same policy concerns apply, due to the closely related nature of these strict liability theories in this factual context, as we next explain.

*610 B

Public Policy and Doctrinal Concerns

The Supreme Court’s holding in Brown, supra, 44 Cal.3d 1049, was that “a manufacturer is not strictly liable for injuries caused by a prescription drug so long as the drug was properly prepared and accompanied by warnings of its dangerous propensities that were either known or reasonably scientifically knowable at the time of distribution.” (Id. at p. 1069.) An important public policy considered by the court in reaching that decision was stated as follows: In the cases in which the subject product is used to make work easier or to provide pleasure (e.g., construction machinery, a lawnmower, or perfume) it is not unreasonable to impose strict liability for design defects. However, more protection for a manufacturer is justified where the product is created “to alleviate pain and suffering or to sustain life.” (Id. at p. 1063.) Along these same lines, the court also referred to “other important medical products (wheelchairs, for example),” apparently to illustrate that harm to some users can be avoided for some medical products, which would allow liability to be more freely imposed on a manufacturer. (Ibid.) However, the court stated that “harm to some users from prescription drugs is unavoidable,” and continued: “Because of these distinctions, the broader public interest in the availability of drugs at an affordable price must be considered in deciding the appropriate standard of liability for injuries resulting from their use.” (Ibid.) This led the court to reject a design defect theory of liability against a drug manufacturer for injuries caused by the defective design of a prescription drug, and also to reject an assertion “that a drug manufacturer should be held strictly liable for failure to warn of risks inherent in a drug even though it neither knew nor could have known by the application of scientific knowledge available at the time of distribution that the drug could produce the undesirable side effects suffered by the plaintiff.” (Id. at p. 1065.)

Another important policy concern in the field of strict liability is that the doctrine “was never intended to make the manufacturer or distributor of a product its insurer. ‘From its inception, . . . strict liability has never been, and is not now, absolute liability. . . . [Ujnder strict liability the manufacturer does not thereby become the insurer of the safety of the product’s user. [Citations.]’ [Citation.]” (Anderson v. Owens-Corning Fiberglas Corp. (1991) 53 Cal.3d 987, 994 [281 Cal.Rptr. 528, 810 P.2d 549], italics omitted.)

From these policy statements we are led to believe that the courts may take into account, in evaluating manufacturer liability for a product that seriously implicates concerns of alleviating pain and suffering or sustaining *611 life, whether the product “was properly prepared and accompanied by warnings of its dangerous propensities that were either known or reasonably scientifically knowable at the time of distribution.” (Brown, supra, 44 Cal.3d at p. 1069.) In cases involving products that create significant scientific concerns with respect to research and innovation, more protection for a manufacturer is justified than in cases of “other important medical products (wheelchairs, for example),” in which harm to some users can more readily be avoided, due apparently to their more mechanical nature. (Id. at p. 1063.) This has led to the rejection of design defect immunity for the more complex medical/pharmaceutical products.

Moreover, in reviewing this record for substantial evidence in support of a manufacturing or production defect theory, we must keep in mind the two formulations of the test: A defective product is one that “differs from the manufacturer’s intended result or from other ostensibly identical units of the same product line.” (Barker, supra, 20 Cal.3d at p. 429.) Where do latex gloves fall along the continuum of strict liability theory, in light of public policy? We turn to the record to decide that question.

C

Application of Rules to This Evidentiary Showing

McGinnis first claims the trial court mistakenly evaluated the evidentiary record only in light of the “intended result test,” by finding she failed to produce essential evidence that the high-protein gloves that Baxter produced departed from its own design, specifications, or prototypes. She contends she showed, through the evidence of the research and data collection that Baxter was doing to reduce protein levels, that Baxter had internal standards that it was developing that constituted such evidence of “formal product design, prototype, or specifications.”

McGinnis next argues the trial court failed to recognize that the alternative formulation of the Barker test applied, because substantial evidence was produced to show that the high-protein gloves she used differed from “apparently identical” gloves that Baxter also produced or distributed, due to the different sources from which Baxter obtained its gloves and its evolving production techniques. The buy-in gloves were labeled the same as Baxter-produced gloves, but they were not subject to the same standards at the same times. Also, there were variances in protein levels among Baxter-produced gloves, depending on the lines that manufactured them.

On review of this order granting JNOV, we must “ “resolve any conflict in the evidence and draw all reasonable inferences therefrom in favor of *612 the jury’s verdict.” ’ ” (Hansen v. Sunnyside Products, Inc., supra, 55 Cal.App.4th at p. 1510.). Even when we do so, we have grave concerns that Plaintiff’s evidence fails to make her case, when all the applicable policies are considered. For example, while Plaintiff presented evidence that additional washing and chlorination of the gloves would reduce allergenic protein levels, Baxter presented evidence that these steps might lead to defects in barrier protection such as pinholes, tearing, or a change in texture. (Morson, supra, 90 Cal.App.4th at p. 782.) Each line, glove, and plant was subject to variances due to Une speed, temperature, plant configuration and other conditions. (Id. at p. 781.) Any changes were subject to the need to consider barrier protection issues, “ ‘with an eye toward ensuring that the function of this Ufe-saving medical device will not be compromised.’ ” (Id. at p. 782.)

Other evidence in the record showed that while Baxter began to label its latex gloves for latex content in 1992-1993, it was not required by the FDA to do so until after 1993. The FDA did not require warning labels in the area of latex allergies until 1996. Until 1998, due to concerns about the success of the universal precautions requirement of glove usage, the FDA prohibited glove manufacturers from making comparisons about protein levels of their product. Also until 1998, there were no FDA protein level requirements or standards, such as had already been established for the strength and barrier protection quaUties of the gloves. The 1998 FDA protein level requirements or standards were implemented only for a low-protein line of gloves. Other glove types were not affected. None of this amounts to evidence that Baxter failed to meet externally imposed government product specifications.

Also, the evidence presented- about Baxter’s internal corporate protein reduction techniques, such as between 1992 and 1994, requiring all its gloves to undergo a post-cure rinse and leach to reduce surface protein levels, raises the issue of whether these internal standards and practices can be used by Plaintiff to show earlier manufacturing techniques were defective. It is not disputed that Baxter personnel developed targets or goals for protein levels in the product. However, Plaintiff failed to show these were enforceable standards, departure from which would create a manufacturing defect. Plaintiff’s approach would essentially penalize the manufacturer for doing documented research to respond to product complaints or to improve the product. This would contravene the public policies outlined in Brown, supra, 44 Cal.3d 1049, with respect to the use of these products by health care workers whose employers’ goals are geared toward alleviating pain and suffering, and also toward compliance with the requirements for universal precautions equipment, for protection of those workers and patients.

As stated in Baxter’s respondent’s briefs, it was uncontested at trial “that Baxter intended to, and did, produce and sell gloves with a wide range of *613 protein levels.” These gloves met Baxter’s design specifications as they existed at all the relevant times. There was no set standard for protein levels under either Baxter’s corporate policies or the government regulations. Plaintiff cannot convert these undisputed facts into an adequate showing of a manufacturing defect under the Barker tests.

We also evaluate the evidence in light of the Baxter argument that McGinnis actually tried this case under a design defect approach, and did not change her arguments into a manufacturing defect format until she realized the design defect approach was fatally flawed. Both the traditional definitions of manufacturing defect presuppose that a suitable design is in place, but that the manufacturing process has in some way deviated from that design. (See Dierks v. Mitsubishi Motors Corp., supra, 208 Cal.App.3d 352, 354-355 [focus is on whether the particular product involved in the incident was manufactured in conformity with the manufacturer’s design]; see also Morson, supra, 90 Cal.App.4th at p. 789.) Here, we are unable to separate out the raw material, NRL, from the forming and processing of it, nor does Plaintiff argue we should. The NRL gloves in this case were processed exactly as Baxter intended that they should be, in light of the state of its scientific and manufacturing knowledge at the time. This was true of all the various lines of production, even though testing was ongoing at some lines and not others at times. That later developments showed the product was subject to immense improvement does not necessarily show the products processed earlier were defective, under either formulation of the Barker test. The fact that simultaneously manufactured gloves were subject to different standards at different production lines, due to the status of the manufacturer’s research and development, where scientific knowledge was as inconclusive as is shown by this record, does not require that some items must be deemed defective under a manufacturing defect approach. Rather, such arguments actually deal with design defect evidence, and the jury properly did not receive those instructions in this case. Allowing the Plaintiff’s verdict to stand here would be inconsistent with the applicable public policies as stated above, for lack of any supporting evidentiary showing.

In conclusion, we believe that Plaintiff’s efforts are ineffective to show that the various NRL gloves that were manufactured precisely as intended, that complied with applicable governmental standards, and that fulfilled their primary barrier function, nevertheless had manufacturing defects, due to the existence of evidence of the testing, improvement, research and development efforts, targets and goals of the manufacturer, at different times and locations, reflective of the state of scientific knowledge regarding latex protein levels of exposure that was available to the relevant participants in this health care product context. The products did not differ from the manufacturer’s intended result, nor did they have materially significant differences *614 among identical units from the same product line. (Barker, supra, 20 Cal.3d at p. 429.) The motion for JNOV was properly granted.

II

New Trial Motion

As fully explained by the court in Fountain Valley Chateau Blanc Homeowner’s Assn. v. Department of Veterans Affairs (1998) 67 Cal.App.4th 743, 751 [79 Cal.Rptr.2d 248] (Fountain Valley), ordinarily, the function of a new trial motion is to allow a reexamination of an issue of fact, and accordingly, on review of a ruling on a new trial motion, the abuse of discretion standard will apply. (See, e.g., Jones v. Evans (1970) 4 Cal.App.3d 115, 121 [84 Cal.Rptr. 6].) However, an appellate court has the power to look at the substance of a new trial ruling rather than just its title. (Fountain Valley, supra, 61 Cal.App.4th at pp. 750-753.) If the effect of the ruling is actually closer in nature to a directed verdict or a JNOV, then in such a case, the ruling may be deemed to have been based upon a conclusion of law, and de novo review is appropriate. (See ibid.)

In this case, the effect of the new trial ruling was to allow Baxter to prevail as a matter of law, since the relevant evidence had already been presented on the strict liability theories. Only the manufacturing defect theory was the basis of the rulings, as the court rejected the other grounds argued. We should treat the order as a legal ruling that overturned the jury verdict, and uphold the judgment that disposed of that verdict, if the record allows the issues to be decided as matters of law. (See Fountain Valley, supra, 61 Cal.App.4th at pp. 750-753.) We believe that it does. Although the trial court granted the new trial motion in the alternative to the JNOV, we deem that grant to be an effort to dispose of the case for the same reasons outlined above as applied in the other context. No new trial is appropriate on this record, as it demonstrates that the manufacturing defect theory cannot as a matter of law apply to the existing facts in the record. Nor has Plaintiff demonstrated any way in which different or new evidence could be presented to justify further efforts along those lines. We deem the new trial order to be moot in light of our analysis of the JNOV issues and our affirmance of the JNOV order.

Ill

Design Defect Theory Based on Consumer Expectations

Finally, McGinnis seeks reversal of the judgment on the basis that the court erred in refusing to instruct the jury on a design defect theory based *615 on a consumer expectations approach. She contends she was effectively nonsuited on that theory. As already alluded to, the prior opinion by this court in Morson, supra, 90 Cal.App.4th 775, 784-785, arising out of the general order made in this case that precluded plaintiffs in these coordinated cases in future trials from seeking such instructions, upheld that general order as a proper interpretation of Barker, supra, 20 Cal.3d 413.

McGinnis, however, seeks a different result on the grounds that although the trial court’s order unquestionably applied to her case, and although that order had been upheld in writ proceedings as of the time of the briefing in this matter, it had not yet become final by the time the opening brief was filed. While that may be so, the Morson opinion is now final and we are entitled to regard it as correct and also as binding upon McGinnis.

McGinnis also asserts that Baxter raised heightened expectations of consumer safety by advertising its product as “The Right Choice,” referring to its efforts to reduce protein amounts in NRL gloves, and that this evidence produced at trial goes beyond the record that was considered in Morson, supra, 90 Cal.App.4th 775. We believe the analysis in Morson sufficiently covers this ground. Moreover, McGinnis has presented nothing in the record or in the applicable authorities to justify a different result. She did not present any alternative design defect theory under a risk-benefit analysis. (Id. at p. 785.) She accordingly failed to place directly before the jury any design defect theory that was not incorrectly based upon the consumer expectations test. In any case, as discussed above, her manufacturing defect evidence closely resembled the evidence she would have gathered for a design defect claim, and it does not support the verdict. We need not further consider this claim, as neither version of the design defect test will legitimately apply under these factual and procedural circumstances.

Disposition

The judgment and order are affirmed. Each party to bear its own costs.

Kremer, P. J., and Nares, J., concurred.

A petition for a rehearing was denied July 15, 2002, and the opinion was modified to read as printed above. Appellant’s petition for review by the Supreme Court was denied October 2, 2002. Kennard, J., and Werdegar, J., were of the opinion that the petition should be granted.

BRECKENRIDGE, Judge.

Eric Magnuson, by his next friend, Judy Mabe, appeals from the grant of a new trial to defendants, Kelsey-Hayes Company and Kelsey-Hayes Canada, Ltd., on a finding by the trial court that it erred in dismissing Chrysler Motors Corporation before trial. Magnuson argues that by his action in voluntarily dismissing Chrysler, he has rendered the issue of the trial court’s dismissal moot. He further argues that the trial court erred in granting a new trial to Kelsey-Hayes and Kelsey-Hayes Canada because they were not prejudiced by the dismissal of Chrysler and had no standing to complain about Chrysler’s dismissal from the case. Chrysler joins in Magnuson’s appeal, briefing both points presented by Magnuson and further challenging the trial court’s grant of a new trial, claiming that Kelsey-Hayes and Kelsey-Hayes Canada waived any objection to the trial court’s dismissal of Chrysler because they did not oppose Chrysler’s motion to dismiss, file a cross-claim against Chrysler or file a motion to rejoin Chrysler as a party.

Kelsey-Hayes and Kelsey-Hayes Canada claim that the issue is not moot or, if it is found to be, that the trial court erred in denying their motions for judgment notwithstanding the verdict because Magnu-son failed to make a submissible case. Kelsey-Hayes and Kelsey-Hayes Canada also contend that should this court overrule the trial court’s order, the cause should be remanded for consideration of the issue of remittitur.

On September 23, 1987, four and one-half-year old Eric Magnuson was injured when he was struck by a wheel and tire which had broken loose from a passing Dodge pick-up truck. The wheel involved in the accident had been designed by Kelsey-Hayes and Chrysler. It was manufactured by Kelsey-Hayes Canada under the direction of Kelsey-Hayes which had established quality assurance standards and policies for the process. Eric Magnuson was struck in the head. He suffered a contre-coup brain injury and went into a coma. Upon waking he suffered temporarily from *451 bilateral cortical blindness, leaving him with residual visual perception defects. As a result of his injuries, Eric Magnuson suffers from permanent right side hemipare-sis; retarded growth of his right leg, arm and hand; auditory impairment; aphasia; fine motor skills defects; poor balance; and diminished cognitive thinking, including a lower I.Q., defective abstract thinking, memory loss, poor concentration and emotional behavior problems. Expert testimony established that Eric Magnuson is unlikely to be employed in a competitive market. His projected lost earnings, reduced to present value, are in excess of $1,130,000.00.

On April 3, 1989, Eric Magnuson filed suit against Kelsey-Hayes, Chrysler and Raytown Dodge. 1 Chrysler filed a motion to dismiss pursuant to § 537.762, RSMo Supp.1987, which was denied by the trial court. Magnuson was given time to conduct discovery in an attempt to establish facts supporting a claim against Chrysler.

On November 3, 1989, Magnuson filed his first amended petition, adding Kelsey-Hayes Canada as a defendant. Chrysler renewed its motion to dismiss which was not opposed by Kelsey-Hayes or Kelsey-Hayes Canada. The trial court entered its order on August 22, 1990, granting Chrysler’s motion to dismiss without prejudice.

Magnuson proceeded to trial against Kelsey-Hayes and Kelsey-Hayes Canada on theories of strict liability and negligence, claiming that the wheel was defectively manufactured and that defendants failed to warn. Magnuson abandoned his earlier claim of design failure, having conceded during discovery that he had no evidence to support a claim of design defect.

Dr. Robert Bohl, Eric Magnuson’s expert metallurgist, testified that the steel forming the wheel had a defective microstruc-ture and that had Kelsey-Hayes microscopically examined the microstructure of the steel, it could have detected the flaw. Dr. Bohl concluded that the design of the wheel was reasonable and that failure was only apt to occur in the case of poor metallurgical structure. He testified that the defective microstructure was related to cooling the steel too slowly during the rolling of the steel. Dr. Bohl also expressed the opinion that the steel was improperly thinned during the manufacturing process.

The Kelsey-Hayes defendants did not object to Dr. Bohl’s testimony that the steel was improperly thinned. The Kelsey-Hayes defendants cross-examined Dr. Bohl extensively on the thinning issue. At the close of Magnuson’s evidence the defendants moved for mistrial on the grounds that Magnuson was pursuing a claim based on the design of the wheel and that Chrysler was substantially responsible for that design. The court denied the motion. Magnuson submitted the case to the jury on theories of manufacturing defect and failure to warn. The jury returned a $4,750,000.00 verdict against Kelsey-Hayes and Kelsey-Hayes Canada on claims of strict liability manufacturing defect, strict liability failure to warn and negligent manufacturing defect. The jury found for defendants Kelsey-Hayes and Kelsey-Hayes Canada on the claim of negligent failure to warn. The trial court entered judgment based on the jury’s verdict on April 9,1991.

On April 19, 1991, Kelsey-Hayes and Kelsey-Hayes Canada filed alternative post-trial motions for judgment notwithstanding the verdict (JNOV), new trial and remittitur. Following argument on these motions, the trial court granted the defendants’ motions for new trial “for the reason that it was error for the Court to dismiss defendant Chrysler Motors Corporation.” The trial court further ordered that its previous order, dismissing Chrysler, be set aside.

On July 19, 1991, Magnuson filed his notice of appeal in this court. Chrysler filed its notice of appeal on July 25, 1991. Kelsey-Hayes and Kelsey-Hayes Canada also filed a notice of appeal, attempting to *452 appeal from the denial of their JNOV motions. 2

On October 24, 1991, Eric Magnuson, pursuant to Rule 67.01, filed a voluntary dismissal of Chrysler from the case. On November 15, 1991, Magnuson and Chrysler filed a stipulation of dismissal. Kelsey-Hayes and Kelsey-Hayes Canada objected to the dismissal as inappropriate under Rule 67.01. This court reserved any ruling on motions pertaining to the dismissal until the submission of the case.

Eric Magnuson’s dismissal of Chrysler under Rule 67.01 was proper and an order of this court dismissing Chrysler from the case is not required. Rule 67.01 provides, in pertinent part:

A civil action may be dismissed by the plaintiff without prejudice without order of court any time prior to the introduction of evidence at the trial. After the introduction of evidence is commenced, a plaintiff may dismiss his action without prejudice only by leave of court or by written consent of the adverse party.

In the instant case, Chrysler consented to the dismissal. Kelsey-Hayes and Kelsey-Hayes Canada argue that Rule 67.01 is available only if no evidence has been introduced and consent of the adverse party has first been obtained. This is not what the Rule requires, however, as Rule 67.01 is written in the disjunctive. A plaintiff may dismiss after the introduction of evidence “by leave of court or by written consent of the adverse party.” Rule 67.01 (emphasis added). See Reed v. Mirts, 437 S.W.2d 719, 721 (Mo.App.1969).

The Kelsey-Hayes defendants further argue that their permission is necessary as they are adverse parties under the rule. They are not adverse parties under Rules 67.04 and 67.05 as to Chrysler, having not availed themselves of the opportunity to bring Chrysler into the case by third-party action or by a cross-claim. See Camden v. St. Louis Public Serv. Co., 206 S.W.2d 699 (Mo.App.1947). Having obtained the consent of Chrysler, Magnuson’s dismissal is permitted by Rule 67.01. Rule 67.01 is made applicable to this court under Rule 41.01(a)(1) which provides that Rules 41 through 101 govern civil actions pending in the Court of Appeals.

The practical effect of Magnuson’s dismissal of Chrysler is to render the trial court’s order moot since it was specifically based upon the trial court’s reasoning that “it was error for the Court to dismiss defendant Chrysler Motors Corporation.” This is the only ground specified in the trial court's order. Where a trial court grants an order on a specific basis all of the other grounds asserted in the motion are deemed overruled. Roux v. City of St. Louis, 690 S.W.2d 448, 450 (Mo.App.1985). Whether or not the motion for new trial should have been sustained on any of the other grounds specified in the motion for new trial is not a matter under review because the Kelsey-Hayes defendants do not argue this point on appeal. Id.

Magnuson cannot be forced to maintain its suit against Chrysler at Kelsey-Hayes and Kelsey-Hayes Canada’s behest. Magnuson has the right to pursue and collect from any tortfeasor of his choosing regardless of any right to contribution that the Kelsey-Hayes defendants may or may not have. See Missouri Pac. R. Co. v. Whitehead & Kales Co., 566 S.W.2d 466 (Mo. banc. 1978). The court in Whitehead & Kales, after discussing relative fault among joint tortfeasors concluded that a “[pjlaintiff continues free to sue one or more concurrent tortfeasors as he sees fit_” Mat 474. A defendant may not object that plaintiff has not sued another tortfeasor or that the case against that tortfeasor has been dismissed. Stith v. J.J. Newberry Co., 336 Mo. 467, 79 S.W.2d 447, 462 (1934).

The order of new trial, along with the reversal of the interlocutory order dismissing Chrysler, had the effect of reinstating Magnuson’s claim against Chrysler, a claim that Magnuson does not wish to pursue. *453 Magnuson s subsequent voluntary dismissal terminated its cause of action against Chrysler making it “as if the suit had never been filed.” Oney v. Pattison, 747 S.W.2d 137, 139 (Mo. banc 1988). There is nothing left to be re-tried; Magnuson’s claims against Kelsey-Hayes and Kelsey-Hayes Canada have already been adjudicated.

Kelsey-Hayes and Kelsey-Hayes Canada maintain that the order granting a new trial is not moot because “[a] cause of action is moot when the question presented for decision seeks a judgment upon some matter which, if judgment were rendered, would not have any practical effect upon any then existing controversy.” Kracman v. Ozark Elec. Co-Op., Inc., 816 S.W.2d 688, 690 (Mo.App.1991) (quoting Bank of Washington v. McAulliffe, 676 S.W.2d 483, 487 (Mo. banc 1984)). Kelsey-Hayes and Kelsey-Hayes Canada argue that the outcome of this appeal has a practical effect on the rights of the Kelsey-Hayes defendants.

Magnuson has already sued the Kelsey-Hayes defendants. The defendants cannot force Magnuson to sue Chrysler. The only existing controversies between the Kelsey-Hayes defendants and Magnuson are as to submissibility and remittitur. Any additional existing controversies would be between the Kelsey-Hayes defendants and Chrysler on the issues of indemnity and contribution.

Kelsey-Hayes and Kelsey-Hayes Canada contend that even if Magnuson is shown to have made a submissible case on any of his claims, the trial court properly granted the motion for new trial because the dismissal of Chrysler was prejudicial to them in that it deprived them of their right to have Chrysler’s relative degree of fault considered by the jury. The Kelsey-Hayes defendants claim that they were entitled to have Chrysler’s fault compared at trial because this was really a design case and Chrysler was co-designer of the wheel. They point to evidence as to the thinning of the steel forming the wheel claiming that this thinning was a design flaw in that no minimum specification covered that thinning.

If Kelsey-Hayes and Kelsey-Hayes Canada had desired a relative determination of Chrysler’s fault, they had the option to file a cross-claim prior to Chrysler’s dismissal or a third-party petition after Chrysler was dismissed. Whitehead & Kales, 566 S.W.2d at 473. They did neither. The only mention of any type of apportionment requested by Kelsey-Hayes and Kelsey-Hayes Canada is in their answers to Magnuson’s first amended petition where each “specifically requests the Court to determine the comparative fault of the plaintiff and all other parties to this action_” (emphasis added). After its original dismissal Chrysler was no longer a party to this action, so even under the theory espoused by the Kelsey-Hayes defendants, the language in said answers was not sufficient to obtain a determination of relative fault. Additionally, a claim for contribution constitutes “a separate independent claim from the underlying tort claims....” Asher v. Broadway-Valentine Center, Inc., 691 S.W.2d 478, 481 (Mo.App.1985). A defendant who has not filed a cross-claim may not complain of errors when the only effect of such errors is to absolve a co-defendant of liability, thereby preventing a joint verdict. Id.

The Kelsey-Hayes defendants contend, however, they were specially prejudiced by Magnuson’s representation that he was proceeding on a manufacturing defect while Magnuson’s evidence at trial supported the theory of design defect. They allege prejudice resulted because Chrysler was not a party and, although the Kelsey-Hayes defendants could not force Magnu-son to join Chrysler, the misleading nature of the representations of theories of recovery precluded their filing a third-party petition against Chrysler. They claim this misrepresentation prevented them from exercising their right to have the same jury that found liability and determined damages consider the apportionment of fault and determine contribution.

The Kelsey-Hayes defendants allowed the presentation of evidence, now claimed to be on design defect, without objection. They stood mute while the evidence was *454 admitted, they extensively cross-examined the witness, Dr. Bohl, on the issue and then on the next day of trial claimed prejudice and requested a mistrial. The Kelsey-Hayes defendants stated that the only remedy they were pursuing was a mistrial. They cannot permit the introduction of such evidence and then argue it injected issues which prejudiced them. Cargill v. Armocido, 476 S.W.2d 506, 508 (Mo. banc 1972). By their conduct, the Kelsey-Hayes defendants are precluded from relief, particularly the drastic remedy of a mistrial.

Additionally, the Kelsey-Hayes defendants cannot show how they were prejudiced as they still can maintain a separate action against Chrysler for contribution. The Kelsey-Hayes defendants claim that they are arguably barred from taking such a step because Chrysler would defend as an “innocent seller” under § 537.762, RSMo Supp.1987. The trial court vacated its ruling that Chrysler was an innocent seller when it vacated its dismissal of Chrysler. Therefore, there is no judicial determination that Chrysler was an innocent seller upon which Chrysler could base a claim of res judicata.

Whether or not Chrysler could prevail, in a separate action for contribution, on an anticipated claim that it is an innocent seller, is a question not properly before this court. Nor is the prospect that Chrysler, in an independent action, will attempt to claim that it is not liable for contribution because the jury verdict was based upon its finding that the wheel contained a “manufacturing defect” and it did not manufacture the wheel. The language of the instruction submitted to the jury does not resolve the question of whether the testimony at trial was in fact of a design defect or a manufacturing defect. It is noted that under Missouri law the required elements for submission of a claim of manufacturing defect mirror the elements for submission of a claim of design defect. The verdict-directing instruction MAI No. 25.04 is intended to be used in both manufacturing and in design defect cases. No distinction is made between the proof necessary to support either claim. Although this appears to result in a blurring of the two theories of recovery, it may well have been intentional, or at least condoned, as evidenced by the MAI instruction. In any event, the question of whether Chrysler will be liable for contribution to the Kelsey-Hayes defendants on any theory of recovery is relegated to a subsequent action.

In additional points, Kelsey-Hayes and Kelsey-Hayes Canada have challenged the submissibility of Magnuson’s claims. They are allowed to make this challenge under the authority of Boyer v. Grandview Manor Care Center, Inc., 793 S.W.2d 346, 347 (Mo. banc 1990), which explains:

In Robbins v. Jewish Hospital of St. Louis, 663 S.W.2d 341 (Mo.App.1983), the jury returned a verdict favoring plaintiff. The trial court denied defendant’s motion for judgment notwithstanding the verdict, but sustained the motion for new trial because it erroneously admitted evidence and erred in giving certain instructions. Plaintiff appealed the order granting the new trial; defendant appealed the denial of its motion for judgment notwithstanding the verdict. The court dismissed defendant’s appeal.

The granting of defendant’s motion for new trial erased the judgment against defendant; therefore, neither a final appealable judgment exists nor is defendant an aggrieved party with standing to appeal ... Nevertheless, where a plaintiff appeals asserting error by the trial court, the defendant may contest the issue of submissibility of plaintiff’s case, an issue inherent in every appeal.

(Emphasis added). Id. at 344. The court went on to rule the submissibility question despite its dismissal of defendant’s appeal.

Kelsey-Hayes and Kelsey-Hayes Canada claim that Magnuson failed to make a submissible case of strict liability or negligent manufacturing defect. Where the submissibility of a plaintiff’s case is challenged on appeal, the evidence is viewed in the light most favorable to the verdict together with all reasonable inferences drawn from the evidence. Coulter v. *455 Michelin Tire Corp., 622 S.W.2d 421, 426 (Mo.App.1981).

The Missouri Supreme Court, in addressing strict liability product claims, instructs that “the sole subject of inquiry is the defective condition of the product and not the manufacturer’s knowledge, negligence, or fault.” Elmore v. Owens-Illinois, Inc., 673 S.W.2d 434, 438 (Mo. banc. 1984). Strict liability in tort imposes upon a manufacturer a duty not to introduce an unreasonably dangerous product into commerce, whether the danger arises from defective manufacture, defective design, or a failure to warn of danger. Peters v. Johnson & Johnson Products, Inc., 783 S.W.2d 442, 444 (Mo.App.1990). Breach of the manufacturer’s duty occurs with the introduction of that product into commerce. Id.

In the instant case Magnuson has made a submissible case on the strict liability theory, demonstrating through the testimony of Dr. Bohl, that the wheel was defectively manufactured. Dr. Bohl testified that the steel used in manufacturing the wheel contained carbide films, making the steel brittle, subject to fracture and thereby dangerous. He also testified that the steel was formed too thin, some 30% thinner than original steel blank. Dr. Bohl’s testimony makes a submissible case for the jury under Missouri law; that the steel forming the wheel was dangerously defective and that the wheel contained the defect when Kelsey-Hayes and Kelsey-Hayes Canada introduced it into commerce.

Kelsey-Hayes and Kelsey-Hayes Canada rely on cases from other jurisdictions, most notably Lombard v. Centrico, Inc., 161 A.D.2d 1071, 557 N.Y.S.2d 627, 628 (App.Div.1990), for the proposition that in order to make a submissible case as to a manufacturing defect, the plaintiff must establish that the product does not conform to design specifications. The Kelsey-Hayes defendants attempt to engraft such a requirement in Missouri relying on Duke v. Gulf & Western Mfg. Co., 660 S.W.2d 404, 411 (Mo.App.1983), wherein this court observes:

In a manufacturing defect case the jury can rather easily determine whether a single product conforms to the intended design. The jury in a design defect case, however, may have no external standard by which to measure the design, or at least the standard with which the design should have complied is not always naturally apparent.

This observation does not mandate that a showing of non-conformity to design be made in a strict liability manufacturing defect case. It is illustrative of one form of proof which may be presented, contrasting that to the lack of external standards in design cases.

As previously noted, Missouri does not require such proof as a threshold of sub-missibility. The Missouri Supreme Court adopted 2 Restatement Torts, Second, § 402 A 3 in Keener v. Dayton Elec. Mfg. Co., 445 S.W.2d 362 (Mo.1969). Review of Missouri case law reveals that this is the standard applied in strict liability cases. 4 Magnuson made a submissible case showing that the Kelsey-Hayes defendants manufactured the wheel; it was in a defective condition unreasonably dangerous when put to a reasonably anticipated use; was in fact used in such a manner; and *456 that his injuries are a direct result of such defect as existed when the wheel was sold.

The Kelsey-Hayes defendants point out that the wheel had seen heavy use over a period of some eleven years. They contend that, as a matter of law, this use precludes a finding of manufacturing defect. This is a factual question for the jury. Magnuson provided evidence that the flaw in the mi-crostructure of the wheel, not the wear and tear from use, was the cause of the accident. From this the jury could find that the defect occurred before the wheel left the Kelsey plant. The cases cited by the Kelsey-Hayes defendants to support their contention are not on point as they are based upon the relationship between the wear and tear on a product and the accident. 5 The evidence in the instant case supports the theory that the wheel was defective at the time of its manufacture and that the defect, and not simply wear and tear, caused the wheel to fracture.

The Kelsey-Hayes defendants further challenge the denial of their JNOY motions claiming Magnuson failed to make a sub-missible case on both negligent manufacturing defect and strict liability failure to warn. Review of these claims of error is rendered unnecessary. This court’s holding that Magnuson made a submissible case on his strict liability manufacturing defect theory supports the award of damages. The alternative theories of negligent manufacturing defect and strict liability failure to warn are extraneous issues. This is in contrast to the general principle that when submitting to a jury multiple theories of recovery upon a single injury, there must be a submissible case on each theory of liability. Johnson v. Bush, 418 S.W.2d 601, 606 (Mo.App.1967); MAI 1.02. The verdict form under MAI No. 36.01 necessitates such, because it does not reveal upon which of the theories of recovery the jury reached its consensus. The non-MAI verdict form utilized in this case, however, required that the jury make a separate finding on each theory of recovery, clearly establishing a finding by the jury of liability upon the Kelsey-Hayes defendants under the strict liability manufacturing defect theory. 6 The amount of damages awarded by the jury would not be impacted whether the jury found for Magnuson upon one theory of recovery or upon three.

Finally, upon a reversal of the order granting a new trial, the Kelsey-Hayes defendants ask this court to remand the cause to the trial court for consideration of their motions for remittitur. They claim that when the trial court granted their motions for new trial, their alternate motions for a grant of remittitur became moot. Magnuson cites Burke v. Moyer, 621 5.W.2d 75, 82 (Mo.App.1981), for the proposition that when a new trial is granted on a specified basis, that ruling constitutes an overruling of all of the other grounds asserted by the movant. In dicta in Burke, this court addressed the failure of the trial court to expressly rule on the motion for remittitur, granting a new trial on other grounds. The view expressed was that the trial court’s failure to rule on remittitur “results in that motion being deemed to have been overruled.” Id. at 82. It is noted that the question of remittitur was not briefed in Burke.

Kelsey-Hayes and Kelsey-Hayes Canada argue that the motions for remittitur were not considered by the trial court because the motions for remittitur became moot upon the granting of a new trial. This position, although contrary to the discussion in Burke, follows the rationale of the case law cited therein. Both Smith v. Courier, 531 S.W.2d 743, 747 (Mo. banc 1976), and Hoehn v. Hampton, 483 S.W.2d 403, 407 (Mo.App.1972), discuss *457 the general rule that the grant of a new trial on a specified basis constitutes the overruling of any other grounds asserted in the motion for new trial. Remittitur is a separate motion; it is not a ground for new trial. The two are mutually exclusive. The trial court could completely overrule the motion for new trial but still order remittitur. A consideration of one of the motions is, therefore, not necessarily a consideration of the other. It should not be assumed that the trial court took the opportunity to consider the issue of remittitur.

Remittitur is a construct with a separate identity than that of excessiveness as a ground for new trial. In Blevins v. Cushman Motors, 551 S.W.2d 602, 614-15 (Mo. banc 1977), the Missouri Supreme Court distinguished these two objections to a jury’s verdict, saying:

[Tjhere are two different objections which can be made against a jury verdict. One is that it is merely excessive, and in that case the verdict may be cured by remittitur. The second type of objection is that the verdict is excessive as the product of bias and prejudice on the part of the jury; and in this case the exces-siveness cannot be cured by remittitur, but requires a new trial.

Although excessiveness of the verdict as grounds for a new trial was deemed overruled by the trial court’s action in the instant case, the motion for remittitur was not, as remittitur involves a process not amounting to a new trial.

Upon this court’s determination that the order granting a new trial would be reversed and the judgment reinstated, the motions for remittitur were, in fact, ripe for consideration. It was necessary for this court to address whether the motions for remittitur should be ruled by this court on appeal or by the trial court on remand. It is generally preferable for the trial court to initially consider the issue, even though under § 537.068, RSMo Cum.Supp.1991, this court has the authority to rule without remand. The statute provides:

A court may enter a remittitur order if, after reviewing the evidence in support of the jury’s verdict, the court finds that the jury’s verdict is excessive because the amount of the verdict exceeds fair and reasonable compensation for plaintiff’s injuries and damages. A court may increase the size of a jury’s award if the court finds that the jury’s verdict is inadequate because the amount of the verdict is less than fair and reasonable compensation for plaintiff’s injuries and damages.

Section 537.068, RSMo Cum.Supp.1991.

The trial court did not reach the question of whether remittitur is proper under the facts of the instant case and thus did not exercise its “broad discretion” on the matter. See Hall v. Superior Chem. & Fertilizer, Inc., 819 S.W.2d 422, 425 (Mo.App.1991). This court declined to consider the matter before the trial court had the opportunity to do so, because this court has not been adequately briefed on the matter. Additionally, neither party asked this court to consider remittitur. The Kelsey-Hayes defendants, in fact, asserted that the trial court was the appropriate tribunal to decide the question.

In order to promote judicial economy and avoid a second appeal, this court, while retaining jurisdiction of the appeal and pri- or to issuing an opinion, remanded the matter to the trial court to allow the trial court to rule on the remittitur motions. On August 31, 1992, the trial court entered its order overruling the motions. The parties were thereafter permitted to raise additional issues on appeal to address the trial court’s order denying remittitur.

The Kelsey-Hayes defendants argue, in the one point on appeal raised in their supplemental brief, that the trial court erred in denying their motions for remitti-tur because the verdict is clearly excessive and exceeds fair and reasonable compensation for Magnuson’s injuries.

In reviewing the trial court’s denial of remittitur, this court must afford the trial judge broad discretion. Hall v. Superior Chemical & Fertilizer, 819 S.W.2d at 425. The trial court is in the best position to weigh the evidence regarding the Kelsey-Hayes defendants’ motions for remitti-tur. Id. On appeal, the trial court’s ruling *458 should not be disturbed unless it constitutes an arbitrary abuse of discretion. Id.

Whether an award resembles awards given in other cases with similar facts is no longer the sole consideration in evaluating the excessiveness of an award. Larabee v. Washington, 793 S.W.2d 357, 360 (Mo.App.1990). Each case must be viewed in light of its own facts and in light of the following factors: (1) present and future loss of income, (2) medical expenses, (3) plaintiff’s age, (4) the nature and severity of the injuries, (5) economic factors, (6) awards given in similar cases, and (7) the superior opportunity of the trial court and jury to appraise plaintiff’s injuries and other damages. Id.

In the case at hand, both the judge and the jury viewed Magnuson at trial, an opportunity which this court did not have. The jury awarded Magnuson $4,750,000 and the trial court, on remand, agreed that this amount was fair and reasonable compensation for Magnuson’s injuries. The Kelsey-Hayes defendants have argued extensively that the jury’s award is excessive when compared with other cases with similar facts, however, that is not the sole test. The factors in Larabee are to be considered in addition to such a comparison. The trial court was in the best position to weigh the evidence regarding such factors. This court will not disturb the trial court’s ruling because it does not constitute an arbitrary abuse of discretion.

The order for new trial is reversed and the cause is remanded to the trial court with directions that the judgment in favor of Magnuson and against Kelsey-Hayes and Kelsey-Hayes Canada be reinstated.

All concur.

*49 WIEAND, Judge.

In this product liability action, the sole issue is the sufficiency of the circumstances surrounding a malfunction of an automobile to establish prima facie the existence of a manufacturing defect. The trial court determined the evidence to be insufficient and entered summary judgment in favor of the manufacturer. After careful review, we reverse.

An order granting summary judgment is appropriate when a review of all the interrogatories, affidavits and depositions of record indicates that there are no genuine issues of material fact and the moving party is entitled to judgment as a matter of law. Buckno v. Penn Linen & Uniform, Service, Inc., 428 Pa.Super. 568, 565, 631 A.2d 674, 675 (1993). However, “[a]n entry of summary judgment may be granted only in cases where the right is clear and free of doubt.” Musser v. Vilsmeier Auction Co., Inc., 522 Pa. 367, 370, 562 A.2d 279, 280 (1989). When considering whether summary judgment is proper, the record must be examined in the light most favorable to the non-moving party, with all doubts resolved against the moving party. Denlinger, Inc. v. Dendler, 415 Pa.Super. 164, 170, 608 A.2d 1061, 1064 (1992).

On November 23, 1984, Wilma Ducko was driving a newly purchased, 1985 Chrysler Fifth Avenue southwardly on the Atlantic City Expressway, in New Jersey, when the vehicle suddenly jerked to the right. Although she tried with all her strength to straighten the course of her vehicle, she said, the steering felt as though it had locked. When she attempted to apply the brakes, they also failed to respond. The vehicle, which had been moving at a speed of 55 m.p.h., travelled across the highway, down an embankment and into a group of trees. Ducko received serious injuries, including a broken back, and the vehicle was totalled. Prior to the accident, the vehicle had been driven 1,655 miles; it had been purchased less than two months before. The road surface at the time of the accident was dry.

An expert employed by the plaintiffs found no specific defect in the vehicle. He opined that Mrs. Ducko’s accident *50 had been caused by a transient malfunction of the system providing power to the steering and brakes. Chrysler’s expert, however, observed that both steering and brakes were operational, and he found no abnormalities in any of the car’s systems. He said that at a speed of 55 m.p.h. even a temporary power failure would not have rendered the steering uncontrollable. He expressed the opinion that the accident was a result of operator error.

When advancing a theory of strict product liability, a plaintiff has the burden of showing that the product was defective, that the defect was the proximate cause of his or her injuries and that the defect existed at the time the product left the manufacturer. Woodin v. J.C. Penney Co., Inc., 427 Pa.Super. 488, 490, 629 A.2d 974, 975 (1993); Vernon v. Stash, 367 Pa.Super. 36, 48, 532 A.2d 441, 447 (1987); Swartz v. General Elec. Co., 327 Pa.Super. 58, 66, 474 A.2d 1172, 1176 (1984). In certain cases of alleged manufacturing defects, however, the plaintiff need not present direct evidence of the defect. When proceeding on a 'malfunction theory, the plaintiff may “present a case-in-chief evidencing the occurrence of a malfunction and eliminating abnormal use or reasonable, secondary causes for the malfunction.” O’Neill v. Checker Motors Corp., 389 Pa.Super. 430, 435, 567 A.2d 680, 682 (1989). See also: Knight v. Otis Elevator Co., 596 F.2d 84, 89 (3d Cir.1979). From this circumstantial evidence, a jury may be permitted to infer that the product was defective at the time of sale. Vernon v. Stash, supra 367 Pa.Super. at 48, 532 A.2d at 448. This was summarized in Woodin v. J.C. Penney Co., Inc., supra, as follows:

Generally, a plaintiff will produce direct evidence of a product’s defective condition. In those cases where the plaintiff is unable to prove the precise nature of the product’s defect, however, he may, in some cases, rely on the “malfunction theory” of product liability. Rogers v. Johnson & Johnson Products, Inc., 523 Pa. 176, 182, 565 A.2d 751, 754 (1989). The malfunction theory allows the plaintiff to use circumstantial evidence to establish a defective prod *51 uct. In Rogers v. Johnson & Johnson Products, Inc., supra, the court stated:

[The malfunction theory] permits a plaintiff to prove a defect in a product with evidence of the occurrence of a malfunction and with evidence eliminating abnormal use or reasonable, secondary causes for the malfunction. [The plaintiff is relieved] from demonstrating precisely the defect yet it permits the trier-of-fact to infer one existed from evidence of the malfunction, of the absence of abnormal use and of the absence of reasonable, secondary causes.

Id. (citations omitted). Although proof of a specific defect is not essential to establish liability under this theory, the plaintiff cannot depend upon conjecture or guesswork. “The mere fact that an accident happens, even in this enlightened age, does not take the injured plaintiff to the jury.” Stein v. General Motors Corp., 58 D. & C.2d 193, 203 (Bucks 1972), aff'd, 222 Pa.Super. 751, 295 A.2d 111 (1972).

Id., 427 Pa.Super. at 492, 629 A.2d at 975-976. The malfunction theory, thus, does not relieve the burden of establishing a defect. Harkins v. Calumet Realty Co., 418 Pa.Super. 405, 418, 614 A.2d 699, 706 (1992). See also: Sochanski v. Sears, Roebuck & Co., 689 F.2d 45, 50 (3d Cir.1982). However, “[t]he malfunction itself is circumstantial evidence of a defective condition, ...” D’Antona v. Hampton Grinding Wheel Co., Inc., 225 Pa.Super. 120, 124, 310 A.2d 307, 309 (1973). When a party relies on the malfunction of a product to prove that it was defective, testimony identifying the exact nature of the alleged defect is not essential. “Although it is helpful for a plaintiff to have direct evidence of the defective condition which caused the injury or expert testimony to point to that specific defect, such evidence is not essential in a strict liability case based on § 402A [of the Restatements (Second) of Torts].” Cornell Drilling Co. v. Ford Motor Co., 241 Pa.Super. 129, 136, 359 A.2d 822, 825 (1976) (emphasis in original). See also: Kuisis v. Baldwin-Lima-Hamilton Corp., 457 Pa. 321, 329, 319 A.2d 914, 920 (1974). Where the *52 alleged malfunction occurs shortly after the product has been delivered to the user, the inference that the defect originated with the manufacturer is stronger. Cornell Drilling Co. v. Ford Motor Co., supra 241 Pa.Super. at 140, 359 A.2d at 827.

In MacDougall v. Ford Motor Co., 214 Pa.Super. 384, 257 A.2d 676 (1969), the plaintiff had purchased her car a little more than a month before and had driven the vehicle only one hundred and forty-three (143) miles prior to the accident. According to the plaintiff’s testimony, while driving on the Pennsylvania Turnpike, at approximately fifty (50) or sixty (60) miles per hour, the steering on her vehicle failed to respond, and her car veered off the road and onto the berm. The plaintiffs expert examined the vehicle and observed irregularities in the steering mechanism which, he said, might have been the cause of the problem described by the plaintiff. He was unable, however, to confirm positively that the irregularities had been the cause of the accident. After the close of all evidence, the jury returned a verdict in favor of the plaintiff. The defendant’s motion for judgment n.o.v. was denied by the trial court. On appeal, the defendant argued that the plaintiff had not shown any specific mechanical defect which had caused the accident. The Superior Court, concluding that the plaintiff had met her burden of proof through circumstantial evidence, said, “proof of the specific defect in construction or design causing a mechanical malfunction is not an essential element in establishing breach of warranty.” Id. at 389, 257 A.2d at 679. Therefore, the Court held that Mrs. MacDougall’s testimony of the bizarre steering action, prior to and at the time of the accident, “established] a mechanical malfunction in the absence of abnormal use which prevented her from maintaining control of the car.” Id. at 391, 257 A.2d at 680. As to the significance of her expert’s opinion, the Court said, “[A]lthough proof of specific steering defects was not a prerequisite to establishing liability, we note that [plaintiffs expert’s] testimony ... lends definite additional support to this inference.” Id. at 391-392, 257 A.2d at 680 (emphasis added).

Later, in Agostino v. Rockwell Mfg. Co., 236 Pa.Super. 434, 345 A.2d 735 (1975), the Superior Court again considered the *53 burden of proof required under the malfunction theory. In that case, the plaintiff had purchased a power saw equipped with a blade guard less than a month before his injury and had used it only ten times. On the day of the accident he successfully cut three boards and was attempting to rest the saw on a work bench when the saw blade caught his trousers and cut his thigh. The plaintiff testified that the blade guard had appeared to be jammed. At trial, the court granted a compulsory non-suit in favor of the defendant, holding that there was no evidence to establish that the saw had malfunctioned. The Superior Court reversed and remanded for a new trial. Because the plaintiff had said that the blade guard appeared to have been jammed, the Court held that “[t]he jury could have properly inferred from [the plaintiffs] testimony that the [blade] guard did not function properly.” Id. at 444, 345 A.2d at 740.

In Brill v. Systems Resources, Inc., 405 Pa.Super. 603, 592 A.2d 1377 (1991), the Superior Court again considered a plaintiffs evidentiary burden under the malfunction theory. There the plaintiff had been injured when the chair in which he had been sitting suddenly collapsed. At trial the plaintiff offered his own testimony concerning the events of his fall and the testimony of an eyewitness. The defendant, on the other hand, presented an expert who had examined the chair and opined that it was not defective. He concluded that the accident must have been the result of abnormal use by the plaintiff. The trial court refused to instruct the jury on the malfunction theory, and the jury returned a verdict for the defendant. On appeal, the Superior Court reversed and remanded for a new trial. Because the jury could have inferred a defect from the plaintiffs testimony, the Court held that a jury instruction on the malfunction theory was required. It said:

Thus, the record contains conflicting evidence regarding abnormal use. The determination of which version of events to credit, of course, was for the jury. See Commonwealth v. Paquette, 451 Pa. 250, 301 A.2d 837 (1973) (credibility determinations within exclusive province of finder of *54 fact). If the jury chose to credit appellant’s testimony, it could have found that there was no abnormal use or other reasonable secondary cause for the chair to collapse---We therefore conclude that appellant was entitled to have the jury instructed that if it found no abnormal use, it could find that the chair malfunctioned.

Id. 405 Pa.Super. at 607-608, 592 A.2d at 1379.

In granting appellee’s motion for summary judgment in the instant case, the trial court relied upon the deposition testimony and reports submitted by Chrysler’s expert. This was error. Mrs. Ducko’s testimony of the erratic performance of the vehicle’s steering and braking systems, under the circumstances of this case, was sufficient to make out a prima facie case of a manufacturing defect in the vehicle. The issue of strict liability, therefore, was a disputed issue for the jury. Although a jury, after considering the testimony of appellee’s expert witnesses, may find that the vehicle was not defective and that the accident was caused by operator error, it was improper for the trial court to make such a determination summarily and as a matter of law.

Our decision, as we have attempted to demonstrate, is compelled by prior decisions of the Superior Court. It is also consistent with decisions in other jurisdictions which have adopted the malfunction theory and allow proof of a defectively manufactured product by circumstantial evidence. See, e.g.: Sochanski v. Sears, Roebuck & Co., 621 F.2d 67 (3d Cir.1980); Stewart v. Ford Motor Co., 553 F.2d 130 (D.C.Cir.1977); Higgins v. General Motors Corp., 287 Ark. 390, 699 S.W.2d 741 (1985); Wakabayashi v. Hertz Corp., 66 Haw. 265, 660 P.2d 1309 (1983); Gillespie v. R.D. Werner Co., Inc., 71 Ill.2d 318, 17 Ill.Dec. 10, 375 N.E.2d 1294 (1978); Stackiewicz v. Nissan Motor Corp. in U.S.A., 100 Nev. 443, 686 P.2d 925 (1984); Moraca v. Ford Motor Co., 66 N.J. 454, 332 A.2d 599 (1975); Brownell v. White Motor Corp., 260 Or. 251, 490 P.2d 184 (1971).

The order of the trial court is reversed and the case is remanded for further proceedings consistent with this opinion. Jurisdiction is not retained meanwhile.

Donovan, Judge.

{¶ 1} Plaintiff-appellants, Donegal Mutual Insurance Company, David Nearon, and Susan Nearon (“plaintiffs”), appeal from a decision of the Darke County Court of Common Pleas that partially sustained a motion for judgment notwithstanding the verdict (“JNOV”) filed by defendant-appellee, White Consolidated Industries, Inc. (“White”) after a jury trial held on September 13, 14, and 15, 2004. Plaintiffs also appeal from the trial court’s decision to reduce the amount of damages awarded them by the jury as a result of its partial grant of JNOV. Lastly, plaintiffs argue that the trial court erred when it overruled their motion for prejudgment interest.

{¶ 2} In its cross-appeal, White argues that the trial court erred when it refused to strike the testimony of plaintiffs’ expert, Bernard Doran. Additionally, White asserts that the trial court erred when it denied its motion for directed verdict and its motion for JNOV with respect to plaintiffs’ manufacturing-defect claim, because plaintiffs failed to present sufficient evidence on that cause of action. In its final cross-assignment, White contends that the trial court erred by overruling its motion for new trial.

I

{¶ 3} Initially, it should be noted that this is the third occasion in which the instant matter has appeared before us.

{¶ 4} On May 11, 1995, a fire destroyed the residence of David and Susan Nearon. After an investigation concerning the cause of the fire, Donegal Mutual Insurance, the insurer of the home, paid the Nearons $130,712.48 for the damages that they had suffered.

*575 {¶ 5} In May 1997, the Nearons filed a lawsuit against White, in which they claimed that a Frigidaire electric stove had caused the fire that destroyed their home. After the trial court ordered that Donegal be joined as a party to assert any existing subrogation interests, the Nearons voluntarily dismissed the cause without prejudice on May 11,1999.

{¶ 6} The Nearons refiled the action on May 11, 2000. After determining that they had failed to abide by the terms of a pretrial scheduling order, the trial court sanctioned plaintiffs by dismissing their cause of action. In Donegal Mut. Ins. Co. v. White Consol. Industries (Oct. 26, 2001), Darke County App. No.2001 CA 1549, 2001 WL 1295362, we reversed that dismissal and remanded the cause for trial.

{¶ 7} In early September 2002, the first trial was held, and the jury found in favor of White with respect to plaintiffs’ product-liability claims. Plaintiffs moved for a new trial, arguing that White’s expert witnesses were not qualified to testify in the state of Ohio. The trial court denied the motion, and plaintiffs appealed the matter to this court. We reversed the decision of the trial court and remanded the cause for a new trial. See Donegal Mut. Ins. Co. v. White Consol. Industries (2003), 153 Ohio App.3d 619, 2003-Ohio-4202, 795 N.E.2d 133.

{¶ 8} The cause proceeded to trial a second time on September 13, 2005. At the close of the trial, the jury found in favor of plaintiffs on both a design-defect and a manufacturing-defect theory of product liability with respect to the allegedly malfunctioning electric stove built by White. On the design-defect claim, the Nearons were awarded $15,000, and Donegal was awarded $37,000. The jurors separately awarded the Nearons $15,000 and awarded Donegal $37,000 on the manufacturing-defect claim, for a grand total of $104,000 in damages.

{¶ 9} White subsequently moved for a new trial and for a JNOV, while the Nearons requested prejudgment interest. As mentioned above, the trial court overruled White’s motion for a new trial but sustained its motion for JNOV with respect to the Nearons’ design-defect claim. Pursuant to this ruling, the trial court reduced the monetary award to appellants by $15,000 for the Nearons and $37,000 for Donegal. Further, the trial court denied the Nearons’ motion for prejudgment interest. Accordingly, the trial court entered judgment for the Nearons for a total of $15,000 and in favor of Donegal for a total of $37,000.

{¶ 10} It is from this judgment that the Nearons and Donegal appeal. It is from the same judgment that White cross-appeals.

Plaintiffs’ Appeal

II

{¶ 11} Plaintiffs’ first assignment of error is as follows:

*576 {¶ 12} “The trial court erred in granting defendant’s motion for judgment notwithstanding the verdict as to the jury’s finding of a design defect in that the court failed to consider all of the evidence most favorable to the non-moving party.”

{¶ 13} In their first assignment, plaintiffs contend that the trial court abused its discretion when it sustained White’s motion for JNOV with respect to the claim for design defect. Specifically, plaintiffs assert that in light of the evidence presented at trial on the issue, the jury reasonably found the existence of a design defect in the control switch on the electric stove. Thus, it was error for the trial court to set aside the jury’s finding that the electric stove suffered from a design defect that caused the fire that destroyed the Nearons’ residence. We agree.

{¶ 14} The standard for reviewing a trial court’s decision concerning a JNOV is the same as that applicable to a motion for directed verdict. Nickell v. Gonzalez (1985), 17 Ohio St.3d 136, 137, 17 OBR 281, 477 N.E.2d 1145. In ruling upon a motion for JNOV, “the court must construe the evidence most strongly in favor of the party against whom the motion is made. Where the evidence is such that reasonable minds may derive differing conclusions therefrom, the motion must be denied. The court shall not consider the weight of the evidence or the credibility of the witnesses in determining its ruling on the motion. In short, every effort must be made to uphold the verdict if reasonably possible.” Motorists Mut. Ins. Co. v. Hamilton Twp. Trustees (1986), 28 Ohio St.3d 13, 15, 28 OBR 77, 502 N.E.2d 204. The nonmovant is also given the benefit of all reasonable inferences from the evidence. Ruta v. Breckenridge-Remy Co. (1982), 69 Ohio St.2d 66, 68, 23 O.O.3d 115, 430 N.E.2d 935. Appellate review of a motion for JNOV is de novo.

{¶ 15} In order to establish the elements of a products-liability design-defect claim, a plaintiff must show that (1) there was a defect in the product manufactured and sold by the defendant, (2) the defect existed at the time the product left the defendant’s control, and (3) the defect was the direct and proximate cause of the plaintiffs injuries or losses. State Farm Fire & Cas. Co. v. Chrysler Corp. (1988), 37 Ohio St.3d 1, 5-6, 523 N.E.2d 489.

{¶ 16} At the close of trial, the court instructed the jury to apply the “consumer-expectation test” pertaining to design defect. Under the consumer-expectation test, a product is defective in design or formulation when it is “more dangerous than an ordinary consumer would expect when used in an intended or reasonably foreseeable manner.” Former R.C. 2307.75(A)(2), Sub.S.B. No. 108, 149 Ohio Laws, Part I, 382, 438. Moreover, the question of what an ordinary consumer expects in terms of the risks posed by the product is generally one for *577 the trier of fact. Welch Sand & Gravel, Inc. v. O & K Trojan, Inc. (1995), 107 Ohio App.Bd 218, 225, 668 N.E.2d 529, 533-534.

{¶ 17} Plaintiffs presented evidence of unsafe, unexpected product performance when they alleged that Susan Nearon merely set the electric stove to self-clean mode before it acted defectively and caused the fire that destroyed the Nearons’ residence. This is sufficient evidence from which a trier of fact may infer the existence of a product defect. However, satisfaction of the consumer-expectation test is not all a plaintiff must prove in a product-liability action premised on a design defect. As we' previously stated, a plaintiff must also present evidence that the defect was the direct and proximate cause of his injuries or losses.

{¶ 18} White argues that plaintiffs were unable to establish the element of proximate cause at trial because the court found that their expert, Bernard Doran, was not qualified to opine as to whether the electric stove was designed improperly and precluded him from testifying that the proximate cause of the fire was a design defect in the switch on the stove. Thus, White asserts that without expert testimony to that effect, plaintiffs failed to produce sufficient evidence demonstrating causation, and the trial court was correct in granting the JNOV with respect to the design-defect claim.

{¶ 19} In order to establish the third element of a design-defect claim, proximate causation, the plaintiff must prove by a preponderance of the evidence that some aspect of the challenged design rendered the product’s performance less safe than the ordinary consumer would expect, resulting in injury. Atkins v. Gen. Motors Corp. (1999), 132 Ohio App.3d 556, 725 N.E.2d 727, citing State Farm Fire & Cas. Co., 37 Ohio St.3d 1, 523 N.E.2d 489. Proof of causation must be by probability. Shumaker v. Oliver B. Cannon & Sons, Inc. (1986), 28 Ohio St.3d 367, 369, 28 OBB, 429, 504 N.E.2d 44. Thus, a plaintiff must offer proof that the allegedly defective condition was the most probable cause of his injuries or losses. Fogle v. Cessna Aircraft Co. (Jan. 16,1992), Franklin County App. No. 90AP-977, 1992 WL 10272. The proof necessary to adequately demonstrate that the defect was the proximate cause of the plaintiffs injuries may be either direct or circumstantial. State Farm Fire & Cas. Co., supra.

{¶ 20} Although Doran was ultimately not allowed to testify as to the exact nature of the design defect that was the proximate cause of the fire, plaintiffs presented abundant circumstantial evidence from which a trier of fact could readily infer that an electrical fault within the switch on the stove had caused the fire that destroyed the Nearons’ residence. In particular, the trial court allowed Doran to testify in detail that an electrical fault within the switch caused the fire. The fact that he was not allowed to testify as to the exact nature of the design *578 defect that caused the electrical fault that led to the fire is irrelevant. Insofar as the evidence suggests that an electrical fault within the switch on the Nearons’ stove caused the fire, a jury could properly conclude that the switch had been defectively designed.

{¶ 21} Construing the evidence most strongly in favor of the nonmoving party, the plaintiffs in this case, we find that the JNOV was improvidently granted. When plaintiffs are afforded the benefit of all reasonable inferences from the evidence, it is clear that the trial court erred, and the jury’s verdict should be reinstated with respect to the claim for design defect.

{¶ 22} Plaintiffs’ first assignment of error is sustained.

III

{¶ 23} Plaintiffs’ second assignment of error is as follows:

{¶ 24} “The trial court erred in reducing the jury’s verdict to $52,000.00 as a result of the court’s decision to partially grant defendant’s motion for JNOV, because even if there was insufficient evidence to support a design defect the jury’s finding of manufacturing defect independently supports the jury’s total verdict of $104,000.00.”

{¶ 25} In light of our holding with respect to plaintiffs’ first assignment, we find it unnecessary to address the merits of their second assignment, which is rendered moot. Having reversed the trial court’s decision granting White’s JNOV with respect to plaintiffs’ design-defect claim, we hold that the original monetary award totaling $104,000 is reinstated, and the trial court’s decision reducing that amount is hereby set aside.

IV

{¶ 26} Plaintiffs’ third assignment of error is as follows:

{¶ 27} “The trial court erred in denying plaintiffs’ motion for prejudgment interest because the court’s ruling is contrary to law.”

{¶ 28} In their third and final assignment, plaintiffs contend that the trial court abused its discretion when it denied a motion for prejudgment interest on the damages award. Specifically, plaintiffs argue that when it denied the motion, the trial court applied the incorrect test when determining whether to award prejudgment interest. Plaintiffs assert that an award of prejudgment interest is mandatory in the present case. We disagree.

{¶ 29} Decisions on the grant or denial of prejudgment interest are within the discretion of the trial court and will not be disturbed absent an abuse of that discretion. Wagner v. Midwestern Indemn. Co. (1998), 83 Ohio St.3d 287, *579 292, 699 N.E.2d 507. An abuse of discretion is “more than an error of law or judgment; it implies that the court’s attitude is unreasonable, arbitrary or unconscionable.” Blakemore v. Blakemore (1983), 5 Ohio St.3d 217, 219, 5 OBR 481, 450 N.E.2d 1140. Moreover, an abuse of discretion is a “perversity of will, passion, prejudice, partiality, or moral delinquency.” Pons v. Ohio State Med. Bd. (1993), 66 Ohio St.3d 619, 621, 614 N.E.2d 748. Under this standard, an appellate court may not merely substitute its judgment for that of the trial court. Id.

{¶ 30} In support of their position, plaintiffs cite this court’s decision in Heinz v. Steffen (1996), 112 Ohio App.3d 174, 678 N.E.2d 264, for the proposition that the trial court has no discretion to disallow an award of prejudgment interest when that award is necessary to make the aggrieved party whole. While this statement is certainly true, plaintiffs argue that this interpretation applies in both contract and tort cases. Heinz dealt with prejudgment interest in a contract action filed by a judgment debtor to collect a judgment pursuant to R.C. 1343.03(A).

{¶ 31} The holding in Heinz, however, does not extend to the grant or denial of prejudgment interest in tort cases. The trial court determines whether to award prejudgment interest in a tort case pursuant to R.C. 1343.03(C), which states:

{¶ 32} “Interest on a judgment, decree, or order for the payment of money rendered in a civil action based on tortious conduct and not settled by agreement of the parties, shall be computed from the date the cause of action accrued to the date on which money is paid if, upon motion of any party to the action, the court determines at a hearing held subsequent to the verdict or decision in the action that the party required to pay the money failed to make a good faith effort to settle the case and that the party to whom the money is to be paid did not fail to make a good faith effort to settle the case.”

{¶ 33} This statute was enacted to promote settlement efforts, prevent parties who engage in tortious conduct from frivolously delaying the ultimate resolution of cases, and encourage good faith efforts to settle controversies outside a trial setting. Kalain v. Smith (1986), 25 Ohio St.3d 157, 159, 25 OBR 201, 495 N.E.2d 572. A party has not “failed to make a good faith effort to settle” under R.C. 1343.03(C) if it has (1) fully cooperated in discovery proceedings, (2) rationally evaluated its risks and potential liability, (3) not attempted to unnecessarily delay any of the proceedings, and (4) made a good faith monetary settlement or responded in good faith to an offer from the other party. Id. If a party has a good faith, objectively reasonable belief that it has no liability, it need not make a monetary settlement offer. Id. The decision as to whether a party’s settlement efforts indicate good faith is generally within the sound discretion of the trial court. Id., citing Huffman v. Hair Surgeon, Inc. (1985), 19 Ohio St.3d *580 83, 19 OBR 123, 482 N.E.2d 1248. The burden of proof is on a party seeking prejudgment interest to establish the lack of good faith of the nonmoving party. Kalain, supra, at syllabus.

{¶ 34} After a thorough review of the record, we hold that the trial court followed the proper standard pursuant to R.C. 1343.03(C) and did not abuse its discretion when it denied plaintiffs’ motion for prejudgment interest. As previously mentioned, this is the third occasion on which this action has come before us on appeal. In the first trial in this matter in 2002, the jury rendered a verdict in favor of the defendants. Thus, it was reasonable for White to possess a good faith, objectively reasonable belief that it had no liability. Plaintiffs merely requested that White make a settlement offer to them. Additionally, plaintiffs never made an offer to settle with White. Lastly, plaintiffs have utterly failed to establish any facts that demonstrate bad faith on the part of White. Plaintiffs have not met their burden of proof pursuant to R.C. 1343.03(C). Accordingly, the trial court did not abuse its discretion when it overruled plaintiffs’ motion for prejudgment interest.

{¶ 35} Plaintiffs’ final assignment of error is overruled.

White’s Cross-Appeal

V

{¶ 36} White’s first assignment of error is as follows:

{¶ 37} “The trial court erred in denying the motion to strike the testimony of expert Bernard Doran.”

{¶ 38} In its first assignment of error, White contends that the trial court erred when it overruled the motion to strike the testimony of plaintiffs’ expert, Bernard Doran. Initially, White asserts that Doran’s opinions were unreliable because he was not competent to testify as an expert in manufacturing defects. White further asserts that Doran’s testimony should have been stricken because he had failed to test his hypothesis. Lastly, White argues that Doran’s opinions were unreliable because he did not offer them within a reasonable degree of scientific certainty.

{¶ 39} The trial court has broad discretion in determining the admissibility of expert testimony, and we may reverse only if the trial court abused its discretion. See Kumho Tire Co. v. Carmichael (1999), 526 U.S. 137, 152-153, 119 S.Ct. 1167, 143 L.Ed.2d 238. In general, courts should admit expert testimony whenever it is relevant and satisfies Evid.R. 702. State v. Nemeth (1998), 82 Ohio St.3d 202, 207, 694 N.E.2d 1332. Thus, the trial court must perform a “gatekeeping” role to insure that expert testimony is sufficiently (a) relevant and (b) reliable to justify its submission to the jury. Daubert v. Merrell Dow Pharma *581 ceuticals, Inc. (1993), 509 U.S. 579, 589, 113 S.Ct. 2786, 125 L.Ed.2d 469. In performing this role, the trial court must follow Evid.R. 702, which states:

{¶ 40} “A witness may testify as an expert if all of the following apply:

{¶ 41} “(A) The witness’ testimony either relates to matters beyond the knowledge or experience possessed by lay persons or dispels a misconception common among lay persons;

{¶ 42} “(B) The witness is qualified as an expert by specialized knowledge, skill, experience, training, or education regarding the subject matter of the testimony;

{¶ 43} “(C) The witness’ testimony is based on reliable scientific, technical, or other specialized information. To the extent that the testimony reports the result of a procedure, test, or experiment, the testimony is rehable only if all of the following apply:

{¶ 44} “(1) The theory upon which the procedure, test, or experiment is based is objectively verifiable or is validly derived from widely accepted knowledge, facts, or principles;

{¶ 45} “(2) The design of the procedure, test, or experiment reliably implements the theory;

{¶ 46} “(3) The particular procedure, test, or experiment was conducted in a way that will yield an accurate result.”

{¶ 47} In order to qualify as an expert, the witness need not be the best witness on the subject. Scott v. Yates (1994), 71 Ohio St.3d 219, 221, 643 N.E.2d 105. The expert must demonstrate some knowledge on a particular subject that is superior to that an ordinary juror possesses. Id., citing State Auto Mut. Ins. Co. v. Chrysler Corp. (1973), 36 Ohio St.2d 151, 160, 65 O.O.2d 374, 304 N.E.2d 891.

{¶ 48} The first issue in White’s motion to strike concerns whether Doran was qualified to present expert testimony on the manufacturing-defect claim. Doran’s testimony clearly established his qualifications in the field of fire investigation and forensic engineering. Doran also earned a bachelor’s degree in physics and has been a licensed professional engineer in Ohio since 1969. Doran further testified that he has an associate’s degree in electronic engineering. Moreover, his testimony dealt with a complex area of expertise of which a lay person would commonly have no knowledge. Thus, the requirements of Evid.R. 702(A) and (B) have been satisfied, and our review is focused solely upon the reliability of Doran’s testimony pursuant to Evid.R. 702(C).

{¶ 49} From Doran’s qualifications and testimony, it is clear that Doran was competent to provide reliable testimony with respect to plaintiffs’ manufacturing- *582 defect claim. Although Doran conceded that he was not an expert in the manufacture of electrical stoves, his background in electrical and forensic engineering provided him with the requisite expertise necessary to opine concerning the cause and origin of the fire in the stove, which he unequivocally stated was an electrical fault in the switch.

{¶ 50} Next, White argues that the motion to strike Doran’s testimony should have been granted because he had failed to test his hypothesis that the fire had originated in the switch as a result of an electrical fault that ignited contaminants that entered the switch through holes in the back of the switch itself. White contends that Doran’s- failure to perform any physical testing utilizing exemplar switches in order to approximate the alleged cause of the fire in the stove should have disqualified him as an expert in the case. We disagree.

{¶ 51} When questioned concerning his failure to perform physical testing, Doran stated:

{¶ 52} “Q: Okay. And what you’re telling us is that the NFPA 921, the current condition of the guide put forth by an organization to which you are a member recommends that fire cause and origin investigators utilize the scientific method. Is that what you are telling us?

{¶ 53} “A: Wait a minute. Scientific method is different than empirical testing. Testing, as they define it, can simply be a thought process, a thinking through of the evidence presented and evaluated mentally in arriving at a conclusion. It doesn’t necessarily mean buy a hundred stoves or switches or other items and testing them for years on end.”

{¶ 54} With respect to the procedure he followed, Doran testified that he took the subject stove apart and compared the failed switch with an exemplar switch that he also dismantled. Based on his knowledge and experience as a fire investigator and forensic engineer, Doran concluded that the fire had begun in the switch. After listening to arguments from both parties, the trial court found that Doran’s failure to test, while going to the weight of his testimony, was not determinative on the issue of its admissibility. Even White concedes in its brief that Doran’s failure to perform that testing was not determinative with respect to the admissibility of his testimony. Thus, the jury was left to decide how reliable Doran’s testimony was, in light of his failure to perform physical testing with exemplar switches like the one in the subject stove. Accordingly, the trial court did not abuse its discretion in this regard.

{¶ 55} Lastly, White argues that Doran’s testimony should be stricken because it was not offered within a reasonable degree of scientific certainty. White asserts that Doran “established nothing more than that his theory was merely a possibility.” We disagree.

*583 {¶ 56} During the direct examination of Doran, the following exchange occurred:

{¶ 57} “Q: Mr. Doran, I would like to ask you a couple of questions. Any time in this testimony that you’re asked to express an opinion, I’d like you to express that opinion to a reasonable degree of engineering probability as the standard. Can we agree to that?

{¶ 58} “A: Yes, of course.”

{¶ 59} White points out that Doran testified that contaminants entering through holes in the back of the switch were the probable cause of the electrical fault that, in turn, caused the fire in the stove. White argues that the trial court should have excluded that testimony because Doran could not explain the exact nature of the alleged electrical fault to any degree of certainty. However, Doran testified that absent a manufacturing defect in the switch, the fire would not have occurred. This is evident in the following exchange:

{¶ 60} “Q: We’ll get to that in a minute. I want to ask you if you have an opinion as to whether or not that switch had an electrical defect there?

{¶ 61} “Q: (Defense Counsel) Objection. Lack of foundation.

{¶ 62} “The Court: Overruled. Go ahead.

{¶ 63} “A: One would infer that the switch was defective in manufacture as it failed.

{¶ 64} “ * * *

{¶ 65} “Q: Is there any question in your mind that there was an electrical failure in that switch?

{¶ 66} “A: No. None whatsoever.

{¶ 67} “Q: No question.

{¶ 68} “A: No.

{¶ 69} “Q: And did the electrical failure in the switch cause the fire at the Nearon house?

{¶ 70} “A: Yes, sir.”

{¶ 71} In light of the foregoing testimony, it is clear that Doran expressed his opinions concerning the origin and cause of the fire within a reasonable degree of scientific certainty. Thus, it was not an abuse of discretion for the trial court to overrule White’s motion to strike the testimony of plaintiffs’ expert, Bernard Doran.

{¶ 72} White’s first assignment of error is overruled.

*584 VI

{¶ 73} White’s second assignment of error is as follows:

{¶ 74} “The trial court erred in denying the motion for directed verdict and motion for judgment notwithstanding the verdict on the manufacturing defect claim.”

{¶ 75} In its second assignment, White contends that the trial court erred when it denied its motions for directed verdict and JNOV with respect to plaintiffs’ manufacturing-defect claim. Specifically, White argues that plaintiffs presented no evidence that the range had been manufactured defectively. White also asserts that plaintiffs presented no evidence that the alleged defect was present when the stove left White’s control. Lastly, White argues that plaintiffs presented insufficient evidence that the alleged defect was the cause of the damage.

{¶ 76} A product is defective in manufacture or construction if, when it leaves the control of its manufacturer, it deviates in a material way from the design specifications, formula, or performance standards of the manufacturer, or from otherwise identical units manufactured to the same design specifications, formula, or performance standards. R.C. 2307.74. A product may be defective in manufacture or construction even if its manufacturer has exercised all possible care in its manufacture or construction. Id.

{¶ 77} In manufacturing-defect cases, evidence of unsafe, unexpected performance of a product is sufficient to infer existence of defect. State Farm Fire & Cas. Co. v. Chrysler Corp. (1988), 37 Ohio St.3d 1, 523 N.E.2d 489. However, that evidence must be accompanied by proof that the alleged defect was present when the product left the hands of the manufacturer and that it proximately caused the plaintiffs injuries. Id.

{¶ 78} In the instant case, plaintiffs presented evidence through the testimony of their expert, Doran, that the electric stove manufactured by White was the source of the fire that had burned down the Nearons’ residence. As previously stated, Doran testified at length that the origin of the fire was the left-hand control switch on the stove and that an electrical failure inside the switch caused the fire. Doran also testified that one could infer the existence of a manufacturing defect in the switch because it had failed.

{¶ 79} With respect to the element of proximate causation, Doran provided unequivocal testimony that the failure of the switch had caused the fire. He testified that heat from the failed switch had ignited the insulation on wires near the switch and that the fire had then spread to wood paneling behind the stove. That the fire was caused by the switch in the stove is further evidenced by the *585 testimony of White’s expert, John Heltman, who agreed that fire patterns on the back panel of the range indicated that hot gas trapped inside the panel left the rear of the panel through vents on the back of the stove. Plaintiffs argued at trial that this demonstrates that the fire had originated in the back of the stove where the failed switch was located. Thus, sufficient evidence was adduced at trial to demonstrate that the faulty switch in the stove was the proximate cause of the fire.

{¶ 80} Lastly, plaintiffs provided sufficient evidence that the defect in the switch was present when the product left the hands of the manufacturer, White. Absent substantial change in the condition in which the product was sold, it may be inferred that the defect was present when it left the hands of the manufacturer. State Farm Fire & Cas. Co., 37 Ohio St.3d 1, 523 N.E.2d 489. This can be established by demonstrating that the product was not tampered with. Atkins v. Gen. Motors Corp. (1999), 132 Ohio App.3d 556, 725 N.E.2d 727, citing McDonald v. Ford Motor Co. (1975), 42 Ohio St.2d 8, 11, 71 O.O.2d 4, 326 N.E.2d 252, 254.

{¶ 81} At trial, Susan Nearon testified that in the 18 months after she had purchased the electric stove, no repairs were made to the product, nor had anyone tampered with the switch that was ultimately found to have caused the fire. Additionally, Doran testified that the stove had been manufactured in such a way that the faulty switch was obscured by the back panel of the stove. Mrs. Nearon testified that to her knowledge, the back panel of the stove had never been removed. Clearly, plaintiffs presented sufficient circumstantial evidence from which it could be inferred that the range was in the same condition as when it was first assembled. Thus, the trial court did not err when it denied White’s motion for directed verdict and JNOV with respect to plaintiffs’ manufacturing defect claim.

{¶ 82} White’s second assignment of error is overruled.

VII

{¶ 83} White’s final assignment of error is as follows:

{¶ 84} “The trial court erred in denying the motion for a new trial.”

{¶ 85} In its final assignment, White contends that the trial court abused its discretion when it overruled the motion for new trial. Specifically, White argues that the jury verdict in favor of Donegal, the Nearons’ insurer, is contrary to law because Donegal failed to bring suit in its own name as the real party in interest before the two-year statute of limitations had expired. We disagree.

{¶ 86} As previously mentioned, the Nearons commenced the initial action against White in May 1997. On May 11, 1999, the Nearons voluntarily dismissed *586 their claims against White. On May 11, 2000, the Nearons refiled their action, with Donegal joined as a coplaintiff. It is undisputed that Donegal had paid the Nearons approximately $141,193.22 to repair their home and replace its contents.

{¶ 87} White argues that pursuant to R.C. 2305.10(A), Donegal is barred by the statute of limitations from litigating the instant matter because, as the real party in interest, Donegal should have asserted its claims in the first filing of this case or within two years after the cause of action arose. In support of this contention, White cites the Ohio Supreme Court case Shealy v. Campbell (1985), 20 Ohio St.3d 23, 20 OBR 210, 485 N.E.2d 701, which held that when an insurer has paid all damages on behalf of its insured, the insurer becomes the real party in interest and must bring suit in its own name.

{¶ 88} Because Donegal did not enter the litigation until May 11, 2000, well after the statute of limitations had expired, White contends that it is barred from doing so now. White argues that the jury’s verdict in favor of Donegal was contrary to law and that the trial court should have granted White a new trial based on the claims of the Nearons. We disagree.

{¶ 89} In Holibaugh v. Cox (1958), 167 Ohio St. 340, 4 O.O.2d 461, 148 N.E.2d 677, the Ohio Supreme Court held that when part, but not all, of an insured’s claim has been paid by the insurer, the insurer may be properly joined as a party if the insured’s action was timely with respect to the statute of limitations. We have a similar situation in the case before us. The Nearons assert, and White does not dispute, that although Donegal paid them approximately $141,193.22, their claims were not fully paid. Consequently, the Nearons remained the real party in interest. We agree with the trial court that the claims of Donegal were derivative claims that could be pursued on behalf of the Nearons but only to the extent of its subrogation interest. Because the Nearons voluntarily dismissed their action on May 11, 1999, they were required to refile by May 11, 2000. When the Nearons refiled on that date within the statute of limitations, Donegal was properly joined as a coplaintiff in the action. Thus, the trial court did not abuse its discretion when it denied White’s motion for new trial.

{¶ 90} White’s final assignment of error is overruled.

VIII

{¶ 91} Plaintiffs’ first assignment of error having been sustained, the judgment of the trial court granting White’s JNOV with respect to plaintiffs’ claim for design defect is reversed, and the jury’s verdict finding for plaintiffs in that regard is hereby reinstated. Additionally, plaintiffs’ second assignment of error is rendered moot, and the jury’s original verdict awarding a total of $104,000 to plaintiffs is reinstated. Plaintiffs’ third and final assignment of error, as well as *587 each of White’s assignments in its cross-appeal, are hereby overruled in their entirety, and the judgment of the trial court is affirmed in those respects.

Judgment reversed in part and affirmed in part.

STEPHENS, Justice.*

This is a products liability case. Four-year-old Richard Nichols was badly burned while playing with matches when his T-shirt caught fire. Through his father, as next friend, he sued Union Underwear Company, Inc., the manufacturer and seller of the shirt. The basis of the suit was strict liability for design defect. Following a jury trial in the Franklin Circuit Court, a verdict was returned for Union Underwear and judgment was entered on the verdict.

Nichols appealed to the Court of Appeals, which affirmed the judgment. Because of the importance of the question presented, we granted discretionary review. The sole issue to be determined is whether the trial court erred in instructing the jury on the definition of “unreasonably dangerous” which appears in comment i of section 402A of the second Restatement of Torts.

Because we are concerned with the instruction given, we need not burden this opinion with an extensive description of the evidence presented to the jury. Nichols concedes that the evidence was of such a nature that the jury had a right to believe that presented by Union Underwear. However, it is relevant here that the proof presented concerned itself with the following major areas: (1) the flammability of the fabric of the shirt; (2) the risk of clothing-inflicted burns to children; (3) the availability of commercially feasible, alternative designs and fabrics for this particular article of clothing; (4) the extent of *431consumer awareness of the danger inherent in flammable children’s clothing; (5) the alleged lack of such awareness of danger by this child and his mother; and (6) the significance of the fact that the fabric complies with applicable federal statutory standards for flammability. It is clear that the jury was given ample evidence on which to base its decision.

The doctrine of strict liability was adopted in this state in the case of Dealers Transport Co. v. Battery Distributing Co., Ky., 402 S.W.2d 441 (1966). While the court recognized that strict liability had obtained “substantial acceptance” throughout the country, it hung its judicial hat on the American Law Institute’s revised restatement of the law of torts. The court adopted section 402A, which provides:

(1) One who sells any product in a defective condition unreasonably dangerous to the user or consumer or to his property is subject to liability for physical harm thereby caused to the ultimate user or consumer, or to his property, if

(a) the seller is engaged in the business of selling such a product, and

(b) it is expected to and does reach the user or consumer without substantial change in the condition in which it is sold.

(2) The rule stated in Subsection (1) applies although

(a) the seller has exercised all possible care in the preparation and sale of his product, and

(b) the user or consumer has not bought the product from or entered into any contractual relation with the seller.

Restatement, Second, Torts § 402A (1965) (emphasis added).

Since the Dealers Transport case, there have been a substantial number of products liability cases before us. In Jones v. Hutchinson Manufacturing, Inc., Ky., 502 S.W.2d 66 (1973), we specifically applied the rule to liability for defective product design. In every case, however, the golden thread that holds the rule together is the Restatement, Second, Torts § 402A, quoted above.

A careful reading of section 402A reveals that, as a condition precedent to strict liability becoming operative in a particular case, the product sold must be “unreasonably dangerous” to the user or consumer or to his property. This court has never specifically addressed the question of how that concept should be presented in the instructions to the jury in a defective desjgn case. Because the trial judge chose to present it in the form of a definition in the instructions in this case and because Nichols claims prejudicial error resulted, we are now faced with that problem.

The instructions given were based on section 402A of the Restatement, as follows:

Instruction # 1

If you believe from the evidence that an ordinarily prudent manufacturer of children’s T-shirts should have foreseen that a child wearing its T-shirts substantially in the manner it was being worn at the time of the accident and that a child, while wearing one of its T-shirts might expose it to fire and in that event suffer injury from the shirt burning, either because of the composition of the shirt and/or because of his manner of wearing it, then the Court instructs you that the defendant had these duties:

A. Not to manufacture and sell a product that was unreasonably dangerous for children, including Ricky Nichols; and

B. To exercise that degree of care as would be expected of an ordinarily prudent manufacturer to manufacture and sell a product that would be reasonably safe for children including Ricky Nichols.

Instruction # 2

If you believe the defendant failed to perform either of the duties set out in Instruction # 1 hereof, and such failure or failures was a substantial factor in bringing about the injury to the Plaintiff, then you will find for the plaintiff; but unless you so believe, you will find for the defendant.

No objection was raised by Nichols to these instructions which simply set out the doctrine of strict liability provided in sec*432tion 402A as applied to the facts developed in the present case. Then, over the strenuous and continuing objection of Nichols, the court gave an instruction defining “unreasonably dangerous.”

Instruction # 4

A product is ‘unreasonably dangerous’ only if it is dangerous to an extent beyond that which would be contemplated by an ordinary adult purchaser thereof, with ordinary knowledge as to its inherent characteristics.

Nichols contends that this instruction was erroneous and prejudicial. He claims that danger beyond an ordinary purchaser’s contemplation is only one of several factors to be considered in determining if a product is, in fact, unreasonably dangerous. Further, he argues that a product's danger does not become reasonable simply because it is within the contemplation or actual awareness of the average consumer. Finally, he contends that no definition of the term “unreasonably dangerous” should have been given, but, as one was given, it should not have singled out one factor (consumer awareness) but, instead, should have brought all factors relevant to that determination to the attention of the jury.

The issue of what constitutes unreasonable dangerousness has been the subject of growing controversy. Much has been written by courts and legal scholars. See, e.g., Darling, The Patent Danger Rule: An Analysis and A Survey of its Vitality, 29 Mercer L.Rev. 583 (1978); Wade, On the Nature of Strict Tort Liability for Products, 44 Miss.L.J. 825 (1973); W. Prosser, Handbook of the Law of Torts ch. 17 (4th ed. 1971); and cases cited therein.

The instruction given was based on comment i to section 402A of the Restatement, which provides, in pertinent part: “The article sold 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.”

It is clear that instruction four limited the jury to finding the product unreasonably if, and only if, it was more dangerous than an ordinary adult would expect it to be. In effect, the product cannot be unreasonably dangerous if the omnipresent and elusive “reasonable man” — ordinary adult— knows about the danger. Under this instruction, the obviousness of the danger becomes the sole determinant of the reasonableness of a danger, rather than simply being one of many factors.

The effect of this instruction is to insulate a product from liability simply because it is patently dangerous, or because it is no more dangerous than would be anticipated by the ordinary person. Some seventeen jurisdictions adhere to this rule, eighteen have repudiated it, and sixteen, including Kentucky, have not addressed the issue. See Darling, supra, at 604-09. We now join those which have considered and rejected “patent danger” or “consumer expectation” as an absolute defense to strict liability for defective design. As Dean Wade, Reporter for the Restatement, Second, Torts, put it:

In many situations, particularly involving design matters, the consumer would not know what to expect, because he would have no idea how safe the product could be made.

Wade, supra, at 829.

We are immediately met with the difficult problem 1 of describing the standard to which the fact finder should compare the *433product to decide whether it was in a “defective condition unreasonably dangerous” when sold. In Ulrich v. Kasco Abrasives Company, Ky., 532 S.W.2d 197, 200 (1976), we pointed out that the inquiry is to be made from the perspective of a “prudent manufacturer of similar products fully apprised of the condition and tendencies of the product when he put it into the stream of commerce . . . .” This is because strict liability makes unnecessary proof by the plaintiff of what a prudent manufacturer exercising ordinary care actually should have discovered and foreseen as in a negligence action. This idea is inherent in the concept of defectiveness.

Dean John Wade and Dean Page Keeton have provided the most widely adopted definition of defectiveness as it applies to strict products liability. The strict liability standard is no different from that of negligence, they say, except that the seller is presumed to have knowledge of the actual condition of the product when it leaves his hands. In other words, as long as the product is shown to be in a defective condition when sold, the plaintiff need not go further and prove that the seller either knew or should have known of the defect.

Phillips, The Standard for Determining Defectiveness in Products Liability, 46 U.Cin. L.Rev. 101, 103 (1977).

We note, however, that this “knowledge” characteristic is not as significant in design defect cases as in manufacturing defect cases. In design defect cases liability is founded upon the premise that the design itself selected by the manufacturer amounted to a defective condition which was unreasonably dangerous; actual knowledge of the design is not the question. “Probably the main difference between design defects and construction flaws is that with respect to design defects the feasibility of making a safer product is usually in issue, while feasibility is not generally an issue for construction flaws.” Phillips, supra, at 104-05. This observation was central to our reasoning in Jones v. Hutchinson Manufacturing, Inc., Ky., 502 S.W.2d 66, 69-70 (1973), when we stated, “We think it apparent that when the claim asserted is against a manufacturer for deficient design of its product the distinction between the so-called strict liability principle and negligence is of no practical significance so far as the standard of conduct required of the defendant is concerned. In either event the standard required is reasonable care.” Thus, the fact finder in a design defect case must decide whether the manufacturer that placed in commerce the product made according to an intended design acted prudently, i.e., was the design a defective condition which was unreasonably dangerous.

We believe that consumer knowledge, the factor considered below, is only one of the factors that should be before the jury in determining whether a product is unreasonably dangerous. We will not set out an exclusive list of the factors which lead to this determination. In Jones, supra, we discussed deviation from industry standard as a factor; in Kasco Abrasives, supra, we recognized the obviousness of the danger and presence of a warning as relevant. Noted commentators have suggested many factors.2 But the facts of the individual case will determine what is relevant to each action.

In the event of another trial, the jury should be instructed as follows:

You will find for the plaintiff only if you are satisfied from the evidence that the material of which the T-shirt was made created such a risk of its being accidentally set on fire by a child wearing it that an ordinarily prudent company engaged in the manufacture of clothing, being fully aware of the risk, would not have put it on the market; otherwise, you will find for the defendant.

The decision of the Court of Appeals and the judgment of the Franklin Circuit Court are reversed and the cause is remanded for *434a new trial consistent with the views herein expressed.

LUKOWSKY, Justice,

concurring.

I agree with the opinion of the majority as far as it goes. However, the opinion leaves the law in products liability design defect cases amorphous. It fails to identify the gut issue.

I believe that whether a design is unreasonably dangerous must be determined by a social utility standard — risk versus benefit. If the benefits to be gained by the consuming public outweigh the risks of danger inherent in a particular design, such a product cannot be “unreasonably dangerous.” See Barker v. Lull Engineering Co., Inc., 20 Cal.3d 413, 573 P.2d 443, 143 Cal.Rptr. 225 (1978); Bowman v. General Motors Corp., E.D.Pa., 427 F.Supp. 234 (1977). The bottom line is that the trier of fact is required to balance two pairs of factors existing at the time of manufacture: (1) the likelihood that the product would cause the claimants harm or similar harms, and the seriousness of those harms; against (2) the manufacturer’s burden of designing a product that would have prevented those harms, and the adverse effect that alternative design would have on the usefulness of the product. That is to say that the manufacturer is not liable unless at the time of manufacture the magnitude of the danger to the claimant outweighed the utility of the product to the public. 72 C.J.S. Supp. Products Liability § 13; W. Kimble and R. Lesher, Products Liability section 55 (1979). See also, Louisville & Jefferson Co. Bd. of Health v. Mulkins, Ky., 445 S.W.2d 849, 851-52 (1969). The ultimate inquiry is risk versus benefit.

In the event of another trial, I believe the jury should be instructed as follows:

You will find for the Plaintiff if you are satisfied from the evidence that at the time of the manufacture of the cotton and polyester T-shirt the risk of harm from its being accidentally set on fire while being worn by a child outweighed the benefit to the public from its availability in the marketplace. Otherwise, you will find for the defendant.

STEPHENSON, Justice,

dissenting.

In my opinion the only error on the part of the trial court was the failure to give a directed verdict to the respondent here.

Ordinarily in “products liability” cases, I think of design defects as failure to provide proper safety measures in design. For example: lawn mowers, farm machinery, etc. In the machinery cases it is the product itself that has the propensity to cause harm to the user.

Here the T-shirt was manufactured in conformity with applicable federal standards for flammability. In the circumstances of this case, it is absurd to have a jury decide whether the T-shirt is “unreasonably dangerous.” Had the manufacturer used highly flammable materials in the garment, I could understand submission of the case to the jury, but not in this situation.

Justice KITTREDGE.

This is a direct appeal in a product liability case tried to a jury in Hampton County. The jury awarded the plaintiff $16,000,000 in actual damages and $15,000,000 in punitive damages. We affirm in part, reverse in part and remand for a new trial.

I.

This product liability action involves a 1987 Ford Bronco II 4x2, manufactured in 1986. Cheryl Hale (or her husband) purchased the 1987 Ford Bronco in June of 1999 for a nominal sum.1 At the time of sale, the Bronco had 137,500 miles on it.

On June 17, 2001, Hale was driving her Bronco along Cromwell Road in Colleton County. Hale was driving several children to her house. Hale’s daughter was seated in the front passenger seat. Plaintiff Jesse Branham, III, was riding in the backseat. Hale recalled that the children were “all excited.” No one was wearing a seatbelt.

The weather was clear and, according to Hale, she was not speeding. Hale admittedly took her eyes off the road and turned to the backseat to ask the children to quiet down. When she took her eyes off the road, the Bronco veered towards the shoulder of the road, and the rear right wheel left *209the roadway. When Hale realized that her inattention resulted in the vehicle leaving the roadway, she responded by overcorrecting to the left. Hale’s overcorrection led to the vehicle “shaking.” The vehicle rolled over. Branham was thrown from the vehicle and was injured.

Branham filed this lawsuit against Ford Motor Company and Hale in Hampton County. At trial,2 Branham did not seriously pursue the claim against Hale. The case against Ford was based on two product liability claims, one a defective seatbelt sleeve claim and the other, a “handling and stability” design defect claim related to the vehicle’s tendency to rollover. Both of these claims were pursued in negligence and strict liability.3 Ford denied liability and, among other things, asserted Hale’s negligence caused the accident. The jury, in a general verdict,4 found both Ford and Hale responsible and awarded Branham $16,000,000 in actual damages and $15,000,000 in punitive damages. Only Ford appeals. The direct appeal is before us pursuant to Rule 204(b), SCACR, certification.

II.

A.

The Seatbelt Sleeve Negligence Claim

Branham alleged Ford was negligent “[i]n selling the Bronco II with a defective rear occupant restraint system.” The amended complaint contains no specifications of Ford’s purported negligence. At trial, Branham claimed Ford was negligent in failing to adequately test the seatbelt sleeve, but he did not challenge the seatbelt sleeve design. Branham filed a companion strict liability claim concerning the seatbelt *210sleeve. Ford successfully moved for a directed verdict on the strict liability seatbelt sleeve claim.

The trial court dismissed the strict liability claim on the ground that the seatbelt sleeve was not as a matter of law in a defective condition unreasonably dangerous to the user at the time of manufacture. Based on this premise, Ford contends the companion negligence claim must fail, for all products liability actions, regardless of the stated theory, have common elements. Madden v. Cox, 284 S.C. 574, 579, 328 S.E.2d 108, 112 (Ct.App.1985) (“In a products liability action the plaintiff must establish three things, regardless of the theory on which he seeks recovery: (1) that he was injured by the product; (2) that the product, at the time of the accident, was in essentially the same condition as when it left the hands of the defendant; and (3) that the injury occurred because the product was in a defective condition unreasonably dangerous to the user.”). Ford, therefore, concludes that the negligence claim (which required Branham to prove that the seatbelt sleeve was in a defective condition unreasonably dangerous to the user) should have been dismissed. We agree. When an element common to multiple claims is not established, all related claims must fail.

A negligence theory imposes the additional burden on a plaintiff “of demonstrating the defendant (seller or manufacturer) failed to exercise due care in some respect, and, unlike strict liability, the focus is on the conduct of the seller or manufacturer, and liability is determined according to fault.” Bragg v. Hi-Ranger, Inc., 319 S.C. 531, 539, 462 S.E.2d 321, 326 (Ct.App.1995). The fault-based element is of no moment where, as here, there is no showing in the first instance of a product in a defective condition unreasonably dangerous to the user.

In addition, Ford asserts there is no separate “failure to test claim” apart from the duty to design and manufacture a product that is not defective and unreasonably dangerous. We agree, for if a product is not in a defective condition unreasonably dangerous to the user, an alleged failure to test cannot be the proximate cause of an injury. The failure to establish that the seatbelt sleeve was in a defective condition unreasonably dangerous to the user for purposes of the strict *211liability claim requires the dismissal of the companion negligence claim.

Relying on Bragg, the trial court determined it appropriate to grant a directed verdict on the strict liability claim, while at the same time allowing the negligence claim to go forward. We find the trial court’s reliance on Bragg misplaced.

In Bragg, the trial court directed a verdict in favor of the manufacturer with respect to the strict liability claim, but refused to grant a directed verdict on the negligence claims. 319 S.C. at 538, 462 S.E.2d at 325. Bragg alleged two negligence claims: negligence “in failing to place appropriate warnings” on the product and another negligence claim “in supplying [a product] that was defectively] [designed].” Id. at 537-38, 462 S.E.2d at 325. The jury returned a verdict against Bragg on the negligence claims.

Bragg appealed the dismissal of the strict liability claim, “contending] the court’s decision to grant the motion for directed verdict on strict liability, while denying the motion for directed verdict on negligence, was logically inconsistent and reversible error because those claims are virtually identical and require the same proof.” Id. at 538, 462 S.E.2d at 325. The court of appeals in Bragg affirmed the trial court and noted that “[s]trict liability and negligence are not mutually exclusive theories of recovery; that is, an injury may give rise to claims that can be established either under principles of strict liability or negligence, and failure to prove one theory does not preclude proving the other.” Id. at 539, 462 S.E.2d at 326.

While we agree that strict liability and negligence are not mutually exclusive theories of recovery, we caution against a broad reading of Bragg in this regard. An analytical framework that turns solely on whether strict liability and negligence are mutually exclusive theories of recovery may miss the mark. As noted, the negligence claim must have a fault-based element, which is not required for a strict liability claim. Where one claim is dismissed and a question arises as to the continuing viability of the companion claim, the critical inquiry is to ascertain the basis for the dismissal. If one claim is dismissed and the basis of the dismissal rests on a common *212element shared by the companion claim, the companion claim must also be dismissed.

In the present case, because the strict liability claim was dismissed due to the absence of an element shared by the companion negligence claim, the negligence claim should have been dismissed as well.

The trial court determined as a matter of law that the seatbelt sleeve was not in a defective condition unreasonably dangerous to the user. Consequently, the absence of this common, shared element required the dismissal of the strict liability claim and the companion negligence claim.5 The trial court erred in failing to direct a verdict as to the negligence seatbelt sleeve claim.6

B.

The “Handling and Stability” Design Defect Claim

The “handling and stability” design defect claim (strict liability and negligence) is the gravamen of Branham’s case. Branham alleged a design defect related to the rollover propensity of the Bronco. Ford appeals from the denial of its motions to dismiss the strict liability and negligence design defect claims. Viewing the evidence in a light most favorable to Branham, we find no error in the submission of these design defect claims to the jury. Pye v. Estate of Fox, 369 S.C. 555, 564, 633 S.E.2d 505, 509 (2006) (stating that on appeal from the denial of a directed verdict motion, the evidence must be viewed in a light most favorable to the nonmovant).

We begin with an overview of the technical information involved in the design defect claims. Ford uses the term “stability index” to describe the overall stability of a vehicle. The stability index is a comparison of the height and width of *213the vehicle, expressed in a numerical term. A closely connected term is the center of gravity. A vehicle’s center of gravity relates to what one usually thinks of as “top heavy” or “stable.” The lower the center of gravity in a vehicle, the more stable it is. Conversely, the higher the center of gravity (top heavy), the less stable the vehicle is.

The stability of a vehicle is related in part to its suspension. According to Branham’s expert, Dr. Melvin Richardson, a vehicle with a stable suspension is able to make a turn in the road, and “as the vehicle goes around the curve, it leans over some and ... the tires stay the same distance apart where they touch the ground.” A vehicle with an unstable suspension will cause the tires to “scrub” the ground during a turn, which “cuts down friction, [and] increases tire wear,” causing the vehicle to handle poorly. When a vehicle is turning and the tires begin to scrub, “you lose some of [the tire’s] capabilities to keep the vehicle going in the right direction and lose some of the ability to control the vehicle.”

Ford primarily employed two engineering tests as a means of determining whether the Bronco II was ready for manufacturing. The first test is called a “J” turn. In this test, as described by Dr. Richardson, the vehicle is driven down a roadway, and “as quickly as possible the driver turns [the wheel to a] predetermined angle and just holds it there” for the remainder of the turn.

The second test is called an accident avoidance maneuver test. This is where the vehicle is turned in an abrupt fashion one way, like in the “J” turn, but with the added maneuver of an immediate turn back in the opposite direction. With these engineering concepts in mind, we turn to the design defect evidence presented.

Thomas Feaheny, a former vice president at Ford, testified for Branham. Feaheny described the marketing forces and engineering insights that led to the development of the Bronco II. The genesis of the Bronco II spawned from the YUMA Program, which came into being in the late 1970s. YUMA was Ford’s code name for the study of small trucks, which eventually resulted in the Ford Ranger, and later the Bronco II. The YUMA prototypes initially had a MacPherson front suspension, which, according to Feaheny, is a “type of inde*214pendent front suspension that is used on a lot of small cars and trucks.” Ford’s engineers requested the MacPherson front suspension for the Bronco II when communicating with management on how best to address the Bronco II’s handling and stability concerns raised during the prototype stage.

Feaheny opined that the MacPherson strut was the “best, most feasible suspension from a functional standpoint and also from a cost and weight standpoint.” However, there was a divergence in viewpoints between corporate executives and engineers, as Ford’s engineers advocated the use of the MacPherson strut for the small truck program. Since the mid-1960s Ford had employed a Twin I-Beam suspension on its bigger trucks. Feaheny testified that “there was a belief that [Ford] should adapt [the] Twin I-Beam suspension to the new small trucks.”

The engineers at Ford believed the MacPherson suspension the better choice and “opposed [the Twin I-Beam suspension] because it was directionally wrong from the standpoint of steering, handling and rollover propensity and other characteristics.” Because the Twin I-Beam suspension was physically larger than the MacPherson suspension, using it required the entire vehicle to be lifted higher. This had a cascading effect on the composite makeup of the vehicle, which detrimentally moved the center of gravity higher off the ground. To make room for the Twin I-Beam suspension, the engine had to be raised “two to three” inches. With the engine raised a few inches, the transmission had to be raised, which caused the hood to be raised, which then caused the seating to be raised. The net effect of this was a higher center of gravity, “which add[ed] a rollover propensity.”

Feaheny also noted that the Twin I-Beam had a tendency for “jacking.” Feaheny stated that jacking is a term used to describe an occurrence when the “vehicle will slide out in a severe handling maneuver. The outboard wheel would tend to dig into ... the suspension arm, which was strong and stiff, [and it] would have to move with that wheel and the inner pivot would go up in the air.” When a vehicle jacks, there is an instantaneous raising of the center of gravity, which further “increase[s] the propensity for rollover.”

*215Use of the Twin I-Beam and its attendant safety concerns came to a head in the late 1970s. A group of engineers approached Feaheny and recommended that Ford use either the MacPherson suspension or the SLA (short long arm) suspension for the YUMA prototypes. The engineers made it clear that they were “very concerned” with the Twin I-Beam. Feaheny directed the engineers to one of his colleagues, Jim Capalongo, and Feaheny later met with Capalongo to discuss the engineers’ concerns. After this meeting, alternative suspension designs were discussed and tested for “about a year” but the Twin I-Beam was still selected.

The reason the Twin I-Beam was selected in the face of engineering concerns was that it served a “major marketing advantage,” as Ford had promoted this form of suspension on its full size trucks since the mid-1960s. In the minds of the marketing executives, the Twin I-Beam was part and parcel of a tough truck, and it made business sense to carry that suspension into the smaller trucks.

The testimony of Dr. Richardson buttressed the evidence supplied by Feaheny and Ford’s internal documents. Dr. Richardson opined that the use of the Twin I-Beam suspension led to the Bronco II being unreasonably dangerous. Dr. Richardson described three common suspension systems referenced above: (1) the SLA; (2) the MacPherson; and (3) the Twin I-Beam. It was through Dr. Richardson that Branham introduced many of Ford’s internal documents showing the competing concerns and interests of the engineers and management over the proper suspension.

The Bronco II was designed from the existing “bones” of the Ford Ranger. Dr. Richardson opined that using the Ranger as the design platform was an appropriate engineering decision, and that it gave Ford the advantage of using components that had already been made.

Dr. Richardson testified to a Ford document dated February 5, 1981, and titled “Revised Stability Index for Utility.” The stated objective of the document was to “review alternatives to increase stability index.” Reading from the document, he stated that, “a study of methods to improve the stability index for the Bronco II has resulted in several design alterna*216tives to achieve an improvement ... from 1.85 to maximum achievable of 2.25 without a totally new concept vehicle.”

The document made a general assessment about improving the stability index. “In order to improve stability index substantially, the following are required: widen track width, and lower center of gravity achieved by raising the wheel center lines with respect to body with trade-offs in ground clearance and vehicle package.” The document also made five proposals to achieve a higher stability index. The first two proposals did not jeopardize the target release date for the Bronco II, but the latter three did. Only one of the proposals would have achieved a stability index of 2.25 for the Bronco II, but it was not selected.

Ford selected what is referred to as “proposal two,” and it had a target stability index of 2.02. Dr. Richardson pointed out that proposal two saved Ford money. None of the proposals on this document argued for a change in the suspension system. But Dr. Richardson opined that had Ford opted to use an SLA or a MacPherson suspension system, then it could have achieved a stability index of 2.25. At that point, however, Ford had already decided to employ the Twin I-Beam suspension notwithstanding its engineers’ criticisms.

Dr. Richardson testified to Plaintiffs exhibit 31, dated March 17, 1982, which discussed “J” turn testing for the Bronco II. In relevant part, the document stated the following:

Engineering sign-off for the Bronco II is scheduled for 7/9/82. Minimal development DVP&R7 testing has been completed because the suspension and steering system designs have not been finalized for improved roll characteristics during the “J” turn maneuver.

A decision is required to solidify the steering and suspension designs. Development recommends pursuing items 1, 2, and 3 below if a small improvement in roll characteristics during a “J” turn maneuver is deemed acceptable, or pursuing item 4 below if a major improvement is required. *217Incorporation of item 4 would most likely cause a delay in Job # 1 [the release date of the Bronco II].

(emphasis in original). Dr. Richardson testified that to his knowledge none of the recommendations set out in the document were adopted.

Dr. Richardson testified to a Ford document dated May 4, 1982. The document identified the current stability index of the Bronco II at 2.03. Dr. Richardson noted that any change to the Bronco II after the date of this document “had to be very small if [Ford] w[as] going to still put [the Bronco II] on the market in the beginning of [1983].” He went on to testify that in the state the Bronco II was then in, with a stability index of 2.03 the vehicle would be “dangerously unstable.”

Branham introduced a Ford document from September 14, 1982, with the following stated purpose: “To identify advanced engineering projects that will be undertaken to provide for continued improvement, Bronco II handling, during its cycle life.” Dr. Richardson responded to the document as follows:

The vehicle should have been made reasonably safe when it was first designed and built. There was time to do that, the discussions in the engineering documents, to me as an engineer, show me that the engineers knew how to do that, could have done it, and that should have been done. To release it without it being reasonably safe then subjects those people who buy it to risk. Now, if it is released in that configuration, it certainly should be improved as time goes along because it shouldn’t be left that way.

Following up on his expert’s opinion, Branham asked whether improvements were ever made to correct the problems in the Bronco II when it was released. Dr. Richardson responded, “there were no improvements made that would correct this defect.”

The rollover propensity in the Bronco II 4x4, as reflected in the stability index and elevated center of gravity, was increased in the Bronco II 4x2. The two-wheel drive Bronco was lighter than its four-wheel version, resulting in reduced stability and an even higher center of gravity. The Bronco II involved in this litigation is a 4x2.

The foregoing is not an exhaustive review of the evidence presented by Branham, but it serves to support the able trial *218judge’s determination that Branham presented sufficient evidence of a design defect known to Ford at or prior to the date of manufacture to withstand a directed verdict motion. We make this determination without having to rely on the further body of evidence of the Bronco II’s rollover tendencies found in the substantial post-distribution evidence which the trial court allowed.8

C.

We next address Ford’s two-fold argument that: (1) Bran-ham failed to prove a reasonable alternative design pursuant to the risk-utility test; and (2) South Carolina law requires a risk-utility test in design defect cases to the exclusion of the consumer expectations test.

For a plaintiff to successfully advance a design defect claim, he must show that the design of the product caused it to be “unreasonably dangerous.” Madden v. Cox, 284 S.C. 574, 579-80, 328 S.E.2d 108, 112 (Ct.App.1985). In South Carolina, we have traditionally employed two tests to determine whether a product was unreasonably dangerous as a result of a design defect: (1) the consumer expectations test and (2) the risk-utility test.

In Claytor v. General Motors Corp., this Court phrased the consumer expectations test as follows: “The test of whether a product is or is not defective is whether the product is unreasonably dangerous to the consumer or user given the conditions and circumstances that foreseeably attend use of the product.” 277 S.C. 259, 262, 286 S.E.2d 129, 131 (1982).

The Claytor Court articulated the risk-utility test in the following manner: “[N]umerous factors must be considered [when determining whether a product is unreasonably dangerous], including the usefulness and desirability of the product, the cost involved for added safety, the likelihood and potential seriousness of injury, and the obviousness of danger.” Id. at 265, 286 S.E.2d at 132.

Later, in Bragg v. Hi-Ranger, Inc., our court of appeals phrased the risk-utility test as follows: “[A] product is unreasonably dangerous and defective if the danger associated with *219the use of the product outweighs the utility of the product.” 319 S.C. 531, 543, 462 S.E.2d 321, 328 (Ct.App.1995). The Bragg court went on to list the above factors set forth in Claytor as the relevant inquiry when weighing the danger of the product versus its utility. Id. at 543-44,462 S.E.2d at 328.

Ford contends Branham failed to present evidence of a feasible alternative design. Implicit in Ford’s argument is the contention that a product may only be shown to be defective and unreasonably dangerous by way of a risk-utility test, for by its very nature, the risk-utility test requires a showing of a reasonable alternative design.9 Branham counters, arguing that under Claytor he may prove a design defect by resort to the consumer expectations test or the risk-utility test. Bran-ham also argues that regardless of which test is required, he has met both, including evidence of a feasible alternative design. We agree with Branham’s contention that he produced evidence of a feasible alternative design. Branham additionally points out that the jury was charged on the consumer expectations test and the risk-utility test.

As discussed above, Branham challenged the design of the Ford Bronco II by pointing to the MacPherson suspension as a reasonable alternative design. A former Ford vice president, Thomas Feaheny, testified that the MacPherson suspension system would have significantly increased the handling and stability of the Bronco II, making it less prone to rollovers. Branham’s expert, Dr. Richardson, also noted that the MacPherson suspension system would have enhanced vehicle stability by lowering the vehicle center of gravity. There was further evidence that the desired sport utility features of the Bronco II would not have been compromised by using the MacPherson suspension. Moreover, there is evidence that use of the MacPherson suspension would not have increased costs. Whether this evidence satisfies the risk-utility test is ultimately a jury question. But it is evidence of a feasible alternative design, sufficient to survive a directed verdict motion.

*220While the consumer expectations test fits well in manufacturing defect cases, we do agree with Ford that the test is ill-suited in design defect cases. We hold today that the exclusive test in a products liability design case is the risk-utility test with its requirement of showing a feasible alternative design. In doing so, we recognize our Legislature’s presence in the area of strict liability for products liability.

In 1974, our Legislature adopted the Restatement (Second) of Torts § 402A (1965), and identified its comments as legislative intent. S.C.Code Ann. §§ 15-73-10-30 (2005). The comments in section 402A are pointed to as the basis for the consumer expectations test.10 Since the adoption of section 402A, the American Law Institute published the Restatement (Third) of Torts: Products Liability (1998). The third edition effectively moved away from the consumer expectations test for design defects, and towards a risk-utility test. We believe the Legislature’s foresight in looking to the American Law Institute for guidance in this area is instructive.

The Legislature has expressed no intention to foreclose court consideration of developments in products liability law. For example, this Court’s approval of the risk-utility test in Claytor yielded no legislative response. We thus believe the adoption of the risk-utility test in design defect cases in no manner infringes on the Legislature’s presence in this area.

Some form of a risk-utility test is employed by an overwhelming majority of the jurisdictions in this country.11 Some *221of these jurisdictions exclusively employ a risk-utility test,12 while others do so with a hybrid of the risk-utility and the consumer expectations test, or an explicit either-or option.13 States that exclusively employ the consumer expectations test *222are a decided minority.14

We believe that in design defect cases the risk-utility test provides the best means for analyzing whether a product is designed defectively. Unlike the consumer expectations test, the focus of a risk-utility test centers upon the alleged defectively designed product. The risk-utility test provides objective factors for a trier of fact to analyze when presented with a challenge to a manufacturer’s design. Conversely, we find the consumer expectations test and its focus on the consumer ill-suited to determine whether a product’s design is unreasonably dangerous.15

We believe the rule we announce today in design defect cases adheres to the approach the trial and appellate courts in this state have been following. In reported design defect cases, our trial and appellate courts have placed their imprimatur on the importance of showing a feasible alternative design. See Claytor v. Gen. Motors Corp., 277 S.C. 259, 265, 286 S.E.2d 129, 132 (1982) (adopting the risk-utility test); Kennedy v. Custom Ice Equip. Co., 271 S.C. 171, 176, 246 S.E.2d 176, 178 (1978) (affirming verdict in favor of plaintiff by noting that plaintiff presented evidence of a design alternative); Mickle v. Blackmon, 252 S.C. 202, 234-35, 166 S.E.2d 173, 187-88 (1969) (discussing a manufacturer’s decision to use *223one type of inferior material as a component part one year, but a superior material the following year — that is, a design alternative); Bragg v. Hi-Ranger, Inc., 319 S.C. 531, 546, 462 S.E.2d 321, 330 (Ct.App.1995) (affirming defense verdict and noting that plaintiff failed to present evidence of a feasible alternative design); Sunvillas Homeowners Ass’n v. Square D Co., 301 S.C. 330, 334, 391 S.E.2d 868, 870 (Ct.App.1990) (affirming a defense directed verdict and noting that plaintiffs expert failed to discuss design alternatives); Gasque v. Heublein, Inc., 281 S.C. 278, 283, 315 S.E.2d 556, 559 (Ct.App. 1984) (affirming a plaintiffs verdict and noting in detail existence of alternative design evidence).

In Kennedy v. Custom Ice Equipment Co., this Court specifically pointed to evidence that the challenged industrial ice machine would have been safer had the manufacturer installed a protective cover. 271 S.C. at 176, 246 S.E.2d at 178. In Gasque v. Heublein, Inc., our court of appeals acknowledged the importance of a reasonable alternative design in a product liability design defect case wherein it noted evidence of alternative designs in an opinion affirming an award for the plaintiff. 281 S.C. at 283, 315 S.E.2d at 559. In like manner is the case of Sunvillas Homeowners Ass’n v. Square D Co., where the court of appeals upheld a directed verdict in favor of a manufacturer, noting that plaintiffs did not produce any evidence of design alternatives. 301 S.C. at 334, 391 S.E.2d at 870. And more recently, in Bragg, our court of appeals again noted the absence of alternative design evidence in affirming a defense verdict. 319 S.C. at 546, 462 S.E.2d at 330. The very nature of feasible alternative design evidence entails the manufacturer’s decision to employ one design over another. This weighing of costs and benefits attendant to that decision is the essence of the risk-utility test.

This approach is in accord with the current edition of the Restatement of Torts:

A product ... is defective in design when the foreseeable risks of harm posed by the product could have been reduced or avoided by the adoption of a reasonable alternative design by the seller or other distributor, or a predecessor in the commercial chain of distribution, and the omission of the alternative design renders the product not reasonably safe.

*224Restatement (Third) of Torts: Products Liability § 2(b) (1998). Concerning the framework for the risk-utility test, we agree with Professor David G. Owen, who observed:

[T]he basic liability test should be congruent with the basic issue that in most cases must be proved. In design defect litigation, that basic issue involves the following fundamental ... question: whether the manufacturer’s failure to adopt a particular design feature proposed by the plaintiff was, on balance, right or wrong. A congruence between this central issue and the liability test requires that the test focus squarely on the issue of what, in particular, allegedly was wrong with the manufacturer’s design decision. More specifically, this inquiry asks whether the increased costs (lost dollars, lost utility, and lost safety) of altering the design — in the particular manner the plaintiff claims was reasonably necessary to the product’s safety — would have been worth the resulting safety benefits.

David G. Owen, Toward a Proper Test for Design Defectiveness: “Micro-Balancing” Costs and Benefits, 75 Tex.L.Rev. 1661, 1687 (1997).

In every design defect case the central recurring fact ■will be a product that failed causing damage to a person or his property. Consequently, the focus will be whether the product was made safe enough. This inquiry is the core of the risk-utility balancing test in design defect cases, yet we do not suggest a jury question is created merely because a product can be made safer. We adhere to our longstanding approval of the principle that a product is not in a defective condition unreasonably dangerous merely because it “can be made more safe.” As we observed in Marchant v. Mitchell Distributing Co.:

Most any product can be made more safe. Automobiles would be more safe with disc brakes and steel-belted radial tires than with ordinary brakes and ordinary tires, but this does not mean that an automobile dealer would be held to have sold a defective product merely because the most safe equipment is not installed. By a like token, a bicycle is more safe if equipped with lights and a bell, but the fact that one is not so equipped does not create the inference that the bicycle is defective and unreasonably dangerous.

*225There is, of course, some danger incident to the use of any product.

270 S.C. 29, 35-36, 36, 240 S.E.2d 511, 513, 514 (1977).

In sum, in a product liability design defect action, the plaintiff must present evidence of a reasonable alternative design. The plaintiff will be required to point to a design flaw in the product and show how his alternative design would have prevented the product from being unreasonably dangerous. This presentation of an alternative design must include consideration of the costs, safety and functionality associated with the alternative design.16 On retrial, Branham’s design defect claim will proceed pursuant to the risk-utility test and not the consumer expectations test.

III.

Notwithstanding the existence of ample evidence to withstand a directed verdict motion on the handling and stability design defect claim, we reverse and remand for a new trial. There are three reasons we reverse and remand the finding of liability and award of actual damages. First, this case implicates two evidentiary rules related to products liability cases. The first rule provides that whether a product is defective must be measured against information known at the time the product was placed into the stream of commerce. When a claim is asserted against a manufacturer, post-manufacture evidence is generally not admissible. The second rule provides that evidence of similar incidents is admissible where there is a substantial similarity between the other incidents and the accident in dispute tending to prove or disprove some fact in controversy. Evidence was introduced that violated both of these rules. Third, Branham’s closing argument was a *226direct appeal to the passion and prejudice of the jury. And although not a standalone ground for reversal, we find that because Ford and Hale were joint tortfeasors, it was error to require the jury to apportion responsibility between the defendants.

A.

Post-distribution evidence

In order for a plaintiff to prove his case in a product liability action, he must show that the “product was in a defective condition at the time that it left the hands of the particular seller ... and unless evidence can be produced which will support the conclusion that it was then defective, the burden is not sustained.” Claytor v. Gen. Motors Corp., 277 S.C. 259, 264, 286 S.E.2d 129, 131-32 (1982) (emphasis added) (quoting Restatement (Second) of Torts § 402A, cmt. g. (1965) adopted as legislative intent via S.C.Code Ann. § 15-73-30 (2005)); see also Bragg v. Hi-Ranger, Inc., 319 S.C. 531, 548-49, 462 S.E.2d 321, 331 (Ct.App.1995) (recognizing that the “product must be ‘measured against a standard existing at the time of sale’ ” and that “ ‘hindsight opinions by [... ] experts suggesting that more should have been done ... are insufficient to discredit the conclusion that the manufacturer met the standard of care’ ”) (quoting Sexton ex rel. Sexton v. Bell Helmets, Inc., 926 F.2d 331, 337 (4th Cir.1991) and Doe v. Miles Labs., Inc., Cutter Labs. Div., 927 F.2d 187, 193 (4th Cir.1991)); Restatement (Third) of Torts: Products Liability § 2, cmt. a. (1998) (“[F]or the liability system to be fair and efficient, the balancing of risks and benefits in judging product design ... must be done in light of the knowledge of risks and risk-avoidance techniques reasonably attainable at the time of distribution.”). Because the claim here is against the manufacturer, the “time of distribution” is the time of manufacture.

While we find Branham presented sufficient evidence to create a jury question on his design defect claim, we further find Ford was prejudiced by Branham’s unrelenting pursuit of post-distribution evidence on the issue of liability. Given the extent of the improper post-distribution evidence introduced, the error cannot be considered harmless.

*227We first clarify what is post-distribution evidence. Simply defined, post-distribution evidence is evidence of facts neither known nor available at the time of distribution. When assessing liability in a design defect claim against a manufacturer, the judgment and ultimate decision of the manufacturer must be evaluated based on what was known or “reasonably attainable” at the time of manufacture.17 See Restatement (Third) of Torts: Products Liability § 2, cmt. a. (1998). The use of post-distribution evidence to evaluate a product’s design through the lens of hindsight is improper. See Gregory v. Cincinnati, Inc., 450 Mich. 1, 538 N.W.2d 325, 326 (1995) (“Evidence of conduct after the date of manufacture improperly shifts the focus from the premanufacturing decision and has the potential to taint any finding of liability.”).

Hale’s Ford Bronco II 4x2 was manufactured in 1986. The following is a sampling of the post-manufacture (or post-distribution) evidence.

Branham introduced a memorandum dated April 14, 1989, dealing with a meeting that three Ford engineers had with “six people from Consumers Report.” The memorandum stated that:

Our objective was to “give it our best shot” at diffusing a very negative story on the Bronco II in the June issue____ The magazine has done a comparative test of the Chevy S-10 Blazer, Geo Tracker, Dodge Raider and Bronco II. As the result of several calls from a Consumer Report writer, we were led to believe that the story could be nearly as negative as last summer’s Suzuki Samurai story. Plus, NHTSA is currently conducting an engineering analysis of the Bronco II which creates a negative cloud. And, FARS [Fatal Analysis Reporting System] data shows Bronco II to *228have a higher fatal rollover rate relative to certain competitors.

The memorandum went on to note the following: “Our data are not terribly favorable. Our rollover rate is three times higher than the Chevy S-10 Blazer.” This evidence of the Bronco II’s rollover rate is post-manufacture evidence.

Later in the same 1989 memorandum, as the engineers discuss how they thought they did, this comment is made: “We think, however, that we have clouded their minds, loosened some conclusions they may have reached prior to our meeting and sent them off to search for additional information that could work to our advantage.” The “clouded their mind” comment became a mantra for Branham on the issue of liability and otherwise.

Through Branham’s expert, Dr. Richardson, a 1989 film was introduced. Counsel emphasized this film, taped in 1989, comparing the S-10 Blazer and the Bronco II. As reflected in Plaintiffs exhibit 54A, which is the corresponding report to the videotape, Ford requested “additional ‘J’ turn tests” on May 17, 1989 for various vehicles, including a 1989 Bronco II 4x4. The tape (post-manufacture evidence) revealed that the 1989 Bronco II did not handle as well as the S-10 Blazer.

Dr. Richardson also testified to a document, Plaintiffs exhibit 168, referencing post-manufacture evidence that compared a 1989 Bronco II (referred to in the document as BII) to the UN46 prototype, now known as the Ford Explorer. This exhibit shows the additional evidence of the rollover tendency of the Bronco II that came to light after 1986:

Current “strategies” for development of utility vehicle stability have changed over the past few years due (sic) the increased availability of rollover accident data and analyses. Previous strategies were partially driven by the Insurance Institute tests of the Jeep CJ7 in the early 80’s which emphasized risk from rollovers caused by extreme (rate and magnitude) steering inputs in emergency maneuvers. Independent DOT, GM and Ford studies have confirmed that rollovers directly induced by extreme steering inputs are rare for any Utility vehicle (including the CJ7). The following quote from GM’s recent SAE Paper (Reconstruction of Rollover Collisions, SAE 890857) summarizes current wis*229dom. “A common pre-rollover maneuver is an off-road path by the car, followed by heavy steer correction back towards the road leading to a side slide, and, ultimately, a trip followed by the rollover.” Based on this new information, the UN46 was developed using a handling philosophy notably different from the BII.18

This post-distribution exhibit concludes:

Based on an analysis of FARS accident summaries and BII & Competitive handling characteristics, it is impossible to identify any type of vehicle “defect” that could explain the BII FARS performance. It is most likely that the handling strategy used during the development of the BII, which fully exploited the vehicles (sic) inherent quickness (due to its short wheelbase), encourages aggressive driving and makes the vehicle more sensitive to the large steering wheel “over-corrections” that seem to be part of most rollover scenarios. This sensitivity is aggravated by the fact the (sic) most operators in rollover accidents are either inexperienced drivers, under the influence of alcohol or both. The UN46, designed with the benefit of the FARS experience for all utility vehicles, has been intentionally developed to resolve these issues.

Yet another example of post-distribution evidence is found in a March 3, 1989 memorandum addressing an accident caused while testing a prototype anti-lock braking system (ABS) at the Dearborn Proving Grounds (DPG). The memorandum revealed that on February 28, 1989, a “demonstration was conducted on an ice pad located on the DPG East-West runway” and that the “accident involved a Kelly-Wayes Company owned 1989 Bronco II with prototype ABS.” The goal was to test the efficacy of the ABS system when running partially on ice and partially on dry ground. During the test procedure the Bronco II rolled over. The rollover occurred on ice.

There are other examples of post-manufacture evidence, but the few examples cited illustrate the inherent prejudice that flows from post-distribution evidence. It is *230good when a manufacturer continues to test and evaluate its product after initial manufacture. As additional information is learned, changes may be made that improve product safety and function. As a matter of policy, the law should encourage the design and manufacture of safe, functional products. In holding manufacturers accountable for unreasonably dangerous products pursuant to a fair system, products liability law serves that goal. Moreover, the law should encourage manufacturers to continue to improve their products in terms of utility and safety free from prior design decisions judged through the lens of hindsight.

Whether the 1987 Ford Bronco II was defectively designed and in a defective condition unreasonably dangerous must be determined as of the 1986 manufacture date of the vehicle. Ford’s 1986 design and manufacture decision should be assessed on the evidence available at that time, not the increased evidence of additional rollover data that came to light after 1986.

B.

Other Similar Incidents

In Whaley v. CSX Transportation Inc., this Court recognized that similar accidents are admissible if they “tend[ ] to prove or disprove some fact in dispute.” 362 S.C. 456, 483, 609 S.E.2d 286, 300 (2005). The Court also recognized that this type of evidence has the potential to be “highly prejudicial.”19 Id. at 483, 609 S.E.2d at 300. Accordingly, it set forth a stringent standard for admissibility: “ ‘[A] plaintiff must present a factual foundation for the court to determine that the other accidents were substantially similar to the accident at issue.’ ” Id. at 483, 609 S.E.2d at 300 (quoting Buckman v. Bombardier Corp., 893 F.Supp. 547, 552 (E.D.N.C.1995)); see also Atkinson v. Orkin Exterminating Co., 361 S.C. 156, 604 S.E.2d 385 (2004) (recognizing that “unless Orkin’s past conduct is ‘similar’ to the conduct directed at the [plaintiffs], it is inadmissible”).

*231Before addressing the “substantially similar” test, we resolve Ford’s twin challenge to the post-manufacture evidence of supposed similar incidents. Even assuming a plaintiff satisfies the Whaley “substantially similar” test, such evidence must not run afoul of the rule in products liability cases that prohibits post-distribution evidence to establish liability. Whaley is instructive in this regard.

Whaley was employed by CSX Transportation. Whaley became ill, allegedly due to work conditions, with heat-related symptoms first reported on May 24, 2000. Whaley introduced evidence that “between 1984 and 2000, CSX had received ninety-seven employee complaints about heat. In addition, the trial judge permitted Whaley to introduce evidence that, between 1993 and 2000, eighteen CSX employees had suffered heat stroke.” Whaley, 362 S.C. at 483, 609 S.E.2d at 300. Because “Whaley did not establish that the reported complaints and injuries stemmed from the same or similar circumstances as his injuries[,]” it was error to admit the evidence. Id. at 483-84, 609 S.E.2d at 300. Yet Whaley never attempted to introduce evidence of other incidents that occurred after the 2000 injury date.

On the issue of liability, Branham presented voluminous evidence of post-manufacture rollover data. The post-manufacture evidence of purported similar incidents was error, even if the “substantially similar” threshold was met.20 Post-manufacture evidence of similar incidents is not admissible to prove liability.

This Court recently revisited Whaley in a products liability setting, Watson v. Ford Motor Co., 389 S.C. 434, 699 S.E.2d 169 (2010) (Shearouse Adv. Sh. No. 10 at 37). In Watson, we repeated that “[ejvidence of similar accidents, transactions, or happenings is admissible in South Carolina where there is *232some special relation between the accidents tending to prove or disprove some fact in dispute.” Id. at 50. In imposing a burden on plaintiffs to demonstrate “that the other accidents were substantially similar to the accident at issue[,]” the Court “set forth factors to support a claim that the present accident was caused by the same defect: (1) the products are similar; (2) the alleged defect is similar; (3) causation related to the defect in the other incidents; and (4) exclusion of all reasonable secondary explanations for the cause of the other accidents.” Id. at 51 (citing Buckman v. Bombardier, 893 F.Supp. at 552).

We turn now to the evidence of pre-manufacture rollover data. Branham introduced evidence of rollover accidents involving the Bronco II and other vehicles in the same class that was known at or prior to the 1986 manufacture of Hale’s Bronco II. Ford claims the pre-manufacture comparative evidence of rollover accidents violates the Whaley-Watson “substantially similar” test because there was no showing that the cause of the other accidents was similar to the cause of the rollover accident at issue.

In commenting on this evidentiary dispute, we must be careful not to foreclose the discretion of the trial court in ruling on objections during the course of the retrial. This is especially true with “other similar incidents” evidence because of its potential to be “highly prejudicial,” thereby implicating Rule 403, SCRE. Our discussion, therefore, is intended as a general guideline, as we recognize a host of factors can arise during the course of a trial that impact a trial court’s decision to admit or exclude evidence.

With that caveat, on the record before us, we disagree with Ford. Admittedly, a showing of comparative rollover accident rates does not establish the manner in which any particular accident occurred. But Ford misconstrues the essence of Branham’s design defect claim. To the extent Branham is able to establish (at or prior to the manufacture date of the subject vehicle) the rate or number of rollover accidents of the Bronco II was greater as compared to other vehicles in its class, such evidence may well be relevant on whether the Bronco II was unreasonably dangerous.

*233We do agree with Ford that if the cause of an accident is known and the cause is not substantially similar to the accident at issue, evidence of the other accident should be excluded. Yet, where the precise cause of an accident is not known, Bronco II rollover accident data has relevance when compared to rollover accident data of other vehicles in class. This relevance is linked directly to Branham’s claim that the design of the Bronco II caused it to have an unreasonably dangerous tendency to rollover.

Like the trial court, we are persuaded by neither Ford’s general argument that many accidents may be attributable to inexperienced or impaired drivers, nor its specific reference to Hale’s inattention as the cause of the June 17, 2001 accident.

First, as referenced in a Ford document (Plaintiffs exhibit 168), Ford recognized the tendency of the Bronco II to roll over, describing it as driving “sensitivity” which is “aggravated by the fact [that] most operators in rollover accidents are either inexperienced drivers, under the influence of alcohol or both.” Assuming a number of rollover accidents are caused by inexperienced or impaired drivers, there is no suggestion in this record that inexperienced or impaired drivers disproportionately favored the Bronco II, thus skewing the comparative rollover accident data. It is inferable that rollover accidents caused by inexperienced or impaired drivers are shared by all vehicles in the class, not just the Bronco II. While Ford’s position may have appeal as a jury argument, it is of little moment on the admissibility question in the record before us.

Second, there may be little or no doubt as to Hale’s negligence, but that misses the point in terms of the admissibility of comparative rollover accident data. A car manufacturer must design and produce vehicles that are not in a defective condition unreasonably dangerous to the user. Cars are designed with utility and safety in mind, and careless driving is a foreseeable reality. The general nature of the alleged negligent driving on the part of Hale was (or should have been) part of the evaluative process that culminated in the ultimate decision of Ford to design, manufacture and market the Bronco II to the driving public. Ford had a duty to design and manufacture the Bronco II as a reasonably safe vehicle.

*234We believe our consideration of the admissibility of the premanufacture rollover accident data necessarily flows from the risk-utility test for products liability design defect cases.

C.

Closing Argument

It is improper for counsel to make a “closing argument to the jury ... calculated to arouse passion or prejudice.” Gathers v. Harris Teeter Supermarket, Inc., 282 S.C. 220, 231, 317 S.E.2d 748, 755 (Ct.App.1984). The closing argument of Branham’s counsel was designed to inflame and prejudice the jury. Closing argument excerpts include:

1. “This is how Ford looks at this. That little bit of thirty people being killed every year didn’t matter. Those thirty people, those thirty extra people getting killed in a year didn’t matter to them because it was just a little bitty number.”

2. “It does matter about those people getting killed. Those thirty people do count. Those thirty people— that’s thirty more people that got killed that year. If you expect these vehicles to last about twenty years, that’s six hundred more people getting killed using this vehicle as opposed to a Chevy S-10 Blazer. That’s serious.”

3. “And that doesn’t count the paralyzed people, the quadriplegics, the people with serious injuries, the thousands of people that have been in these events because of this rollover propensity of this vehicle that they knew about, and they knew it since day one but they chose profit over safety every time because they looked at it as numbers. They didn’t look at it as lives, as people.”

4. “I submit to you that the evidence is that they did it because they thought it was a little, small number.... [T]hey did not look at it as thirty lives a year[ ], they didn’t look at it as six hundred lives. That’s how they should have looked at it, but that was not how they did it.”

5. “They got together at the highest levels of Ford Motor Company and they made a judgment that rather than *235delaying and improving the Bronco II, they were going to sell the vehicle as it was and that they were going to risk people’s lives and they were going to risk serious injuries like we have here today. They were going to risk people’s brains.”

6. “Jesse Branham is here today with a brain injury and six hundred other people, or however many it is, lost their lives, and numerous others have brain injuries or are paralyzed, quadriplegic, have extremely serious injuries. We believe that you should tell Ford Motor Company what you think about this kind of thing.”

It is unmistakable that the closing argument relied heavily on inadmissible evidence. In addition, as will be discussed below, much of the prejudice resulting from the improper evidence was merged in closing argument with Branham’s pursuit of punitive damages in requesting that the jury punish Ford for harm to Branham and others. The closing argument invited the jury to base its verdict on passion rather than reason. The closing argument denied Ford a fair trial. Scoggins v. McClellion, 321 S.C. 264, 269, 468 S.E.2d 12, 15 (Ct.App.1996) (“The test for granting a new trial on the basis of improper closing argument by opposing counsel is whether the complaining party was prejudiced to the extent that he or she was denied a fair trial.”).

D.

The Verdict Form

Over Ford’s objection, the trial court required the jury to apportion liability between Ford and Hale. This was error. S.C. Dep’t of Transp. v. First Carolina Corp, of S.C., 372 S.C. 295, 303, 641 S.E.2d 903, 907-08 (2007) (“[A] special verdict question may be so defective in its formulation that its submission results in a prejudicial effect which constitutes reversible error.”). Whether Ford was prejudiced by the improper verdict form is speculation, but we address the issue in light of the remand for a new trial.

Ford and Hale were alleged joint tortfeasors. The accident occurred in 2001. In 2001, multiple tortfeasors were jointly and severally responsible for all damages. Concerning a *236plaintiffs ability to collect on a judgment, there could be no apportionment of fault among joint tortfeasors.21 The trial court used a verdict form that is standard in comparative negligence cases where a defendant alleges the plaintiffs own negligence caused the accident and resulting injuries. Question five of the verdict form asked the following:

Taking the combined negligence that proximately caused the parties’ injuries as one hundred percent (100%), what percentage of that negligence is attributable to Defendant Ford Motor Company and what percentage is attributable to the Defendant Cheryl Jane Hale?

The jury apportioned fault 55% Ford and 45% Hale. Allocating fault between Ford and Hale served no legitimate purpose. Our comparative system for allocating liability between a plaintiff and a defendant is in no manner implicated where fault lies, if at all, among multiple defendants. Since the Nelson v. Concrete Supply22 decision adopting comparative negligence (between a plaintiff and a defendant), this Court has reaffirmed the applicability of joint and several liability among joint tortfeasors. Summer v. Carpenter, 328 S.C. 36, 48, 492 S.E.2d 55, 61 (1997); Am. Fed. Bank, FSB v. No. One Main Joint Venture, 321 S.C. 169, 175-76, 467 S.E.2d 439, 443 (1996); see also Fernanders v. Marks Constr. of S.C., Inc., 330 S.C. 470, 478, 499 S.E.2d 509, 513 (Ct.App.1998).

The trial court justified the apportionment question on the basis of a need to ensure that any punitive damage award was supported by a negligence cause of action, and not the strict liability claim. The trial court’s reasoning is not persuasive. If there were genuine concern regarding the basis of a plain*237tiffs verdict, the easy solution was a verdict form tailored to that concern, just as Ford requested.

A detailed verdict form would have specified whether a finding of negligence against Ford was based on the seatbelt sleeve claim or the design defect claim, or both. A proper verdict form would have avoided the confusion caused by having the jury apportion blame between jointly and severally liable defendants. More to the point, Ford’s requested special verdict form would have avoided the very real risk that the jury (unaware of joint and several liability principles) would take the cue from the apportionment question and inflate the actual damage award to ensure Branham received a full recovery from the one deep-pocket defendant. The actual damage award causes genuine concern as to the effect on the jury of the improper verdict form.

IY.

Ford challenges the jury’s award of $16,000,000 actual damages and $15,000,000 punitive damages. Because of our directive for a new trial, we decline to address Ford’s contentions that these awards are excessive.

A.

Actual Damages

Ford contends the $16,000,000 actual damage award is grossly excessive. “When a verdict is ‘grossly excessive and the amount awarded is so shockingly disproportionate to the injuries as to indicate that the jury acted out of passion, caprice, prejudice, or other consideration not founded on the evidence, it becomes the duty of this Court, as well as the trial court, to set aside the verdict.’ ” Sanders v. Prince, 304 S.C. 236, 238, 403 S.E.2d 640, 642 (1991) (quoting Small v. Springs Indus., Inc., 292 S.C. 481, 487, 357 S.E.2d 452, 455 (1987)). In light of the remand for a new trial, it is unnecessary to resolve Ford’s claim that the actual damage award is grossly excessive.

B.

Punitive damages

The issue of punitive damages was properly submitted to the jury. Ford, however, contends that the $15,000,000 puni*238tive damage award cannot withstand constitutional scrutiny. We agree. Because of the necessity of a new trial, we address two issues: Branham’s reliance on “harm to others” and the evidence of compensation of Ford’s executives.

The pervasive prejudice resulting from the improper post-manufacture eyidence on the issue of liability was compounded in Branham’s pursuit of punitive damages. Perhaps the manifestation of this error is most easily seen in counsel’s request that the jury punish Ford for harming others beyond Bran-ham. See Durham v. Vinson, 360 S.C. 639, 653, 602 S.E.2d 760, 767 (2004) (reversing an award of punitive damages because the trial court allowed the jury to “punish” the defendant for a “bad act unrelated” to the defendant’s action toward the plaintiff); see also Philip Morris USA v. Williams, 549 U.S. 346, 350, 353, 127 S.Ct. 1057, 166 L.Ed.2d 940 (2007) (reversing a punitive damages award where plaintiffs counsel asked the jury to “think about how many other Jesse Williams in the last 40 years in the State of Oregon there have been,” and holding that “the Constitution’s Due Process Clause forbids a State to use a punitive damages award to punish a defendant for injury that it inflicts upon nonparties or those whom they directly represent, ie., injury that it inflicts upon those who are, essentially, strangers to the litigation”).

As outlined above, punishing Ford in this case for harming all Bronco II rollover victims was a central theme in counsel’s closing argument.23 The trial court charged the jury not to punish Ford for other “conduct.” The charge violated the “harm to others” prohibition set forth in Durham v. Vinson *239and Philip Moms USA v. Williams. By focusing on conduct, as opposed to harm to Branham, the charge invited the jury to punish Ford for all Bronco rollover deaths and injuries — the very harm Durham and Philip Morris forbid.

We next examine the admission of financial data regarding Ford. Unless the United States Supreme Court holds otherwise under the Due Process Clause, we adhere to South Carolina law that “the wealth of a defendant is a relevant factor in assessing punitive damages.” Welch v. Epstein, 342 S.C. 279, 307, 536 S.E.2d 408, 423 (Ct.App.2000). This is frequently described as the “ability to pay” factor. But this factor is not without boundaries. “Punitive damages pose an acute danger of arbitrary deprivation of property. Jury instructions typically leave the jury with wide discretion in choosing amounts, and the presentation of evidence of a defendant’s net worth creates the potential that juries will use their verdicts to express biases against big businesses, particularly those without strong local presences.” Honda Motor Co. v. Oberg, 512 U.S. 415, 432, 114 S.Ct. 2331, 129 L.Ed.2d 336 (1994); see also State Farm Mut. Auto. Ins. Co. v. Campbell, 538 U.S. 408, 427, 123 S.Ct. 1513, 155 L.Ed.2d 585 (2003) (“[Preference to [the defendant’s] assets ... ha[s] little to do with the actual harm sustained by the [plaintiff]. The wealth of a defendant cannot justify an otherwise unconstitutional punitive damages award.”).

Branham presented evidence of Ford’s net worth, income, revenues and cash flow. In 2005, Ford’s net worth was $12,597,000,000. Also in 2005, Ford had $1,986,000,000 in income, $177,000,000,000 in revenue and $21,000,000,000 in cash flow. Branham extrapolated these figures to “per week,” “per day,” and “per hour.” For example, concerning Ford’s cash flow, “[t]hat’s $416 Billion per week,24 $59 Million per day, [and] $2,474 Million per hour.”

This Court has approved the use of a defendant’s net worth as a proper guide in assessing the “ability to pay” factor. Hicks v. Herring, 246 S.C. 429, 437, 144 S.E.2d 151, 154 (1965) (noting that “the wealth of a defendant is a relevant factor in *240assessing punitive damages”). This Court has not, however, addressed the propriety of extrapolating financial data in the manner introduced at trial. Our court of appeals has directly addressed this issue and found no abuse of discretion in the admission of per day earnings of a defendant, Orangeburg Sausage Co. v. Cincinnati Ins. Co., 316 S.C. 331, 344, 450 S.E.2d 66, 73-74 (Ct.App.1994), nor has it found an abuse of discretion in admitting the per day operating revenue or per day net income. Bryant v. Waste Mgmt., Inc., 342 S.C. 159, 169-70, 536 S.E.2d 380, 385-86 (Ct.App.2000). These court of appeals decisions have not been tested since the 2003 Supreme Court opinion of State Farm v. Campbell.

Because the United States Supreme Court has discovered that a state court’s punitive damages award implicates federal substantive due process, this Court is not the final arbiter of determining what financial evidence is proper in assessing punitive damages. Evidence concerning net worth appears the safest harbor. Honda Motor speaks directly to “net worth.” 512 U.S. at 432, 114 S.Ct. 2331. Consideration of a defendant’s net worth is well-rooted in the common law of punitive damages. State Farm v. Campbell’s cautionary observation that “reference to [the defendant’s] assets ... ha[s] little to do with the actual harm sustained by the [plaintiff]” militates against venturing beyond net worth and extrapolations from net worth. State Farm, 538 U.S. at 427, 123 S.Ct. 1513. The retrial shall be confined to such evidence.

While the United States Supreme Court’s foray into punitive damages law is, to be sure, confusing, there can be no serious doubt concerning financial evidence of the salaries and compensation of a defendant corporation’s officers. Such evidence introduces an arbitrary factor in a jury’s consideration and assessment of punitive damages.

Branham went far beyond the pale in submitting evidence of Ford’s senior management compensation, including the following: In 2005, the Ford Chairman and Chief Executive Officer was compensated by stock options worth $5,300,000; the Ford President and Chief Operating Officer received a salary $1,458,000 in 2005; the Ford Executive Vice President received a salary of $972,000 in 2005; the Ford Chief Financial Officer received a salary of $916,000 in 2005; a *241former Ford chairman received $880,000 in 2005. Additional testimony revealed that “[i]n 2005, they didn’t pay bonuses. In 2004, they did pay bonuses: $2 Million, $1 Million, $1 Million and $1 Million plus stock and other compensation. This gives you a picture of Ford Motor Company’s financial condition.” The admission of this evidence was error and highly prejudicial.

V.

Alignment of Parties

We reach Ford’s final challenge assigning error to the trial court’s failure to realign defendant Hale as a plaintiff. Ford requested realignment of Hale as a plaintiff so that Ford would not have to share its allotment of preemptory jury strikes25 with Hale.26 We find the issue not preserved for review, but we address this issue in the hope that our speaking to the matter will aid the bench and bar. This is a novel issue in our modern jurisprudence.

Hale and her counsel sat on the plaintiffs side throughout the trial, beginning with jury selection. We recognize that Hale filed a cross-claim against Ford, but that claim was severed from this trial. Hale’s counsel declined to cross-examine any witness called by Branham but one. The one witness Hale cross-examined was Branham’s economic expert. The question: “[H]ow many millions are in a billion?”

The only bona fide defendant in this case was Ford. The following is the totality of Branham’s closing argument concerning Hale:

I want to first talk with you just a minute about what occurred in the "wreck.... Ms. Hale was going down *242Cromwell Road at a relatively slow speed, 35 miles an hour, she looked in the back, went to the edge of the road, she made — went to the right, made a steer to the left.

Nobody knows what that steering was because nobody has a picture of it. We could argue about that from now to eternity and nobody would know, because nobody can know. But she was driving ordinary, she wasn’t doing anything— she wasn’t out there doing any reckless driving out there that day.

Here, Ms. Hale looked in the backseat; there’s no question about that, she took her eyes off the road. But did she do something that was wrong. She did what all reasonable drivers would do, which was she tried to get back on the road. She made the turn to do it and the vehicle rolled over, at 35 miles an hour, under those circumstances.

Trial judges in South Carolina have the authority to realign parties. Beyond a court’s inherent authority to manage and conduct a trial, our Rule 21, SCRCP, regarding joinder of parties is identical to the federal rule, Rule 21, FRCP.

Misjoinder of parties is not ground for dismissal of an action. Parties may be dropped or added by order of the court on motion of any party or of its own initiative at any stage of the action and on such terms as are just. Any claim against a party may be severed and proceeded with separately.

Rule 21, SCRCP.

Federal courts rely on Rule 21 as authority to realign parties. See In-Tech Mktg. Inc. v. Hasbro, Inc., 685 F.Supp. 436, 442 n. 19 (D.N.J.1988) (noting that Rule 21 “permits [the District] Court, sua sponte to re-align any party at any time”); First Nat’l Bank of Shawnee Mission v. Roeland Park State Bank & Trust Co., 357 F.Supp. 708, 711 (D.Kan. 1973) (noting that the District Court “may order a realignment of the parties ‘on such terms as are just’ ” pursuant to Rule 21). Our sister state of Georgia relies on Rule 21 in recognizing that “a trial court does have the discretion, ‘at any stage of the action and on such terms as are just,’ ... to realign the *243parties.” Cawthon v. Waco Fire & Cas. Ins. Co., 259 Ga. 632, 386 S.E.2d 32, 33 (1989) (citing its codified version of Rule 21).

The Cawthon decision is instructive. At trial, the Cawthons made a motion to have a co-defendant realigned as a plaintiff. The “Cawthons were concerned that they would be forced to share jury strikes with the [codefendant whose interests were aligned with the plaintiffs].” Id. at 33. The trial court recognized the unfairness of the present alignment, “but [stated] that it had no authority to realign the parties.” Id. Relying on its own version of Rule 21, the Georgia Supreme Court found the trial court erred in “concluding it did not have discretion to realign the parties,” and affirmatively held that trial courts have the right to realign parties in the interests of justice. Id.

We adopt the reasoning of Cawthon, including the authority of a trial court to realign parties “at any stage of the action.” Id. The decision whether to realign the parties lies within the sound discretion of the trial court and will not be disturbed on appeal absent a showing of an abuse of discretion and resulting prejudice.

VI.

The judgment of the trial court is affirmed in part, reversed in part and the case is remanded for a new trial.

AFFIRMED IN PART; REVERSED IN PART; AND REMANDED.

Justice PLEICONES.

I concur in part and dissent in part. While I agree that the risk-utility test is the appropriate test for design defect cases, I do not believe this Court has the power to simply discard the consumer expectations test, expressly adopted by the General Assembly in S.C.Code Ann. §§ 15-73-10 through 30. Furthermore, in my opinion, much of the evidence the majority views as improper “post-manufacture” evidence was properly *244admissible to prove (1) the foreseeable risk of harm posed by the Bronco II as produced, (2) that the proposed alternative designs could have reduced or avoided the foreseeable risk, and (3) that the Bronco II, absent the alternative designs, was not reasonably safe.

I. Risk-Utility Test

As the majority notes, the General Assembly adopted the Restatement (Second) of Torts § 402A in 1974. See S.C.Code Ann. §§ 15-73-10 through 30. The comments to § 402A, which form the basis for the consumer expectations test, were expressly adopted as legislative intent. S.C.Code Ann. § 15-73-30. The majority then notes that the American Law Institute has, in the Restatement (Third) of Torts, moved away from the consumer expectations test for design defects in favor of the risk-utility test and proposes that this Court do the same. While I agree with the majority that the risk-utility test is the appropriate test for design defect cases, I do not believe that this Court has the authority to simply reject the General Assembly’s chosen test, even if we believe that body would approve of the change. See Benat v. State Farm Mut. Ins. Co., 286 S.C. 132, 333 S.E.2d 57 (Ct.App.1985) (“It is the duty of this court to interpret the law. We have no legislative authority and cannot vary a statutory scheme and this is true no matter how logical the basis of the variance.”).

However, I believe that this Court may effect the same result under the existing statute by interpreting the consumer expectations test in the specific context of design defect cases. S.C.Code Ann. § 15-73-10 provides that “[o]ne who sells any product in a defective condition unreasonably dangerous to the user or consumer or to his property is subject to liability....” Comments to § 402A explain that a product is in a “defective condition unreasonably dangerous to the user or consumer or to his property” when the product is in “a condition not contemplated by the ultimate consumer, which will be unreasonably dangerous to him.” Restatement (Second) of Torts § 402A cmt. g. A product is unreasonably dangerous when it is “dangerous to an extent beyond that which would be contemplated by the ordinary consumer.” Restatement (Second) of Torts § 402A cmt. i. These comments form the basis for the consumer expectations test.

*245In my view, given the complexity of many modern products, a consumer’s expectations are not directed to any specific characteristic of the design, but rather to the manufacturer’s design decision. The ordinary consumer expects that the manufacturer will weigh the foreseeable risks against the benefits and only offer a product for sale to the general public if the latter outweighs the former. See 60 S.C.L.Rev. 1101.

Accordingly, I concur in the majority’s decision to apply risk-utility principles to design defect claims. However, in my view, such change must be achieved within the framework of existing statutory provisions.

II. Post-Manufacture Evidence

I respectfully disagree with the majority’s stance on “post-manufacture” evidence. The majority reverses the jury verdict based, in part, on its finding that “Ford was prejudiced by Branham’s unrelenting pursuit of post-manufacture evidence on the issue of liability.” The opinion defines “post-manufacture evidence” as “evidence of facts neither known nor available at the time of manufacture.” Such evidence is, in the majority’s view, inadmissible because “[wjhen assessing liability in a design defect claim, the judgment and ultimate decision of the manufacturer must be evaluated based on what was known at the time of manufacture.” I believe the majority’s rule sweeps too broadly and absorbs within its ambit evidence which is properly admissible in a design defect case.

I note at the outset that the majority opinion may be read as barring any evidence created after the date of manufacture. If this is the majority’s view, I strongly disagree. In my view, such a rule would deprive the fact finder of relevant evidence regarding what the manufacturer knew or should have known, design alternatives, and the risk inherent in the manufacturer’s design.

In a products liability action, the plaintiff must prove (1) that he was injured by the product; (2) that the injury occurred because the product was in a defective condition, unreasonably dangerous to the user; and (3) that the product at the time of the accident was in essentially the same condition as when it left the hands of the defendant. See Bragg v. Hi-Ranger, Inc., 319 S.C. 531, 539, 462 S.E.2d 321, *246326 (Ct.App.1995). Under the risk-utility test for design defect cases, a plaintiff must prove the second element, product defect, by showing that “the foreseeable risks of harm posed by the product could have been reduced or avoided by the adoption of a reasonable alternative design by the seller ... and the omission of the alternative design renders the product not reasonably safe.” Restatement (Third) of Torts: Products Liability § 2(b) (1998).

In seeking to meet his burden, the plaintiff introduced the following evidence which the majority finds objectionable:

(1) A memo dated April 14, 1989, dealing with a meeting that three Ford engineers had with representatives from Consumer Reports, discussing a comparative test of the Bronco II and other similar cars, showing the Bronco II to have a higher fatal rollover rate than the other cars.

(2) A film, taped in 1989, showing “J-tests” comparing the Chevy Blazer and Bronco II, and demonstrating that the Bronco II did not handle as well as the Blazer.

(3) A document that compared a 1989 Bronco II to a prototype of the Ford Explorer, showing that the handling strategy of the Bronco II makes the vehicle more sensitive to steering over-corrections that seem to be part of most rollover scenarios.

In finding the above evidence improper and inadmissible, the majority notes that:

Whether the 1987 Ford Bronco II was defectively designed and in a defective condition unreasonably dangerous must be determined as of the 1986 manufacture date of the vehicle. Ford’s 1986 design and manufacture decision should be assessed on the evidence available at that time, not the increased evidence of additional rollover data that came to light after 1986.

While I agree in general with the majority’s proposition, I note that when the reports were generated or tests conducted is of little consequence, since testimony established that the vehicles tested were substantially the same as the model involved in the accident, the testing methods were available to Ford prior to the date of manufacture, and the rollover risk was known to Ford prior to the date of manufacture. In *247short, the date on which the evidence was created is of little utility in determining the relevance of the evidence and a broad rule barring evidence created “post manufacture” actually serves to defeat the goals of the risk-utility test.

First, I believe the evidence was admissible to show foreseeable risk. The risk-utility test, as set forth in the Restatement (Third) of Torts, speaks not in terms of evidence of risk of which the manufacturer was actually aware, but in terms of foreseeable risk. No party disputes that Ford had the ability to test the 1987 Bronco II in the same way as was done in the disputed evidence mentioned above. In fact, Ford conducted such tests, the results of which led some Ford engineers to conclude that the wheel base design was flawed.27 Consequently, in my view, the memo, film, and document were properly admissible to show foreseeable risk, an essential element of the plaintiffs burden of proof in a design defect case.28

Second, I believe the video was admissible to show the viability of the proposed reasonable alternative design. To satisfy the risk-utility test, the plaintiff must prove, in most instances, that the foreseeable risk could have been avoided by the adoption of a reasonable alternative design. Restatement (Third) of Torts: Products Liability, § 2 cmt. d. The alternative design must be one that could have been practically adopted at the time of the sale. Id.

The plaintiff proposed reasonable alternative designs that were available at the time of manufacture, i.e. the MacPherson Strut suspension system and SLA suspension system, and he was entitled to an opportunity to show that the alternative designs could have reduced or avoided the foreseeable risk. Testimony at trial established that the Blazer used the SLA suspension system and the video demonstrated that the SLA *248system remedied the alleged rollover propensity of vehicles using the Twin I-Beam suspension system. The video was therefore, in my view, properly admitted.

Finally, I believe the memo, film, and document were properly admissible to aid the plaintiff in proving the final element of the risk-utility test: that “the omission of the alternative design renders the product not reasonably safe.” Though the “post-manufacture” evidence dealt with Bronco II vehicles manufactured between 1987 and 1989, testimony at trial established that there were no major changes to the Bronco II after 1987. The vehicles’ rollover propensities are therefore relevant to the issue of the reasonableness of Ford’s choice of the Twin I-Beam suspension system over the SLA or MacPherson system.

For the reasons stated above, I concur in part and dissent in part.

Connor T. Hansen, J.

The sole issue on this appeal is whether the complaint states a cause of action against the Esther Williams All-Aluminum Swimming Pool Company and the Banner Builders, Inc., and their insurers. We conclude that it does not. Therefore, we affirm the trial court in its decision sustaining the demurrer.

In Dippel v. Sciano,1 this court adopted sec. 402A of Restatement, 2 Torts 2d,2 pertaining to strict liability in *330tort. Under this section, where the plaintiff proves he was injured by a product “in a defective condition unreasonably dangerous to the user" and establishes the other requisite elements listed in the section, he is relieved of the burden of proving specific acts of negligence by the manufacturer who is then deemed negligent per se.3 Where a plaintiff is unable to prove the elements necessary to recovery under a theory of strict liability, the manufacturer or seller may still be liable under a' negligence theory where the plaintiff is able to prove specific negligent conduct. However, even under negligence law, the plaintiff still must prove that the product causing the injury was dangerous and defective.

Comment g to sec. 402A of Restatement, 2 Torts 2d, defines “defective condition" in part as follows:

“g. Defective condition. The rule stated in this Section applies only where the product is, at the time it leaves the seller’s hands, in a condition not contemplated by the ultimate consumer, which will be unreasonably dangerous to him." 4

*331The particular defect in the design of the swimming pool, as alleged in the complaint, is the absence of a self-latching and closing gate to prevent entry to the pool. We are satisfied that the swimming pool is not defective in this respect because, as a matter of law, the swimming pool was as safe as it reasonably could be since it did contain a retractable ladder, which unfortunately was allegedly left down and led to the injury of the small child.

Even if a product is defective, it must be shown to be unreasonably dangerous to the user or consumer. Comment i to sec. 402A of the Restatement defines “unreasonably dangerous” in part as follows:

“i. Unreasonably dangerous. The rule stated in this Section applies only where the defective condition of the product makes it unreasonably dangerous to the user or consumer. Many products cannot possibly be made entirely safe for all consumption, and 3jiy food or drug necessarily involves some risk of harm, if only from over-consumption. Ordinary sugar is a deadly poison to diabetics, and castor oil found use under Mussolini as an instrument of torture. That is not what is meant by ‘unreasonably dangerous’ in this Section. The article sold 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.” (Emphasis supplied.) 5

This emphasized language has been cited with approval in the strict liability cases of Netzel v. State Sand & Gravel Co.6 and Arbet v. Gussarson.7

In Arbet the court upheld as against a demurrer the complaint of two automobile accident burn victims alleging that their car was unreasonably dangerous and de*332fective because of a plastic gas line apparatus retaining gasoline in the passenger compartment. The court commented that the alleged defect in the car was a latent defect:

“. . . It must be noted also that the design characteristics complained of in the instant case were hidden dangers, not apparent to the buyer of the car, and not the subject of a manufacturer’s warning. This is a different case, therefore, than a case where a plaintiff sues the manufacturer of a Volkswagen and complains that the car was designed too small to be safe. Such a defect could hardly be said to be hidden. . . . since the ordinary consumer would expect a Volkswagen to be less safe in an accident than, say, a Cadillac, the smallness of the car with the attendant danger would not per se render it inherently dangerous. Rather it must contain a dangerous defect whose presence an ordinary consumer would not reasonably expect.”8

Thus, the test in Wisconsin of whether a product contains an unreasonably dangerous defect depends upon the reasonable expectations of the ordinary consumer concerning the characteristics of this type of product. If the average consumer would reasonably anticipate the dangerous condition of the product and fully appreciate the attendant risk of injury, it would not be unreasonably dangerous and defective. This is an objective test and is not dependent upon the knowledge of the particular injured consumer, although his knowledge may be evidence of contributory negligence under the circumstances.9 In Schuh v. Fox River Tractor Co.10 for example, the court held that the positioning of the clutch *333lever on a crop blower machine constituted an unreasonably dangerous defect because a potential user might be misled as to its function. However, the court held the particular injured plaintiff’s contributory negligence greater than any negligence of the manufacturer because the plaintiff was an experienced operator of the machine and knew of the potential dangers, yet failed to exercise due care.

Based upon the principles discussed above, we conclude that the swimming pool described in plaintiffs’ complaint does not contain an unreasonably dangerous defect. The lack of a self-latching gate certainly falls within the category of an obvious rather than a latent condition. Equally important, the average consumer would be completely aware of the risk of harm to small children due to this condition, when the retractable ladder is left in a down position and the children are left unsupervised. We conclude, therefore, that plaintiffs’ second amended complaint fails to state a cause of action.

By the Court. — Judgment and order affirmed.

Wilkie, C. J.

(dissenting). The first problem I find in the majority opinion is that it holds that there was no defect here as a matter of law. I would hold that this was a factual determination to be tried out by the trier of fact. There is a question here for the trier to determine whether the swimming pool (which did not have a self-latching and closing gate) was unavoidably unsafe as, for example, knives, baseball bats, alcohol, small foreign cars, and, therefore, not defective. Then, too, it would be a question of fact whether rendering the product safe by incorporating other safety features would destroy the usefulness of the product, or would be far too costly. On this the defective swimming pool manufacturer ¡would have the burden of proof.

The additional holding of the majority ruling out liability where the defect is obvious and apparent — as here *334—is really based upon the concept of assumption of risk which Dippel, supra, held was not an absolute bar to recovery but rather a matter of contributory negligence. I would, therefore, drop the requirement that a product must be affirmatively shown to be unreasonably dangerous and I would regard the introduction of this element as a factor of contributory negligence to be compared with the negligence on the part of the manufacturer.

The requirement of showing “unreasonable danger” was dropped by the California Supreme Court in the 1972 companion cases of Cronin v. J. B. E. Olson Corp.1 and Luque v. McLean.2 In Cronin the court held that a plaintiff no longer need show that the product injuring him was unreasonably dangerous, but rather merely that it was defective. The court did not precisely define what the term “defective” standing alone is now to mean,3 but it did explain its reasons for abandoning the “unreasonable danger” limitation:

“. . . Prosser, the reporter for the Restatement, suggests that the ‘unreasonably dangerous’ qualification was *335added to foreclose the possibility that the manufacturer of a product with inherent possibilities for harm (for example, butter, drugs, whiskey and automobiles) would become ‘automatically responsible for all the harm that such things do in the world.’ (Prosser, Strict Liability to the Consumer in California, 18 Hastings L. J. (1966), 9, 23.)

“The result of the limitation, however, has. not been merely to prevent the seller from becoming an insurer of his products with respect to all harm generated by their use. Rather, it has burdened the injured plaintiff with proof of an element which rings of negligence. As a result, if, in the view of the trier of fact, the ‘ordinary consumer’ would have expected the defective condition of a product, the seller is not strictly liable regardless of the expectations of the injured plaintiff. If, for example, the ‘ordinary consumer’ would have contemplated that Shopsmiths posed a risk of loosening their grip and letting the wood strike the operator, another Greenman might be denied recovery. In fact, it has been observed that the Restatement formulation of strict liability in practice rarely leads to a different conclusion than would have been reached under laws of negligence. . . . Yet the very purpose of our pioneering efforts in this field was to relieve the plaintiff from problems of proof inherent in pursuing negligence . . . and warranty . . . remedies, and thereby ‘to insure that the costs of injuries resulting from defective products are borne by the manufacturers

“. . . We think that a requirement that a plaintiff also prove that the defect made the product ‘unreasonably dangerous’ places upon him a significantly increased burden and represents a step backward in the area pioneered by this court.

“We recognize that the words ‘unreasonably dangerous’ may also serve the beneficial purpose of preventing the seller from being treated as the insurer of its products. However, we think that such protective end is attained by the necessity of proving that there was a defect in the manufacture or design of the product and that such defect was a proximate cause of the injuries. Although the seller should not be responsible for all injuries involving the use of its products, it should be liable for all *336injuries proximately caused by any of its products which are adjudged ‘defective.’ ”4

In Luque the court held that in strict liability cases the obviousness of the defect is not a bar to recovery, but that this factor could be an element of contributory negligence or unreasonable assumption of risk.

Thus, I would conclude that the obviousness of a defect is not a total bar to recovery, but merely a matter pertaining to contributory negligence.

There is an additional difficulty here that the small child is less than seven years of age and therefore incapable of negligence on his own. In swimming pools of this type, which are obviously intended for adult use, their use by children is anticipated but always under the immediate supervision of adults. In a products liability case, as here, I would therefore consider the negligence on the part of the parents for improper supervision as a matter of contributory negligence to be imputed to the child, and accordingly offset against the negligence of the manufacturer.

Further, the manufacturer here would still have the additional defense of showing that its negligence in producing a' defective product was not causal.

But in any event, the complaint does state a cause of action and the demurrer should have been overruled and the defendants ordered to answer.

I am authorized to state that Mr. Justice Heffernan joins in this dissent.

OPINION OF THE COURT

Titone, J.

Are the elements of New York’s causes of action for strict products liability and breach of implied warranty always coextensive? If not, can the latter be broader than the former? These are the core issues presented by the questions that the United States Court of Appeals for the Second Circuit has certified to us in this diversity action involving an allegedly defective vehicle. On the facts set forth by the Second Circuit, we hold that the causes of action are not identical and that, under the circumstances presented here, it is possible to be liable for breach of implied warranty even though a claim of strict products liability has not been satisfactorily established.

L

As stated by the Second Circuit, this action arises out of a June 9, 1986 accident in which plaintiff Nancy Denny was severely injured when the Ford Bronco II that she was driving rolled over. The rollover accident occurred when Denny slammed on her brakes in an effort to avoid a deer that had walked directly into her motor vehicle’s path. Denny and her spouse sued Ford Motor Co., the vehicle’s manufacturer, asserting claims for negligence, strict products liability and breach of implied warranty of merchantability (see, UCC 2-314 [2] [c]; 2-318). The case went to trial in the District Court for the Northern District of New York in October of 1992.

*252The trial evidence centered on the particular characteristics of utility vehicles, which are generally made for off-road use on unpaved and often rugged terrain. Such use sometimes necessitates climbing over obstacles such as fallen logs and rocks. While utility vehicles are traditionally considerably larger than passenger cars, some manufacturers have created a category of down-sized "small” utility vehicles, which are designed to be lighter, to achieve better fuel economy and, presumably, to appeal to a wider consumer market. The Bronco II in which Denny was injured falls into this category.

Plaintiffs introduced evidence at trial to show that small utility vehicles in general, and the Bronco II in particular, present a significantly higher risk of rollover accidents than do ordinary passenger automobiles. Plaintiffs’ evidence also showed that the Bronco II had a low stability index attributable to its high center of gravity and relatively narrow track width. The vehicle’s shorter wheel base and suspension system were additional factors contributing to its instability. Ford had made minor design changes in an effort to achieve a higher stability index, but, according to plaintiffs’ proof, none of the changes produced a significant improvement in the vehicle’s stability.

Ford argued at trial that the design features of which plaintiffs complained were necessary to the vehicle’s off-road capabilities. According to Ford, the vehicle had been intended to be used as an off-road vehicle and had not been designed to be sold as a conventional passenger automobile. Ford’s own engineer stated that he would not recommend the Bronco II to someone whose primary interest was to use it as a passenger car, since the features of a four-wheel-drive utility vehicle were not helpful for that purpose and the vehicle’s design made it inherently less stable.

Despite the engineer’s testimony, plaintiffs introduced a Ford marketing manual which predicted that many buyers would be attracted to the Bronco II because utility vehicles were "suitable to contemporary life styles” and were "considered fashionable” in some suburban areas. According to this manual, the sales presentation of the Bronco II should take into account the vehicle’s "suitability] for commuting and for suburban and city driving.” Additionally, the vehicle’s ability to switch between two-wheel and four-wheel drive would "be particularly appealing to women who may be concerned about driving in snow and ice with their children.” Plaintiffs both testified that *253the perceived safety benefits of its four-wheel-drive capacity were what attracted them to the Bronco II. They were not at all interested in its off-road use.

At the close of the evidence, the District Court Judge submitted both the strict products liability claim and the breach of implied warranty claim, despite Ford’s objection that the two causes of action were identical. With respect to the strict products liability claim the court told the jury that "[a] manufacturer who places a product on the market in a defective condition is liable for injury which results from use of the product when the product is used for its intended or reasonably foreseeable purpose.” Further, the court stated:

"A product is defective if it is not reasonably safe. * * * It is not necessary for the plaintiffs to prove that the defendant knew or should have known of the product[’]s potential for causing injury to establish that the product was not reasonably safe. Rather, the plaintiffs must prove by a preponderance of the evidence that a reasonable person * * * who knew of the product’s potential for causing injury and the existence of available alternative designs * * * would have concluded that such a product should not have been marketed in that condition. Such a conclusion should be reached after balancing the risks involved in using the product against the product[’]s usefulness and its costs against the risks, usefulness and costs of the alternative design as compared to the product defendant did market.”

With respect to the breach of implied warranty claim, the court told the jury:

"The law implies a warranty by a manufacturer which places its product on the market that the product is reasonably fit for the ordinary purpose for which it was intended. If it is, in fact, defective and not reasonably fit to be used for its intended purpose, the warranty is breached.

"The plaintiffs claim that the Bronco II was not fit for its ordinary purpose because of its alleged *254propensity to rollover and lack of warnings to the consumer of this propensity.”1

Neither party objected to the content of these charges.

In response to interrogatories, the jury found that the Bronco II was not "defective” and that defendant was therefore not liable under plaintiffs’ strict products liability cause of action. However, the jury also found that defendant had breached its implied warranty of merchantability and that the breach was the proximate cause of Nancy Denny’s injuries. Following apportionment of damages, plaintiff was awarded judgment in the amount of $1.2 million.

Ford subsequently moved for a new trial under rule 59 (a) of the Federal Rules of Civil Procedure, arguing that the jury’s finding on the breach of implied warranty cause of action was irreconcilable with its finding on the strict products liability claim. The trial court rejected this argument, holding that it had been waived and that, in any event, the verdict was not inconsistent.

On defendant’s appeal, a majority at the Second Circuit held that defendant’s trial conduct had not resulted in a waiver of the inconsistency issue. Reasoning that the outcome of the appeal depended upon the proper application of New York law, the court certified the following questions for consideration by this Court pursuant to article VI, § 3 (b) (9) of the State Constitution and rule 500.17 of the Rules of the Court of Appeals (22 NYCRR 500.17): (1) whether the strict products liability claim and the breach of implied warranty claim are identical; (2) whether, if the claims are different, the strict products liability claim is broader than the implied warranty claim and encompasses the latter; and (3) whether, if the claims are different and a strict liability claim may fail while an implied warranty claim succeeds, the jury’s finding of no product defect is reconcilable with its finding of a breach of warranty.

II

In this proceeding, Ford’s sole argument is that plaintiffs’ strict products liability and breach of implied warranty causes of action were identical and that, accordingly, a defendant’s verdict on the former cannot be reconciled with a plaintiff’s verdict on the latter. This argument is, in turn, premised on both the intertwined history of the two doctrines and the close *255similarity in their elements and legal functions. Although Ford recognizes that New York has previously permitted personal injury plaintiffs to simultaneously assert different products liability theories in support of their claims (see, Victorson v Bock Laundry Mach. Co., 37 NY2d 395, 400), it contends that the breach of implied warranty cause of action, which sounds in contract, has been subsumed by the more recently adopted, and more highly evolved, strict products liability theory, which sounds in tort. Ford’s argument has much to commend it. However, in the final analysis, the argument is flawed because it overlooks the continued existence of a separate statutory predicate for the breach of warranty theory and the subtle but important distinction between the two theories that arises from their different historical and doctrinal root.

When products liability litigation was in its infancy, the courts relied upon contractual warranty theories as the only existing means of facilitating economic recovery for personal injuries arising from the use of defective goods (e.g., Mendel v Pittsburgh Plate Glass Co., 25 NY2d 340, overruled on other grounds Victorson v Bock Laundry Mach. Co., supra; Blessington v McCrory Stores Corp., 305 NY 140; see, Heller v U. S. Suzuki Motor Corp., 64 NY2d 407, 410). Citing statutory authority (UCC 2-314, 2-715 [2] [b]; former Personal Property Law § 96 [1]), the courts posited the existence of an implied warranty arising as an incident of the product’s sale and premised a cause of action for consequential personal injuries based on breaches of that warranty (see, Heller v U. S. Suzuki Motor Corp., supra, at 410).

Eventually, the contractually based implied warranty theory came to be perceived as inadequate in an economic universe that was dominated by mass-produced products and an impersonal marketplace. Its primary weakness was, of course, its rigid requirement of a relationship of privity between the seller and the injured consumer — a requirement that often could not be satisfied (see, Martin v Dierck Equip. Co., 43 NY2d 583, 589-590). Some courts (including ours) recognized certain narrow exceptions to the privity requirement in an effort to avoid the doctrine’s harsher effects (e.g., Greenberg v Lorenz, 9 NY2d 195; see, Heller v U. S. Suzuki Motor Corp., supra, at 410; Prosser and Keeton, Torts § 96, at 682 [5th ed]). However, the warranty approach remained unsatisfactory, and the courts shifted their focus to the development of a new, more flexible tort cause of action: the doctrine of strict products liability (Martin v Dierck Equip. Co., supra, at 590; Micallef v Miehle Co., 39 NY2d 376; Victorson v Bock Laundry Mach. Co., supra, *256 at 402; see, Codling v Paglia, 32 NY2d 330; Goldberg v Kollsman Instrument Corp., 12 NY2d 432, 436; see also, MacPherson v Buick Motor Co., 217 NY 382).

The establishment of this tort remedy has, as this Court has recognized, significantly diminished the need to rely on the contractually based breach of implied warranty remedy as a means of compensating individuals injured because of defective products (see, Heller v U. S. Suzuki Motor Corp., supra, at 411; Martin v Dierck Equip. Co., supra, at 590). Further, although the available defenses and applicable limitations principles may differ, there is a high degree of overlap between the substantive aspects of the two causes of action (see, Victorson v Bock Laundry Mach. Co., supra, at 405). Indeed, on an earlier occasion, this Court observed, in dictum, that "strict liability in tort and implied warranty in the absence of privity are merely different ways of describing the very same cause of action” (Mendel v Pittsburgh Plate Glass Co., supra, at 345; accord, Gumbs v International Harvester, 718 F2d 88 [3d Cir]; Sterno Aero AB v Page Airmotive, 499 F2d 709, 712 [10th Cir]; Larsen v Pacesetter Sys., 74 Haw 1, 837 P2d 1273; 1 Frumer and Friedman, Products Liability § 2.03, at 2-28; 2 Frumer, op. cit., § 9.04 [1], at 9-42, 9-44; Clark and Smith, Product Warranties Tí 12.03 [1], at 12-7).

Nonetheless, it would not be correct to infer that the tort cause of action has completely subsumed the older breach of implied warranty cause of action or that the two doctrines are now identical in every respect (see, Di Prospero v Brown & Sons, 110 AD2d 250, 251). The continued vitality of the warranty approach is evidenced by its retention and expansion in New York’s version of the Uniform Commercial Code (UCC 2-314 [2] [c]; 2-318). The existence of this statutory authority belies any argument that the breach of implied warranty remedy is a dead letter (see, Heller v U. S. Suzuki Motor Corp., supra, at 411-412).2

Although the products liability theory sounding in tort and the breach of implied warranty theory authorized by the UCC coexist and are often invoked in tandem, the core element of "defect” is subtly different in the two causes of action. Under *257New York law, a design defect may be actionable under a strict products liability theory if the product is not reasonably safe. Since this Court’s decision in Voss v Black & Decker Mfg. Co. (59 NY2d 102, 108), the New York standard for determining the existence of a design defect has required an assessment of whether "if the design defect were known at the time of manufacture, a reasonable person would conclude that the utility of the product did not outweigh the risk inherent in marketing a product designed in that manner” (see also, Cover v Cohen, 61 NY2d 261, 270; Robinson v Reed-Prentice Div. of Package Mach. Co., 49 NY2d 471, 479). This standard demands an inquiry into such factors as (1) the product’s utility to the public as a whole, (2) its utility to the individual user, (3) the likelihood that the product will cause injury, (4) the availability of a safer design, (5) the possibility of designing and manufacturing the product so that it is safer but remains functional and reasonably priced, (6) the degree of awareness of the product’s potential danger that can reasonably be attributed to the injured user, and (7) the manufacturer’s ability to spread the cost of any safety-related design changes (Voss v Black & Decker Mfg. Co., supra, at 109). The above-described analysis is rooted in a recognition that there are both risks and benefits associated with many products and that there are instances in which a product’s inherent dangers cannot be eliminated without simultaneously compromising or completely nullifying its benefits (see, Prosser and Keeton, op. cit, § 99, at 699). In such circumstances, a weighing of the product’s benefits against its risks is an appropriate and necessary component of the liability assessment under the policy-based principles associated with tort law.

The adoption of this risk/utility balance as a component of the "defectiveness” element has brought the inquiry in design defect cases closer to that used in traditional negligence cases, where the reasonableness of an actor’s conduct is considered in light of a number of situational and policy-driven factors.3 While efforts have been made to steer away from the fault-*258oriented negligence principles by characterizing the design defect cause of action in terms of a product-based rather than a conduct-based analysis (see, e.g., Voss v Black & Decker Mfg. Co., supra, at 107; Barker v Lull Eng’g Co., 20 Cal 3d 413, 418, 573 P2d 443; Prosser and Keeton, op. cit., § 96, at 689), the reality is that the risk/utility balancing test is a "negligence-inspired” approach, since it invites the parties to adduce proof about the manufacturer’s choices and ultimately requires the fact finder to make "a judgment about [the manufacturer’s] judgment” (Birnbaum, Unmasking the Test for Design Defect: From Negligence [to Warranty] to Strict Liability to Negligence, 33 Vand L Rev 593, 610, 648; see, e.g., Sage v FairchildSwearingen Corp., 70 NY2d 579, 587; cf., Enright v Lilly & Co., 77 NY2d 377, 387 [failure to warn claim "though * * * couched in terms of strict liability, is indistinguishable from a negligence claim”]). In other words, an assessment of the manufacturer’s conduct is virtually inevitable, and, as one commentator observed, "[i]n general, * * * the strict liability concept of 'defective design’ [is] functionally synonymous with the earlier negligence concept of unreasonable designing” (Schwartz, New Products, Old Products, Evolving Law, Retroactive Law, 58 NYU L Rev 796, 803, citing United States v Carroll Towing Co., 159 F2d 169, 173 [Hand, J.]; see, e.g., Gauthier v AMF, Inc., 788 F2d 634, 637 [9th Cir] [Mont law]; Birchfield v International Harvester Co., 726 F2d 1131, 1139 [6th Cir] [Ohio law]; St. Germain v Husqvarna Corp., 544 A2d 1283, 1285 [Me]; 1 Frumer and Friedman, op. cit., § 2.02, at 2-14 — 2-16; § 2.04, at 2-35 — 2-36; Henderson and Twerski, Doctrinal Collapse in Products Liability: The Empty Shell of Failure to Warn, 65 NYU L Rev 265, 271-272).

It is this negligence-like risk/benefit component of the defect element that differentiates strict products liability claims from UCC-based breach of implied warranty claims in cases involving design defects. While the strict products concept of a product that is "not reasonably safe” requires a weighing of the product’s dangers against its over-all advantages, the UCC’s concept of a "defective” product requires an inquiry only into whether the product in question was "fit for the ordinary purposes for which such goods are used” (UCC 2-314 [2] [c]).4 The latter inquiry focuses on the expectations for the perfor*259manee of the product when used in the customary, usual and reasonably foreseeable manners. The cause of action is one involving true "strict” liability, since recovery may be had upon a showing that the product was not minimally safe for its expected purpose — without regard to the feasibility of alternative designs or the manufacturer’s "reasonableness” in marketing it in that unsafe condition.

This distinction between the "defect” analysis in breach of implied warranty actions and the "defect” analysis in strict products liability actions is explained by the differing etiology and doctrinal underpinnings of the two distinct theories. The former class of actions originates in contract law, which directs its attention to the purchaser’s disappointed expectations; the latter originates in tort law, which traditionally has concerned itself with social policy and risk allocation by means other than those dictated by the marketplace.

The dissent takes issue with the foregoing conclusion, arguing, in essence, that any residual distinction that exists between the two causes of action should be eliminated and that the analysis for "defect” in implied warranty claims should be deemed to encompass the risk/utility analysis that has previously been incorporated in tort causes of action. This argument is predicated on the dissent’s view that the common history of the two causes of action and the perceived advantages of risk/ utility analysis counsel in favor of the use of a unitary standard. The dissent has even gone so far as to suggest that the breach of implied warranty cause of action should be treated like a tort claim despite the fact that it is based on the provisions of the Uniform Commercial Code.

What the dissent overlooks is that, as long as that legislative source of authority exists, we are not free to merge the warranty cause of action with its tort-based sibling regardless of whether, as a matter of policy, the contract-based warranty claim may fairly be regarded as a historical relic that no longer has any independent substantive value. Rather, we must construe and apply this separate remedy in a manner that remains consistent with its current roots in contract law (see, Codling v Paglia, supra [recognizing a tort cause of action to avoid stretching the breach of implied warranty theory to the point where it no longer reflects its origin as part of the bargain between the consumer and seller]).

*260To the extent that the dissent advocates a merger of the common-law and statutory causes of action through the use of a single analytical standard, its argument is undermined by an examination of what other jurisdictions have done. In most of the cases where the courts have pronounced the merger of breach of warranty with the other products liability theories sounding in tort, they were relying on specific State statutory schemes that were enacted to govern products liability litigation, contain express preemptive language and also specifically define "product liability claim” as one encompassing breach of express or implied warranty as well as negligence and strict liability in tort (see, e.g., Philpott v Robbins Co., 710 F2d 1422 [applying Ore Rev Stat § 30.905]; Chamberlain v Schmutz Mfg. Co., 532 F Supp 588 [applying Kan Stat Ann § 60-3301]; Daily v New Britain Mach. Co., 200 Conn 562, 512 A2d 893 [applying Conn Gen Stat Ann § 52-572m]; Washington Water Power Co. v Graybar Elec. Co., 112 Wash 2d 847, 774 P2d 1199 [applying Wash Rev Code Ann § 7.72.010]; see also, McWilliams v Yamaha Motor Corp., 780 F Supp 251, revd on other grounds 987 F2d 200; but see, Grinnel v Pfizer & Co., 21A Cal App 2d 424, 432, 79 Cal Rptr 369). Indeed, the proposed Model Uniform Product Liability Act, which was issued by the Commerce Department in 1979 (reprinted in 3B Frumer and Friedman, op. cit., Appendix B; see, 44 Fed Reg 62721), embodies precisely the kind of doctrinal merger that the dissent advocates. New York, of course, has not adopted the Model Act or any other such unifying measures.5

Contrary to the dissent’s suggestion, the current version of UCC 2-318 is not the equivalent of these uniform product liability provisions, nor does it manifest an intention by our State’s Legislature to engraft a tort cause of action onto a UCC article that concerns itself principally with the contract-based obligations (see, dissenting opn, at 272). Indeed, the Law Revision Commission Staff Notes, which the dissent cites, clearly state that the proposed amendments to UCC 2-318 "would * * * allow recovery by the [strict products liability] plaintiffs on a different cause of action” (Bill Jacket, L 1975, ch 774, Mem of NY Law Rev Commn, Staff Notes relating to A-3070 [emphasis supplied]). Similarly, the Sponsoring Memorandum on which the dissent relies states that the bill’s purpose was to *261"extend more intelligently the warranty provided to a purchaser of goods under the UCC” (Mem of Assemblyman Silverman, reprinted in 1975 NY Legis Ann, at 110). In fact, it is evident from the legislative materials accompanying the bill’s passage that its purpose was to expand the class of plaintiffs who can avail themselves of the Code’s warranty remedies and not to transform those remedies into a new tort cause of action (see, 1A ULA 558 [Master ed], UCC 2-318, Official Comment).

Moreover, the dissent’s novel proposal that the contract-based consumer-expectation test should be abandoned for the tort-based risk/utility approach even for contract-based warranty claims has not been embraced or even suggested by any of the risk/utility advocates that the dissent cites. For example, although the drafters of the Third Restatement have endorsed risk/utility analysis for design defect cases sounding in tort, they also have made clear that claims based on warranty theories are "not within the scope” of the newly drafted section and are, in fact, "unaffected by it” (Restatement [Third] of Torts: Products Liability [Tent Draft No. 2, Mar. 13, 1995] § 2, comment m, at 42). Further, the drafters have noted that "[w]arranty law as a body of legal doctrine separate from tort may impose legal obligations that go beyond those set forth” in the Restatement of Torts (id., comment q, at 46).

Similarly, while the commentators on which the dissent relies criticize the consumer-expectation-based tests for product defect and argue instead for the use of a risk/utility approach, their arguments are addressed to tort causes of action alone. One of the cited commentators, for example, argues that the consumer expectation test is a "blunt instrument” "when it comes to recognizing and maximizing the * * * goals, objectives, interests and values important to modern tort law” (Kennedy, The Role of the Consumer Expectation Test under Louisiana’s Products Liability Tort Doctrine, 69 Tul L Rev 117, 152 [emphasis supplied]). The same commentator also acknowledges that different standards might be appropriate for different theories of recovery where other objectives and values are pertinent (id.). Another commentator cited by the dissent contends that the risk/utility analysis should be used in place of a consumer-expectation test, but the argument is, once again, premised on the assumption that the latter "is not a tort way of looking at the problem of product defect” (Birnbaum, op. cit., at 646 [emphasis supplied]). This commentator also affirmatively criticizes courts that have failed "to separate conceptually the notions of strict liability, negligence, warranty, and absolute liability” (id., at 601).

*262Significantly, the consumer-expectation test has its advocates as well as its critics. In fact, the proposed Model Uniform Products Liability Act has itself been criticized on the ground that it does what the dissent urges, i.e., it eliminates consumer expectation as a test for tort claims (Twerski and Weinstein, A Critique of the Uniform Products Liability Law — A Rush to Judgment, 28 Drake L Rev 221, 230-233; accord, 1 Frumer and Friedman, op. cit., § 1.08 [2] [c] [ii]). Such criticisms stem from recent expressions by "courts and commentators [of] considerable support for a threshold test which does not require that the complexities of risk-utility analysis be undertaken in every design defect case” (Twerski and Weinstein, op. cit., at 230-231). In view of the "rigors of the risk-utility test,” it has been suggested that it is "worthwhile” to retain the consumer-expectation test and "explor[e] solutions to [its] subjectivity problem” rather than simply abandoning it (id., at 232).6

In any event, while the critics and commentators may debate the relative merits of the consumer-expectation and risk/utility tests, there is no existing authority for the proposition that the risk/utility analysis is appropriate when the plaintiffs claim rests on a claimed breach of implied warranty under UCC 2-314 (2) (c) and 2-318. Further, the absence of authority for the dissent’s position is not surprising since the negligence-like risk/ utility approach is foreign to the realm of contract law.

As a practical matter, the distinction between the defect concepts in tort law and in implied warranty theory may have little or no effect in most cases. In this case, however, the nature of the proof and the way in which the fact issues were litigated demonstrates how the two causes of action can diverge. In the trial court, Ford took the position that the design features of which plaintiffs complain, i.e., the Bronco II’s high center of gravity, narrow track width, short wheel base and specially tailored suspension system, were important to preserving the vehicle’s ability to drive over the highly irregular terrain that typifies off-road travel. Ford’s proof in this regard was relevant to the strict products liability risk/utility equation, which required the fact finder to determine whether the Bronco II’s value as an off-road vehicle outweighed the risk of the rollover accidents that could occur when the vehicle was used for other driving tasks.

*263On the other hand, plaintiffs’ proof focused, in part, on the sale of the Bronco II for suburban driving and everyday road travel. Plaintiffs also adduced proof that the Bronco II’s design characteristics made it unusually susceptible to rollover accidents when used on paved roads. All of this evidence was useful in showing that routine highway and street driving was the "ordinary purpose” for which the Bronco II was sold and that it was not "fit” — or safe — for that purpose.

Thus, under the evidence in this case, a rational fact finder could have simultaneously concluded that the Bronco II’s utility as an off-road vehicle outweighed the risk of injury resulting from rollover accidents and that the vehicle was not safe for the "ordinary purpose” of daily driving for which it was marketed and sold. Under the law of this State such a set of factual judgments would lead to the concomitant legal conclusion that plaintiffs’ strict products liability cause of action was not viable but that defendant should nevertheless be held liable for breach of its implied promise that the Bronco II was "merchantable” or "fit” for its "ordinary purpose.” Importantly, what makes this case distinctive is that the "ordinary purpose” for which the product was marketed and sold to the plaintiff was not the same as the utility against which the risk was to be weighed. It is these unusual circumstances that give practical significance to the ordinarily theoretical difference between the defect concepts in tort and statutory breach of implied warranty causes of action (see, e.g., McLaughlin v Michelin Tire Corp., 778 P2d 59, 66-67 [Wyo]; accord, 1 Madden, Products Liability § 5.11, at 160 [2d ed]).

From the foregoing it is apparent that the causes of action for strict products liability and breach of implied warranty of merchantability are not identical in New York and that the latter is not necessarily subsumed by the former. It follows that, under the circumstances presented, a verdict such as the one occurring here — in which the manufacturer was found liable under an implied warranty cause of action and not liable under a strict products cause of action — is theoretically reconcilable under New York law. Whether the particular verdict produced by the jury in this case was reconcilable in light of the charge and in accordance with case law applying rule 59 (a) of the Federal Rules of Civil Procedure is a question of Federal procedure which we are not well positioned to *264resolve.7 Hence, we construe the third certified question as posing only the theoretical question of whether this jury’s verdict is hypothetically possible under New York’s governing legal principles.

Accordingly, certified question No. 1 should be answered in the negative, certified question No. 2 in the negative and certified question No. 3 in the affirmative.

Simons, J.

(dissenting). I agree with the majority that causes of action in strict products liability and breach of implied warranty are not identical. In my view, however, the strict products liability claim is substantively broader than and encompasses the implied warranty claim and, thus, the jury’s verdict of no defect in the products liability cause of action is not reconcilable with its finding of breach of implied warranty. Accordingly, I would answer the first two questions certified to the Court no and yes and find it unnecessary to answer the third question.

I

Liability without fault may be imposed against a manufacturer or supplier of a defective product and in favor of one injured by the product. The product may be defective because it is improperly made, because its design is defective or because the manufacturer’s warnings against foreseeable risks in using it are inadequate. The members of the Court agree that strict products liability and implied warranty are similar in the sense that both causes of action require that, before plaintiff may recover, the product be defective, i.e., there must be something wrong with it. We disagree, however, over how defectiveness is determined. The question does not appear to have been previously addressed by the Court in the context of personal injury litigation.

The majority concludes that the implied warranty and strict products liability causes of action are different because the existence of an actionable defect is determined by two different *265analyses. Viewing implied warranty from a contract perspective, it would define defectiveness by whether the product lived up to the consumer’s expectations whereas defectiveness, for strict products liability purposes, is determined by application of the risk/utility standard. In my judgment, the consumer expectation standard, appropriate to commercial sales transactions, has no place in personal injury litigation alleging a design defect and may result in imposing absolute liability on marketers of consumers’ products. Whether a product has been defectively designed should be determined in a personal injury action by a risk/utility analysis.

A

Logically, there is no substantive difference for testing liability in the two causes of action. Recovery in each depends upon establishing that the product was defective because improperly designed. But the word "defect” has no clear legal meaning. In this case, the court defined defect in its strict products liability charge but did not attempt to define it otherwise; in the warranty cause of action the meaning had to be found in the court’s instructions describing the nature of the cause of action. Nevertheless, the predicate for recovery in both claims was the same.

The court charged the jury that to recover in strict products liability the plaintiffs had to prove that the Bronco II was "defective” when it was placed on the market. A product is defective, the court said, if it is "not reasonably safe” when used for "its intended or reasonably foreseeable purpose.” That charge was consistent with settled New York law which holds that a manufacturer or supplier may be strictly liable for injuries sustained when a product is used for its intended purpose or for an unintended but reasonably foreseeable purpose (see, Lugo v LTN Toys, 75 NY2d 850, 852; Micallef v Miehle Co., 39 NY2d 376, 385-386; Biss v Tenneco, Inc., 64 AD2d 204, 206). The court charged the jury that to recover for breach of implied warranty the plaintiff was required to establish that the Bronco II was not "reasonably fit for the ordinary purpose for which it was intended.” That instruction is consistent with language found in UCC 2-314 (2) (c).

When these two definitions are compared, it is apparent that a defect for strict products liability purposes is broader than a defect for implied warranty purposes. The vehicle could not have been defective when used for its ordinary and intended purpose (warranty), but not defective and reasonably safe when *266used for its "intended or for an unintended but reasonably foreseeable purpose” (strict products liability). As the Court of Appeals observed, foreseeable use "certainly includes all uses that are 'ordinary’ [and] perhaps some that are not 'ordinary’ ” (see, Denny v Ford Motor Co., 42 F3d 106, 112). The jury having concluded that the Bronco II was not defective for strict products liability purposes, could not logically conclude that it was defective for warranty purposes.

B

Nor is there any legal reason to distinguish the two causes of action in this respect. Breach of implied warranty and strict liability in tort developed from separate legal doctrines but are not materially different when applied to personal injury claims involving design defects. While breach of implied warranty retains its contractual law characteristics when applied to commercial transactions, it has been consistently recognized that it is a tort when applied to personal injury litigation and that tort principles should apply. To introduce a new test of defectiveness into tort litigation — one based on contract principles — can only destabilize the well-settled law in this area. Both causes of action are torts and defectiveness for both should be determined by the same standard.

The law imposing liability without fault against those making and marketing consumer products evolved in stages, progressing from negligence to implied warranty and eventually to the adoption in New York of a new cause of action known as strict products liability. Implied warranty has been generally associated with the law of contracts (although the Restatement advises us warranty was originally a matter of tort liability), but if implied warranty ever was a contract doctrine, it is now something very different from the warranty cause of action used in commercial transactions (see, Restatement [Second] of Torts § 402 A, comment m; 5 Harper, James and Gray, Torts § 28.27, at 540 [2d ed]; Prosser and Keeton, Torts § 97, at 691 [5th ed]; 1 Weinberger, New York Products Liability § 15:03). Indeed, the idea that there could ever be a claim for breach of implied warranty without privity is a concept entirely foreign to contract law. Moreover, the liability currently imposed in the name of warranty goes far beyond any liability based upon conventional contract notions and encompasses such tort concepts as consequential damages and contributory fault. As Dean Prosser has said: "[T]his warranty, if that is the name for it * * * is something separate and *267distinct which sounds in tort exclusively, and not at all in contract; which exists apart from any contract between the parties; and which makes for strict liability in tort” (Prosser, Spectacular Change: Products Liability in General, 36 Cleveland Bar Assn J 149, 167-168).

Finally, there can be no doubt about how this Court has viewed the action. We have repeatedly recognized not only that breach of implied warranty when asserted to recover for personal injuries is a tortious wrong (see, Victorson v Bock Laundry Mach. Co., 37 NY2d 395, 402; Velez v Craine & Clark Lbr. Corp., 33 NY2d 117, 124 [converting an action in implied warranty to one for strict products liability]; Codling v Paglia, 32 NY2d 330, 340, quoting Singer v Walker, 39 AD2d 90; Goldberg v Kollsman Instrument Corp., 12 NY2d 432, 436), but also "that strict liability in tort and implied warranty in the absence of privity are merely different ways of describing the very same cause of action” (Mendel v Pittsburgh Plate Glass Co., 25 NY2d 340, 345).

Nevertheless, the idea that contractual principles inhere in breach of implied warranty claims for personal injuries has persisted, producing conceptual difficulties and anomalies when the courts tried to apply the cause of action in a tort setting {see, Prosser and Keeton, Torts § 97, at 692 [5th ed]). In Codling v Paglia (32 NY2d 330, supra), we were confronted with a claim in implied warranty seeking to impose liability against a manufacturer in favor of a nonuser bystander injured by a defective automobile. We had long since abandoned the privity requirement in many personal injury claims based on implied warranty and incrementally extended the duty of manufacturers and suppliers not only to purchasers and users, but to users’ family members (see, Greenberg v Lorenz, 9 NY2d 195), to remote purchasers (Randy Knitwear v American Cyanamid Co., 11 NY2d 5), to an airline passenger suing the manufacturer of a defective component part of an airplane (Goldberg v Kollsman Instrument Corp., 12 NY2d 432, supra), and to rescuers suing the manufacturer of a defective oxygen mask (Guarino v Mine Safety Appliance Co., 25 NY2d 460). In Codling we recognized the difficulties in adopting implied warranty principles in personal injury claims and, abandoning privity entirely, recognized a new cause of action under the broad principle of strict products liability, as other courts before us had done, to hold the manufacturer liable to the bystander.

This new cause of action was not separate from implied warranty but an amalgam which had been constructed by the *268courts to establish a cause of action for liability without fault by merging warranty concepts (to avoid fault analysis) with negligence concepts (to avoid privity) (see, Victorson v Bock Laundry Mach. Co., 37 NY2d 395, 401, supra; Restatement [Third] of Torts: Products Liability [Tent Draft No. 2] § 1). The new cause of action recognized products liability as a discrete area of tort law, which borrows from both negligence and warranty, and attempts to avoid the confusion spawned by trying to categorize the various claims and remedies under prior law (id.). It imposes strict liability as a matter of social policy predicated on the idea that defendants ought "to pay for the costs attributable to damaging events caused by defects of a kind that made the product more dangerous than it would otherwise be”, concerns that had little to do with conventional contract principles (see, Prosser and Keeton, Torts § 98, at 692 [5th ed]). A difficulty has arisen, however, because in recognizing a cause of action for strict products liability, the courts have not had "a clear notion about the 'meaning of defect’ ”, especially in the context of defective design cases (id.).

In sum, although procedural distinctions may remain because mandated by the Legislature’s enactment of various provisions of the Uniform Commercial Code (see, Heller v U. S. Suzuki Motor Corp., 64 NY2d 407, 411), strict products liability and breach of implied warranty causes of action are substantively similar and impose liability without fault (see, Martin v Dierck Equip. Co., 43 NY2d 583, 589-590; Mendel v Pittsburgh Plate Glass Co., 25 NY2d 340, 345, supra; Ryion v Len-Co Lbr. Corp., 152 AD2d 978; Dickey v Lockport Prestress, 52 AD2d 1075, 1076). It makes little sense, therefore, to perpetuate a legal distinction between them based upon the method for determining defectiveness, particularly when the flaws in the consumer expectation standard for measuring defectiveness are recognized.

II

The majority has not attempted to define the consumer expectation standard, nor did the District Court use the phrase in its charge. Under one formulation, however, the standard provides that a product is defective, i.e., it is unreasonably dangerous, if it is "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” (see, Restatement [Second] of Torts § 402 A, comment i; see also, Kennedy, The Role of the *269 Consumer Expectation Test under Louisiana’s Products Liability Tort Doctrine, 69 Tul L Rev 117, 120 [1994]). The consumer expectation standard originated from the sales notion that a seller could agree, expressly or impliedly, to indemnify a buyer if the purchased product did not satisfy the buyer’s purposes. The obligation to "indemnify” applied only to the parties to the sale, those in privity, and did not "run with the goods” (see, 5 Harper, James and Gray, op. cit., § 28.16, at 454). As evolving social policy sought to hold manufacturers and sellers liable for personal injuries caused by defective products, however, the requirement of privity was narrowed and then eliminated, and the courts extended liability as far as social policy required (id., at 455-456). With these developments, it made little sense to think in terms of the buyer’s bargain or expectations. In many, if not most, cases the buyer was not litigating.

By contrast, the standard usually employed to determine design defectiveness in strict products liability claims requires a balancing of the risks attendant on using the product with the utility of the product when used as intended. As we stated in Robinson v Reed-Prentice Div. of Package Mach. Co. (49 NY2d 471, 479): "Where a product presents an unreasonable risk of harm, notwithstanding that it was meticulously made according to detailed plans and specifications, it is said to be defectively designed. This rule, however, is tempered by the realization that some products, for example knives, must by their very nature be dangerous in order to be functional. Thus, a defectively designed product is one which, at the time it leaves the seller’s hands, is in a condition not reasonably contemplated by the ultimate consumer and is unreasonably dangerous for its intended use; that is one whose utility does not outweigh the danger inherent in its introduction into the stream of commerce” (see also, Voss v Black & Decker Mfg. Co., 59 NY2d 102, 107-108; Rainbow v Elia Bldg. Co., 79 AD2d 287, 291, affd 56 NY2d 550).

Although some jurisdictions have recognized the consumer expectation standard, or some variation of it, in tort litigation* New York has never done so and its utility for resolving claims of design defects has been widely criticized by commentators (see, Birnbaum, Unmasking the Test for Design Defect: From Negligence [to Warranty] to Strict Liability to Negligence, 33 *270Vand L Rev 593, 611-618 [1980]; Kennedy, The Role of the Consumer Expectation Test under Louisiana’s Products Liability Tort Doctrine, 69 Tul L Rev 117, 143-150; Fischer, Products Liability — The Meaning of Defect, 39 Mo L Rev 339, 348-350 [1974]; and see, authorities cited in 5 Harper, James and Gray, Torts § 28.32A, at 576). They contend that the test is ambiguous because it does not clearly refer to the expectations of the actual plaintiff or to those of ordinary consumers; in practice it has been applied inconsistently and, from a social policy standpoint, it produces bad results.

If the test is applied to determine the actual buyer’s expectations, as in contract law, it can result in imposing absolute liability upon manufacturers and sellers making them insurers of the product’s safety merely because the product did not live up to the consumer’s subjective expectations. If the test is used objectively, it is beyond the experience of most lay jurors to determine what an "ordinary consumer” expects or "how safe” a sophisticated modern product could or should be made to satisfy those expectations unless the jury is allowed to consider the cost or impracticality of alternative designs or, indeed whether any alternative design for the product was available.

The test can also produce bad results. For example, if the risk is one that is easily understood and appreciated by the average consumer, the manufacturer might not be liable even if the defect could be eliminated by available and inexpensive design changes. Conversely, if the defect was not apparent, liability might attach even if the product was in fact state of the art.

Moreover, the consumer expectation test is unworkable when applied in cases involving design defects. In claims involving manufacturing defects, a consumer may reasonably expect a product to be made in accordance with the manufacturer’s standards and expect to be compensated for injuries resulting from the manufacturer’s failure to meet them. The product is reasonably held defective because the manufacturer has not made the product as it intended. However, in design defect cases the plaintiff contends that the product has been made precisely as intended but is nevertheless defective because the design is defective. But unless some external standard, such as available alternative designs and risk/utility analysis is employed, how is the jury to measure the propriety of the design? The consumer cannot reasonably expect a design to be changed if the cost of doing so far outweighs the utility of the product or if there is no alternative design available. Some *271products are inherently dangerous, knives was the illustration we used in the Robinson case (supra), and when that is so, policy concerns mandate that the responsibility for risks that cannot reasonably be designed out of a product should be transferred to the consumer, the party who has the choice of using them or not. (Restatement [Third] of Torts: Products Liability [Tent Draft No. 2] § 2, comment a, at 16.) The method for determining just what products fall within that group is the risk/utility analysis.

Because of these and other shortcomings, one commentator has stated that, when it comes to measuring defectiveness, the consumer expectation test applied without a risk/utility analysis is "a blunt instrument” (Kennedy, op. cit., at 150). Few courts have relied solely on it as a measure of defectiveness (see, Birnbaum, op. cit, at 615).

No New York court has recognized the consumer expectation standard to determine defectiveness in personal injury actions grounded on implied warranty — at least the parties and the majority have not cited any decision doing so — and I can see no persuasive policy reasons why we should do so now. If the test is unworkable when applied in tort causes of action grounded on strict products liability, it is equally unworkable when applied in tort causes of action grounded on breach of implied warranty. The correct standard in strict liability claims, according to the Third Restatement, should include a balancing of the risk of danger against the utility of the product as designed. In its words, "consumer expectations do not constitute an independent standard for judging the defectiveness of product designs” (Restatement [Third] of Torts: Products Liability [Tent Draft No. 2] § 2, comment f, at 29). They are "not determinative of defectiveness” because they do not take into account "whether the proposed alternative design could be implemented at reasonable cost, or whether an alternative design would provide greater overall safety”, i.e., the test does not take into consideration risk/utility factors (id.). Consumer expectations only value is when used as a factor in determining the reasonableness of alternative designs or how the product is portrayed and perceived by the public, i.e., whether the risk was foreseeable. As we stated in Robinson v Reed-Prentice Div. of Package Mach. Co. (supra) the conditions contemplated by "the ultimate consumer” must be taken into account, but the risk/utility analysis remains a necessary part of the equation for determining defectiveness in products liability cases (Restatement [Third] of Torts, op. cit; see also, Birnbaum, op. cit, at 617).

*272Ill

The majority maintains, however, that the consumer expectation standard must be applied because breach of implied warranty is a statutory cause of action and the Court is not free to ignore the statute’s provisions or draw a distinction between its application to commercial claims and personal injury claims.

Implied warranties have been a part of our statutory law since at least 1911, long before any serious attempt was made to base tort liability on them {see, former Personal Property Law § 96, now UCC 2-314). Section 96, and its successor provisions in the Uniform Commercial Code, were enacted to address problems arising in commercial transactions. For many years they had no significant impact upon personal injury litigation because of the rules of privity. However, in 1975, shortly after Codling v Paglia (supra) was decided, section 2-318 of the Uniform Commercial Code was amended to harmonize it with existing case law by eliminating the requirement of privity in personal injury claims (see, 1975 NY Legis Ann, at 110; Heller v U. S. Suzuki Motor Corp., 64 NY2d 407, 411, supra). The amendment had no relevance to commercial claims; it was proposed by the Legislature, and widely supported, because it acknowledged and encouraged the judicial development of a separate category of warranty providing a tort remedy for personal injuries (see, 1975 NY Legis Ann, at 110; see also, Bill Jacket, L 1975, ch 774, Mem of State Consumer Protection Board, July 14, 1975; Mem of NY Law Rev Commn, Staff Notes Relating to A-3070; Mem of New York State Trial Lawyers Assn, May 12, 1975). The Legislature’s recognition of a distinction between the statutory cause of action for personal injury claims and commercial claims based on implied warranty is further manifested by the Legislature’s decision to adopt alternative B of the three formulations proposed by the National Conference of the Commissioners on Uniform State Laws, the alternative which removed the requirement of privity in personal injury claims based upon implied warranty, rather than alternative C which extends the rule (abolishing privity) to warranty claims other than those dealing with injuries to the person {see, 1A ULA 558 [Master ed], UCC 2-318, Official Comment 3).

Moreover, no words in the statute either before or after the amendment, provide that the defectiveness of the product in tort claims, or commercial claims for that matter, is to be measured by the consumer’s expectations. That standard has been *273developed by the courts. It may accurately assess the terms and conditions of the bargain between the parties to a sale but it can hardly extend beyond them to address defectiveness in the sense that something is "wrong” with the product. The thing "wrong” with the product in the consumer expectation test is that it has not lived up to the consumer’s expectations and this is so even if the design of the product is perfection itself. The standard may retain some vitality when applied to commercial transactions but its individualized concept of injury is entirely foreign to tort doctrine underlying this area of law which is based upon the broad concept of enterprise responsibility to protect the public at large from harm.

Moreover, the statutory formulation of implied warranty has never restricted us in developing the tort remedy before. Long before the statute eliminated the requirement of privity for recovery, the courts narrowed and then eliminated it altogether. We did not feel inhibited by the statute in doing so: policy, not language, controlled the interpretation and application of the statute. Nor have the courts been constrained by the statute’s provisions when eliminating the UCC’s requirement of notice in tort actions (see, Fischer v Mead Johnson Labs., 41 AD2d 737; Kennedy v Woolworth Co., 205 App Div 648) or when shaping the law of disclaimers to apply them neutrally to personal injury cases (see, Velez v Craine & Clark Lbr. Corp., supra; see also, Walsh v Ford Motor Co., 59 Misc 2d 241; see also, 5 Harper, James and Gray, Torts § 28.25 [2d ed]).

The warranty claim in this case was for tortious personal injury and rests on the underlying "social concern [for] the protection of human life and property, not regularity in commercial exchange” (see, Restatement [Third] of Torts, op. cit., § 2, comment q, at 46). As such, it should be governed by tort rules, not contract rules. Nothing has prevented us in the past from construing and applying the provisions of the Uniform Commercial Code to supplement and advance the policy concerns underlying strict products liability generally, and we should not construe the statute now to establish a standard for determining defectiveness which is inconsistent with the present law in this area (see generally, UCC 1-103).

Accordingly, I dissent.

Chief Judge Kaye and Judges Bellacosa, Smith, Levine and Ciparick concur with Judge Titone; Judge Simons dissents in a separate opinion.

Following certification of questions by the United States Court of Appeals for the Second Circuit and acceptance of the *274questions by this Court pursuant to section 500.17 of the Rules of the Court of Appeals (22 NYCRR 500.17), and. after hearing argument by counsel for the parties and consideration of the briefs and the record submitted, certified question No. 1 answered in the negative, certified question No. 2 answered in the negative, and certified question No. 3 answered in the affirmative.

CHIEF JUSTICE RICE

delivered the Opinion of the Court.

¶1 In this products liability case, we consider whether the trial court erred when it gave a jury instruction that allowed the jury to apply either the consumer expectation test or the risk-benefit test to determine whether a driver’s- car seat was unreasonably dangerous due to -a design defect. The court of appeals concluded that'the trial court did err by instructing the jury separately on the consumer expectation test,’because the test already comprises an element of the risk-benefit test. Walker v. Ford Motor Co., 2015 COA 124, ¶¶ 26-28, — P.3d —, -.

¶2 We now affirm the court of appeals, albeit on different grounds. This court determined more than thirty years ago that the risk-benefit test is the appropriate test to assess whether a product is Unreasonably dangerous due to a design defect where the dangerousness of the design is “defined primarily by technical, scientific information.” Ortho Pharm. Corp. v. Heath, 722 P.2d 410, 414 (Colo. 1986), overruled on other grounds by Armentrout v. FMC Corp., 842 P.2d 176, 183 (Colo. 1992). We have found the consumer expectation- test, by contrast, “not suitable” in such a case. Id. at 416. Here, the jury was tasked -with determining whether a car seat was unreasonably dangerous due to a design - defect — a determination that, as evidenced by the extensive expert testimony at trial, required consideration of technical, scientific, information. Thus, the proper test under which to assess the design’s dangerousness was the risk-benefit test, 1 not -the consumer expectation test. We therefore hold that the trial court erred by instructing the jury on both tests, thereby allowing it to base its verdict on the consumer expectation test alone. We hold further that the jury’s separate finding of negligence-did not render the instructional error harmless. Accordingly, we affirm the court of appeals on different grounds and remand the ease for further proceedings consistent with this opinion.

I.

¶3 Forrest Walker was rear-ended while driving his 1998 Ford Explorer in Boulder, Colorado. Upon impact, Walker’s car accelerated forward and his car seat yielded rearward. Walker asserts that he sustained head and neck injuries in the crash, and he sued the other driver and Ford Motor Company (“Ford”) to recover for those injuries. Walker settled with the other driver, but he proceeded to trial against Ford on theories of strict liability and negligence. He claimed that the seat was defective in its design, and that Ford was negligent for failing to take reasonable care in the design and manufacture of its product so as to prevent an unreasonable risk of harm.

¶4 Walker’s case was tried to a jury in 2013. During trial, Ford and Walker offered extensive testimony from biomechanical and seat-design experts on the design characteristics of the car seat. Ford’s experts explained the concept behind yielding seats, saying they absorb energy that would otherwise impact the driver in a crash, and testified to the benefits of such seats in collisions like Walker’s. Ford also presented data from testing that it claims proved the benefits of Ford’s seat design, showed that the seat performed better in rear-end collision testing than its 1998 competitors, and demonstrated that the crash forces Walker experienced did not exceed injury thresholds. Walker’s experts testified that, although car seats should have some yield, the seat in the 1998 Ford Explorer needed to be stronger, and that it was technologically and economically feasible in 1998 to build a stronger seat with a better head restraint. Walker’s experts also testified that the seat was not state-of-the-art in 1998 and gave examples of feasible design alternatives. During closing arguments, Walker’s attorney appealed to the jury to use “common sense” and suggested it could conclude the seat was unreasonably dangerous by “look[ing] at what happened” -to the seat, “without having to decide who’s right among the experts on the liability issues.”

¶5 At the end of trial, as relevant here, Ford asked that the jury be instructed to assess the dangerousness of the car seat using the risk-benefit test. Walker requested the consumer expectation test. The trial court gave the jury an instruction, based on the Colorado pattern jury instraction at the time, allowing it to apply either test. Specifically, the instruction stated:

A product is unreasonably dangerous because of a defect in its design if it creates a risk of harm to persons or property that would not ordinarily be expected or is not outweighed by the benefits to be achieved from such design.

A product is defective in its design, even if it is manufactured and performs exactly as intended, if any aspect of its design makes the product unreasonably dangerous.

Jury Instr. No. 18; see also CJI-Civ. 4th 14:3 (2016). 2 The court also gave the jury a separate instruction listing seven non-exclusive factors it could consider in “weighing the risks versus the benefits of a product design.” Jury Instr. No. 19. 3 Of these seven factors, “factor six” stated that the jury could consider “the user’s anticipated awareness of dangers inherent in the product and their avoidability because of general public knowledge of the obvious condition of the product, or of the existence of suitable warnings or instructions.” Id. The jury ultimately found for Walker on both his strict-liability and negligence claims-, and it awarded him nearly $3 million plus interest. After Ford’s motion for a new trial or judgment notwithstanding the verdict, was denied when the trial court did not rule on it during the allotted time, Ford appealed the verdict.

¶6 The court of appeals reversed the jury’s verdict. Walker, ¶ 3. The court held’that the trial court erred by instructing the jury separately on the consumer expectation test, because the test is included as an element of the risk-benefit test. Id. at ¶¶ 14, 26. Specifically, the court summarily concluded that factor six of the risk-benefit test “is merely a rephrasing of the consumer expectation test,” Id..at ¶ 19. The court went on to hold that the instructional error was not harmless, because it allowed the jury to consider the consumer expectation test twice and to find for Walker even if it failed to consider the other elements of the risk-benefit test. Id. at ¶¶ 30 -33. The court thus reversed the verdict and remanded the case for a new trial, with directions for the trial court to omit the separate consumer expectation test from the jury instructions. Id. at ¶ 34.

¶7 Walker petitioned this court for review and we granted certiorari. 4 We now affirm the court of appeals, albeit on different grounds, and remand the case for further proceedings consistent with this opinion.

II.

¶8 First, we consider whether the trial court erred by instructing the jury on the consumer expectation test, and we conclude that it did. We then address whether the jury’s separate finding of negligence rendered the instructional error harmless, and we conclude that it did not.

A.

¶9 We review de novo whether a jury instruction states the law correctly, and we review the trial court’s decision to give a particular jury instruction for an- abuse of discretion. Day v. Johnson, 255 P.3d 1064, 1067 (Colo. 2011).

¶10 This court has looked to the doctrine of strict products liability as set forth in section 402A of the Restatement (Second) of Torts. Camacho v. Honda Motor Co., 741 P.2d 1240, 1244 (Colo. 1987). Under section 402A, a manufacturer may be held strictly liable for harm caused by “any product in a defective condition unreasonably dangerous to the user or consumer.” Restatement (Second) of Torts. § 402A (Am. Law. Inst. 1965); see also Camacho, 741 P.2d at 1244. A product may be in such a condition due to a manufacturing defect, which causes the product to fail to conform to the manufacturer’s specifications, or due to a failure to warn or a design defect that renders the product unreasonably dangerous despite the fact that it was manufactured exactly as intended. Camacho, 741 P.2d at 1247. Additionally, a motor-vehicle manufacturer may be held liable for injuries sustained in a motor-vehicle accident “where a manufacturing or design defect, though not the cause of the accident, caused, or enhanced the injuries.” Id at 1242-13. The plaintiff must prove that a product is defective and unreasonably dangerous in order to establish liability under section 402A. See Fibreboard Corp. v. Fenton, 845 P.2d 1168, 1175 (Colo. 1993) (citing Camacho, 741 P.2d at 1245).

¶11 In making the determination as to whether a product’s design is unreasonably dangerous, we have recognized two tests: the consumer expectation test and the risk-benefit analysis. Ortho, 722 P.2d at 413. The consumer expectation test asks whether a product performed as safely as an ordinary consumer would expect, and it derives largely from comment i to section 402A. Id Comment i states that a product must’ be “dangerous to an extent beyond that which would be contemplated by the ordinary consumer” and provides examples, including, whiskey containing fuel oil, and butter contaminated with poisonous fish oil. § 402A cmt. i.

¶12 The risk-benefit test asks a different question: whether the benefits of a particular design outweigh the risks of harm it presents to consumers. See Ortho, 722 P.2d at 413; Armentrout, 842 P.2d at 182-84 (stating the test and overruling Ortho' to the extent it placed the burden of proof on the manufacturer). This court first applied the risk-benefit test to assess whether a product’s design was defective in Ortho, 722 P.2d at 413-14, in which we also listed seven factors that may be considered in weighing the risks and benefits of a design, id. at 414 (citing John W. Wade, On the Nature of Strict Tort Liability for Products, 44 Miss. L.J. 825, 837-38 (1973)). 5 We have since clarified that 'those seven factors, while illustrative of considerations that may be helpful in determining whether a design is defective; are riot exclusive and need not be strictly applied in every case. Armentrout, 842 P.2d at 184. For example, in Armentrout, we stated that the existence of a feasible design alternative may be a'factor in the analysis of the dangerousness of a product design, 6 Id. at 185 & n.11. Noting that the Armentrouts had presented evidence of a feasible design alternative, we declared' in that case that such evidence would be considered on retrial. See id. at 185.

¶13 In this case, Walker arid Ford agree that Instruction 18 allowed the jury to use either the consumer expectation test or the risk-benefit test to determine whether the car seat was unreasonably dangerous. The consumer expectation test was encompassed in the phrase “creates a risk of harm to persons or property that would not ordinarily be expected,” and the risk-benefit test in the phrase “a risk of harm ,.. not outweighed by the benefits to be achieved from such design.” Jury Instr. No. 18. The trial court also instructed the jury that it could consider and weigh, "among other things,” the seven factors enumerated in Ortho. Jury Instr. No. 19; Ortho, 722 P.2d at 414. Walker claims that Instruction 18 stated Colorado law correctly, and the jury’s verdict sho.uld stand. We disagree.

¶14 This court has stated repeatedly that the risk-benefit test, not the consumer expectation test, is the proper test to use in assessing whether a product like the car seat at issue here is unreasonably dangerous due to a design defect, In Ortho, this court considered both tests in the context of a claim that a .prescription drug was unreasonably dangerous and defective, in design. .722 P.2d at 413-14. We concluded that the risk-benefit test was the. “appropriate standard” under which to. “measure the reasonableness of [the] danger” presented by the drug. Id. In reaching our conclusion, we noted that the dangerousness of the drug was “defined primarily by technical, scientific information,” Id. at 414. “The consumer expectation test,” we stated, “fails to address adequately [that] aspect of the problem,” id, and is “a test not suitable in prescription drug cases when the actionable product is alleged to be unsafe by design notwithstanding its production in precisely the manner intended,” id at 415. We found the risk-benefit test, by contrast, to properly “focus[ ] on the practical policy issues characteristic” of such a product. Id. at 414; Accordingly, we held that the trial court had erred by failing to instruct on the risk-benefit test. Id. at 416.

¶15 We reiterated our conclusions regarding the two tests in Camacho. There, the underlying issue was whether a motorcycle was defectively designed and unreasonably dangerous because it was not equipped with crash bars to protect a rider’s legs.. 741 P.2d at 1241-42.. We held that the lower .courts had erred, by using a consumer expectations standard to assess the motorcycle’s dangerousness, id. at 1242, reasoning that “the consumer contemplation concept embodied in comment i [to section 4Q2A] ... does not provide a satisfactory test for determining whether particular products are in a defective condition unreasonably dangerous to the user or consumer,” id.’at 1246.'We also pointed to this court’s recognition in Ortho that “exclusive reliance upon consumer expectations” is “particularly inappropriate” where both the dangerousness of the design and “the efficacy of alternative designs ... must be defined primarily by technical, scientific information.” Id. at 1246 — 47.-

¶16 Notably, we éxpoúnded upon the reasons for using the risk-benefit test, rather than the consumer expectation test, in design-defect cases involving technical,’ complex product designs.-We noted that products-liability law has developed in part to “encourage manufacturers to use .information gleaned from testing, inspection and-data analysis” to help avoid product accidents. Id. at 1247. Using the risk-benefit test to assess the dangerousness .of products-helps further' this objective, as it. directs fact-finders to consider the manufacturer’s ability- to minimize or eliminate risks and the effect such an alteration-would have on the product’s utility, other safety aspects, or affordability. See,' e.g., id. at 1247-48 (listing risk-benefit factors). Moreover, “manufacturers of such complex products as motor vehicles invariably have greater access than do ordinary- consumers to the- information necessary to' reach informed decisions concerning the efficacy of potential safety measures.” Id. at 1247.. Finding that, the motorcycle’s status required “interpretation of mechanical engineering data derived from research and testing — interpretation which necessarily:includes the application of scientific and technical principles,” id. at 1248, we remanded the- case for further proceedings using the proper risk-benefit standard, id. at 1242,1249.

¶17 Turning to the ease before us, we -conclude that the trial court erred by instructing the jury on the consumer expectation test. This case,, like those outlined above, concerns an alleged design defect, and the dangerousness of the car seat design is “defined primarily by technical, scientific information.” Ortho, 722 P.2d at 414. Indeed, like the motorcycle without crash bars in Camacho, the dangerousness of the car seat design required- the “interpretation ;of mechanical engineering data derived from research and testing.” Camacho, .741 P,2d at 1248. The parties provided the jury with extensive.expert testimony regarding testing data, seat rigidity, restraint systems, risks and benefits, and feasible design alternatives. Applying such evidence “necessarily includes the application of scientific and technical principles:” Id. Therefore, the proper test for the'jury to use in its determination of the dangerousness of the seat design was the risk-benefit test.

¶18. We are .not persuaded by Walker’s argument that.there is no inconsistency in applying and, instructing a jury on both tests, as was done here. As developed above, our precedent holds the consumer expectation test to be improper in design-defect .cases involving scientific, technical information. Here, Instruction 18 allowed the jury to apply either the consumer expectation test or the risk-benefit, test. Jury Instr. No. 18. (stating that.the jury could find the seat unreasonably. dangerous based on a risk that “would not ordinarily be expected or. is not outweighed, £>y the benefits” of the design (emphasis added)). Therefore, the instruction authorized; the jury to apply .the consumer expectation test alone. Counsel for Walker stressed this fact to the jury during closing arguments, urging the jury to use “common sense” and suggesting it could conclude that the seat was unreasonably dangerous by “looking] at what happened” to the seat, “without having to decide 'who’s right among the' experts on the liability issues.” Instruction 18 was contrary to our precedent, and the trial court erred by giving it. 7

¶19. Because we find that Instruction 18 was inconsistent with our precedent, we need not reach the issue upon which the court of appeals rested its opinion, namely, whether the consumer expectation test is subsumed into factor six of the risk-benefit test. Instruction 19, which listed the seven factors from our Ortho decision, has not been challenged in this case, and may be used on remand.

B.

¶20 Walker argues that, even if Instruction 18 was erroneous, the jury’s separate finding of negligence renders the error harmless.

¶21 This court will deem an error harmless, and thus will not reverse a judgment, unless the error resulted in substantial prejudice to a party. See Armentrout, 842 P.2d at 186; accord C.A.R. 35(c) (“The appellate court may disregard any error or defect not affecting the substantial rights of the parties.”). With respect to jury instructions, “the giving of an erroneous instruction constitutes reversible error once prejudice is shown.” Mile Hi Concrete, Inc. v. Matz, 842 P.2d 198, 204 (Colo. 1992).

¶22 Walker asserts that his negligence claim is an alternative basis for the jury’s verdict. Emphasizing that the jury found for him in both strict liability and negligence, he argues that the negligence finding renders harmless any error in the trial court’s strict liability instructions. Walker argues further that the jury was properly instructed on the elements of negligence and on the separate nature of his claims, and we must assume the jury followed instructions. He thus contends that the court of appeals erred in finding reversible error, and the jury’s verdict should stand. We disagree.

¶23 A manufacturer is not negligent for designing a reasonably safe product. Accordingly, regardless of whether a design-defect claim is based in strict liability or negligence, in order to properly return a verdict for the plaintiff, a fact-finder must determine that the product at issue is unreasonably dangerous. See, e.g., Camacho, 741 P.2d at 1245 (“[W]hen a product is not reasonably safe a products liability action may be maintained”); Mile Hi, 842 P.2d at 206 (“Regardless of whether a product liability action is grounded in negligence or strict liability, a plaintiff must prove that the product was defective.”). For all of the reasons laid out in this opinion, such a determination in this case should have been made using the risk-benefit test.

¶24 Moreover, in a design-defect case such as this, the risk-benefit test essentially subsumes the issue of negligence. See, e.g., Keller v. Koca, 111 P.3d 445, 447-48 (Colo. 2005) (explaining that a determination of negligence requires consideration of multiple factors, including “the risk involved, the foreseeability of the injury weighed against the social utility of the actor’s conduct,” and the burden of guarding against harm). Reasonableness is a negligence concept. See Camacho, 741 P.2d at 1245. Thus, as this court has recognized, the risk-benefit test “includes language which is rooted in negligence.” Fibreboard Corp., 845 P.2d at 1173.

¶25 Here, because of the instructional error, the jury was permitted to rely upon an improper standard in making its negligence determination. The trial court instructed the jury that, in order to find for Walker on his negligence claim, it must find that Ford “was negligent by failing to exercise reasonable care to prevent the driver seat and restraint system in the 1998 Ford Explorer from creating an unreasonable risk of harm.” Jury Instr. No. 23. “[U]nreasonable risk of harm,” however, was not separately defined. And when the jury asked the court what it meant by “reasonably dangerous,” the court referred the jury to Instruction 18 — the instruction containing both the risk-benefit and consumer expectation tests. Trial Tr. 2:14-25, March 21, 2013. Consequently, the jury was permitted to use a consumer expectation standard to assess the reasonableness of the risk presented by the seat design and find Ford negligent, even if — indeed, especially if — it followed the trial court’s instructions. Accordingly, we cannot conclude that the tidal court’s error in giving Instruction 18 was harmless. 8

III.

¶26 For the reasons stated above, we affirm the court of appeals on different grounds, and remand the case for further proceedings consistent with this opinion.

OPINION

By the Court,

Maupin, L:

Leo Gasse was killed and Robin Lewis catastrophically injured due to carbon monoxide poisoning during an overnight outing in a Sea Ray pleasure boat at the Lake Mead National Recreation Area. Lewis, along with Gasse’s heirs, Teresa Rae Webb and Tricia Marie Gasse, brought suit against Sea Ray Boats, Inc., alleging that Sea Ray is strictly liable in tort in connection with the incident. A jury returned a verdict in favor of Sea Ray, finding that the boat was not a defective or unreasonably dangerous product. This appeal followed.

Appellants’ primary contention centers on the district court’s failure to adopt appellants’ proffered instructions on their theory of liability; that warnings concerning the risk of carbon monoxide migration secondary to use of the boat’s air conditioning system were inadequate. Because we conclude that appellants were entitled to more specific instructions with regard to the warnings issue, we reverse the district court’s judgment and remand this matter for a new trial.

FACTS

In May 1991, Leo Gasse and Jimmy Paxson purchased a used Sea Ray pleasure boat from a Las Vegas area Sea Ray dealership. In addition to gasoline propulsion engines, the boat contained a small gasoline generator, which powered the boat’s accessories, including the air conditioner.

On May 29, 1993, during a weekend cruise on Lake Mead, Gasse and Lewis “side-tied” the boat to a beach and went to sleep in the boat’s cabin, leaving the gasoline generator running to power the air conditioner. The next morning, Anthony Caro, Jr., a friend who was staying at the beach, knocked on the cabin door and received no response. He returned later that afternoon, boarded the boat, and found Gasse dead and Lewis barely breathing. Mr. Caro testified that the engines were not running when he first checked on the couple and when he returned.

Subsequent investigation confirmed that the generator, rather than the engines, was the source of the carbon monoxide, a tasteless odorless gas. This proposition was bolstered by other trial tes *103 timony that, had engine exhaust been the source, the couple may have been able to detect the problem because of the distinctive odor of exhaust fumes.

Two warnings regarding carbon monoxide poisoning accompanied the sale of this type of boat in 1981, one written by ONAN, the generator manufacturer, 1 and the other by the National Marine Manufacturers’ Association (NMMA). 2 Sea Ray provided boat purchasers with an assortment of other manuals, none of which are relevant to this case. Both warnings primarily addressed the danger of carbon monoxide exposure from engine exhaust.

When Gasse and Paxson purchased the boat, the Sea Ray dealership service manager, George Schenk, and the salesman, Curt Snouffer, warned of the danger of exhaust fumes and carbon monoxide, and the necessity of ventilating the boat to remove hazardous fumes. Schenk and Snouffer demonstrated this process by opening a window and the hatch to allow for flow-through ventilation, and explained the need to have the rear door remain open when running the main propulsion engines. Lastly, Schenk indicated that idling the engine with the front hatch closed could cause accumulations of carbon monoxide.

Appellants theorized that a process described as ‘ ‘migrating carbon monoxide’ ’ caused the accident. The process occurs when carbon monoxide, although safely exhausted from the boat’s gasoline generator into the open air, is blown back into the boat by wind, entering the passenger cabin through small openings. Sea Ray’s ex *104 pert agreed with this theory of causation, but noted that such a phenomenon is quite rare and for carbon monoxide to accumulate to dangerous levels, passenger cabin ventilation must have been obstructed.

Sea Ray’s expert testified regarding the safety of sleeping with the air conditioner running. He admitted that although boaters will often sleep with the air conditioner running unless warned not to do so, certain precautions should be taken. These include: (1) posting a watch, since in 1981, the year the boat was manufactured, no carbon monoxide detection devices were available; (2) anchoring the boat from the bow rather than the side, so that any wind currents would blow away from the stern; or (3) creating flow-through ventilation before going to sleep. The expert conceded that Sea Ray’s manual contained no such instructions or warnings, but stressed that no incidents of this type resulting in death had ever been reported in connection with the particular pleasure boat model involved in this case. Sea Ray’s expert also voiced his opinion that the warnings given were adequate with regard to carbon monoxide exposure, and that the risk of ‘ ‘migrating” carbon monoxide from on-board generators was not a known hazard when the boat was originally purchased in 1981.

Sea Ray’s expert additionally relied upon a Nevada Department of Wildlife booklet found on the boat after the incident. The booklet discussed the hazards of exhaust fumes, warned that carbon monoxide itself is tasteless and odorless, that plenty of air flow should be maintained because exhaust fumes can blow back into a boat when running downwind, and that adequate ventilation was required when using catalytic heaters for warmth.

The warnings that are the subject of this appeal specifically addressed the danger of carbon monoxide exposure from exhaust fumes, generally addressed dangers attendant to carbon monoxide exposure, and only inferentially addressed dangers in connection with generator fumes. All of this is important because, as noted, the discrete odor from engine exhaust would arguably alert the passengers to the presence of noxious fumes, while emissions from the generator probably would not.

Jury instructions on “adequate warning”

Appellants submitted a proposed jury instruction regarding legal requirements for an “adequate warning” based on Pavlides v. Galveston Yacht Basin, Inc., 3 a Fifth Circuit case applying a three-factor test under Texas law 4 for determining whether a product warning was adequate. The proposed instruction read as follows:

*105 A warning must (1) be designed so it can reasonably be expected to catch the attention of the consumer; (2) be comprehensible and give a fair indication of the specific risks involved with the product; and (3) be of an intensity justified by the magnitude of the risk.

The district court rejected this proposed instruction and instead gave the following two instructions:

First:

Although you are to consider only the evidence in the case in reaching a verdict, you must bring to the consideration of the evidence your everyday common sense and judgment as reasonable men and women. Thus, you are not limited solely to what you see and hear as the witnesses testify. You may draw reasonable inferences from the evidence which you feel are justified in the light of common experience, keeping in mind that such inferences should not be based on speculation or guess.

Second:

The question of whether or not a given warning is legally sufficient depends upon the language used and the impression that such language is calculated to make upon the mind of the average user of the product.

The first instruction is a stock instruction that the jury should simply use its common sense in evaluating and drawing inferences from the evidence introduced at trial. The second instruction is generally worded, containing partial excerpts from Pavlides. 5

During deliberations, the jury sent a note to the trial judge, requesting a definition of an “adequate warning.” Appellants proposed an instruction taken from a products liability treatise to the district court. 6 The district court rejected this instruction, as well as again rejecting appellants’ proposed Pavlides instruction. Consequently, the district court simply reread the two instructions it had previously given on the issue to the jury.

After the trial judge reread the instructions, the jury foreman informed the judge that the reading did not assist the jury in its delib *106 erations. The district court again sought a definition of “adequate warning” from the parties. Appellants reoffered the treatise definition, arguing that it was essentially consistent with Nevada case authority. 7 The district court again rejected the treatise definition, and refused to instruct the jury further, despite the confusion. Soon after the rereading of the jury instructions, one juror was replaced during deliberations for unspecified reasons. Shortly thereafter, the jury returned a verdict in favor of Sea Ray. This appeal followed.

DISCUSSION

Failure to give appellants’ proposed “adequacy of warnings’’ instruction

Respondent contends that warnings instructions in cases such as this one should be generally worded and that the adequacy of warnings should be left to the common sense of the finder of facts. Appellants contend that the district court erred by not instructing the jury with their more specific definition of “adequate warning.” We agree with appellants.

In American Casualty Co. v. Propane Sales & Service, we held that a party is entitled to have the jury instructed on all of his theories of the case that are supported by the evidence, 8 and that general, abstract or stock instructions on the law are insufficient if a proper request for a specific instruction on an important point has been duly proffered to the court. 9 We reversed in American Casualty Co. because the jury was left to guess “from general ‘stock’ instructions” discrete elements of proof “in the rather unusual context of a gas explosion case.” 10 However, in American Casualty Co., we also observed that “[i]n some instances a requested instruction, although proper, will not be essential to the jury’s understanding of the case.” 11

*107 Under Nevada law, 12 “strict liability may be imposed even though the product is faultlessly made if it was unreasonably dangerous to place the product in the hands of the user without suitable and adequate warning concerning safe and proper use.” 13 Inherent in this doctrine is that “a product must include a warning that adequately communicates the dangers that may result from its use or foreseeable misuse.” 14 More particularly, in Fyssakis v. Knight Equipment Corp., we held that adequacy of warnings was an issue of fact for the jury where an industrial strength soap manufacturer’s warnings did not alert the user that the soap could cause blindness. 15 In Allison v. Merck and Company, 16 a district court entered summary judgment in favor of a manufacturer of a childrens vaccine. We reversed in light of our conclusion that the drug manufacturer was required to adequately warn parents of possible side effects of immunization, including blindness, deafness or mental retardation. Accordingly, we held that a general warning that an inoculated child could encounter rashes and possible brain inflammation was arguably inadequate and issues of fact remained as to the sufficiency of the warnings given. 17 In remanding the Allison case for trial on the adequacy of the warnings, we rejected the notion that a drug manufacturer could, via a general warning, avoid liability as a matter of law, even where the product was either reasonably or unavoidably unsafe. 18

In the instant matter, the purchasers of the boat were comprehensively warned about the dangers of carbon monoxide poisoning from exhaust fumes, fumes characterized by a distinctive odor. Here, however, the injuries sustained by Gasse and Lewis were not caused by exhaust fumes; they were caused by odorless and tasteless carbon monoxide fumes from the generator that *108 powered the boat’s air conditioner. Whether the warnings described above, which generally addressed dangers and symptoms of carbon monoxide poisoning and specifically addressed carbon monoxide exposure secondary to engine exhaust and running the heater, sufficiently apprised Gasse and his co-owner of carbon monoxide poisoning from use of the air conditioner remained the primary issue of fact throughout the trial below. Thus, the text of the “warnings” instruction became critical to the jury’s fact-finding mission.

Here, the district court’s “warnings” instructions provided very little in the way of guidance, other than to generally state that whether a warning is legally sufficient depends upon the “impression” that the warnings language “is calculated to make upon the mind of the average user of the product,’ ’ and that the jury should use its common sense in resolving the issue. This instruction was not sufficient to assist the jury in resolving the liability issues based upon Sea Ray’s alleged failure to warn. First, in Fyssakis and Allison, we refused to exonerate products manufacturers as a matter of law from strict tort liability based upon general warnings language. Second, these instructions left lay jurors, persons in much the same position as the users of the product at issue, to search their imaginations to test the adequacy of the warnings. Third, given that experts testified in this case to the nature and quality of the warnings that were given and their supposed behavioral impact, the jurors were entitled to more specific guidance as to the law governing the duty to warn in connection with consumer products.

We therefore embrace the rule of law stated in the Pavlides instructions offered by appellants below, and hold that Nevada trial courts should advise juries that warnings in the context of products liability claims must be (1) designed to reasonably catch the consumer’s attention, (2) that the language be comprehensible and give a fair indication of the specific risks attendant to use of the product, and (3) that warnings be of sufficient intensity justified by the magnitude of the risk.

The district court’s failure to instruct the jury as suggested by appellants mandates reversal for a new trial.

Applicable law to be applied on remand

Appellants argue that the district court improperly applied admiralty law instead of Nevada law to the proceedings below. We agree.

*109 In Jerome B. Grubart, Inc. v. Great Lakes Dredge & Dock Co., 19 the United States Supreme Court established a two-part “location” and “connection” test for determining when the exercise of federal maritime jurisdiction is appropriate. 20 Under the “location” leg of this test, a court must determine whether “the tort occurred on navigable water or . . . [the] injury suffered on land was caused by a vessel on navigable water.” 21 Although the location prong of the test is satisfied in this instance, the “connection” prong is not. One feature of the “connection” test requires an analysis of the general features of the incident causing the injury to determine whether the incident has “ ‘a potentially disruptive impact on maritime commerce.’ ” 22

We conclude that the incident in question here had no potential for disruption of maritime commerce on the Lake Mead Reservoir. Gasse and Lewis were occupants of a single pleasure boat moored in an isolated location at night. Thus, there is no basis to apply admiralty law to this controversy in lieu of Nevada law. 23

CONCLUSION

The district court’s warnings instructions merely admonished the jury to use its common sense in resolving the sufficiency of the warnings, guided only by a general and partial definition of “adequate warning” under Pavlides. Although Fyssakis and Allison do *110 not delineate how juries are to be instructed on this issue, when read together with American Casualty Co., they impliedly require a more specific instmction on the adequacy of warnings than given here. Under our adoption of the Pavlides instmction, appellants are entitled to have their instmction on the definition of “adequate warning” submitted to the jury. That the jury ultimately became engaged in a dialogue with the district court on this very issue, and that the jury foreman indicated repetition of prior instmctions was not helpful to the jury’s deliberations, underscores the insufficiency of the instmc-tions that were given.

We therefore reverse the judgment of the district court and remand this matter for new trial proceedings 24 conducted in accordance with this opinion. 25

Per Curiam.

Plaintiff appeals as of right an order granting defendants’ motion for summary disposition pursuant to MCR 2.116(C)(10) in this products liability action arising out of an industrial fire caused by the use of a glue solvent.

Plaintiff, as subrogee, brought this action against the manufacturer of the glue solvent, *542Ralph Wilson Plastics Company (rwpc), and the seller, Plywood-Detroit, Inc. (pdi), alleging negligence and breach of implied warranty based on defendants’ failure to warn. Plaintiff is the property and casualty insurer for its subrogor, National Seating Company (nsc).

Nsc is a manufacturer of tables and chairs. Foam rubber, glue, and glue solvents are commonly used in nsc’s manufacturing process. Beginning in 1980, nsc began purchasing fifty-five-gallon drums of Lokweld 110, a glue solvent, from pdi at intervals of eight to ten weeks. Shortly after nsc began these purchases, pdi provided nsc with the relevant material safety data sheets (msds) required by the Occupational Safety and Health Act (miosha), MCL 408.1001 et seq.; MSA 17.50(1) et seq., which nsc, as an employer, was obligated to disseminate to its employees pursuant to MCL 408.1011(c); MSA 17.50(11)(c).

According to the msds, Lokweld is a highly flammable, even explosive solvent, with a flash point of ten to fifteen degrees Fahrenheit. Lokweld is to be stored and used only in well-ventilated areas. All electrical equipment is to be explosion proofed, and spills are to be immediately sponged with absorbent rags that must be placed in a closed metal container.

The fifty-five-gallon drums in which Lokweld was sold by pdi carried red and yellow warning labels on the top and sides, proclaiming Lokweld to be a flammable liquid, with a second label announcing "danger” in inch-high letters, beneath which were these warnings:

EXTREMELY FLAMMABLE

VAPORS MAY CAUSE FLASH FIRES!

HARMFUL OR FATAL IF SWALLOWED!

TAKE THE FOLLOWING PRECAUTIONS BEFORE REMOVING lid:

*543DO NOT SMOKE—BE SURE WORK AREA IS WELL VENTILATED—OPEN WINDOWS—EXTINGUISH ALL FLAMES, PILOT LIGHTS—TURN OFF STOVES, HEATERS, ELECTRIC MOTORS—READ WARNINGS AND FIRST AID INSTRUCTIONS ON SIDE LABELS—POST NUMBER OF DOCTOR OR POISON CONTROL CENTER AT NEAREST TELEPHONE.

Another label contained the following warnings:

Prevent buildup of vapors—open all windows and doors—use only with cross ventilation. Keep away from heat, sparks, and open flame. Do not smoke, extinguish all flames and pilot lights, and turn off stoves, heaters, electric motors, and other sources of ignition during use and until all vapors are gone. Avoid prolonged contact with skin or repeated breathing of vapors.

Close container tightly after use. Store in a cool, well-ventilated area.

Nsc’s management was well aware that the foam rubber used in the chair manufacturing process is routinely subject to a build up of static electricity. On May 5, 1987, David Swift, a new, part-time employee of nsc, used a rag soaked in Lokweld to clean excess glue from his work table. Some of the Lokweld spilled on the floor. Swift used a piece of foam rubber to soak up the solvent. When Swift picked up the foam rubber, a flash fire erupted. Nsc sustained extensive property damage, for which plaintiff reimbursed nsc. Fortunately, no one was injured. Plaintiff contends, and defendants do not dispute, that the solvent was ignited by static electricity.

Plaintiff brought suit in February 1990 contending that the warnings about the flammability of Lokweld were inadequate. During the course of discovery, plaintiff’s expert testified in a deposition that the warning labels on the drums were inadequate because they failed to warn of the specific *544danger from static electricity. The expert contended that hazard studies and literature in the chemical industry regarding volatile products had recognized fire hazards from static electricity for many years, but that the danger of fire from static electricity is a little-known phenomenon among the lay public. According to the expert, most people would not include static electricity in the term "spark.” However, neither the American National Standards Institute nor the Federal Hazardous Substances Act, 15 USC 1261 et seq., distinguish among sparks by virtue of the generating cause. Plaintiffs expert conceded that nsc, having purchased the product for a number of years, was probably experienced in using Lokweld, but noted that long-term use can result in carelessness.

David Swift testified that neither he nor his coworkers were ever given any information about using Lokweld by nsc managers. He could not remember whether warning labels were on the drum but stated that if there were labels, he did not read them. He had never heard of a material safety data sheet, and had never seen any written information about Lokweld. Although he assumed Lokweld to be flammable, and knew that foam rubber is always charged with static electricity, he claimed to have no idea that a static spark could start a fire.

In moving for summary disposition under MCR 2.116(C)(10), a pdi representative submitted a supporting affidavit, asserting that nsc continued to purchase Lokweld through January 1991. Nsc’s president and owner stated, in an opposing affidavit, that he had no knowledge of the particular risk from static electricity sparks, and claimed he had never been warned that static electricity could ignite Lokweld. He asserted that nsc had taken precautions against other sources of ignition, and *545that if nsc had known of the danger from static electricity, it would not have used Lokweld.

In seeking summary disposition, defendants contended that they owed no duty to warn, because the product was placed in the hands of a sophisticated user; plaintiff contended that its subrogor was not a sophisticated user of the product. Defendants further asserted that, if they had a duty, their warnings were adequate, but even if their warnings were inadequate, plaintiff could not show proximate cause. Defendants noted that nsc had been cited by the Michigan Occupational Safety and Health Administration on February 2, 1987, for failing to provide its employees with information regarding hazardous substances and failing to maintain and make available msds records for its employees.

In a written opinion, the circuit court concluded that defendants had a duty to warn, and that the warnings given adequately encompassed the risk. Absent inadequate warnings, the circuit court reasoned that plaintiff could not possibly show proximate cause.

Plaintiff argues that the warnings given by defendants were inadequate as a matter of law because they did not warn of the risk of ignition by static electricity and that the failure to warn was a proximate cause of nsc’s damages. We disagree.

If a manufacturer had to list all sources of friction, or all sources of sparks, as a means of warning of a flammability hazard, its warning label would have to be of epic or encyclopedic proportions. Even then, the manufacturer could not be certain that it had covered every possibility. The combinations of circumstances or materials that could create a spark or friction would be almost limitless. This Court has previously recognized that excessive warnings on product labels *546may be counterproductive, causing "sensory overload” that literally drowns crucial information in a sea of mind-numbing detail. Dunn v Lederle Laboratories, 121 Mich App 73, 81; 328 NW2d 576 (1982).

We agree with defendants that a prominent warning, as here, that a product is flammable, and that either open flame or sparks can cause it to ignite, is adequate for the task of warning even unsophisticated users of the product that such sources of combustion must be kept from the vicinity of the product. To adopt plaintiffs argument would effectively make it impossible to ever adequately warn of flammability hazards in a manner sufficient to avoid liability where, as here, users of a product admit that no label would have sufficed, because the warning label that was provided with the product was ignored. A spark is a spark, and a possible source of ignition of a highly flammable compound, irrespective of the means by which the spark was generated, whether by static electricity or otherwise. Accordingly, as a matter of law, defendant rwpc’s warning, which did specify sparks as a potential hazard, was not inadequate. This conclusion makes it unnecessary to determine whether a prominent warning of flammability generally might obviate the need to detail the types of ignition sources, at least in terms of including such obvious causes as flame and sparks.

We are also of the opinion that the "sophisticated user” or "knowledgeable user” doctrine properly applies, in general, to plaintiffs subrogor as a bulk user of Lokweld. Commercial enterprises that use materials in bulk must be regarded as sophisticated users, as a matter of law. Higgins v E I DuPont de Nemours & Co, Inc, 671 F Supp 1055, 1058-1059 (D Md, 1987). Thus, the affidavit of the nsc president, wherein he stated that he failed to *547appreciate the real nature of the flammability hazard despite defendants’ warning labels and material safety data sheets, must be rejected because nsc, and its management, must be charged as a matter of law with sufficient expertise to have knowledge of how to use this product in a safe manner. Hall v Ashland Oil Co, 625 F Supp 1515, 1521 (D Conn, 1986).

This is particularly true where, as here, nsc, as the employer, had an obligation under miosha, as it existed before May 25, 1986, to "[m]ake available to employees in a conspicuous location in the area where a chemical substance or mixture which is used in a quantity capable of creating a hazard is being manufactured or mixed in the manufacturing process, a material safety data sheet or other informational sheet or system” that specifies, inter alia, "[t]he flash point, autoignition temperature, . . . and any unusual fire or explosion hazards of the chemical substance or mixture.” MCL 408.1011(1)(c)(v); MSA 17.50(11)(1)(c)(v).

Michigan jurisprudence has recognized the sophisticated user doctrine at least since Antcliff v State Employees Credit Union, 414 Mich 624; 327 NW2d 814 (1982). The case law is consistent on this point with 2 Restatement Torts, 2d, § 388, comment k, pp 306-307. See Mascarenas v Union Carbide Corp, 196 Mich App 240; 492 NW2d 512 (1992). Those with a legal obligation to be informed concerning the hazards of materials used in manufacturing processes must be relied upon, as sophisticated users, to fulfill their legal obligations, thereby absolving manufacturers in some circumstances of the duty to warn the users of chemical products, where such use is in the course of employment for a sophisticated bulk user. Any other rule would mean that " '[m]odern life would be intolerable unless one were permitted to rely to *548a certain extent on others’ doing what they normally do, particularly if it is their duty to do so.’ 2 Restatement Torts, 2d, § 388, comment n, p 308.” Tasca v GTE Products Corp, 175 Mich App 617, 624; 438 NW2d 625 (1988).

Plaintiffs further contention, that its subrogor would not have used Lokweld had it appreciated the flammability hazard from static electricity, is belied by defendants’ uncontradicted affidavit indicating that nsc continued to use Lokweld through January 1991. Regarding the claim that this creates a genuine issue of material fact, a prerequisite for avoiding summary disposition under MCR 2.116(C)(10), whether the discrepancy between nsc’s stated intent and its actions is "material,” we think it is clearly not "genuine.” Summary disposition cannot be avoided by conclusory assertions that are at odds either with prior sworn testimony of a party or, as here, actual historical conduct of a party. Gamet v Jenks, 38 Mich App 719, 726; 197 NW2d 160 (1972). Here, nsc’s actions after the disastrous episode of May 5, 1987, stand in eloquent contradiction to the affidavit of its president that, had it known of the flammability hazards from static electricity, it would never have used Lokweld. Thus, again, the requisite proximate cause between the asserted inadequacy of defendants’ warning and plaintiff’s damage is lacking.

Having addressed plaintiffs arguments from every possible perspective, we are forced to conclude that the circuit court correctly granted the motion for summary disposition. Accordingly, the decision of the circuit court is affirmed.

Connor, J.

(dissenting). I agree that, as a sophisticated user, National Seating Company (nsc), had the duty to comply with workplace safety laws and have its employees read and follow all instructions *549that defendants supplied about the safe use of Lockweld 110. Therefore, if the information supplied had been adequate to alert someone to the danger posed by sopping up a spill with foam rubber, or if defendants had instructed nsc about a safe method of sopping up spills that would have precluded the use of foam rubber, defendants would not be liable for the damage caused by the fire.

However, under "Precautions for Safe Handling and Use,” the material safety data sheet said only:

Spills should be cleaned up immediately. Soak up spill with absorbent material or wipe up with rags. Absorbent material or rags should be put into closed container.

Foam rubber is an "absorbent material.” Thus, the fire at nsc started while nsc’s employee was acting in accordance with defendants’ instructions for safe handling.

The only warning defendants supplied nsc to alert it to the danger that soaking up spills with foam rubber could cause a fire was the label warning to keep the solvent away from sparks. I would hold the adequacy of the warnings defendants supplied to be a question for the trier of fact to decide.1

I dissent.

SNELL, Justice.

Appellant, Prosoeo, Inc. (Prosoeo) appeals from a jury verdict and judgment entered against it in a products liability action brought by David Olson and members of his family (Olson). The district court entered judgment against Prosoeo in an amount exceeding $735,000, including $42,000 in consortium damages. Olson based his claim for damages on theories of strict liability and negligence. We affirm.

Numerous errors are assigned by Prosoeo as grounds for reversal. Our standard of review in this case is for errors of law. Iowa R.App.P. 4.

I. Facts

David Olson is a bricklayer foreman employed by Seedorf Masonry Company (See-dorf). Late on the afternoon of December 15, 1988, Olson spotted a fifteen gallon drum of mortar cleaner sitting on the ground. To prevent the cleaner drum from freezing to the ground, he picked it up and moved it onto a nearby pallet. When Olson dropped the drum on the pallet, the bung closure popped out of the drum, splashing hydrochloric acid based cleaner into his right eye. Despite extensive medical care, Olson eventually lost sight in his right eye. In April 1991 doctors fitted Olson with an artificial eye.

The mortar cleaner, called “Sure Klean 600,” is manufactured and packaged by Pro-soco. The fifteen gallon drum into which Prosoeo packages the cleaner is manufactured by Delta Drum Corporation (Delta Drum). The bung closures used in the fifteen gallon drums are manufactured by Rieke Corporation (Rieke). Olson initially named Rieke and Delta Drum in this lawsuit. Rieke and Delta Drum settled their cases with Olson. Olson sued Prosoeo under several theories of strict liability and negligence. 1 Prosoeo requested a state-of-the-art defense jury instruction with regard to Ol *287 son’s strict liability and negligence theories. The jury found Prosoco one-hundred percent at fault for Olson’s injuries under both theories.

II. Strict Liability and Negligence Claims

Prosoco claims the court erred by submitting the case on both strict liability and negligence theories. We have said that courts must give requested jury instructions when they correctly state the law applicable to the facts of the case and if the legal concept is not embodied in other instructions. Stover v. Lakeland Square Owners Ass’n, 434 N.W.2d 866, 868 (Iowa 1989). However, even instructions correctly stating the law should not give undue emphasis to any particular theory, defense, stipulation, burden of proof, or piece of evidence. Dickman v. Truck Transp., Inc., 224 N.W.2d 459, 464 (Iowa 1974). Error in giving or refusing to give instructions is reversible, only if prejudicial. Stover, 434 N.W.2d at 868.

Prosoco contends the submission of instructions on both strict liability and negligence theories was duplicative and confusing, resulting in prejudicial error. Prosoco asks this court to build on our holding in Hillrichs v. Avco Corp. in which we held that a plaintiffs strict liability claim depended on the same elements of proof as his negligence claim. See Hillrichs v. Avco Corp., 478 N.W.2d 70, 75-76 (Iowa 1991). On remand, we held that only the negligence claim should be retried. Id.

Olson contends the submission of a strict liability instruction does not preclude liability based on a negligence theory. He stresses that our decision in Hillrichs was limited strictly to the facts of that case and claims that in the case at bar the strict *288 liability and negligence instructions submitted do not depend on the same elements of proof.

In Hawkeye-Security Insurance Co. v. Ford Motor Co., 174 N.W.2d 672, 684 (Iowa 1970), we adopted section 402A of the Restatement (Second) of Torts, recognizing a strict liability cause of action in products liability cases. We distinguished strict liability/design defect claims from negligence claims in Alter v. Rodgers Machinery Manufacturing Co., 268 N.W.2d 830, 835 (Iowa 1978). In Alter the alleged design defect involved a saw which the plaintiff claimed was unreasonably dangerous when used in the normal, anticipated manner intended by the seller. Id. at 832. We held the presence of the “unreasonably dangerous” element in our strict liability/design defect analysis did not inject considerations of negligence into the strict liability case. Id. at 835. We noted that in strict liability, the plaintiff’s proof concerns the condition of the product which is designed or manufactured in a particular manner. Id. In contrast, negligence concerns the reasonableness of the manufacturer’s conduct in designing and selling the product as it did. Id.

Some courts have held that a product may be unreasonably dangerous and thus “defective,” triggering strict liability, if the seller fails to give adequate directions or warning regarding its use. See Tucson Indus., Inc. v. Schwartz, 108 Ariz. 464, 501 P.2d 936, 940-41 (1972); Anderson v. Owens-Corning Fiberglas Corp., 53 Cal.3d 987, 281 Cal.Rptr. 528, 536-38, 810 P.2d 549, 557-59 (1991); Woodill v. Parke Davis & Co., 79 Ill.2d 26, 37 Ill.Dec. 304, 306, 402 N.E.2d 194, 196 (1980); Owens-Illinois, Inc. v. Zenobia, 325 Md. 420, 601 A.2d 633, 641 (1992); Mauch v. Manufacturers Sales & Serv., Inc., 345 N.W.2d 338, 346 (N.D.1984); Phillips v. Kimwood Mach. Co., 269 Or. 485, 525 P.2d 1033, 1038 (1974). Our court has never explicitly adopted this principle. See Iowa Civil Jury Instruction 1000.6 cmt. (1991). In Cooley v. Quick Supply Co., 221 N.W.2d 763, 765 (Iowa 1974), a case involving a prematurely exploding dynamite fuse, the jury considered both a strict liability count and a negligence count. In the strict liability count the only product defect submitted was that the fuse did not appear to be ignited when it was in fact ignited. Id. The sole count of negligence alleged that the defendant failed to warn as to the safe and proper use of the product. Id. We affirmed the jury’s verdict for the plaintiff. Id. at 773.

Generally, there are two competing views regarding the failure to warn/strict liability question. The first is that there is little, if any, difference between strict liability and negligence in failure to warn eases. Flaminio v. Honda Motor Co., 733 F.2d 463, 467 (7th Cir.1984) (Wisconsin); Standhardt v. Flintkote Co., 84 N.M. 796, 803-04, 508 P.2d 1283, 1290-91 (1973); Opera v. Hyva, Inc., 86 A.D.2d 373, 450 N.Y.S.2d 615, 618 (1982); Hardiman v. Zep Mfg. Co., 14 Ohio App.3d 222, 470 N.E.2d 941, 944 (1984); Castrignano v. E.R. Squibb & Sons, Inc., 546 A.2d 775, 782 (R.I.1988). Opposing this view are cases that apply varying forms of a strict liability analysis in failure to warn cases. Some jurisdictions impose strict liability by imputing knowledge of a product’s propensity to injure as it did to a defendant-manufacturer, and then asking the jury: With such knowledge would the defendant have been negligent in selling the product without a warning? See Tucson Indus., 501 P.2d at 941; Ulrich v. Kasco Abrasives Co., 532 S.W.2d 197, 200 (Ky.1976); Phillips, 525 P.2d at 1039; Berkebile v. Brantly Helicopter Corp., 462 Pa. 83, 337 A.2d 893, 902-03 (1975); Ayers v. Johnson & Johnson Baby Prods. Co., 117 Wash.2d 747, 818 P.2d 1337, 1343-46 (1991).

Other jurisdictions apply strict liability by requiring plaintiffs to prove defendants knew or should have known of the danger. See Anderson, 281 Cal.Rptr. at 537-38, 810 P.2d at 558-59; Woodill, 37 Ill.Dec. at 308, 402 N.E.2d at 198; Zenobia, 601 A.2d at 641; Mauch, 345 N.W.2d at 345. Usually, these courts require plaintiffs to prove that the defendant, “because of the ‘present state of human knowledge,’ knew or should have known of the danger presented by the use or consumption of a product.” Woodill, 37 Ill. Dec. at 308, 402 N.E.2d at 198 (citing Restatement (Second) of Torts, § 402A cmt. k (1965)); see also Zenobia, 601 A.2d at 641; Anderson, 281 Cal.Rptr. at 537, 810 P.2d at 558 (To prove strict liability in failing to *289 warn, plaintiffs must prove that defendant did not adequately warn of risk “known or knowable in light of the generally recognized and prevailing best scientific knowledge at the time of manufacture and distribution.”).

A different analysis is made in some cases to distinguish strict liability from negligence concepts in failure to warn cases on the ground that in negligence the focus is on the defendant’s conduct, while in strict liability, the focus is on the condition of the product. See Anderson, 281 Cal.Rptr. at 537, 810 P.2d at 558; Woodill, 37 Ill.Dec. at 308, 402 N.E.2d at 198; Mauch, 345 N.W.2d at 346; Ayers, 818 P.2d at 1345.

After reviewing the authorities and comments on the failure to warn question, we believe any posited distinction between strict liability and negligence principles is illusory. We fail to see any distinction between negligence and strict liability in the analyses of those jurisdictions injecting a knowledge requirement into their strict liability/failure to warn equation. See, e.g., Woodill, 37 Ill.Dec. at 308, 402 N.E.2d at 198. The standard applied by these “strict liability” jurisdictions is exactly the same in practice as holding defendants to an expert standard of care under a negligence theory. The burden on plaintiffs is the same. They must prove a defendant knew or should have known of potential risks associated with the use of its product, yet failed to provide adequate directions or warnings to users. With regard to those jurisdictions imputing to defendants knowledge of its product’s propensity to injure as it did, we have refused in the past to impose a duty upon manufacturers to warn of unknowable dangers. Moore v. Vanderloo, 386 N.W.2d 108, 116 (Iowa 1986).

We also find the product/conduct distinction made by several jurisdictions to justify maintaining a strict liability/failure to warn theory of little practical significance. See, e.g., Anderson, 281 Cal.Rptr. at 537, 810 P.2d at 558. According to courts taking stock in this distinction, under a strict liability theory the focus is on the unreasonably dangerous condition of the product. See id.; Woodill, 37 Ill.Dec. at 308, 402 N.E.2d at 198; Mauch, 345 N.W.2d at 346. In contrast, these courts hold the question in negligence cases is whether the defendant’s conduct breached a duty to exercise reasonable care. Woodill, 37 Ill.Dec. at 308, 402 N.E.2d at 198; Mauch, 345 N.W.2d at 346. In practice, the courts basing the application of a strict liability theory on this distinction cannot help but slip back into the type of analyses virtually identical to those employed in negligence cases. See Woodill, 37 Ill.Dec. at 308-09, 402 N.E.2d at 198-99 (holding the requirement that plaintiff plead and prove defendant knew or should have known of danger that caused injury, and faded to warn is “entirely consistent with the principles of strict liability.”); Mauch, 345 N.W.2d at 345 (holding defendants strictly liable for failure to warn of “dangers involved in a use which can be reasonably anticipated.”). Inevitably the conduct of the defendant in a failure to warn case becomes the issue. As one commentator notes regarding the produeVconduet distinction, “[i]t is easy enough to assert the manufacturer’s conduct is not in issue, but the ‘condition of the product’ test is hardly self-executing.” Keith Miller, Design Defect Litigation in Iowa: The Myths of Strict Liability, 40 Drake L.Rev. 465, 481 (1991).

Maintaining the distinction to justify submission of failure to warn claims under both strict liability and negligence theories is a vain effort. We hold it was error to submit instructions regarding Prosoco’s failure to warn under both negligence and strict liability theories. Instruction No. 43 and Instruction No. 26 were duplicative. Both instructions essentially required the jury to determine whether Prosoco negligently failed to warn users of the dangers in moving or using Sure Klean 600 in fifteen gallon containers. See Hillrichs, 478 N.W.2d at 76.

We believe that the correct submission of instructions regarding a failure to warn claim for damages is under a theory of negligence and the claim should not be submitted as a theory of strict liability. In testing the defendant’s liability for negligence in failing to warn, the defendant should be held to the standard of care of an expert in its field. See West v. Broderick & Bascom Rope Co., 197 N.W.2d 202, 210 (Iowa 1972); Castrignano, 546 A.2d at 782; cf. Zenobia, 601 A.2d at 639. The relevant inquiry therefore is whether the *290 reasonable manufacturer knew or should have known of the danger, in light of the generally recognized and prevailing best scientific knowledge, yet faded to provide adequate warning to users or consumers. See West, 197 N.W.2d at 209; Castrignano, 546 A.2d at 782; cf. Zenobia, 601 A.2d at 641; see also W. Page Keeton, et al., Prosser and Keeton on the Law of Torts § 99, at 697 (5th .ed. 1984) [hereinafter “Keeton”].

In reviewing the instructions given on the theory of negligence, we find that the claim of failure to warn as submitted substantially invoked this standard. Special verdict form No. 1 found Prosoco one-hundred percent liable on the claim of negligence. Also submitted by a special verdict form was the claim of strict liability under which the jury found Prosoco one-hundred percent at fault.

We have reversed and remanded cases where a general verdict of liability resulted from the submission of two theories, one of which contained an error in the instructions. See Erickson v. Wright Welding Supply, Inc., 485 N.W.2d 82, 86 (Iowa 1992); Gordon v. Noel, 356 N.W.2d 559, 565 (Iowa 1984). In these situations, we were unable to determine that the verdict resulted from a theory that was free of error. See Erickson, 485 N.W.2d at 86; Gordon, 356 N.W.2d at 565. However, in the case at bar, the special verdicts are sufficiently insulated from each other to stand on their own. We do not believe the error in submitting a failure to warn instruction as part of the strict liability claim had any prejudicial effect on the jury’s consideration of the same issue contained in the negligence claim. The separation of these theories both in theory and submission insulate them from the prejudice that may result from an undue emphasis of even correct statements of law. As such, we distinguish our decisions of Coker v. Abell-Howe Co., 491 N.W.2d 143, 146 (Iowa 1992), and Rosenau v. City of Estherville, 199 N.W.2d 125, 133 (Iowa 1972), where the errors required a reversal. The error in the strict liability submission regarding a failure to warn in the case at bar does not require reversal.

III. State-of-the-Art Defense

Prosoco contends the district court erred in refusing its request to instruct on its “state-of-the-art” defense with regard to Olson’s claim of negligence. The court did submit a “state-of-the-art” instruction regarding the strict liability claim.

Iowa Code section 668.12 (1991) provides:

In any action brought pursuant to this chapter against an assembler, designer, supplier of specifications, distributor, manufacturer or seller for damages arising from an alleged defect in the design, testing, manufacturing, formulation, packaging, warning, or labeling of a product, a percentage of fault shall not be assigned to such persons if they plead and prove that the product conformed to the state of the art in existence at the time the product was designed, tested, manufactured, formulated, packaged, provided with a warning, or labeled. Nothing contained in this section shall diminish the duty of an assembler, designer, supplier of specifications, distributor, manufacturer or seller to warn concerning subsequently acquired knowledge of a defect or dangerous condition that would render the product unreasonably dangerous for its foreseeable use or diminish the liability for failure to so warn.

Section 668.12 is a complete defense in product defect cases. Fell v. Kewanee Farm Equip. Co., 457 N.W.2d 911, 920 (Iowa 1990). Prosoco argues a state-of-the-art instruction is required by Iowa Code section 668.12 in negligence actions because the statute specifically speaks to the allocation of “fault.” Iowa Code § 668.12. Section 668.1 defines fault, for the purposes of applying chapter 668, as “one or more acts or omissions that are in any measure negligent or reckless toward the person or property of the actor or others, or that subject a person to strict tort liability.” Id. § 668.1.

Olson counters that section 668.12’s use of the term “defect” compels the submission of a state-of-the-art instruction only in strict liability cases. Olson contends it is impossible to have a state-of-the-art defense in a negligence action because if the product com *291 plied with the state-of-the-art, there would be no lack of due care and no negligence.

Counsel for both parties present persuasive arguments on this point. Section 668.12 does concern the allocation of “fault,” a concept defined in the chapter to include negligence principles. See id. § 668.1. On the other hand, the statute is applicable only in situations where “damages aris[e] from ... alleged defects in products.” Id. § 668.-12.

We noted earlier in our discussion of the failure to warn concept, that the focus is on the conduct of the defendant, not the condition of the product. Section 668.12, on the other hand, provides a state-of-the-art defense for product defects. We believe the issue of the adequacy of warnings is a matter quite extraneous to the condition of the product itself. Cf. Anderson, 281 Cal.Rptr. at 537, 810 P.2d at 558. It is also illogical to excuse negligent conduct by a state-of-the-art defense based on the reasoning that all other manufacturers were equally negligent. Thus, the state-of-the-art defense provided by section 668.12 is incongruous when applied to claims based on negligent failure to warn.

We note this does not preclude defendants from presenting state-of-the-art evidence at trial. However, rather than establishing an absolute defense in negligent failure to warn cases, defendant’s evidence would go to rebut the plaintiff’s proof that the defendant breached a duty to exercise the degree of care a reasonable manufacturer would have used in light of generally recognized and prevailing scientific knowledge. We hold that an instruction on a state-of-the-art defense is not required in negligent failure to warn eases. The trial court properly refused to give the instruction requested by Prosoeo on this issue.

IV. Sufficiency of Evidence to Submit Failure to Warn Instruction

There is no duty to warn where risks are known and obvious to the plaintiff. Nichols v. Westfield Indus., Ltd., 380 N.W.2d 392, 401 (Iowa 1985). Prosoeo contends the clumsiness and awkwardness of the drum were conditions of which Olson knew or should have known and were obvious because Olson handled the drums frequently in the course of his employment. Olson argues that the risks of the bung closure popping out when he set the drum down with the lid hand tightened were not known nor were they obvious.

We believe sufficient evidence exists in the record to support the submission of a failure to warn instruction to the jury in this case. Olson testified that he did not know how the lid came off the drum. The closure had never popped out before the accident. Olson did testify that he occasionally had problems with “cross threading” closures when screwing them into the containers. We believe the evidence presented was sufficient to submit a jury question on negligent failure to warn.

V. Failure to Submit Instruction on Unreasonable Failure to Avoid Injury

The doctrine of avoidable consequences states that a party cannot recover damages that result from consequences that a party could reasonably have avoided. Coker, 491 N.W.2d at 149. The doctrine comes into play after a legal wrong occurs, but while some damage may still be averted, and bars recovery only for such damages. Id. The doctrine is akin to mitigation of damages in contract actions when plaintiffs must attempt to reduce damages after an injury occurs. Id.

Prosoeo contends Olson’s failure to check the closure and wear goggles and protective gear occurred after Prosoco’s negligence occurred. It argues the doctrine of avoidable consequences thus bars recovery. Olson contends the doctrine operates only after the injury occurs, but while some damage may still be averted.

These types of questions come down to issues of timing. Id. The doctrine would only operate in this case to preclude damages after the occurrence of Olson’s injury. There is no legal wrong — that is no negligence— without actual injury. See Keeton, § 30, at 165. The factors proposed by Prosoeo that would trigger the operation of the doctrine (i.e. wearing goggles) occurred before Olson’s *292 injury and thus before Prosoco committed any legal wrong toward Olson. Moreover, there is no evidence in the record that shows Olson could do anything to mitigate his damages. We affirm the district court’s refusal to instruct on this issue of defense.

VI. Inconsistent Jury Verdicts

Jury verdicts are to be liberally construed to give effect to the intention of the jury. Hoffman v. National Medical Enters., Inc., 442 N.W.2d 123, 126 (Iowa 1989). If verdicts can be reconciled in any reasonable manner consistent with the evidence and its fair inferences, and in light of the jury instructions, we will uphold them. Id. at 127.

Prosoco contends that because the jury found Rieke, the manufacturer of the bung closures, and Delta Drum, the manufacturer of the drums, zero percent liable, it is impossible that Prosoco could be one-hundred percent liable for Olson’s injuries. Olson contends the jury could have been persuaded that Prosoco was solely liable in this case due to the unique use to which Prosoco put the closures and drums — as storage for hydrochloric acid based mortar cleaner.

We agree with Olson. It was not necessarily inconsistent for the jury to find no liability on the part of the component manufacturers, and to find Prosoco liable. Obviously, the jury believed it was the unique use to which Prosoco put the closures and drums which established Prosoeo’s liability to the exclusion of liability on the part of Rieke and Delta. See generally Keeton, § 100, at 704-06. There was no evidence that the two-inch NPT closure made by Rieke and the fifteen gallon drum made by Delta Drum were unreasonably dangerous in and of themselves. The evidence was sufficient, however, for the jury to conclude that a dangerous product was created when the NPT closure and drum were used to contain fifteen gallons of acid based mortar cleaner. The jury verdicts were consistent and the district court is affirmed on this issue.

VII. Excessive Jury Verdict

When reviewing verdicts for excessive damages, we view the evidence in a light most favorable to the plaintiff. DeBurkarte v. Louvar, 393 N.W.2d 131, 139 (Iowa 1986). We will not disturb a jury verdict for damages unless it is “flagrantly excessive or inadequate, so out of reason so as to shock the conscience, the result of passion or prejudice, or lacking in evidentiary support.” Harsha v. State Sav. Bank, 346 N.W.2d 791, 799 (Iowa 1984). None of these factors are evident in this case.

Prosoco complains of the jury’s award of $20,000 for future medical expenses. Olson’s ocularist testified that the cost of periodically cleaning and replacing Olson’s prosthetic eye would run in excess of $22,000. Prosoco also complains the award of $10,000 for lost future earning capacity is not supported by the record. The gist of Prosoeo’s argument is that Olson has not missed significant amounts of work to date since recovery from his injury. However, several of Olson’s coworkers testified to the substantially lower quality of work Olson performs since his injury. One eoworker had to take responsibility for shabby work done by Olson since returning to Seedorf. We do not believe an award of $10,000 is unreasonable in light of this evidence. We find Prosoeo’s arguments for reduction of the damages for pain and suffering equally without merit. There is support in the record for all the damages the jury awarded Olson.

VIII.Juror Misconduct

To impeach a verdict on the ground of juror misconduct, three conditions must be met:

1. Evidence from the jurors must consist only of objective facts as to what actually occurred in or out of the jury room bearing on misconduct;

2. The acts or statements complained of must exceed tolerable bounds of jury deliberation;

3. It must appear the misconduct was calculated to, and with reasonable probability did, influence the verdict.

State v. Cullen, 357 N.W.2d 24, 27 (Iowa 1984).

Prosoco presents affidavits from two jurors that describe an experiment per *293 formed with the water filled model drums produced at trial. Prosoco contends the conduct exceeded tolerable bounds and “must have been reasonably calculated to and probably did influence the jury.” Olson contends the juror affidavits are incompetent evidence. Even if they are competent evidence, Olson argues the experiment did not exceed tolerable bounds of jury conduct because the experiment did not go beyond any evidence already in the record. Olson also notes the experiments probably did not influence the jury because they were carried out by the only juror to vote contrary to a plaintiffs verdict.

We note that the experiments were cumulative of evidence already presented by expert witnesses at trial. See State v. Thompson, 326 N.W.2d 335, 337 (Iowa 1982). In any event, this issue fails because it falls within the ambit of our decision on juror testimony in Ryan v. Arneson, 422 N.W.2d 491 (Iowa 1988). There we said:

Based on Iowa’s adoption of language identical to Federal Rule of Evidence 606(b), and the policy reasons for insulating the manner in which the jury reaches its verdict, we now adopt the federal rule which protects each of the components of deliberation including juror arguments, statements, discussions, mental and emotional reactions, votes, and any other feature of the process occurring in the jury room. The district court was correct to disregard affidavits of jurors concerning a quotient verdict.

Id. at 495.

IX. Federal Preemption

Prosoco argues that Olson’s state common law tort claims are preempted by the federal Hazardous Materials Transportation Act (HMTA). 49 App.U.S.C. § 1801 et seq. (Supp.1991). In reply, Olson contends the HMTA has no application to cases involving a consumer’s end use. Moreover, Olson contends the preemption doctrine does not prevent states from imposing common law tort standards stricter than those imposed by federal law.

Explicit federal preemption of state law occurs when Congress expressly indicates its intention to preempt state law. Brotherhood of Maintenance of Way Employees v. Chicago & N.W. Transp. Co., 514 N.W.2d 90, 93 (Iowa 1994). Implied preemption occurs if it can be determined that Congress intended to occupy an entire field, excluding state law. Id. Implicit preemption will most likely be found when it is determined that state law conflicts with congressional policy. Id.

We find no merit in Prosoco’s argument on this point. Congress enacted the HMTA in order “to improve the regulatory and enforcement authority of the Secretary of Transportation to protect the Nation adequately against the risks to life and property which are inherent in the transportation of hazardous materials in commerce.” 49 App.tLS.C. § 1801 (1988). There is no indication that Congress intended to preempt common law state tort remedies arising out of a consumer’s end use of a hazardous material. See Hurt v. Coyne Cylinder Co., 956 F.2d 1319, 1323 (6th Cir.1992). The HMTA and its accompanying regulations are intended to regulate the shipment or transport of hazardous materials in intrastate, interstate, and foreign commerce. Id.; see also 49 App. U.S.C. § 1804 (1988). In a hypertechnieal sense, Olson “transported” hazardous materials when he picked up the fifteen gallon drum of Sure Klean 600 and set it on the pallet. The construction of any statute, however, must be reasonably, sensibly, and fairly made with a view of carrying out the obvious intention of the legislature enacting it. In re Estate of Keegan, 369 N.W.2d 447, 450 (Iowa 1985). We do not believe Congress intended to regulate the kind of conduct at issue in this case. Other jurisdictions are in accord. See Hurt, 956 F.2d at 1323; Kemner v. Monsanto Co., 217 Ill.App.3d 188, 160 Ill.Dec. 192, 206, 576 N.E.2d 1146, 1160 (1991); Sawash v. Suburban Welders Supply Co., 407 Mass. 311, 553 N.E.2d 894, 898 (1990).

We have considered all of the arguments and issues raised by Prosoco in this appeal. No reversible error has occurred. The judgment entered by the trial court is affirmed.

AFFIRMED.

JOHNSTONE, Justice.

Michael Ostendorf was severely injured when the Clark forklift he was driving tipped over and pinned his right foot to the ground. The accident occurred when a baggage tug vehicle operated by another employee collided with the forklift Osten-dorf was driving. The accident occurred in October 1994 at the Delta Airlines terminal of the Greater Cincinnati-Northern Kentucky International Airport in Boone County, Kentucky. Ostendorf was employed by Delta at the time.

Ostendorf and his wife filed suit in Kenton Circuit Court against Clark claiming (1) strict product liability, (2) negligent design, (3) breach of duty to retrofit the forklift with operator restraints, (4) negligent conduct of the retrofit campaign, and (5) breach of warranty. In addition to compensatory damages, Ostendorf sought punitive damages. The circuit court granted summary judgment in favor of Clark. Ostendorf appealed to the Court of Appeals, which reversed the summary judgment on the strict liability and negligent design claims, as well as the compensatory and punitive damages relating to those claims. The Court of Appeals further held that Kentucky does not recognize a common law duty by a seller to retrofit an existing product that was not defective at the time it was manufactured. Finally, the Court of Appeals held that Ostendorf failed to present sufficient evidence to impose liability on Clark for negligent performance of a voluntary retrofit campaign. We affirm the Court of Appeals’ decision.

Because this matter was disposed of on a motion for summary judgment, we shall review the facts of this case in the light most favorable to the Appellant. See Steelvest. Inc., v. Scansteel Service Center, Inc., Ky., 807 S.W.2d 476 (1991). The forklift Ostendorf was operating was a 1980 model C-300Y40 manufactured by Clark Equipment Company. It was sold to Western Airlines by an authorized Clark dealer in California. Delta Airlines acquired the forklift when it purchased the assets of Western Airlines. The forklift was designed and manufactured in accordance with the then applicable industry standards, as set out by the Federal Occupational Safety and Health Act (OSHA) *533 and the American National Standards Institute (ANSI). At the time the forklift was manufactured, operator restraints were not required.

Ostendorf presented evidence that Clark and other manufacturers of forklift trucks exercised considerable influence on the formulation of the ANSI standards through their membership in the Industrial Truck Association (ITA). Since the 1960⅛, Clark and the ITA were aware of the tendency of forklift trucks to overturn under certain circumstances, causing injury and death to the operator. However, there was considerable debate throughout the 1970’s regarding the need to install operator restraints. In 1979, a Clark engineer was killed when a forklift that he was test driving overturned. This incident, among other things, prompted Clark to further study the need for operator restraints.

In 1983, Clark developed a new safety seat incorporating an operator restraint for its forklifts. In addition, Clark voluntarily implemented a retrofit program that provided owners of the C-300 and C-500 model forklifts with the opportunity to have the new restraint system installed on their forklifts. Under this program, Clark offered to retrofit qualifying forklifts without cost either for parts or for labor. Clark mailed notices to its customers advising them of the availability of the retrofit. There is a dispute concerning whether Delta received notice of the retrofit program; Clark claims it mailed the notice to Delta, but Delta denies receipt.

On appeal to this Court, Ostendorf argues that (1) Clark had a common law duty to retrofit its forklifts; (2) Clark negligently conducted the retrofit program; (3) Clark failed to meet the standard for an appropriate retrofit campaign under the standard set out in the Restatement (Third) of Torts, which this Commonwealth should adopt; and (4) punitive damages should not require intentional conduct. Clark does not cross-appeal the decision of the Court of Appeals to reinstate the negligence and strict liability claims.

DUTY TO RETROFIT

Ostendorf first argues that a product manufacturer has an affirmative, common law duty to retrofit existing products with safety features that are necessary to make the product reasonably safe. Osten-dorf arrives at this duty based on principles of negligence: foreseeability and reasonable care. In a well-reasoned opinion by Judge Knopf, the Court of Appeals held to the contrary, declaring that Kentucky does not recognize a common law duty by a seller to retrofit an existing product which was not defective at the time it was manufactured with subsequently developed safety features. We agree with the Court of Appeals, with one exception. We think the relevant point in time is not when the product is manufactured, but when it is sold.

In Kentucky, the existence of a duty is a matter of law for the court because “[wjhen a court resolves a question of duty it is essentially making a policy determination.” Mullins v. Commonwealth Life Ins. Co., Ky., 839 S.W.2d 245, 248 (1992). A duty to retrofit is a duty to upgrade or improve a product. Some courts have created such a duty in certain, limited circumstances. For example, in Braniff Airways, Inc. v. Curtiss-Wright Corp., 411 F.2d 451 (2d Cir.1969), plaintiff Braniff purchased an airplane engine from Curtiss-Wright. The engine was installed on one of Braniffs passenger airplanes, which later crashed; it was determined that the engine was the cause. Finding that an airplane engine was a product involving “human safety,” the Sec *534 ond Circuit held: “It is clear that after such a [human-safety] product has been sold and dangerous defects in design have come to the manufacturer’s attention, the manufacturer has a duty either to remedy these or ... at least to give users adequate warnings and instructions concerning methods for minimizing the danger.” Accord Nod v. United Aircraft Corp., 342 F.2d 232, 236-37 (3d Cir.1964) (duty of care owed by the manufacturer of an airplane propeller system, “an instrumentality likely to endanger the public”). In one case, a Texas court created a duty to retrofit where the service representative of a helicopter manufacturer purchased and resold one of the company’s used helicopters but failed to install an improved tail rotor developed to prevent a known flight hazard. See Bell Helicopter Co. v. Bradshaw, 594 S.W.2d 519 (Tex.Civ.App.1979). While the few cases finding a duty to retrofit often involve products that directly implicate human safety, that is not always the case. One court found a duty to retrofit a mining shovel where the total number of units sold was so small, 120, that it was easily feasible for the manufacturer to retrofit the product or warn buyers of the danger. See Readenour v. Marion Power Shovel, 149 Ariz. 442, 719 P.2d 1058 (1986). And one court has even held, without qualification, that “failure to conduct an adequate retrofit campaign may constitute negligence apart from the issue of defective design.” Hernandez v. Badger Const. Equip. Co., 28 Cal.App.4th 1791, 34 Cal.Rptr.2d 732 (1994) (court affirmed jury finding that crane manufacturer could be hable for failing to install later developed safety device on crane that was not defective when sold); but see Tabieros v. Clark Equip. Co., 85 Hawai'i 336, 944 P.2d 1279 (1997) (concluding there is no duty to retrofit and criticizing both the Readenour and Hernandez decisions as wanting “any discernible reason explaining their implicit adoption of a duty to retrofit”).

But the majority of jurisdictions reach a different conclusion: there is no duty to retrofit a product not defective when sold. See 47 A.L.R. 5th 395. We find the majority reasoning persuasive and we find the result appropriate. There are two main reasons militating against a duty to retrofit. First, in many cases, a duty to retrofit is properly the province of an administrative or legislative body. Second, and more importantly, there is no reason to create a duty to retrofit a product not defective when sold — traditional principles of negligence and strict products liability suffice.

A retrofit campaign may be a costly undertaking. It is typically a multi-step, multi-party process that may cost millions of dollars. See Douglas R. Richmond, “Expanding Products Liability: Manufacturers’ Post-Sale Duties to Warn, Retrofit and Recall,” 36 Idaho L.Rev. 7, 10 (1999). The complexity of the decision to retrofit and the ramifications of that decision recommend that courts should not make that determination, but rather should leave it to governmental bodies more suited to the task. See, e.g., Gregory v. Cincinnati Inc., 450 Mich. 1, 538 N.W.2d 325, 334 (Mi.1995); Patton v. Hutchinson Wil-Rich Mfg. Co., 253 Kan. 741, 861 P.2d 1299, 1315 (1993). Professor Schwartz, principal author of The Uniform Product Liability Act, cautions persuasively against a duty to retrofit by judicial fiat:

[C]ourts that impose a post-sale obligation to remedy or replace products already in the marketplace arrogate to themselves a power equivalent to that of requiring product recall. Product recalls, however, are properly the province of administrative agencies, as the federal statutes that expressly delegate recall authority to various agencies suggest. As Congress has recognized, ad *535 ministrative agencies have the institutional resources to make fully informed assessments of the marginal benefits of recalling a specific product. Because the cost of locating, recalling, and replacing mass-marketed products can be enormous and will likely be passed on to consumers in the form of higher prices, the recall power should not be exercised without extensive consideration of its economic impact. Courts, however, are constituted to define individual cases, and their inquiries are confined to the particular facts and arguments in the cases before them. Decisions to expand a manufacturer’s post-sale duty beyond making reasonable efforts to warn product users about newly discovered dangers should be left to administrative agencies, which are better able to weigh the costs and benefits of such action. [Schwartz, The post-sale duty to warn: Two unfortunate forks in the road to a reasonable doctrine, 58 N.Y.U.L.Rev. 892, 901.]

These considerations aside, there is a more fundamental reason to eschew imposition of a duty to retrofit: that duty is superfluous in light of existing negligence and product liability doctrines. See Tabieros, 85 Hawai'i 386, 944 P.2d 1279. A party injured by a product can bring suit for that injury under three different theories: (1) breach of warranty under the Uniform Commercial Code, (2) negligence, or (8) strict liability in tort. See Williams v. Fulmer, Ky., 695 S.W.2d 411 (1985). When the case involves a retrofit, the plaintiff is claiming the product was defectively designed. Here, for example, Ostendorf claims the C-300 forklift was defectively designed because it did not have operator safety restraints. A plaintiff in Kentucky can bring a defective design claim under either a theory of negligence or strict liability. The foundation of both theories is that the product is “unreasonably dangerous.” Ulrich v. Kasco Abrasives Co., Ky., 532 S.W.2d 197, 200 (1976). Whereas negligence examines the conduct of the manufacturer — could the manufacturer foresee the harm to the plaintiff and did the manufacturer act reasonably to prevent that harm — strict liability typically evaluates the condition of the product. But in a negligent design case, even a strict liability claim examines the manufacturer’s conduct. See Nichols v. Union Underwear Co. Inc., Ky., 602 S.W.2d 429 (1980) (distinction between strict liability and negligence “is of no practical significance so far as the standard of conduct required of the defendant”). So under either theory, it is the legal duty of a manufacturer to use reasonable care to protect against foreseeable dangers. In a design defect case, courts use some form of risk-utility analysis to assess the decisions made by manufacturers with respect to the design of their products. See Prentis v. Yale Mfg. Co., 421 Mich. 670, 365 N.W.2d 176, 183 (1984); see also David J. Leibson, 13 Kentucky Practice: Tort Law § 13.7, p. 384 (1995). Because a manufacturer chooses the design of its product, the emphasis is on the manufacturer’s conduct, not the product:

A conscious decision to design a product in a certain manner necessitates that the focus be on conduct rather than the product. Hence, the trier of fact must employ a risk-utility balancing test that considers alternative safer designs and the accompanying risk pared against the risk and utility of the design chosen “to determine whether ... the manufacturer exercised reasonable care in making the design choices it made.”

Gregory, 538 N.W.2d at 329-30 (internal citations omitted). Significantly, the risk-utility test examines what the manufacturer knew or should have known at the time the product was sold.

*536 In Gregory, the Michigan Supreme Court considered the propriety of instituting a duty to retrofit. That case involved a worker injured in 1986 by an industrial machine that was not defective when manufactured in 1964. The injured worker argued that the manufacturer had a duty to retrofit its product with later designed safety devices, but the court refused to create such a duty. As the Gregory court explained, there are two retrofit scenarios: a retrofit for a latent defect or a retrofit because of a technological advance. See Gregory, 538 N.W.2d at 325. A latent defect is one that existed undiscovered at the time of manufacture. Because a manufacturer’s liability for a design defect hinges on what it knew or should have known at the time of sale, a negligence or strict liability theory can assess liability for a latent defect, so a duty to retrofit is unnecessary. If, however, the retrofit results from a post-sale technological advance, then the product was not originally defective, but has become so due only to the later advancement. That poses a problem under the traditional theories because “Hocusing on post[sale] conduct in a negligent design case improperly shifts the focus from point-of-[sale] conduct and considers post[sale] conduct and technology that accordingly has the potential to taint a jury’s verdict regarding a defect.” Id. at 334. The result is that liability for a post-sale defect would have to be premised on some other theory, like an independent duty to retrofit. But there are prevailing reasons not to impose such liability on manufacturers for post-sale advances, chiefly: “[I]mposing a duty to update technology would place an unreasonable burden on manufacturers. It would discourage manufacturers from developing new designs if this could form the bases for suits or result in costly repair....” Id. at 337; see also Restatement (Third) of Torts: Products Liability § 11, Comment a (1998). Placing such an onerous burden on manufacturers would be tantamount to making strict liability absolute liability and making a manufacturer an insurer. See Modelski v. Navistar Int'l. Trans. Corp., 302 Ill.App.3d 879, 236 Ill.Dec. 394, 707 N.E.2d 239, 247 (1999); Jones v. Hutchinson Mfg. Inc., Ky., 502 S.W.2d 66, 70 (1973). This we decline to do.

A duty to retrofit is an example of a post-sale obligation imposed on a manufacturer. Other examples are the duty to warn of later-discovered defects or foreseeable misuses and the duty to recall a defective product. Numerous cases impose a duty to warn of later discovered defects. 1 See, e.g., Gracyalny v. Westinghouse Elec. Corp., 723 F.2d 1311,1318 (7th Cir.1983) (manufacturer’s duty to warn extends to dangers that arise after marketing); LaBelle v. McCauley Indus. Corp., 649 F.2d 46, 49 (1st Cir.1981) (manufacturer’s duty to warn extends to purchaser even if defects are discovered after initial sale); Chrysler Corp. v. Batten, 264 Ga. 723, 450 S.E.2d 208, 211-13 (1994) (duty to warn arises whenever manufacturer knows or reasonably should know of danger arising from product use). When a product is defective, a manufacturer may be subject to one of these post-sale duties. The nature of the defect will dictate the appropriate remedy: a defect that may result in a few minor injuries may only require a warning, whereas a defect that may result in serious injury or death could require more. These remedial measures can be seen as “a continuum of post-sale duties which the law might impose.... ” Gregory, 538 N.W.2d at 341 (Cavanagh, J., dissenting).

*537 One writer has suggested that post-sale claims might provide some advantages to traditional claims of negligence and strict liability:

Post-sale claims afford plaintiffs strategic advantages not available in typical products liability cases. For example, a post-sale failure-to-warn claim may allow a plaintiff to avoid a statute of repose, may negate a state-of-the-art defense, may allow a plaintiff to avoid a defense based on a user’s substantial modification of a product, or may allow before a jury otherwise inadmissible evidence.

Richmond, 36 Idaho L.Rev. at 10. After examining these considerations, we are persuaded that they have little validity in design defect cases in Kentucky. First, Kentucky has no statute of repose for product-related claims. Next, in a design defect case, unlike in a duty to warn case, the state-of-the-art defense is still viable and cannot be circumvented because, as discussed, supra, the critical time period is before the product is sold. The next consideration, a user’s substantial modification of the product that makes the product defective, can be dealt with under existing theories of liability. Finally, in some jurisdictions, evidence of subsequent remedial measures is not admissible to prove the product was defective. See Readenour, 719 P.2d at 1061, citing A.R.S. § 12-686 (in product liability action, subsequent changes in design or state of the art not admissible as direct evidence of a defect). But in product liability actions in Kentucky, evidence of subsequent remedial measures is admissible. See KRE 407. Not only are we unpersuaded by these reasons, but we see no other good reason to create an independent duty to retrofit. 2 Accordingly, if Ostendorf can prove what he claims, that Clark’s product was defective when sold, then he can recover under existing theories of negligence or strict liability.

LIABILITY FOR VOLUNTARY RETROFIT CAMPAIGN

Ostendorf next argues that Clark is liable to him because of its retrofit campaign. Ostendorfs claims are based on the voluntary retrofit program Clark began after developing safety improvements in 1983. Although Clark had begun to retrofit its model C-300 forklifts, the forklift involved in the accident injuring Osten-dorf was never retrofitted. While not clear from Ostendorfs brief in this appeal, it appears from the Court of Appeals’ opinion that Ostendorf claims Clark “failed to adequately notify customers of the availability of and the need for the new safety features, and that it failed to provide its dealers with sufficient incentives to implement the retrofit program.” Clark maintains that it mailed Delta notice of the program in 1985, but Delta denies ever receiving the notice. Ostendorf first contends that, under the common law principle of voluntary assumption of a duty, Clark negligently conducted its retrofit campaign. Ostendorf next asserts that Clark failed to meet the standard for an appropriate retrofit campaign under the Restatement (Third) of Torts: Products Liability § 11 (1998).

Ostendorfs common law argument is that a manufacturer who voluntarily undertakes a retrofit program can be held liable for negligently performing that program. Ostendorf looks, in part, to Bell Helicopter, 594 S.W.2d 519, to support his claim. In that case, the manufacturer be *538 gan a retrofit campaign to remedy a defective tail rotor system in a helicopter, but the helicopter flown by the plaintiffs had not yet been retrofitted when it crashed. The Texas Court of Civil Appeals held that Bell assumed the duty to improve upon the safety of its helicopter by replacing the defective part in its helicopter line and “[o]nce the duty was assumed, Bell had an obligation to complete the remedy [in this particular helicopter] by using reasonable means available to it ....” Bell Helicopter, 594 S.W.2d at 532. But Bell Helicopter is easily distinguishable from the present case. In Bell Helicopter, a Bell-authorized service station purchased the helicopter before the accident but after the retrofit campaign had begun. The service station was aware of the retrofit and the danger a failure to retrofit posed. The service station sold the helicopter without retrofitting it and the crash followed. The court imputed the failure to retrofit to Bell. Unlike that case, the record in the case at bar is devoid of evidence that the defendant in any way had or regained control over the product that led to the plaintiffs injury.

Ostendorf also quotes well-settled Kentucky case law to support his case: “one who volunteers to act, though under no duty to do so, is charged with the duty of acting with due care.” Sheehan v. United Services Auto Ass’n, Ky.App., 913 S.W.2d 4, 6 (1996); see also Estep v. B.F. Saul Real Estate Inv. Trust, Ky.App., 843 S.W.2d 911, 914 (1992). This is an accurate statement of the law, but falls short of attaching liability to Clark. As the Court of Appeals recognized, “traditionally, the purpose of imposing liability upon a party who has assumed a duty to act is premised upon reliance.” See, e.g., Louisville Cooperage Co. v. Lawrence, 313 Ky. 75, 78, 230 S.W.2d 103, 105 (1950). The Restatement (Second) of Torts § 324A (1965), which considers liability based on negligent performance of an undertaking, includes reliance as one of three bases for imposing a duty:

One who undertakes, gratuitously or for consideration, to render services to another which he should recognize as necessary for the protection of a third person or his things, is subject to liability to the third person for physical harm resulting from his failure to exercise reasonable care to protect his undertaking, if

(a) his failure to exercise reasonable care increases the risk of such harm, or

(b) he has undertaken to perform a duty owed by the other to the third person, or

(c) the harm is suffered because of reliance of the other or the third person upon the undertaking.

Under this framework, Ostendorf would have to demonstrate not only that Clark undertook a retrofit campaign on the C-300 forklifts for the protection of third persons, but also that Clark’s negligent performance of that task (a) increased the risk of harm to Ostendorf, (b) was incompatible with the discharge of a duty by Ostendorf s employer, Delta, or (c) caused Ostendorf to suffer harm because either Delta or Ostendorf relied on Clark to complete the retrofit. See Tabieros, 944 P.2d at 1302; see also Flock v. Scripto-Tokai Corp., 2001 WL 34111725, 2001 U.S. Dist. Lexis 23885 (S.D. TX 2001); Morrison v. Kubota Tractor Corp., 891 S.W.2d 422 (Mo.App.1994); but see Blossman Gas Co. v. Williams, 189 Ga.App. 195, 375 S.E.2d 117 (1988). While it is clear that Clark began a retrofit campaign, there is no evidence that any of the three additional conditions existed. In fact, regarding reliance, Delta and Ostendorf deny ever receiving notice of the retrofit program from Clark. As discussed, supra, the existence *539 of a duty is essential to a negligence claim. In the absence of one of these three conditions, Clark owed no duty to Ostendorf based on the voluntary retrofit campaign.

The Restatement (Third) of Torts: Products Liability § 11 (1998), however, does not impose such stringent requirements to establish liability, at least in the context of product recalls. That section provides:

One engaged in the business of selling or otherwise distributing products is subject to liability for harm to persons or property caused by the seller’s failure to recall a product after the time of sale or distribution if:

(a)(1) a governmental directive issued pursuant to a statute or administrative regulation specifically requires the seller or distributor to recall the product; or

(a)(2) the seller or distributor, in the absence of a recall requirement under Subsection (a)(1), undertakes to recall the product; and

(b) the seller or distributor fails to act as a reasonable person in recalling the product.

Ostendorf argues that § 11(a)(2) applies to this case. Ostendorf further alleges that Clark’s voluntary recall was merely an attempt to avoid a government mandated recall or retrofit, as explained in Comment c to § 11:

In the context of products liability, courts appear to assume that voluntary recalls are typically undertaken in the anticipation that, if the seller does not recall voluntarily, it will be directed to do so by a governmental regulator. Having presumably forestalled the regulatory recall directive, the seller should be under a common-law duty to follow through on its commitment to recall.

But § 11 has not been adopted in Kentucky. And Ostendorf fails to cite a single case from any jurisdiction that relies on § 11 to impose liability on a manufacturer for a negligent retrofit campaign. More importantly, adopting the § 11 approach— with its lax requirements — would have the perverse effect of discouraging voluntary retrofits and recalls. A retrofit campaign is a complex process requiring an abundance of technical, administrative, and legal coordination. Imposing liability on a company for a good faith — but perhaps incomplete — effort to undertake that task might dissuade that company from acting until required to by a government directive. Moreover, a company that would have begun a voluntary campaign in the absence of such strict liability but refrains in the face of such liability might never be compelled to perform that campaign. 3 The result is delay in the first instance and complete omission in the second, and a heightened risk of injury to consumers either way. In the interest of the safety of the citizens of this Commonwealth, we decline to adopt § 11.

We think the better course is to impose liability under the dictates of the Restatement (Second) § 324A, which requires proof of reliance, action inconsistent with a duty owed by another party, or increased risk of harm. As discussed, supra, and by the Court of Appeals: “If a product was defective at the time of [sale], then the seller may be hable for injuries resulting from that defect. However if the product was not defective at the time of [sale], the seller’s post[sale] conduct must contribute to the injury before liability will be imposed.” Based on the facts in this case, *540 we hold that by initiating a voluntary retrofit campaign, Clark did not assume a duty sufficient to impose liability for Os-tendorf s injuries.

CONSTITUTIONAL ISSUES

Ostendorfs last argument raises questions about the constitutionality of KRS 411.184, which deals with punitive damages in Kentucky. Ostendorf raises two concerns, one involving the mens rea required to support punitive damages — intentional conduct or gross negligence — and the other involving the standard of proof required to prove such damages, clear and convincing or preponderance of the evidence. Because the trial court granted summary judgment on the underlying claims, it did not touch upon the constitutional issues. In the absence of a ruling by the trial court, the Court of Appeals appropriately declined to rule on the matter, even though its decision reinstated claims upon which punitive damages could be founded. We likewise decline to consider constitutional arguments unaddressed by the lower courts.

For the foregoing reasons, we affirm the decision of the Court of Appeals.

ROTH, Circuit Judge:

On January 12, 2015, Heather Oberdorf returned home from work, put a retractable leash on her dog, and took the dog for a walk. Unexpectedly, the dog lunged, causing the D-ring on the collar to break and the leash to recoil back and hit Oberdorf's face and eyeglasses. As a result, Oberdorf is permanently blind in her left eye.

Oberdorf bought the collar on Amazon.com. As a result of the accident, she sued Amazon.com, including claims for strict products liability and negligence. The District Court found that, under Pennsylvania law, Amazon was not liable for Oberdorf's injuries. In its opinion, the District Court emphasized that a third-party vendor-rather than Amazon itself-listed the collar on Amazon's online marketplace and shipped the collar directly to Oberdorf. Those facts were the basis for the District Court's two main rulings.

First, the District Court found that Amazon is not subject to strict products liability claims because Amazon is not a "seller" under Pennsylvania law. Second, the District Court found that Oberdorf's claims are barred by the Communications Decency Act (CDA) because she seeks to hold Amazon liable for its role as the online publisher of third-party content.

I

Both issues in this case pertain to Amazon's role in effectuating the sale of products offered by third-party vendors. Therefore, we begin by describing the anatomy of a sale on Amazon.com. 1

Amazon Marketplace

Amazon is the world's most valuable retail company. 2 Its website is an online marketplace where Amazon retails its own products as well as those of more than one million third-party vendors. 3 These third-party vendors decide which products to sell, the means of shipping, and product pricing. For its part, Amazon lists the products on the Amazon Marketplace, collects order information from consumers, and processes payments. In exchange for these services, Amazon collects fees from each third-party vendor.

In order to use Amazon's services, a third-party vendor must assent to Amazon's Services Business Solutions Agreement. This Agreement governs every step of the sales process.

Once a third-party vendor has assented to the Agreement, the vendor chooses which product or products it would like to sell using Amazon's website. This choice is, with some notable exceptions, left to the discretion of the vendor. Among the exceptions are products that Amazon determines are illegal, sexually explicit, defamatory, or obscene.

When the third-party vendor has chosen a product that it wants to offer on Amazon's website, the vendor provides Amazon with a description of the product, including its brand, model, dimensions, and weight. Pursuant to the Agreement, the vendor must also provide Amazon with digital images of the product, as well as other information such as shipping and handling options, product availability, in-stock status, and any other information reasonably requested by Amazon.

Based on this information, Amazon formats the product's listing on its website. This function, too, is provided for in the Agreement, by which Amazon retains the right in its sole discretion to determine the content, appearance, design, functionality, and all other aspects of the Services, including by redesigning, modifying, removing, or restricting access to any of them. In fact, the Agreement grants Amazon a royalty-free, non-exclusive, worldwide, perpetual, irrevocable right and license to commercially or non-commercially exploit in any manner, the information provided by third-party vendors.

The third-party vendor can then choose which, if any, of Amazon's other services it will use in conjunction with listing its product on Amazon's website. For example, Amazon offers "Amazon Clicks," an advertising service in which Amazon highlights and promotes the vendor's product to customers. Amazon also offers a "Fulfillment by Amazon" service, in which it takes physical possession of third-party vendors' products and ships those products to consumers. Otherwise, the vendor itself is responsible for shipping products directly to consumers.

The listed price for the product is chosen by the third-party vendor, subject to one exception: Vendors may not charge more on Amazon than they charge in other sales channels. Nor, according to the Agreement, may third-party vendors offer inferior customer service or provide lower quality information about products than in other sales channels. To the extent that third-party vendors need to communicate with customers regarding their orders on Amazon, they must do so through the Amazon platform.

With these preliminaries completed, Amazon lists the product online and sales begin. As customers make purchases on Amazon's website, Amazon collects payment and delivers order information to the third-party vendor. At checkout, the customer can choose any shipping method offered by the third-party vendor, and any promises made by the vendor with respect to shipping date must be met. Amazon ensures compliance with this obligation by requiring the vendor to send Amazon shipping information for each order. In addition, vendors have a powerful interest in providing quality products and ensuring timely delivery, as Amazon allows shoppers to publicly rate the vendors and their products.

In exchange for its role in the transaction, Amazon collects two types of fees: one is a commission, typically between seven and fifteen percent of the overall sales price; the other is either a per-item or monthly fee, depending on the third-party vendor's preference. At least once every two weeks, Amazon remits all sales proceeds, minus fees, to the vendor. Pursuant to the Agreement, Amazon is classified as the third-party vendor's "agent for purposes of processing payments, refunds, and adjustments ... receiving and holding Sales Proceeds on your behalf, remitting Sales Proceeds to Your Bank Account, charging your Credit Card, and paying Amazon and its Affiliates amounts you owe ...." 4

Throughout each step of the sales process, Amazon may at any time cease providing any or all of the Services at its sole discretion and without notice, including suspending, prohibiting, or removing any listing. Amazon also retains other important privileges. For example, Amazon can require vendors to stop or cancel orders of any product. If Amazon determines that a vendor's actions or performance may result in risks to Amazon or third parties, it may in its sole discretion withhold any payments to the vendor. Furthermore, Amazon requires that its vendors release it and agree to indemnify, defend, and hold it harmless against any claim, loss, damage, settlement, cost, expense, or other liability.

The Dog Collar

On December 2, 2014, Heather Oberdorf logged onto Amazon's website. She typed search information for dog collars into Amazon's search terms box. She decided to purchase the dog collar at issue, which was sold by a third-party vendor, "The Furry Gang." The Furry Gang shipped the dog collar directly from Nevada to Oberdorf, who put the collar on her dog, Sadie. Then, on January 12, 2015, while Oberdorf was walking Sadie, the D-ring on the collar broke and the retractable leash recoiled into Oberdorf's eyeglasses, injuring her and permanently blinding her in her left eye.

Neither Amazon nor Oberdorf has been able to locate a representative of The Furry Gang, which has not had an active account on Amazon.com since May 2016.

Procedural History

Oberdorf filed a complaint in the United States District Court for the Middle District of Pennsylvania, bringing claims for strict product liability, negligence, breach of warranty, misrepresentation, and loss of consortium. 5 Oberdorf propounds two separate theories of strict product liability: (1) failure to provide adequate warnings regarding the use of the dog collar, and (2) defective design of the dog collar. She also asserts a variety of negligence theories, namely that Amazon was negligent in (1) distributing, inspecting, marketing, selling, and testing of the dog collar in an unreasonable manner; (2) allowing the dog collar to enter the stream of commerce in a dangerous condition; (3) failing to conduct a proper hazard analysis; (4) failing to follow the guidelines of the "safety hierarchy"; and (5) failing to provide the product with features, elements, precautions, or warnings that would have made it safer.

The District Court granted Amazon's motion for summary judgment, finding that (1) Amazon cannot be sued under Pennsylvania's strict products liability law because it does not constitute a "seller" within the meaning of Pennsylvania strict liability law, and (2) Oberdorf's claims are barred by the CDA because she seeks to hold Amazon liable for its role as the online publisher of a third party's content.

II

The District Court had jurisdiction pursuant to 28 U.S.C. § 1332 . We have jurisdiction pursuant to 28 U.S.C. § 1291 . Because our review of a district court's grant of summary judgment is plenary, we affirm only where "there is no genuine dispute as to any material fact and the movant is entitled to judgment as a matter of law." 6 In determining whether summary judgment is appropriate, we view all facts and make all reasonable inferences in favor of the non-moving party, in this case, the Oberdorfs. 7

III

We begin our analysis by addressing Amazon's contention that it is not subject to Oberdorf's strict products liability claims.

Because our subject matter jurisdiction stems from the parties' diverse citizenship, we apply Pennsylvania law in deciding whether the District Court properly dismissed Oberdorf's strict products liability claim. 8 The Pennsylvania Supreme Court has made clear that the Second Restatement of Torts § 402A applies to Pennsylvania strict products liability claims. 9 Section 402A specifically limits strict products liability to "sellers" of products. 10 Amazon relies on this limitation as its defense, claiming that it is not a "seller" because it merely provides an online marketplace for products sold by third-party vendors. We disagree. 11

A

Amazon relies heavily on the Pennsylvania Supreme Court's decision in Musser v. Vilsmeier Auction Co, Inc. 12 to support its contention that it is not a "seller." Although Musser is a significant case to which we look for guidance, it does not command the result that Amazon seeks.

The plaintiff in Musser was injured by a tractor that his father had bought at an auction house. Following his injury, he sought to hold the auction house strictly liable as a "seller" of the allegedly defective tractor. The Pennsylvania Supreme Court held that the auction house could not be considered a "seller," and thus that the plaintiff must prove that the auction house acted unreasonably (i.e., bring a negligence claim) in order to hold it liable. 13 In making this ruling, the court relied on the policy rationale articulated in comment f of § 402A of the Second Restatement of Torts:

The basis of the rule is the ancient one of the special responsibility for the safety of the public undertaken by one who enters into the business of supplying human beings with products which may endanger the safety of their persons and property, and the forced reliance upon that undertaking on the part of those who purchase such goods. This basis is lacking in the case of the ordinary individual who makes the isolated sale, and he is not liable to a third person or even to his buyer in the absence of his negligence. 14

The court noted that, when the above policy rationale "will not be served, persons whose implication in supplying products is tangential to that undertaking will not be subjected to strict liability for the harms caused by defects in the products." 15 Therefore, because "[t]he auction company merely provided a market as the agent of the seller," the court concluded that applying strict liability doctrine to the auction house would not further the doctrine's underlying policy justification. 16

In its opinion, the Pennsylvania Supreme Court made clear that courts later tasked with determining whether an actor is a "seller" should consider whether the following four factors apply:

(1) Whether the actor is the "only member of the marketing chain available to the injured plaintiff for redress";

(2) Whether "imposition of strict liability upon the [actor] serves as an incentive to safety";

(3) Whether the actor is "in a better position than the consumer to prevent the circulation of defective products"; and

(4) Whether "[t]he [actor] can distribute the cost of compensating for injuries resulting from defects by charging for it in his business, i.e., by adjustment of the rental terms." 17

We consider below each of the four factors articulated in Musser .

1

The first factor is whether Amazon "may be the only member of the marketing chain available to the injured plaintiff for redress." 18 In Musser , the court found that this factor failed to support a finding that the auction house was a "seller" because in an auction there is a vendor, for whom the auctioneer is the agent and who may be amenable to suit under § 402A for negligence or breach of warranty. 19 In other words, the plaintiff in Musser could sue the other parties in the sales distribution chain.

Amazon contends that, just as every item offered at an auction house can be traced to a seller who may be amenable to suit, every item on Amazon's website can be traced to a third-party vendor. However, Amazon fails to account for the fact that under the Agreement, third-party vendors can communicate with the customer only through Amazon. This enables third-party vendors to conceal themselves from the customer, leaving customers injured by defective products with no direct recourse to the third-party vendor. There are numerous cases in which neither Amazon nor the party injured by a defective product, sold by Amazon.com, were able to locate the product's third-party vendor or manufacturer. 20

In this case, Amazon's Vice President of Marketing Business admitted that Amazon generally takes no precautions to ensure that third-party vendors are in good standing under the laws of the country in which their business is registered. In addition, Amazon had no vetting process in place to ensure, for example, that third-party vendors were amenable to legal process. After Oberdorf was injured by the defective collar, neither she nor Amazon was able to locate The Furry Gang. As a result, Amazon now stands as the only member of the marketing chain available to the injured plaintiff for redress.

The first factor weighs in favor of imposing strict liability on Amazon. 21

2

The second factor we consider is whether "imposition of strict liability upon the [actor would] serve[ ] as an incentive to safety." 22 In Musser , the Pennsylvania Supreme Court "fail[ed] to see how the imposition of strict liability [on the auction house] would be more than a futile gesture in promoting the manufacture and distribution of safer products," chiefly because the auction house was "not in the business of designing and/or manufacturing any particular product or products." 23 Amazon asserts that it does not have a relationship with the designers or manufacturers of products offered by third-party vendors. Therefore, it contends that imposing strict liability would not be an incentive for safer products. Again, we disagree with Amazon.

Although Amazon does not have direct influence over the design and manufacture of third-party products, Amazon exerts substantial control over third-party vendors. Third-party vendors have signed on to Amazon's Agreement, which grants Amazon "the right in [its] sole discretion to ... suspend[ ], prohibit[ ], or remov[e] any [product] listing," 24 "withhold any payments" to third-party vendors, 25 "impose transaction limits," 26 and "terminate or suspend ... any Service [to a third-party-vendor] for any reason at any time." 27 Therefore, Amazon is fully capable, in its sole discretion, of removing unsafe products from its website. Imposing strict liability upon Amazon would be an incentive to do so.

The second factor favors imposing strict liability on Amazon. 28

3

The third factor we consider is whether Amazon is "in a better position than the consumer to prevent the circulation of defective products." 29

In Musser , the court indicated that the auctioneer was not in a better position than the consumer to prevent the circulation of defective products because it lacked an "ongoing relationship with the manufacturer from which some financial advantage inures to [its] benefit ...." 30 Similarly, in Nath v. National Equipment Leasing Corp. , 31 the Pennsylvania Supreme Court held that, because financing agencies perform only a "tangential" role in the sales process, "their relationship with a particular manufacturer does not, in the normal course, possess the continuity of transactions that would provide a basis for indirect influence over the condition and the safety of the product." 32 Here, while Amazon may at times lack continuous relationships with a third-party vendor, the potential for continuing sales encourages an on-going relationship between Amazon and the third-party vendors.

Moreover, Amazon is uniquely positioned to receive reports of defective products, which in turn can lead to such products being removed from circulation. Amazon's website, which Amazon in its sole discretion has the right to manage, serves as the public-facing forum for products listed by third-party vendors. In its contract with third-party vendors, Amazon already retains the ability to collect customer feedback: "We may use mechanisms that rate, or allow shoppers to rate, Your Products and your performance as a seller and Amazon may make these ratings and feedback publicly available." 33 Third-party vendors, on the other hand, are ill-equipped to fulfill this function, because Amazon specifically curtails the channels that third-party vendors may use to communicate with customers: "[Y]ou may only use tools and methods that we designate to communicate with Amazon site users regarding Your Transactions ...." 34

The third factor also weighs in favor of imposing strict liability on Amazon. 35

4

The fourth factor we consider is whether Amazon can distribute the cost of compensating for injuries resulting from defects.

In Musser , the court "acknowledge[d] that it would be possible for the auctioneer to pass on the costs of imposing strict liability upon him; possibly as [the injured plaintiff] suggests, by indemnity agreements between the auctioneer and the seller." 36 However, although the court found that extending the meaning of "seller" to include the auctioneer would provide another remedy for injured customers, the court demurred, stating that this would "only marginally" promote the "purpose of the policy considerations" underlying § 402A. 37

In this case, however, Amazon has already provided for indemnification by virtue of a provision in the Agreement:

You release us and agree to indemnify, defend, and hold harmless us, our Affiliates, and our and their respective officers, directors, employees, representatives, and agents against any claim, loss, damage, settlement, cost, expense, or other liability (including, without limitation, attorneys' fees) .... 38

Moreover, Amazon can adjust the commission-based fees that it charges to third-party vendors based on the risk that the third-party vendor presents.

Amazon's customers are particularly vulnerable in situations like the present case. Neither the Oberdorfs nor Amazon has been able to locate the third-party vendor, The Furry Gang. Conversely, had there been an incentive for Amazon to keep track of its third-party vendors, it might have done so.

The fourth factor also weighs in favor of imposing strict liability on Amazon. Thus, although the four-factor test yielded a different result when applied by the Musser court to an auction house, all four factors in this case weigh in favor of imposing strict liability on Amazon. 39

B

We do not rely exclusively upon the four-factor test to reach our conclusion that Amazon is subject to strict products liability claims for sales involving third-party vendors. Our reasoning is consistent with that in other Pennsylvania cases.

Notably, in Hoffman v. Loos & Dilworth, Inc. , 40 the Pennsylvania Superior Court decided that a sales agent was a "seller" under § 402A, and thus subject to strict product liability under Pennsylvania law. 41 Although Hoffman predates Musser , its holding remains valid, as neither Musser nor any subsequent decision by the Pennsylvania Supreme Court has called Hoffman 's holding into question. 42 Moreover, Hoffman addresses Amazon's main argument: Amazon claims that it cannot be considered a "seller" because it does not take title to or possession of the products sold by third-party vendors. The court held in Hoffman that under Pennsylvania law a participant in the sales process can be held strictly liable for injuries resulting from defective products, even if the participant does not take title or possession of those products. 43

Hoffman involved bulk sales of linseed oil. The manufacturer's sales agent, E.W. Kaufmann Co., would transmit orders for linseed oil from the packager to the distributor. That was Kaufmann's only role in the sales process. As part of the summary judgment briefing in Hoffman , Kaufmann submitted an affidavit from its principal executive, stating that it did not take title, possession, or ownership of any of the relevant linseed oil during the distribution or sales process. 44 Nonetheless, the court made clear that strict liability in Pennsylvania is properly extended "to anyone 'who enters into the business of supplying human beings with products which may endanger the safety of their persons and property.' " 45 Because Kaufmann's tasks amounted to being "in the business of selling or marketing merchandise," rather than performing a "tangential" role, it could be held strictly liable for injuries resulting from defects in that merchandise. 46

In reaching this conclusion, the court discussed two prior Pennsylvania Supreme Court cases, both of which also inform our judgment that Amazon is subject to strict liability. The first of these is Francioni v. Gibsonia Truck Corp. , 47 in which the Pennsylvania Supreme Court decided that the term "seller," as used in § 402A, does not limit strict products liability to the context of sales; that is to say, the term "seller" can also extend to lessors and bailors. The court held that strict products liability should be applied broadly to those who market products, "whether by sale, lease or bailment, for use and consumption by the public." 48

Four years later, in Nath v. National Equipment Leasing Corp. , 49 the Pennsylvania Supreme Court declined to extend the application of strict product liability to financial lessors because the financial lessor's "participation in the chain of events was tangential," in such a way that it "was not able to, nor would it have been in a position to, effect or oversee the safety of the product." 50 The core of the court's logic was that "[a] finance lessor is not in the business of selling or marketing merchandise," but rather it "is in the business of circulating funds." 51

In this case, Amazon's role extends beyond that of the Hoffman sales agent, who in exchange for a commission merely accepted orders and arranged for product shipments. Amazon not only accepts orders and arranges for product shipments, but it also exerts substantial market control over product sales by restricting product pricing, customer service, and communications with customers. 52 Amazon's involvement, in other words, resembles but also exceeds that of the sales agent labeled a "seller" in Hoffman .

At oral argument, Amazon contended that it should not be likened to a sales agent because it lists products and collects payment on behalf of various third-party vendors, whereas a sales agent typically represents a single seller or manufacturer. This is a distinction without a difference. Pennsylvania state courts have repeatedly found that large retailers who offer a range of different products are "sellers" within the meaning of § 402A. 53 Amazon is not exempted from strict products liability simply because its website offers a variety of products. 54

C

Amazon's remaining arguments similarly fail to demonstrate that it is not subject to strict product liability in Pennsylvania.

For example, Amazon asks that we look to dictionary definitions of the word "seller" for support. However, comment f to § 402A makes clear that the term "seller" is not limited by its dictionary definition, as it "applies to any manufacturer of such a product, to any wholesale or retail dealer or distributor, and to the operator of a restaurant." 55 Amazon contends that we should construe "seller" as a person who transfers a thing that she owns to another in exchange for something of value, usually money. This concept runs squarely against Pennsylvania case law that does not require an actor to possess or hold title to an item in order to be considered a "seller" for purposes of § 402A. 56

Amazon also relies heavily on non-controlling case law from jurisdictions other than Pennsylvania. However, in deciding whether Amazon is a "seller" within the meaning of § 402A, we must predict what the Pennsylvania Supreme Court would decide under Pennsylvania law interpreting the Second Restatement of Torts. 57 It is of little consequence whether Amazon is a "seller" for purposes of other states' statutes, as each of those statutory schemes is based on distinct language and policy considerations. 58

Therefore, having concluded that Amazon should be considered a "seller" under § 402A of the Second Restatement of Torts, we hold that under Pennsylvania law, Amazon is strictly liable for consumer injuries caused by defective goods purchased on Amazon.com.

IV

The second issue in this appeal is whether Oberdorf's claims, both for negligence and for strict liability, including failure to provide adequate warnings regarding the use of the dog collar, are barred by § 230 of the CDA. 59 Unlike the first issue, this is a question of federal law. We conclude that the CDA bars some, but not all, of Oberdorf's claims.

The CDA states, in relevant part, that "[n]o provider or user of an interactive computer service shall be treated as the publisher or speaker of any information provided by another information content provider." 60 This section, sometimes referred to as the CDA "safe harbor provision," 61 "precludes courts from entertaining claims that would place a computer service provider in a publisher's role, and therefore bars lawsuits seeking to hold a service provider liable for its exercise of a publisher's traditional editorial functions-such as deciding whether to publish, withdraw, postpone, or alter content." 62 The CDA is intended to allow interactive computer services companies "to perform some editing on user-generated content without thereby becoming liable for all defamatory or otherwise unlawful messages that they didn't edit or delete." 63

The CDA safe harbor provision was passed by Congress in the wake of a controversial New York state court decision allowing defamation claims to proceed against a website host. 64

The crux of Amazon's argument is that Oberdorf's negligence and strict liability claims are barred because she seeks to treat Amazon as the publisher or speaker of material provided by The Furry Gang, an information content provider. 65 Amazon contends Oberdorf's claims are essentially that Amazon should be held liable for letting The Furry Gang post the offer for the dog collar and for failing to police that offer once it was posted. Oberdorf, on the other hand, asserts that her claims do not pertain to Amazon's role in publishing third-party information but rather to its direct role in the actual sale and distribution of the defective product. That is true to a point but Oberdorf also contends that Amazon should have revised the content provided to include warnings to ensure the safe use of the dog collar. The question that we must answer is "Would such an addition to the content be part of the editorial function of the Amazon website?"

Courts throughout the country have interpreted the CDA safe harbor provision broadly. 66 Turning first to our Court, in Green v. America Online , 67 the plaintiff alleged that AOL had failed to properly police its chat rooms to prevent a third party from posting defamatory content that caused the plaintiff emotional distress. 68 We held that the CDA safe harbor provision barred the plaintiff's claims because he was "attempt[ing] to hold AOL liable for decisions relating to the monitoring, screening, and deletion of content from its network-actions quintessentially related to a publisher's role." 69

Other federal appellate courts have addressed related questions. For example, the First Circuit barred sex trafficking claims against a classified advertisement website because the allegations centered on the website's role in failing to regulate third-party content that led to the plaintiffs' injuries. 70 The Fifth Circuit barred negligence claims alleging that an online social network took insufficient precautions to prevent a fifteen-year-old teenager from lying about her age, thereby leading to her being contacted and sexually assaulted. 71 The Seventh Circuit barred housing discrimination claims against an online message board for permitting discriminatory posts. 72

These cases demonstrate that claims are precluded whenever "the duty that the plaintiff alleges the defendant violated derives from the defendant's status or conduct as a 'publisher or speaker.' " 73 Nonetheless, courts have refused to extend the scope of the CDA safe harbor provision "to immunize a party's conduct outside the realm of the Internet just because it relates to the publishing of information on the Internet." 74 In Barnes v. Yahoo!, Inc. , 75 for example, the Ninth Circuit barred the plaintiff's negligence claims against Yahoo for failing to remove posts by her ex-boyfriend containing nude photographs of her. However, the court held that the CDA safe harbor provision did not immunize Yahoo against the plaintiff's related promissory estoppel claim, which was based on Yahoo's promise to remove the injurious content rather than on any editorial function. 76

While we recognize that Amazon exercises online editorial functions, we do not agree that all of Oberdorf's claims seek to treat Amazon as the publisher or speaker of information provided by another information content provider. As previously discussed, Amazon is a "seller" of products on its website, even though the products are sourced and shipped by third-party vendors such as The Furry Gang. 77 Amazon's involvement in transactions extends beyond a mere editorial function; it plays a large role in the actual sales process. This includes receiving customer shipping information, processing customer payments, relaying funds and information to third-party vendors, and collecting the fees it charges for providing these services.

Therefore, to the extent that Oberdorf's negligence and strict liability claims rely on Amazon's role as an actor in the sales process, they are not barred by the CDA. However, to the extent that Oberdorf is alleging that Amazon failed to provide or to edit adequate warnings regarding the use of the dog collar, we conclude that that activity falls within the publisher's editorial function. That is, Amazon failed to add necessary information to content of the website. For that reason, these failure to warn claims are barred by the CDA.

Because the District Court did not parse Oberdorf's claims in order to distinguish between "failure to warn" claims and claims premised on other actions or failures in the sales or distribution processes, we will vacate its holding that Oberdorf's claims are barred by the CDA. To the extent that Oberdorf's claims rely on allegations relating to selling, inspecting, marketing, distributing, failing to test, or designing, they pertain to Amazon's direct role in the sales and distribution processes and are therefore not barred by the CDA safe harbor provision. 78 Those claims will be remanded to the District Court.

V

For the above reasons, we hold that (1) Amazon is a "seller" for purposes of § 402A of the Second Restatement of Torts and thus subject to the Pennsylvania strict products liability law, and (2) Oberdorf's claims against Amazon are not barred by § 230 of the CDA except as they rely upon a "failure to warn" theory of liability. We will therefore affirm the dismissal under the CDA of the failure to warn claims. We will vacate the remainder of the judgment of the District Court and remand this matter for further proceedings consistent with this opinion.

SCIRICA, Circuit Judge, concurring in part and dissenting in part.

This case implicates an important yet relatively uncharted area of law. No Pennsylvania court has yet examined the product liability of an online marketplace like Amazon's for sales made by third parties through its platform. Our task, as a federal court applying state law, is to predict how the Pennsylvania Supreme Court would decide the case. Berrier v. Simplicity Mfg., Inc. , 563 F.3d 38 , 45 (3d Cir. 2009). We must take special care "to apply state law and not ... to participate in an effort to change it." McKenna v. Ortho Pharm. Corp. , 622 F.2d 657 , 663 (3d Cir. 1980) (internal citation omitted). In my view, well-settled Pennsylvania products liability law precludes treating Amazon as a "seller" strictly liable for any injuries caused by the defective Furry Gang collar.

The plaintiffs weigh in detail policy reasons for allowing them to sue Amazon. Plaintiffs' theory would substantially widen what has previously been a narrow exception to the typical rule for identifying products liability defendants sufficiently within the chain of distribution. A "seller" in Pennsylvania is almost always an actor who transfers ownership from itself to the customer, something Amazon does not do for Marketplace sellers like The Furry Gang. 1 For similar reasons, every court to consider the question thus far has found Amazon Marketplace not a "seller" for products liability or other purposes; several of those courts have done so under products liability regimes similar to Pennsylvania's. 2 For these reasons, I respectfully dissent from the majority's disposition of the claims not barred by the Communications Decency Act (CDA).

I.

Amazon is a multinational technology company. Among other ventures, it hosts online sales. Products are offered for sale at Amazon.com in three primary ways. First, Amazon sources, sells, and ships some products as seller of its own goods. Second, third-party sellers sell products through Amazon Marketplace "fulfilled by Amazon," purchasing Amazon's services in storing and shipping their products. Third, at issue here, third-party sellers sell products through Amazon Marketplace without additional "fulfillment" services. These sellers, like The Furry Gang, supply and ship products directly to consumers without ever placing the items in Amazon's possession.

Amazon Marketplace has grown enormously in recent years. Over a million businesses of all sizes sell products on Amazon Marketplace, according to Amazon's own figures, and "small and medium-sized businesses selling in Amazon's stores now account for 58 percent of [Amazon's] sales." Amazon, 2019 Amazon SMB Impact Report 1, 3, https://d39w7f4ix9f5s9.cloudfront.net/61/3b/1f0c2cd24f37bd0e3794c284cd2f/2019-amazon-smb-impact-report.pdf (last accessed June 17, 2019). These businesses and their products are diverse: a recent profile of highly successful Amazon Marketplace sellers included businesses offering beauty products, indoor gardening kits, and an educational toy teaching coding. Kiri Masters, 4 Companies Founded By Millennials That Are Making Millions on Amazon , Forbes (Aug. 9, 2018), http://www.forbes.com/sites/kirimasters/2018/08/09/4-companies-founded-by-millennials-who-are-selling-millions-on-amazon. Amazon Marketplace and other online platforms give consumers the ability to buy from small and large businesses whose products they may never have encountered in an ordinary physical store.

Amazon envisions its Marketplace as an open one. It reserves the right to remove sellers' listings or terminate Marketplace services for any reason and requires sellers to represent they are in good legal standing, but it does not apply a general vetting process to all sellers to identify those who do not in fact meet that standard. Amazon also does not narrow the Marketplace's offerings by limiting the number of sellers who may offer each type of product: any number of sellers may register. In displaying products to customers, Amazon distinguishes products sold through the Marketplace from those sold directly by Amazon, identifying the seller responsible for the item in a "sold by" line placed prominently next to the price and shipping information. App. 211. The seller's name also appears on the order confirmation page, before the customer clicks "place your order" to finalize the purchase. Id. Amazon's conditions of use for customers affirm the distinction, explaining, in Amazon Marketplace purchases from third-party sellers, "you are purchasing directly from those third parties, not from Amazon. We are not responsible for examining or evaluating, and we do not warrant, the offerings of any of these businesses or individuals." Appellee's Br. 4 (citing Amazon, Conditions of Use , http://www.amazon.com/gp/help/customer/display.html/ref=footer% 20cou?ie=UTF8& nodeId=508088 (last updated May 21, 2018)).

A customer on Amazon Marketplace buys a product that has been chosen, sourced, and priced by the third-party seller. The seller contractually commits to "ensure that [it is] the seller of each of [its] Products" listed for sale. App. 164. In exchange for Amazon's services including listing the product and managing payments, the seller is charged a monthly fee, as well as a referral fee of a percentage of each sale made. The relationship reflected in the agreement between Amazon and the seller is one of "independent contractors." Id. at 270. The agreement specifically disclaims other potential relationships: it does not mean to create relationships of "partnership, joint venture, agency, franchise, sales representative, or employment," nor does it create an "exclusive relationship" constraining either Amazon or the third-party seller from other sales relationships. Id.

For each new listing a seller creates, the seller identifies the product it plans to sell, designates a price, and writes a product description. Amazon requires the description include certain minimum information about the product, and requires the seller to offer a price as favorable as the one it offers in any other sales channels. Amazon also automatically formats the information provided into a product listing page matching others on the Marketplace, and it sometimes modifies listings to streamline user experience: for example, by grouping together the pages of multiple sellers who offer the identical product sourced in different ways. The seller may choose to offer, or not to offer, a warranty on its product. Regardless, the agreement makes the seller "responsible for any nonconformity or defect in, or any public or private recall of, any of [its] products." Id. at 173.

II.

When a product is sold on Amazon Marketplace, the third-party seller offering the product for sale, as well as the product's original manufacturer, may each be sued in product liability if the product is defective. This case raises the question whether Amazon, too, can be liable as a "seller" of such a product, because of the assistance it gives to the third-party seller that provided the product to the customer. Plaintiffs answer this question in part by invoking a set of four policy factors laid out in Francioni v. Gibsonia Truck Corp. , 472 Pa. 362 , 372 A.2d 736 , 739 (1977). For the reasons I discuss in Part III, the Francioni factors should come second to an analysis under Pennsylvania law of the defendant's role in supplying the product. See Cafazzo v. Cent. Med. Health Servs. , 542 Pa. 526 , 668 A.2d 521 , 523 (1995). I begin, instead, with the definition of "seller" under Pennsylvania law.

A.

A seller under Pennsylvania product liability law is one "engaged in the business of selling ... a product." Id. at 523 (quoting Restatement (Second) of Torts § 402A(1)(a) (Am. Law Inst. 1965) ). In nearly all cases, "selling" entails something Amazon does not do for Marketplace products: transferring ownership, or a different kind of legal right to possession, from the seller to the customer. Thus, in Pennsylvania, sellers include traditional wholesalers and retailers, as well as those who supply a product through a transaction other than a sale. See, e.g. , Chelton v. Keystone Oilfield Supply Co. , 777 F. Supp. 1252 , 1256 (W.D. Pa. 1991) (wholesaler); Burch v. Sears, Roebuck & Co. , 320 Pa.Super. 444 , 467 A.2d 615 , 618 (1983) (retailer); Francioni , 372 A.2d at 739-40 (lessor); Villari v. Terminix Int'l, Inc. , 663 F. Supp. 727 , 730-31 (E.D. Pa. 1987) (pest control company supplying insecticide as part of service). Though varied, each of these cases holds liable a "seller" who transferred the right to possess the product from itself to the customer.

Transfer of a right to possession is so typical to "sellers" that exceptions are rare. There is one primary exception in Pennsylvania caselaw: the "manufacturer's representative." See Hoffman v. Loos & Dilworth, Inc. , 307 Pa.Super. 131 , 452 A.2d 1349 , 1351 (1982). Hoffman reversed a grant of summary judgment for the defendant and held a "manufacturer's representative" could be strictly liable. Id. at 1354-55. A manufacturer's representative, also termed "sales agent" or "manufacturer's agent," is a salesperson who helps a manufacturer expand sales by representing a product, usually in a particular region for a period of time, promoting the product to retailers or directly to customers. 3 After evaluating this uniquely involved retail role, the Superior Court of Pennsylvania concluded it was much more than "tangential," and the sales agent could be held liable as a seller. Id . at 1354.

But in no other scenario has a Pennsylvania court imposed "seller" liability on a defendant whose role in the sale did not include transferring ownership or possession of the product. For example, the Pennsylvania Supreme Court considered and rejected "seller" liability for an auctioneer who "never owned, operated or controlled the equipment which was to be auctioned." Musser v. Vilsmeier Auction Co. , 522 Pa. 367 , 562 A.2d 279 , 279 (1989). As the court explained, in auctioning off a product owned and controlled by the third-party seller, "[t]he auction company merely provided a market as the agent of the seller. ... Selection of the products was accomplished by the bidders, on their own initiative and without warranties by the auction company." Id. at 282. The auction company's significant role in assisting the sale was nonetheless "tangential" to the core of the transaction, the exchange between the buyer and third-party seller. Id . Similarly "tangential" was the role of the financer of a sale, who, although technically temporary owner and lessor of the supplied product, participated only by "offering the use of money." Nath v. Nat'l Equip. Leasing Corp. , 497 Pa. 126 , 439 A.2d 633 , 636 (1981) (quoting Francioni , 372 A.2d at 740 n.3 ). Pennsylvania courts would not hold liable as sellers such tangential actors as shopping malls renting space to retailers, credit card companies that enable sales transactions, or newspapers or websites hosting classified ads. See generally 3 Summ. Pa. Jur. 2d Torts § 41:61 (2d ed. 1998).

Amazon Marketplace, like the auctioneer in Musser , takes an important part in assisting sales, but is "tangential" to the actual exchange between customer and third-party seller. 562 A.2d at 282 . Like an auctioneer, Amazon Marketplace provides the "means of marketing" to a third-party seller who accomplished the "fact of marketing" when it "chose the products and exposed them for sale." Id . (citing Francioni , 372 A.2d at 738 ). Amazon Marketplace's services to any individual seller for an individual product are not "undertaken specifically," but rather, as with the auctioneer, provided on essentially similar terms to a large catalogue of sellers. Id. ; see id . at 282 n.3. And like an auctioneer, Amazon Marketplace never owns, operates, or controls the product when it assists in a sale. See id. at 279.

Amazon Marketplace's similarities to the auctioneer emphasize it has little in common with Hoffman 's manufacturer's representative, the only kind of "seller" held liable despite not having made a transfer of ownership or possession rights. Hoffman , 452 A.2d at 1354 . Amazon Marketplace does not offer the co-strategizing relationship promised by manufacturers' representatives. Amazon Marketplace is not an outsourced sales force working with individual manufacturers to boost sales: it offers a marketing platform, and it is up to the third-party seller to make best use of the platform to maximize sales.

Under Pennsylvania law, then, Amazon was not the seller of The Furry Gang's product and therefore is not liable for any product defect. Because established Pennsylvania law precludes holding Amazon strictly liable here, I respectfully dissent.

B.

While Pennsylvania law alone dictates this outcome, it is reinforced by "analogous decisions" and "other reliable data." McKenna , 622 F.2d at 663 . Pennsylvania courts have yet to consider whether Amazon is strictly liable for defective products sold through its Marketplace. So just like a Pennsylvania court would, I consider relevant decisions from other jurisdictions. These sources, including two federal appellate decisions so far, confirm what Pennsylvania law already makes clear. Amazon's role in assisting a product's sale does not make it that product's "seller."

The Sixth Circuit found Amazon Marketplace was not a "seller" after consulting a variety of sources to clarify the term's expansive but ambiguous meaning within the Tennessee Products Liability Act. Fox v. Amazon.com, Inc. , 930 F.3d 415 , 422-25 (6th Cir. 2019). The Tennessee statute defines "seller" as "any individual or entity engaged in the business of selling a product," including a "retailer," "wholesaler," "distributor," "lessor," or "bailor." Tenn. Code Ann. § 29-28-102 (7) (West 2012). Considering and rejecting a more "limited construction" proposed by Amazon, Fox , 930 F.3d at 422-23 , the court interpreted this definition to include not only those who transfer title, but "any individual regularly engaged in exercising sufficient control over a product in connection with its sale, lease, or bailment, for livelihood or gain," id. at 425 . The court nonetheless found Amazon Marketplace did not fall within even this more expansive definition. The court held Amazon did not exercise sufficient control to be the product's "seller" because Amazon "did not choose to offer the [product] for sale, did not set the price of the [product], and did not make any representations about the safety or specifications of the [product] on its marketplace." Id. at 425 . It noted that, as in this case, Amazon did not fulfill the product and therefore never possessed it or shipped it to the customer. Id .

Even where Amazon Marketplace did fulfill the product at issue, the Fourth Circuit held Amazon was not the product's "seller" under Maryland common law. Erie Ins. Co. v. Amazon.com, Inc. , 925 F.3d 135 , 144 (4th Cir. May 22, 2019). 4 In that case, unlike this one, Amazon "fulfilled" the product by storing it prior to sale then shipping it to the customer. The court held Amazon Marketplace was not a seller because, despite its role in "fulfilling" the sale, it never performed the basic act of sale: it did not "transfer title to purchasers of [the product] for a price." Id. at 141 . The court explained, relying in part on Pennsylvania precedent, those who "own ... the products during the chain of distribution are sellers," while those who "render services to facilitate that distribution or sale[ ] are not sellers." Id. (citing Musser , 562 A.2d at 283 ). Amazon, because it only rendered services and did not transfer ownership, was not a seller. Id. at 144. 5 Here, Amazon played an even more limited role: it neither stored nor shipped the product.

Other federal courts have reached the same outcome. Courts have declined to treat Amazon Marketplace as a "sales agent," as plaintiffs ask us to do. In dismissing a products liability lawsuit against Amazon essentially identical to this one, the District Court for the Southern District of New York considered a New York case that, like Pennsylvania's Hoffman decision, held a sales agent liable as a "seller." Eberhart v. Amazon.com, Inc. , 325 F. Supp. 3d 393 , 398-99 (S.D.N.Y. 2018) (citing Brumbaugh v. CEJJ, Inc. , 152 A.D.2d 69 , 547 N.Y.S.2d 699 , 700-01 (3d Dep't 1989) ); see Hoffman , 452 A.2d at 1354 . The court found Brumbaugh exceptional and of no help in establishing liability against Amazon. Eberhart , 325 F. Supp. 3d at 398-99 . It reasoned, among New York cases holding defendants liable as sellers, "the vast majority of opinions" involve a defendant who did "at some point, own the defective product." Id . at 398. And while Brumbaugh 's sales agent, given its exclusive role in connecting the manufacturer with local distributors, was "a mandatory link in [the] distributive chain," id. at 399 (quoting Brumbaugh , 547 N.Y.S.2d at 701 ), the court found Amazon Marketplace's role in merely "facilitating purchases" did not rise to this level. Id.

The Northern District of Illinois adopted similar reasoning in rejecting Amazon's liability. Garber v. Amazon.com, Inc. , 380 F.Supp.3d 766 , 2019 WL 1437877 , at *7 (N.D. Ill. Mar. 31, 2019). The court identified and analyzed the few Illinois cases in which parties who never owned the product were nonetheless held liable. It found Amazon Marketplace resembled neither a manufacturer's representative, nor a broker who made direct agreements with customers to sell the product and held exclusive rights to do so. Id. at 776-80, 2019 WL 1437877 , at *7-9 (discussing Hammond v. N. Am. Asbestos Corp. , 97 Ill.2d 195 , 73 Ill.Dec. 350 , 454 N.E.2d 210 , 216 (1983), and Bittler v. White & Co. , 203 Ill.App.3d 26 , 148 Ill.Dec. 382 , 560 N.E.2d 979 , 981 (1990) ). Amazon was not liable for third-party sales because "[the third-party seller], not Amazon, owned the [product], sourced it and listed it for sale on Amazon's marketplace, and sold and shipped it directly to the [customers]." Id. at 776-77, 2019 WL 1437877 , at *7. Moreover, unlike the more involved defendants held to be sellers, Amazon's " 'major role' was providing a venue and marketplace for third-party sellers ... to connect with buyers." Id . at 778, 2019 WL 1437877 , at *8. 6 This consensus among courts analyzing common law analogous to Pennsylvania's confirms that Amazon's limited role in Marketplace sales does not make it a "seller" liable for defective products.

The Third Restatement of Torts' § 20, defining "seller," captures state law trends in Pennsylvania, common also to New York, Illinois, and other states, as federal courts considering product liability suits against Amazon Marketplace have noted. See Restatement (Third) of Torts: Prod. Liab. § 20 (Am. Law Inst. 1998) ; see also, e.g. , Eberhart , 325 F. Supp. 3d at 398 & n.4 (finding its conclusion against product liability for Amazon Marketplace "reinforced" by § 20, though New York had not adopted the provision). Although the Supreme Court of Pennsylvania has not yet had occasion to consider § 20, the provision's incremental clarification of existing law, fully consistent with Pennsylvania case law, is the kind of guidance the Pennsylvania Supreme Court would likely find informative.

In explaining the parameters of seller liability, § 20 reinforces and builds on the Second Restatement's analogous definition. Under the Second Restatement, a seller is one who is "engaged in the business of selling products for use or consumption." Restatement (Second) of Torts § 402A cmt. f. This definition distinguishes the non-liable occasional seller from the liable regular seller, but it offers no help in determining what kind of involvement in a sale might be considered selling. The Third Restatement analyzes intervening case law to specify that a seller is one who "transfers ownership ... either for use or consumption," or, alternately, one who "otherwise distributes a product" by providing that product "in a commercial transaction other than a sale." Restatement (Third) of Torts: Prod. Liab. § 20. 7 Amazon Marketplace does not transfer ownership of third-party products, so it is not a seller. In its commentary, the Third Restatement also addresses the situation of businesses participating in a sale who do not themselves do the work of transferring title. These actors are "product distribution facilitators," meaning those "[p]ersons assisting or providing services to product distributors, while indirectly facilitating the commercial distribution of products." Id. cmt. g. In general, product distribution facilitators are not liable as sellers. Id. reporter's note g (citing Musser , 562 A.2d at 283 ). Amazon Marketplace fits the description of a product distribution facilitator, so it would not qualify as a seller.

The Third Restatement includes the Hoffman exception to the general rule for product distribution facilitators: exclusive sales representatives or exclusive agents are often considered sellers because the agent "trafficks intimately in [the products]," Brumbaugh , 547 N.Y.S.2d at 701 , and is "bound by its exclusive sales representative contract ... to promote the sale of [the] products," Bittler , 148 Ill.Dec. 382 , 560 N.E.2d at 982 . See Restatement (Third) of Torts: Prod. Liab. § 20 reporter's note g; see also Hoffman , 452 A.2d at 1354 . Moreover, where the agent's franchise to sell the product is exclusive, meaning the agent is entitled to coordinate all sales of the product for a period of time in a particular region, then the agent is "a mandatory link in [the] distributive chain," cementing the rationale for its liability. Brumbaugh , 547 N.Y.S.2d at 701 . 8 As noted, see supra pp. 157-58, Amazon Marketplace clearly does not fit this description.

The Pennsylvania Supreme Court explained in Tincher v. Omego Flex, Inc. that it would "adopt[ ] ... sections of a restatement ... if the cause of action and its contours are consistent with the nature of the tort and Pennsylvania's traditional common law formulation." 628 Pa. 296 , 104 A.3d 328 , 354 (2014) (internal quotation mark omitted). 9 Pennsylvania's approach is to evaluate "sections of a restatement" rather than adopting or rejecting the entire document. Id. at 339 . In excluding most sales facilitators from products liability but making an exception for sales agents, Pennsylvania case law matches § 20 of the Third Restatement, meeting Tincher 's standard. Under Pennsylvania law, as in § 20, an actor who assists a sale, but does not directly transfer ownership or possession, is in nearly all cases not a "seller" of the product. Compare 3 Summ. Pa. Jur. 2d Torts § 41:60 -61, with Restatement (Third) of Torts: Prod. Liab. § 20 reporter's note g. Under Pennsylvania law, as in § 20, one kind of sales facilitator may nonetheless be a "seller": a sales agent, who personally represents the manufacturer in advocating for the product's sale, and may have an exclusive right to facilitate the product's sale for a period of time within a geographic area. Hoffman , 452 A.2d at 1354-55 ; Restatement (Third) of Torts: Prod. Liab. § 20 reporter's note g. Because § 20 is "consistent with the nature of the tort and Pennsylvania's traditional common law formulation," it is likely the provision would inform the Pennsylvania Supreme Court's analysis of the question at issue in this case. Tincher , 104 A.3d at 354 .

In Pennsylvania, as in states across the country including New York and Illinois, a business assisting a sale is not a "seller" for products liability purposes unless it takes on the particularly involved retail relationship of sales agent/manufacturer's representative. 10 The Third Restatement's § 20 tracks this pattern. Like every federal court to consider this issue so far, I would find Amazon Marketplace not a seller.

III.

To deem Amazon a "seller" of Marketplace products, plaintiffs rely in large part on a four-factor policy test first articulated in Francioni v. Gibsonia Truck Corp. , 472 Pa. 362 , 372 A.2d 736 , 739 (1977). More recently, though, the Pennsylvania Supreme Court has clarified the Francioni test's purpose and limited applicability, making it evident the test has no role in answering the question at issue here. See Cafazzo v. Cent. Med. Health Servs. , 542 Pa. 526 , 668 A.2d 521 , 525 (1995).

A.

As the Pennsylvania Supreme Court has explained, the Francioni test guides "whether a particular supplier of products, whose status as a supplier is already determined, is to be held liable for damages caused by defects in the products supplied." Id. at 525 . Cafazzo makes clear the Francioni test should not be used to determine whether, in the first place, a particular defendant is a supplier of the product: that question is a "precondition necessary for application of this analysis." Id. That question must be answered by considering "what is being done" by the defendant, and whether the activity constitutes supplying products. Id. at 524 .

No one disputed that the Francioni defendant, who was a lessor of hauling equipment, had supplied the equipment that caused the plaintiff's injury. 372 A.2d at 737 . The issue, instead, was whether supplying equipment via a lease made the defendant a "seller," even though the transaction was not technically a sale. Id. The Pennsylvania Supreme Court identified four policy factors which, in other jurisdictions, had served to justify including lessors as "sellers" for products liability purposes:

(1) In some instances the lessor, like the seller, may be the only member of the marketing chain available to the injured plaintiff for redress; (2) As in the case of the seller, imposition of strict liability upon the lessor serves as an incentive to safety; (3) The lessor will be in a better position than the consumer to prevent the circulation of defective products; and (4) The lessor can distribute the cost of compensating for injuries resulting from defects by charging for it in his business, [i].e., by adjustment of the rental terms.

Id. at 739.

Because the equipment leasing company was in the business of supplying products via lease, and because of the four identified policy factors, the court held the equipment leasing company should be considered a "seller." Id. at 739-40. Similarly, though reaching the opposite outcome, the Pennsylvania Supreme Court applied the Francioni test to a pharmacy defendant, holding, although the pharmacy had supplied drugs to patients, policy factors would counsel against imposing strict product liability on the pharmacy. Coyle v. Richardson-Merrell, Inc. , 526 Pa. 208 , 584 A.2d 1383 , 1387 (1991).

The Pennsylvania Supreme Court has also cited the Francioni test as additional support for its holding, after determining that a potential defendant was not a "supplier." See Musser , 562 A.2d at 281 ; Nath , 439 A.2d at 634 . In each case, the court analyzed the activities of the defendant, and found them too "tangential" to the actual sale of the product to support seller liability. Musser , 562 A.2d at 282 ; Nath , 439 A.2d at 636 . The court then applied the Francioni factors and found, in each case, policy considerations did not support seller liability, either. Musser , 562 A.2d at 282-83 ; Nath , 439 A.2d at 635-36 .

Cafazzo clarifies the relationship between the two holdings present in each of these cases, establishing that policy factors alone cannot create seller liability. In Cafazzo , the court held a hospital and physician were not suppliers of a medical device they provided to a patient, accompanied by a surcharge for the device, because "the relationship of hospital and/or doctor to patients is not dictated by the distribution of such products, even if there is some surcharge on the price of the product." 668 A.2d at 524 .

As Cafazzo makes evident, once a court has determined a defendant is too "tangential" to be considered a supplier of the product at issue, applying the Francioni policy factors is unnecessary. Id. at 523-24 . Courts may nonetheless discuss them in order to demonstrate that, "even assuming that [the defendants] could reasonably be termed sellers ... the policy reasons for strict liability are not present." Id. at 525 . For the reasons already discussed, Amazon Marketplace is too tangential to third-party sales of products to be considered a supplier or seller of those products under Pennsylvania law. The Francioni policy factors therefore cannot establish seller liability.

B.

Even if Amazon Marketplace could be considered a supplier, application of the Francioni factors would not result in liability. Indeed, the policy outcomes produced by liability for Amazon Marketplace closely resemble those produced by liability for an auctioneer, as in Musser . See 562 A.2d at 282-83 . Thus, for reasons very similar to those identified in Musser , the Francioni policy factors do not support strict product liability for Amazon Marketplace.

The first factor, availability of other members of the distribution chain, weighs in Amazon's favor, just as in Musser . The Musser court found the first factor did not support liability for the auctioneer because, "in an auction there is a seller, who is served by the auctioneer." Id. at 282 (footnote omitted). Whereas in Francioni , the defendant lessor leased the product directly to the customer and the court identified no other member of the marketing chain, see 372 A.2d at 739 , all Amazon Marketplace products are sold by third-party sellers who are available to be sued. That seller may be defunct, insolvent, or impossible to locate by the time of suit, just as the seller of an auctioned product may be. But as the Pennsylvania Supreme Court noted in applying this factor, "[t]o assign liability for no reason other than the ability to pay damages is inconsistent with our jurisprudence." Cafazzo , 668 A.2d at 526 . 11

The second two factors, the potential incentive to safety and the defendant's relative ability to prevent circulation of the products, weigh in favor of Amazon, as they did in Musser . In Musser , the court considered the auctioneer's current business model, finding the auctioneer was "not in the business of designing and/or manufacturing any particular product," nor did the auctioneer attempt to create the kind of "ongoing relationship" with any of its large catalogue of sellers "which might equip the auctioneer to influence the manufacturing process." 562 A.2d at 282 . The auctioneer was not the kind of seller who makes it "his business to know the product he sells." Id. at 283. Similarly, Amazon Marketplace is "not in the business" of choosing, monitoring, or influencing third-party sellers' products or their manufacturing processes. Id. 12 Rather, the current model of Amazon Marketplace is an open one. All sellers meeting Amazon's terms may offer their products, and the same general terms apply to all. Although Amazon reserves the right to eject sellers, the company does not undertake to curate its selection of products, nor generally to police them for dangerousness. As it operates now, Amazon Marketplace does not exercise, relative to the consumer, any greater "influence in the manufacture of safer products." Id. Though it is possible to envision, as plaintiffs do, a new model for Amazon Marketplace in which the company researches products for potential defects and polices sellers to ensure they do not offer them, such a model would be fundamentally different from the Amazon Marketplace that exists now. Indeed, nearly any sales facilitator, including an auctioneer, could transform itself by creating a system to research sellers, and excluding those deemed unfit, instead of serving all comers meeting its terms. This kind of transformation, though, would have costs as well as benefits, for small entrepreneurs who might be excluded as too risky, and for consumers whose access to all goods would likely be reduced with greater scrutiny of sellers. Under Musser , Pennsylvania products liability law does not demand a sales facilitator enter a fundamentally new business model simply because it could.

Just as in Musser , the final factor, Amazon's ability to pass on the costs of product liability suits by charging all sellers more, is the only one weighing in favor of imposing liability. But because a variety of market participants could take on this insurer role if they chose, even if only peripherally involved in the sale, the fourth factor is not determinative. As the Musser court explained, where the other three policy factors are not present, imposing strict liability "would be related to the purpose of the policy considerations underlying the [ Second Restatement of Torts § 402A ] only marginally." Id.

Because Amazon Marketplace is not a "supplier" of third-party products, our inquiry must end there under Cafazzo . 668 A.2d at 525 . Even were we to apply the Francioni policy test, though, its four policy factors also counsel against liability.

IV.

For these reasons, I respectfully dissent from the majority opinion as to plaintiffs' products liability claims not barred by § 230 of the CDA and would affirm the District Court's dismissal of those claims. 13

MEMORANDUM AND ORDER

KLEIN, United States Magistrate Judge.

Plaintiffs filed this product liability, personal injury and wrongful death case against defendants, alleging that Ryan Ehlis’s (“Ehlis”) ingestion of Adderall®, a prescription pharmaceutical, was responsible for Ehlis’s actions in taking the life- of his five week-old daughter Tyra. A hearing on the parties’ cross motions for summary judgment was held on August 13, 2002. At the time of the hearing, the court took the matter under advisement. Approximately two weeks later a telephone conference was held for the purpose of scheduling. At that time the court advised the parties of its inclination to grant defendants’ motion for summary judgment, dismissing the action entirely, obviating the need for trial. That inclination was solidified by further review and research, culminating in this order granting defendants’ Motion for Summary Judgment.

Brief Factual Background

Ehlis was a student at the University of North Dakota. Because he was experiencing some difficulty with a class, Ehlis sought the’assistance of Dr. Thomas Peterson, a licensed psychiatrist. Ehlis informed Dr. Peterson during an office visit that as a child he was diagnosed with Attention Deficit Hyperactivity Disorder (ADHD) and had taken Ritalin. After an approximately 45 minute office visit, Dr. Peterson confirmed the diagnosis and prescribed Adderall®. Adderall® is the brand name of a Food and Drug Administration (“FDA”) approved pharmaceutical containing amphetamine salts, manufactured and marketed for the treatment of *1191 ADHD in children and narcolepsy in adults.

Ehlis took his first dose of Adderall® shortly after receiving the prescription. Apparently he took the second dose about four hours later. Ehlis testified that he took the dosage prescribed by Dr. Peterson on two subsequent days, but reduced the dosage because of the “strong” effect the medication had on him. Ehlis further stated that he took no medication on the weekend, and felt normal. Ehlis then testified that he resumed taking the medication as prescribed, until Friday morning of the following week, when he ingested the remaining pills of the 30 day prescription. His significant other and mother of his children, Angie Moreno (“Moreno”), testified somewhat to the contrary as to the amount of medication Ehlis was taking at any given time. Further, she told police that Ehlis was not acting himself the very first day he took Adderall®. She also testified that Ehlis would wake up frightened and she would give • him his Adderall® to calm him down. At no time did either Ehlis or Moreno contact Dr. Peterson to discuss with him the symptoms Ehlis was experiencing from the medication.

Ehlis testified that he began experiencing delusions and hallucinations. He also claims he had several “out-of-body” experiences while taking the medication, including talking to his dead grandfather and talking with God. Ehlis claims that he shot his daughter, and then turned the gun on himself, on direct orders from God. Ehlis was charged with the murder of his daughter, but the charges were dismissed upon confirmation that Ehlis suffered from an “Amphetamine-Induced Psychotic Disorder” and did not have the requisite criminal responsibility.

Plaintiffs commenced this lawsuit against the manufacturers of Adderall®, alleging defendants knew that the drug Adderall® sometimes induces psychosis and failed to properly and adequately warn of the risks associated with its ingestion. In addition, plaintiffs assert that defendants engaged in illegal marketing and advertising of the drug, and promoted the drug containing hidden warnings.

Motion For Reconsideration of Order Granting Amendment

Defendants seek clarification of this court’s order dated June 5, 2002 granting plaintiffs’ motion to amend their complaint to add a claim of strict liability under §§ 519 and 520 of the Restatement (Second) of Torts, asserting that defendants are strictly liable because they engaged in an abnormally dangerous activity. The court ruled in favor of the plaintiffs, citing to Freeman v. Hoffman-La Roche, Inc., 260 Neb. 552, 618 N.W.2d 827, 834-35 (2000), which contains an extensive discussion of the applicability of comment k to § 402 of the Restatement (Second). Defendants respectfully assert the court misconstrued the plaintiffs’ claims and that the court’s opinion addresses unavoidably unsafe products under the Restatement (Second) of Torts § 402A, comment k, whereas the plaintiffs’ motion to amend addresses abnormally dangerous activities under §§ 519 and 520 of the Restatement. The defendants further assert that several courts have held these sections inapplicable to the manufacture and sale of pharmaceutical products, Reeder v. Hammond, 125 Mich.App. 223, 336 N.W.2d 3 (1983); Gaston v. Hunter, 121 Ariz. 33, 588 P.2d 326 (1978), although North Dakota has neither specifically adopted nor rejected the provisions. See Wirth v. Mayrath Industries, Inc., 278 N.W.2d 789 (N.D.1979). Plaintiffs admit as much, stating, “Sections 519-520 of the Restatement (Second) have been adopted by many jurisdictions, and the best Erie prediction that can be made in the case at hand is that such sections *1192 are also applicable in North Dakota.” However, plaintiffs fail to cite to-any cases which have specifically held §§ 519 and 520 are applicable to the sale and manufacture of pharmaceutical drugs.

The court has again reviewed the parties’ pleadings, their submission with respect to plaintiffs’ motion to amend, and the Freemen decision in its entirety. After careful consideration, the court is prepared to rescind its prior order granting the amendment permitting plaintiffs to assert a claim pursuant to §§ 519 and 520 of the Restatement (Second) of Torts. Plaintiffs’ original complaint contained a claim for strict liability under § 402A. This court’s order did nothing more than affirm plaintiffs right to proceed under that section, and suggested that comment k provides an affirmative defense rather than provides absolute immunity, which has been asserted by defendants. Further, defendants are essentially correct that the court’s June 5 order does not specifically address whether §§ 519 and 520 form a basis for liability in this case. It is to this issue the court now turns.

Plaintiffs’ amendment seeks to assert a claim that defendants are strictly liable for engaging in an “abnormally dangerous” or “ultrahazardous” activity pursuant §§ 519 and 520 of the Restatement (Second) of Torts. Section 519 provides:

§ 519. General Principle.

(1) One who carries on an abnormally dangerous activity is subject to liability for harm to the person, land or chattels of another resulting from the activity, although he has exercised the utmost care to prevent the harm.

(2) This strict liability is limited to the kind of harm, the possibility of which makes the activity abnormally dangerous.

Section 520 contains the factors to be considered in determining whether an activity is abnormally dangerous. This section provides:

In determining whether an activity is abnormally dangerous, the following factors are to be considered:

(a)Existence of a high degree of risk harm to persons, land or chattels of others;

(b) Likelihood that the harm that results from it will be great;

(c) Inability to eliminate the risk by the exercise of reasonable care;

(d) Extent to which the activity is-not a matter of common usage;

(e) Inappropriateness of the activity to the place where it is carried on; and

(f) Extent to which its value to the community is outweighed by its dangerous attributes.

As the court recognized in Progar v. The Washington Hospital, there are a multiplicity of- situations in which absolute liability has been found to apply, ranging from blasting, an oil drilling situation not unlike dynamite blasting, dam breaking, the storage of explosives, and even against “one who suffered his baboon to escape.” 49 Pa. D. & C.2d 485, 488-89 (1970). These types of cases all have a common thread. However, the application of absolute liability to the realm of pharmaceutical products is glaringly absent, and has specifically been refused in its application to the production of new drugs. See Gaston v. Hunter, 121 Ariz. 33, 48, 588 P.2d 326, 341 (1978) (The rules relating to “ultrahazardous” or “abnormally dangerous” activities are inapplicable to the “production of new drugs”).

In Reeder v. Hammond, 125 Mich.App. 223, 336 N.W.2d 3, 6 (1983), cited by the defendants, plaintiffs appealed the trial court’s denial of their motion to file a second amended complaint to add a new theory of liability (something defendants *1193 argue plaintiffs are now attempting to do in this case). In Reeder, plaintiffs believed that if they could prove that the manufacture of the pharmaceutical Biphet-amine and birth control pills constitutes an ultrahazardous activity, “then defendants owe a duty to the consumer to warn of any risks associated with the use of their products.” The appellate court held, given the Michigan Supreme Court’s decision in Smith v. ER Squibb & Sons, Inc., 405 Mich. 79, 88, 273 N.W.2d 476, 479 (1979), the district court did not err in refusing the amendment “since the proffered theory of liability was without merit.” Reeder, 125 Mich.App. at 230, 336 N.W.2d at 6. The Michigan Supreme Court in Smith took the analysis one step further and adopted the learned intermediary doctrine, stating that a manufacturer of a prescription drug has a legal duty only to warn the medical professional, not the patient, of any risks inherent in the use of the drug the manufacturer knows or should know exists. Smith, 405 Mich, at 88, 273 N.W.2d at 479. On that basis the plaintiffs’ action was dismissed.

Having thus concluded that the amendment was improvidently granted, the court will now address the pending motions for summary judgment, noting that this court’s adoption of the learned intermediary doctrine supplants the plaintiffs’ motion to amend to assert a cause of action under §§ 519 and 520 in any event. Summary Judgment Standard

Summary judgment is appropriate if there is not a genuine issue of material fact and the moving party is entitled to judgment as a matter of law. Fed. R. Civ. Pro. 56(c); Celotex Corp. v. Catrett, 477 U.S. 317, 322, 106 S.Ct. 2548, 91 L.Ed.2d 265 (1986). Rule 56 of the Federal Rules of Civil Procedure “mandates the entry of summary judgment ... against a party failing to make a showing sufficient to establish the existence of an element essential to that party’s case.” Celotex, 477 U.S. at 322, 106 S.Ct. 2548. If the moving party has supported its motion for summary judgment, the nonmoving party has an affirmative burden placed on it to go beyond the pleadings and show a genuine triable'issue of fact. Commercial Union Ins. Co. v. Schmidt, 967 F.2d 270, 271 (8th Cir.1992). However, the court considering a motion for summary judgment must view the evidence in the light most favorable to the nonmoving party who enjoys “the benefit of all reasonable inferences to be drawn from the facts.” Vacca v. Viacom Broadcasting of Missouri, Inc., 875 F.2d 1337, 1339 (8th Cir.1989) (citation omitted).

Summary judgment is improper if the court finds a genuine issue of material fact; however, “the mere existence of some alleged factual dispute between the parties will not defeat an otherwise properly supported ’ motion for summary judgment. ...” Commercial Union Insurance Co. v. Schmidt, 967 F.2d 270, 271-72 (8th Cir.1992) (citation omitted). The issue is whether “the evidence is sufficient to allow a reasonable jury to return a verdict for the nonmoving party.” Landon v. Northwest Airlines, Inc., 72 F.3d 620, 624 (8th Cir.1995).

Discussion

I. Plaintiffs’ Motion for Partial Summary Judgment.

Plaintiffs assert there are no material facts in dispute and they are entitled to judgment as a matter of law. Plaintiffs contend that Shire is “at fault” pursuant to N.D. Cent.Code § 28-01.3-01; because Ehlis was unaware of the danger posed by Adderall®, the product is unreasonably dangerous. Because the product is unreasonably dangerous, Shire is strictly hable. Alternatively plaintiffs do argue the warnings as to Ehlis were insufficient, asserting that summary judgment on the failure to warn theory would also be appropriate. *1194 Plaintiffs also anticipate defendants’ learned intermediary doctrine defense, asserting the theory “applies only to the failure to warn/marketing defect theory, and has nothing to do with the over-riding concept of product ‘defect’ under other theories of liability.” Memorandum in Support of Plaintiffs’ Motion for Partial Summary Judgment, at 8 n. 6.

Defendants contend plaintiffs have not asserted the product was defectively manufactured, only that the warnings were inadequate, making the product defective. Plaintiffs have postured this case as a failure to warn case, both in terms of negligence and strict liability. However, plaintiffs then argue that “even if the labeling were the sole focus of the case, that would not exonerate Shire in this instance because Dr. Peterson’s deposition is abundantly clear that he was not aware of and did not expect Adderall® to precipitate psychosis in Ryan Ehlis at normal, prescribed doses, which it did.” Id.

In response to plaintiffs’ motion, defendants assert Plaintiffs’ Undisputed Facts are very much in dispute. Defendants dispute that the Adderall® at a prescribed dosage precipitated Ehlis’s psychosis, asserting it was an overdose that precipitated the psychosis. Defendants cite to their medical expert in support of this contention, Dr. Pliszka. Plaintiffs challenge Dr. Pliszka’s testimony and suggest a Daubert motion may be forthcoming, although they already have raised the issue in a Motion in Limine. 1 Further, defendants challenge the testimony of plaintiffs’ experts, arguing that they have never been disclosed as experts and that defendants have never been provided proper reports, a point of considerable contention in later filings. Defendants assert that the only issue is failure to warn, and that plaintiffs “have not presented any evidence whatsoever of any other theories of liability.” Defendant Shire Richwood Inc.’s Memorandum in Opposition to Plaintiffs’ Motion for Partial Summary Judgment, at 9. As such, the focus is on the adequacy of the warnings, “not on whether Adderall® caused Ryan’s psychosis.” Defendant Shire Richwood Inc.’s Memorandum in Opposition to Plaintiffs’ Motion for Partial Summary Judgment, at 9. Finally, the parties address the presumption recognized in Butz v. Werner, 438 N.W.2d 509 (N.D.1989), that had an adequate instruction been given it would have been heeded. Defendants argue the warnings as given were adequate as a matter of law. Defendants further assert there is no evidence that Dr. Peterson would have acted differently if a different warning had been given, that Dr. Peterson was aware from his residency of the potential that amphetamines may cause psychosis, and that he continues to prescribe Adderall® to this day.

Defendants argue that “because the warning on the Adderall label is adequate as a matter of law, it could not have been the proximate cause of the shootings.” Defendant Shire Richwood Inc.’s Memorandum in Opposition to Plaintiffs’ Motion for Partial Summary Judgment, at 21.

Defendants’ final argument against summary judgment in favor of plaintiffs is that “the connection between a defendant’s act or omission and the damage or injury suffered by the plaintiff may be broken by a superseding, intervening event.” Defendants allege Angie Moreno’s failure to take action when the effects of the Adderall® became apparent breaks the chain of causation and relieves Shire of any liability for plaintiffs’ injuries, ultimately defeating plaintiffs’ motion for partial summary *1195 judgment. Defendant argues that this superseding, intervening cause breaks the chain of causation. Defendant Shire Rich-wood Inc.’s Memorandum in Opposition to Plaintiffs’ Motion for Partial Summary Judgment, at 9.

This court views this action as a failure to warn case, despite plaintiffs’ attempts to argue defective product as a result of an inadequate warning. Because this court adopts the learned intermediary doctrine, and further finds that the warning was adequate as to Dr. Peterson, as more,fully discussed below, Plaintiffs’ Motion for Partial Summary Judgment must fail and is therefore DENIED.

II. Defendants’ Motion for Summary Judgment.

Defendants assert no less than five theories by which they contend they are entitled to summary judgment. Although the court finds several of the theories persuasive, including the preemption theory raised by the defendants at the hearing, the court will first address the theory on which it “hangs its hat” in granting defendants’ Motion for Summary Judgment, the learned intermediary doctrine.

A. No Duty to Warn the Ultimate User Under the Learned Intermediary Doctrine.

Plaintiffs assert not only that defendants failed to warn Ehlis of the dangers associated with the ingestion of Adderall®, but that the warnings associated with its use were inadequate as to Dr. Peterson as well. Defendants argue the learned intermediary doctrine obviates any inquiry into the adequacy of the warning to the patient and further that Dr. Peterson was well aware of the risks associated with prescribing Adderall®.

Since there are no North Dakota cases directly on point, the first line of inquiry is whether or not North Dakota would adopt the learned intermediary doctrine. 2 Counsel were specifically asked to advise the court of their positions as to whether North Dakota would adopt the learned intermediary doctrine if directly .faced with the question, or whether certification of the question to the North Dakota Supreme Court would be appropriate. The court has concluded, taking into account the parties’ arguments, the overwhelming majority of jurisdictions that have adopted the doctrine, and the fact that North Dakota has adopted § 402A of the Restatement (Second) of Torts, from which the learned intermediary doctrine derived, that North Dakota would recognize the learned intermediary, doctrine as the rule of law in cases where the adequacy of the warning as to a prescription drug is at issue. See Hill v. Searle Laboratories, 884 F.2d 1064, 1067 (8th Cir.1989) (Recognizing the existence .of the learned intermediary doctrine in Arkansas). This position is further cemented by the court’s holding in Hill wherein the Eighth Circuit Court of Appeals agreed with the district court and defendant Searle that “the Arkansas courts would join the majority of jurisdictions and adopt the defense to strict liability contained in comment k” to section 402A, recognizing the strength of Searle’s argument that “the Arkansas Supreme Court has often referred to the comments of section 402A, implicitly adopting them.” Id. at 1067. North Dakota courts also have adopted several comments to § 402A. See, e.g. Crowston v. Goodyear Tire & Rubber, 521 N.W.2d 401 (N.D.1994). Given the 1 similarity in positions on the Restatement between North Dakota and Arkansas, and the Eighth Circuit Court of Appeal’s willingness to permit the district *1196 court to interpret state law, this court is confident North Dakota would similarly adopt the learned intermediary doctrine.

The learned intermediary doctrine provides the manufacturer has a duty to warn only the physician of the risks involved in the use of the pharmaceutical, and the physician' then acts as the “learned intermediary” between the manufacturer and the patient. Kirsch v. Picker International, Inc., 753 F.2d 670, 671 (8th Cir.1985). Thus, a warning to the doctor is deemed a warning to the patient and there is no duty on the manufacturer to communicate directly with all ultimate users of the prescription drug. Id.

The second line of inquiry is whether Dr. Peterson was sufficiently aware of the risks associated with prescribing Adde-rall® to Ehlis. Courts are generally in agreement that a warning is adequate where it is reasonable under the circumstances. More specifically, courts have held that in order to be adequate the warning must satisfactorily convey the seriousness of the danger such that a reasonable physician would be alerted to the danger. 28 C.J.S., Drugs and Narcotics, § 62.

After careful review of the record, and particularly Dr. Peterson’s deposition testimony, the court is convinced there could be no other conclusion but that Dr. Peterson knew the risks of prescribing Adderall® to Ehlis. Dr. Peterson’s testimony is clear that he “appreciated that stimulants can cause psychosis as a side effect,” although rarely. Deposition of Thomas Peterson, M.D., p.p. 25-26 (hereinafter Peterson Depo.). Dr. Peterson confirmed he reviewed the Physicians Desk Reference (“PDR”) for Adderall® sometime prior to prescribing it to Ehlis and was aware of the potential side effects of Adderall®, including psychosis. Peterson Depo. p.p. 25-26. Dr. Peterson was also aware that substance induced psychosis “is within the DSM-IV.” Peterson Depo. p. 26. Dr. Peterson also testified that he understood that “some people become psychotic without actually taking an overdose of the medication” and that this understanding applied to Adderall® as well as other amphetamines. Peterson Depo. p. 28. Further, Dr. Peterson testified that he was aware that the risk of psychosis was identified in the package insert at the time he prescribed the medication to Ehlis. Peterson Depo. p. 29. When asked about the accuracy of the statement in the package insert indicating that psychotic episodes at recommended doses are rare, Dr. Peterson testified he believed that to be an accurate statement. Peterson Depo. p. 34. Finally, Dr. Peterson testified that he will continue to prescribe Adderall® in the future, believing it to be a good medicine. Peterson Depo. p. 34. 3

*1197 There is simply no question Dr. Peterson understood the risks associated with prescribing Adderall® and voluntarily proceeded with an educated course of conduct. Unfortunately the effects were disastrous beyond anyone’s comprehension, even the physician’s. However, defendants are not legally responsible for the effect, having fulfilled their duty to adequately warn the physician of the risks associated with the use of Adderall®. Defendants’ Motion for Summary Judgment is therefore GRANTED.

B. Food, Drug and Cosmetic Act Does Not Provide a Private Right of Action.

Defendants argue they are entitled to summary judgment because the Food, Drug and Cosmetic Act does not provide for a private right of action. They contend that the plaintiffs’ assertion that the labeling for Adderall® should contain a “black box” warning is an attempt to assert a private right of action under the FDCA, á claim not permitted under the Act. Plaintiffs assert they are not alleging a private right of action, but rather their claim is for negligence under North Dakota common law and strict liability as provided by statute. Plaintiffs further assert that “FDA approval is not shield to liability.” Hill v. Searle, 884 F.2d 1064, 1067 (8th Cir.1989).

The Eighth Circuit, in Brooks v. Howmedica, recently held that a state failure to warn claim was preempted by § 306k of the Medical Device Amendments (“MDA”) to the Food, Drug, and Cosmetics Act, prohibiting the plaintiff from claiming the manufacturer’s instructions were inadequate. 273 F.3d 785, 791 (8th Cir.2001). Defendants assert this preemption would apply with respect to the labeling of Adderall® as well, citing the provisions of 21 C.F.R. § 314.70(c) for the proposition that the label may not be changed without FDA approval, except in limited circumstances, and .then only temporarily if the supplement is not approved by the FDA. While the court agrees that the FDA provision “smacks” of preemption, unfortunately the determination is not nearly so “cut and dried,” as can be seen from the Brooks court’s exhaustive analysis of the Supreme Court’s decision in Medtronic, Inc. v. Lohr, 518 U.S. 470, 116 S.Ct. 2240, 135 L.Ed.2d 700 (1996) wherein the Court held that “State law will be preempted under the MDA when ‘a particular state requirement threatens to-interfere with a specific federal interest.’ ” Brooks, 273 F.3d at 793. The court in Brooks noted that “[t]he Court concluded that Lohr’s failure to warn claim was too general to require preemption,” id., but that “ ‘any state common-law action that would impose a requirement different from, or in addition to’ a specific federal requirement” would be preempted. Id. at 794 (citation omitted). In finding that Brooks’ failure to warn claim would interfere or conflict with the specific federal requirements imposed during the regulation of the medical device, the court noted the extensive regulation of the label, stating “the FDA drafted or approved every word of the Simplex label, and any changes were subject to close FDA scrutiny.” Id. at 795. This was not the ease in Medtronic. Thus, the court reasoned that “[a] jury finding of negligent fáilure to warn would be premised on the fact that the label for Simplex was not written in a particular way or did not contain certain information. This would be equivalent to a state regulation imposing specific label requirements.” Id. at 796. Further, the court noted, “The effect of a jury finding of negligent failure.to warn would be that state law would require Howmedica (defendant) to change the label and package insert for Simplex (the product), but Howmedica may- not unilaterally make such changes under federal law.” Id. at 796.

*1198 As defendants point out, plaintiffs’ claims in this case mirror those brought by Brooks, and although plaintiffs in oral argument attempted to distinguish the FDA from the MDA, the court is not convinced that preemption would not apply in this case. The FDA dictates the contents of the label for Adderall® and defendants were prohibited from changing it without prior approval from the FDA, except in limited circumstances for a limited period of time. This concept sounds in preemption, with the same rationale as adopted by the court in Brooks. Therefore, the court finds that summary judgment on this basis is also properly granted the defendants.

The final arguments asserted by defendants will be summarily addressed for the sake of completeness, given this court’s dismissal on other grounds.

C. Requisite Expert Testimony Lacking.

Defendants argue that plaintiffs have failed their burden of proof as they lack the requisite expert testimony. Defendants argue that the only expert identified by the plaintiffs, Donald H. Marks, M.D., is unable to provide the requisite testimony that the warning is inadequate, or that different labeling would have resulted in a different outcome. Defendants emphasize Dr. Marks’ statement that it is common knowledge that this class of drugs causes psychosis as support for its proposition that Dr. Marks is unable to testify that changes to the warning would have resulted in a different outcome, and that to do so would be speculative.

In their reply brief defendants tie the “lack of requisite expert testimony” to North Dakota’s rebuttable presumption against defects statute and the fact that plaintiffs cannot competently challenge the fact that the labeling for Adderall® complies with either FDA mandated labeling guidelines or industry standards.

The court will not grant defendants’ summary judgment on this basis, concluding that Dr. Marks’ testimony goes more to weight and credibility than it does to admissibility. Failure of proof based on the expert’s testimony would be a question for the jury, if this action were proceeding.

D. Failure to Rebut the Presumption Against Defects-N.D. Cent.Code § 28-01.3-09.

Defendants assert plaintiffs cannot rebut the presumption against defects contained in N.D. Cent.Code § 28-01.30-09. This section provides:

There is a rebuttable presumption that a product is free of any defects or defective condition where the plans, designs, warnings, or instructions for the product or the methods and techniques of manufacturing, inspecting, and testing the product were in conformity with government standards established for that industry or where no government standards exist then with applicable industry standards, which were in existence at the time the plans, designs, warnings, or instructions for the product of the methods and techniques of manufacturing, inspecting, and testing the product were adopted.

N.D. CentCode § 28-01.3-09 (1995).

Defendants argue that plaintiffs have offered no evidence, including the testimony of their expert, that the labeling was anything but in complete compliance with applicable FDA regulations, the government standard. Thus, defendants argue the presumption has not been rebutted and they are entitled to summary judgment in their favor. Plaintiffs argue the labeling falls below the government standard, particularly in light of the fact that there is no warning about Adderall® triggering psychosis, at normal doses, in adult patients who have no prior history of psychosis. Plaintiffs thus argue that the *1199 off-label use of Adderall® for the treatment of ADHD in adults suffices to rebut the presumption. In response, defendants argue that Adderall® is approved for prescription to adults with narcolepsy, and that the section of the warning that Adde-rall® can cause psychosis at recommended doses applies with equal force to children and adults.

The court will not grant summary judgment based on the rebuttable presumption; one that has been sufficiently challenged by raising the off-label prescription of Ad-derall® to adults. Plaintiffs have presented some evidence that defendant knew the drug was being so prescribed and had not sought FDA approval for this additional use. Thus, as to this use, the drug did not comply with the FDA standards and the presumption is thereby rebutted.

E. Injuries Suffered Were the Result of an Intervening, Superseding Cause.

Defendants argue that Angie Moreno’s inaction when she realistically should have recognized the psychosis caused by the Adderall® was a superseding cause, breaking the chain of causation. In the court’s view this is not a basis to deny liability, but may impact the extent of liability, if any. In this case there is no liability given the court’s grant of summary judgment to the defendants on other grounds.

Conclusion

For the foregoing reasons, defendants’ Motion to Clarify is GRANTED (Doc. # 68), plaintiffs’ Motion for Summary Judgment is DENIED (Doc. # 63), and defendants’ Motion for Summary Judgment is GRANTED (Doc. # 61). All other pending motions are hereby MOOT.