dissenting.
This is a case not so much about “market-share” liability as it is about the social utility of childhood vaccines that pose an infinitesimal but ever-present risk of catastrophic side effects. Like most parents, I had no idea of the potential side effects of the whooping-cough vaccine when it was administered to my own children. Knowing its possible consequences today, as an informed parent, I am certain that I would accept the risk as one in the best interests of the child. I have a sense that the product may be unavoidably unsafe. I would have no sense of outrage if, after weighing all of these risks, a child of mine suffered a severe side effect. Of course, I would welcome a no-fault remedy such as is found in the National Childhood Vaccine Injury Act, 42 U.S.C.A. §§ 300aa-1 to -34, but I might not expect the tort-liability system to compensate me or my child for the known potential of loss.
It is this intuitive feeling about the worth of the DPT vaccines, rather than a sense of injustice about the market-*192share theory itself, that drives the Court to reject a “market-share” theory of liability in cases of multiple tortfeasors. Thus, I believe the Court has inverted its priorities. It has rejected the market-share theory by addressing the “unavoidably-unsafe” issue. While the latter defense may be valid in this case, it is not appropriate to resolve that issue on this summary judgment motion. The majority has decided the market-share question on policy grounds more relevant to the question of “duty” than to the “causation” question before us today. In doing so it has deprived the plaintiffs of the opportunity to investigate and present argument concerning the disputed policy question decided by the majority. In addition, by applying the “unavoidably-unsafe” policy questions to the more general market-share question the Court has precluded recovery even for failure-to-warn and negligence theories — a result not warranted even if one were to accept the majority’s policy analysis. See Brown v. Superior Court of California, 44 Cal.3d 1049, 245 Cal.Rptr. 412, 751 P.2d 470 (1988) (holding prescription drug manufacturers not strictly liable for resulting injuries, but still subject to manufacturing defect, negligence, and failure-to-warn claims).
I
A.
What is market-share liability? And why do people say such bad things about it? At first glance, the doctrine appears to offend our notions of causation: How can you hold me responsible for something if you cannot prove I did it? Glenn O. Robinson, in his analysis of the issue, “Multiple Causation in Tort Law: Reflections on the DES Cases,” 68 Va.L.Rev. 713 (1982), suggests that the doctrine is nothing more than a familiar application of tort law’s way of dealing with multiple tortfeasors. When multiple tortfeasors have caused an injury, the plaintiff is ordinarily relieved of the burden of proving which of them has caused the injury. The usual example is the *193case of two people who have set fires, either of which, would have destroyed a farm field. Neither is excused from his or her conduct by saying “I caused you no injury. The other fire would have destroyed your farm anyway.” Restatement (Second) of Torts § 432(2) & comment d, illustrations 3, 4 (1965). Another example is the concert-of-action theory. This liability theory holds defendants liable when, in pursuing a common scheme, any of them breaches a duty owed to a third person regardless of who causes the injury. See Andreassen v. Esposito, 90 N.J.Super. 170 (App.Div.1966) (participant in illegal drag race liable for injuries caused by other driver), certif. denied, 46 N.J. 605 (1966); see W. Prosser, Law of Torts 291-93 (4th ed. 1971); Restatement (Second) of Torts § 876 (1977) (adopting the concert-of-action theory).
“Alternative liability” imposes liability on two or more actors all of whom breached a duty of care but it is not clear which party’s breach caused the injury. The most famous case involving alternative liability is Summers v. Tice, 33 Cal.2d 80, 199 P.2d 1 (1948). Because two hunters were both negligent in shooting in the plaintiff’s direction, and it could not be proven which gun was the source of the plaintiff’s eye and lip injuries, the court shifted the burden to each defendant to exculpate himself or be liable for the entire damages. Id. at 89, 199 P.2d at 5. If Summers is the classic statement of alternative liability, then the elements of the doctrine are: (1) plaintiff is unable to identify which defendant caused the injury; (2) all potentially liable injured parties are joined as defendants; (3) each defendant breached a duty of care owed to plaintiff; and (4) defendants are in a far better position to offer evidence of causation than plaintiff is.
The Restatement (Second) of Torts § 433B(3) (1965) has adopted the doctrine, saying:
Where the conduct of two or more actors is tortious, and it is proved that harm has been caused to the plaintiff by only one of them, but there is uncertainty as to which one has caused it, the burden is upon each such actor to prove that he has not caused the harm.
*194Comment h to this subsection, often quoted by litigants’ counsel, says, in relevant part, that although
[t]he cases thus far decided in which the rule * * * has been applied all have been cases in which all of the actors involved have been joined as defendants. * * * It is possible that cases may arise in which some modification of the rule stated may be necessary because of complications arising from the fact that one of the actors involved is not or cannot be joined as a defendant * * * or because of substantial differences in the character of the conduct of the actors or the risks which they have created.
See Annotation, “Liability of several persons guilty of acts one of which alone caused injury, in absence of showing as to whose act was the cause,” 5 A.L.R.2d 98 (1949).
The Summers alternative-liability principles were applied to an industry to impose “enterprise liability” in Hall v. E.I. Du Pont De Nemours & Co., Inc., 345 F.Supp. 353 (E.D.N.Y.1972). Thirteen children were injured in twelve unrelated accidents by unmarked blasting caps that the children found and detonated. Plaintiffs could not identify the particular manufacturers, so they sued “substantially the entire blasting cap industry and its trade association * * Id. at 386. That court permitted the plaintiffs to establish liability based on a variety of theories all recognizing that the development of “explicit or implicit safety standards, codes, and practices which are widely adhered to in an entire industry * * * could support a finding of joint control of risk and a shift of the burden of proving causation to the defendants” where individual defendant-manufacturers cannot be identified. Id. at 374.
In the touchstone case of “market-share” liability, Sindell v. Abbott Laboratories, 26 Cal.3d 588, 163 Cal.Rptr. 132, 607 P.2d 924, cert. denied, 449 U.S. 912, 101 S.Ct. 285, 286, 66 L.Ed.2d 140 (1980), the California Supreme Court rejected all three of the above theories of liability: (1) concert of action was unsupported by the facts of DES’s development; (2) pure alternative liability was not allowed because not all of the possibly-responsible parties were in the suit; and (3) enterprise liability did not apply because (a) the DES industry was so large, (b) unlike in Hall, supra, 345 F.Supp. 353, there was no *195allegation that the defendants had developed common safety standards through a trade association, and (c) adherence to industry standard was mandated, at least in part, by the government. Id. 26 Cal.3d at 601-610, 607 P.2d at 930-35, 163 Cal.Rptr. at 139-143.
The Sindell Court proposed a new theory. The policy reasons were clearly expressed:
In our contemporary complex industrialized society, advances in science and technology create fungible goods which may harm consumers and which cannot be traced to any specific producer. The response of the courts can be either to adhere rigidly to prior doctrine, denying recovery to those injured by such products, or to fashion remedies to meet these changing needs. * * *
The most persuasive reason for finding plaintiff states a cause of action is that advanced in Summers: as between an innocent plaintiff and negligent defendants, the latter should bear the cost of the injury. [Id. at 610-611, 607 P.2d at 936, 163 Cal.Rptr. at 144.]
The Sindell doctrine is that “[i]f plaintiff joins in the action the manufacturers of a substantial share of the [product allegedly causing injury] which her mother might have taken, * * * the burden of proof [shifts] to defendants to demonstrate that they could not have made the substance which injured plaintiff * * Id. at 611, 607 P.2d at 937, 163 Cal.Rptr. at 145. The difference from alternative liability is that the plaintiff is not required to have before the court each and every manufacturer of the product. However, by the requirement of the substantial share of the market you increase the “likelihood that this comparative handful of producers manufactured the [product] which caused plaintiff’s injuries.” Ibid, 607 P.2d at 937, 163 Cal.Rptr. at 145.
As we know, the Wisconsin Supreme Court has mapped out a variant on these theories of liability in Collins v. Eli Lilly Co., 116 Wis.2d 166, 342 N.W.2d 37, cert. denied, 469 U.S. 826, 105 S.Ct. 107, 83 L.Ed.2d 51 (1984). As in Sindell, the Wisconsin Court rejected the concert-of-action, alternative-, and enterprise-liability theories. The Wisconsin Court went on to recognize “the fundamental fairness of Sindell’s shifting the burden of proof to the defendants,” but concluded that the “unalloyed *196market share theory does not constitute the most desirable course to follow in DES cases because the theory, while conceptually attractive, is limited in practical applicability.” Id. at 188-189, 342 N.W.2d at 48. According to that court, “defining the market and apportioning market share [is] a near impossible task if it is to be done fairly and accurately,” and “a second ‘mini-trial’ to determine market share” would be a “waste of judicial resources.” Id. 188-192, 342 N.W.2d at 48-49. The court reasoned that each company “contributed to the risk of injury” even if all of them did not act in concert.
We conclude that it is better to have drug companies or consumers share the cost of the injury than to place the burden solely on the innocent plaintiff. Finally, the cost of damages awards will act as an incentive for drug companies to test adequately the drugs they place on the market for general medical use. This incentive is especially important in the case of mass-marketed drugs because consumers and their physicians in most instances rely upon advice given by the supplier and the scientific community and, consequently, are virtually helpless to protect themselves from serious injuries caused by deleterious drugs. [Id. at 190-194, 342 N.W.2d at 49-50 (footnote omitted).]
Therefore, the Wisconsin Court ruled that plaintiff could proceed initially against one defendant and must prove the following: (1) the plaintiff’s mother took the product; (2) the product caused the plaintiff’s injury; (3) the defendant-producer marketed the type of product that claimant took, i.e., the color, shape, size, etc.; and (4) the manufacturer’s conduct in producing the product constituted beach of duty to the plaintiff. Id. at 193-194, 342 N.W.2d at 50.
Plaintiff can recover all damages from the one defendant. Ibid, 342 N.W.2d at 50. Plaintiff may, of course, sue as many companies as may be possibly liable and the companies may implead others. Id. at 194-196, 342 N.W.2d at 51. Once the prima facie case is proven, the burden of proof shifts to the defendant to prove it did not produce or market the DES at the relevant time or in the relevant market. Id. at 197-198, 342 N.W.2d at 52. It is up to each to establish that its DES could not have reached plaintiff’s mother. Assigning liability and apportioning damages is then done under the Wisconsin comparative-negligence statute. Id. at 197-199, 342 N.W.2d at 52-53.
*197In the final and the most significant case, a recent decision of the New York Court of Appeals adopted a form of market-share liability with the significant qualification that it would not excuse manufacturers who could show that their product did not injure the plaintiff. Hymowitz v. Eli Lilly and Company, 73 N.Y.2d 487, 541 N.Y.S.2d 941, 539 N.E.2d 1069 (1989).
[F]or essentially practical reasons, we adopt a market share theory using a national market. We are aware that that [sic] the adoption of a national market will likely result in a disproportion between the liability of individual manufacturers and the actual injuries each manufacturer caused in this state. * * * Thus our market share theory cannot be founded upon the belief that, over the run of cases, liability will approximate causation in this State * * *. Nor does the use of a national market provide a reasonable link between liability and the risk created by a defendant to a particular plaintiff * * *. Instead, we choose to apportion liability so as to correspond to the overall culpability of each defendant, measured by the amount of risk of injury each defendant created to the public at large. Use of a national market is a fair method, we believe, of apportioning defendants’ liabilities according to their total culpability in marketing DES for use during pregnancy. Under the circumstances, this is an equitable way to provide plaintiffs with the relief they deserve, while also rationally distributing the responsibility for plaintiffs’ injuries among defendants.
$$**$***
Nevertheless, because liability here is based on the overall risk produced, and not causation in a single case, there should be no exculpation of a defendant who, although a member of the market producing DES for pregnancy use, appears not to have caused a particular plaintiff’s injury. * * *
Finally, we hold that the liability of DES producers is several only, and should not be inflated when all participants in the market are not before the court in a particular case. We understand that, as a practical matter, this will prevent some plaintiffs from recovering 100% of their damages. However, we eschewed exculpation to prevent the fortuitous avoidance of liability, and thus, equitably, we decline to unleash the same forces to increase a defendant’s liability beyond its fair share of responsibility. [Id. 73 N.Y.2d at 512-513, 541 N.Y.S.2d at 950, 539 N.E.2d at 1078 (citations and footnote omitted).]
Although perhaps the most controversial of the market-share decisions, its holding results in the scheme that is most similar to the almost universally-praised federal Vaccine Act.
As noted, Professor Robinson conceives of such alternative liability rules, and especially the Sindell rule, as being variants on the Summers principle. See Robinson, supra, 68 Va. L. Rev. 713. He notes, however, that because in Summers there *198was a one in two probability that either defendant was responsible for the shot that injured the plaintiff, the result is much more acceptable from an intuitive basis. Id. at 724. In contrast, in the DES cases the odds on any given producer “could be less than 1 in 300.” Ibid. If, for example,
90% of the market is divided equally, under the Summers rule of joint and several liability each of the manufacturers risks 100% liability based on a 15% probability that it caused the injury. * * *
In recognition of this disparity, the Sindell case adopted a rule that reduces the disproportion between potential liability and the probability that a defendant caused the injury by imposing liability on producers only according to their respective market shares in the sale of DES. [Id. at 725.]
The Sindell court also imposed a requirement that a “substantial share” of the market be joined, but Robinson finds this aspect unpersuasive and inconsistent with the opinion’s underlying logic. See id. at 725-26.
With these qualifications, Robinson notes that
[i]n fact, the departure from traditional doctrine is not so large as may appear. * * * Sindell itself is essentially an application of Summers v. Tice, modified to include a rule of contribution. To be sure, the application of Summers to an entire industry, as distinguished from two hunters, stretches that precedent, at least as it has been construed heretofore. Yet it does no violence to the underlying principle of Summers, rationalized either in terms of shifting the burden of proof on causation or in terms of a substantive liability rule. Indeed, this application of Summers is conservative insofar as it incorporates a rule of contribution limiting each tortfeasor’s liability commensurate with its particular contribution to the aggregate risk created by the product. [Id. at 768.]
B.
Recognition of a market-share theory of liability would be entirely consistent with New Jersey’s history and traditions. Judge Carton’s provocative dissent in Nopco Chem. Co. v. Blaw-Knox Co., 113 N.J.Super. 19, 24 (App.Div.), rev’d, 59 N.J. 274 (1971), foreshadows the growth of our law. There, one or more of multiple defendants handling a commercial product during its journey from the factory to the purchaser were responsible for its damaged condition. Ibid. Plaintiff could not prove which of the defendants had caused the damage. *199Thus, all defendants argued that they were entitled to a dismissal of plaintiffs claims. Ibid. Judge Carton reasoned:
The law does not compel such a “lame and impotent conclusion.” Reason and ordinary common sense dictate that * * * existing procedures be adapted or a new remedy be devised which will cause those parties most likely to possess knowledge of the occurrence to come forward with the facts peculiarly within their possession. To me it seems indefensible that the court should stand idly by and lend itself to such an obvious thwarting of justice. [Ibid.]
His dissent carried the day in the New Jersey Supreme Court, which held that “[w]e are firmly of the view that the complexity of the situation should not leave plaintiff remediless or require it to sue each defendant separately and successively at its peril simply because there is no precise precedent in this State.” Nopco Chem. Div. v. Blaw-Knox Co., 59 N.J. 274, 282 (1971).
So too, in Anderson v. Somberg, 67 N.J. 291 cert. denied, 423 U.S. 929, 96 S.Ct. 279, 46 L.Ed.2d 258 (1975), a multiple-defendant, surgical-injury case, this Court adopted a form of alternative liability. The Anderson plurality held that the plaintiff must recover from at least one of the defendants since no theory for the cause of the instrument’s breaking in the plaintiff’s spinal canal “was within reasonable contemplation save for the possible negligence of [the doctor] in using the instrument, or * * * [a] defect * * * attributable to a dereliction of duty by the manufacturer, the distributor, the hospital, or all of them.” Id. at 296. A Nopco-like shift in the burden of production was insufficient. Rather, the burden of proof shifted as well. Id. at 300. Since at least one defendant must inevitably fail to meet the burden, a verdict must be returned for the plaintiff. Ibid. In concurring, Justice Jacobs voted to affirm on the basis of the Appellate Division majority opinion which held that the jury should have been instructed to return a verdict against at least one of the defendants because the contrary verdict represented a miscarriage of justice. Id. at 305. The dissent argued that the suit became “trial by lot, or by chance.” Id. at 312.
But the Anderson judgment has stood the test of time. When one of multiple tortfeasors has most probably caused *200plaintiff’s injury, the law does not permit those tortfeasors to exonerate themselves by insisting that the plaintiff’s inability to prove which of them caused the injury is a total bar in law to recovery.
In New Jersey, we have confronted several other situations in which causation is difficult, if not impossible, to prove within traditional standards, and have always rejected a complete bar to redress for wrongful conduct. As the majority has rightly noted:
The torts process, like the law itself, is a human institution designed to accomplish certain social objectives. One objective is to ensure that innocent victims have avenues of legal redress, absent a contrary, overriding public policy. * * *
* * * [W]e strive to ensure that the application of negligence doctrine advances the fundamental purpose of tort law and does not unnecessarily or arbitrarily foreclose redress based on formalisms or technicalisms. [People Express Airlines, Inc. v. Consolidated Rail Corp., 100 N.J. 246, 254-55 (1985).]
Of course, causation-in-fact is far more than a technicalism. But the principle of causation is not an end of the legal system, but rather the means by which the legal system achieves its purposes. Thus, in Evers v. Dollinger, 95 N.J. 399 (1984), we recognized that under the circumstances, it would be impossible for the plaintiff to prove that if her physician had acted non-negligently, she would more likely than not have avoided the spread and recurrence of cancer. However, rather than insulate from liability tortious conduct that may have caused injury, we permitted recovery if the plaintiff could demonstrate within a reasonable degree of medical probability that if the physician had acted non-negligently, there was a substantial chance that her condition might have been avoided. We recognized the same standard in Hake v. Manchester Township, 98 N.J 302 (1985), permitting recovery if the plaintiff could show that there was a substantial chance that the decedent could have survived if rescue had been attempted. We have also relieved plaintiffs of the burden of proving the causal connection between negligence and damages when a health-care pro*201vider has negligently contributed to a condition. Fosgate v. Corona, 66 N.J. 268, 272-73 (1974).
Our approach has been flexible to adapt traditional limitations on causation and recovery to the evolving needs of a complex society. In our recent decision in Ayers v. Jackson Township, 106 N.J. 557 (1987), we eliminated the requirement that a defendant be shown to have caused physical injury as a predicate to recovery. No party could prove that the defendant’s tortious conduct had proximately caused the plaintiff’s medical injury except for an unquantifiable enhanced risk of future disease. However, each party could prove exposure to risk-causing substance and the need for medical surveillance to permit early discovery and treatment of any disease that might develop. In the absence of proof of actual physical injury, we allowed a limited and modified recovery of damages. The opinion recognized the economic reality of risk of injury to human health, and imposed legal consequences.
In each of those cases we were confronted with parties who had breached a duty of care, a victim ostensibly injured as a result of that breach of duty, and a seemingly unprovable causal connection. Yet, rather than permit a whole class of tortious acts and injuries to be unremediable, we recognized a standard of causation appropriate to the circumstances.
In the case of generic drugs, marketed in such a manner as to make them essentially undistinguishable to the party bearing the risk of injury, market-share liability is just such an appropriate principle of liability. Professor Landes and Judge Posner, who advanced the theory that the role of the law is to foster the most economic advantages for society, suggested that “the idea of causation becomes a result rather than the premise of the economic analysis of accidents.” Landes & Posner, “Tort Law as a Regulatory Regime for Catastrophic Personal Injuries,” 13 Journal of Legal Studies 417 (1984). Within that context I believe that the proportionate market-share theory of liability poses the fairest solution to to the *202problem. It is a resolution described by the amici curiae Merck & Co., Inc., Abbott Laboratories, and The Upjohn Company as “the least unfair” of the non-identification liability theories.
II
A.
The majority has not expressed any disagreement with the conceptual underpinnings of market-share liability. See ante at 188. Rather, the majority states that “the thrust of this appeal: [is] the public-policy and public-health considerations that would accompany the imposition of market-share liability in this context.” Ante at 177. It thus appears that it has no objection to the market-share concept of causation, but is solely, and perhaps justifiably, concerned that the imposition of any liability for this clearly beneficial vaccine will be detrimental to society’s interest in the availability of essential pharmaceutical products. While this is undoubtedly a valid concern, even if it properly related to the causation question before us today, surely we have an insufficient record on which to decide the question. Recall that the Law Division addressed one and only one question: whether New Jersey would recognize alternative liability for defective products under any state of facts. Moreover, the majority’s premise does not necessitate its conclusion. Put another way, it is obvious that none of the majority’s expressed concerns would be affected by a sudden discovery by the parties of the actual manufacturer of the DPT dose in this case. Thus, under the majority’s ruling a manufacturer must bear the same risks of liability that the majority seeks to insulate the industry from, except to the extent that the company can issue a product that would be for any reason difficult to distinguish from that of other manufacturers.
The majority rejects the market-share theory of causation for this product based on its perception that DPT is a beneficial drug and that any liability will threaten its supply and the *203efforts to develop a safer alternative. In New Jersey, we have been particularly candid to recognize proximate causation as an expression of legal policy. See Brown v. United States Stove Co., 98 N.J. 155, 173 (1984); Caputzal v. The Lindsay Co., 48 N.J. 69, 77-78 (1966). Nevertheless, the “legal policy” has always addressed the concerns underlying the reasons for a “causation” limitation on liability, i.e., “whether conduct can be considered sufficiently causally connected to accidental harm so as to justify the imposition of liability * * *.” Brown, supra, 98 N.J. at 173. Courts have had to conform the causation requirement to the needs of a changing society, recognizing that although the intentional battery may be the “classic” tort, such simple connections between conduct and result are increasingly less discernible.
The question of whether conduct should be considered tortious (here, whether a product should be considered safe) is also a policy question, but the policy issues are different from those presented for the causation issue. While causation questions are primarily concerned with the difficulty of proving, discovering, or even conceptualizing physical causal relationships, the tortious conduct question concerns even more profound notions of duty and moral responsibility. See Kelly v. Gwinnell, 96 N.J. 538 (1984). Traditionally, a finding of tortious conduct required a conclusion that the actor acted in some way that society has recognized as immoral or wrong. See Brown v. Kendall, 60 Mass. (6 Cush.) 292 (1850). Modern concepts of strict liability have departed from the culpability requirement in favor of more pragmatic concepts, e.g., Michalko v. Cooke Color and Chem. Corp., 91 N.J. 386 (1982), but the notion that a person must have acted “badly” still pervades the doctrine. Thus a manufacturer will not be held responsible for injuries caused by a product unless the plaintiff can prove that the product was “defective.” O’Brien v. Muskin Corp., 94 N.J. 169, 179-80 (1983) (“The necessity of proving a defect in the product as part of the plaintiff’s prima facie case distinguishes *204strict from absolute liability, and thus prevents the manufacturer from also becoming the insurer of a product.”).
In Feldman v. Lederle Laboratories, 97 N.J. 429 (1984), this Court was presented with the argument that all prescription drugs that are unsafe should be immunized from strict liability under comment k to section 402A of the Restatement (Second) of Torts. Id. at 441. Public policy, it was argued, demanded that prescription drugs be considered immune from strict liability because they were unavoidably unsafe. Ibid. We rejected such a wholesale argument:
[W]e see no reason to hold as a matter of law and policy that all prescription drugs that are unsafe are unavoidably so. Drugs, like any other products, may contain defects that could have been avoided by better manufacturing or design. Whether a drug is unavoidably unsafe should be decided on a case-by-case basis; we perceive no justification for giving all prescription drug manufacturers a blanket immunity from strict liability manufacturing and design defects claims under comment k.
Moreover, even if a prescription drug were unavoidably unsafe, the comment k immunity would not eliminate strict liability for failure to provide a proper warning. [Id. at 447.]
Plaintiffs in this case have asserted that safer DPT vaccines were available. They have also asserted, among other claims, that the warnings were inadequate and that the companies were negligent in their decision to market the whole-cell DPT virus. Under Feldman, these causes of action would be permitted to go forward even if the vaccine were unavoidably unsafe. Similarly, even the decision in Brown v. Superior Court, supra, 44 Cal.3d 1049, 1069-70 n. 12, 751 P.2d 470, 483 n. 12, 245 Cal.Rptr. 412, 424 n. 12, which rejected Feldman’s case-by-case approach to the unavoidably-unsafe question and instead imposed a blanket unavoidably-unsafe immunity for all prescription drugs, would permit these other causes of action to proceed. Finally, although the federal Act prohibits recovery in civil actions for injuries from vaccines administered after the effective date of the Act, if the side-effect was unavoidable, it permits suits claiming a failure-to-warn and improper manufacture in appropriate circumstances. 42 U.S.C.A. § 300aa-22(b)(1). By contrast, this Court has applied the policy considerations *205underlying the unavoidably-unsafe determination to impose a much more sweeping immunity.
The Court is quite explicit in limiting its rejection of market-share liability to contexts in which the relevant public policy considerations are present. Nevertheless, the decision gives no real guidance respecting its scope. Does it apply to all vaccines or just children’s vaccines? Or is it applicable to all drugs that, in hindsight, we think are important? Perhaps it is applicable to all drugs that the drug manufacturer might have thought at the time distributed were important. Other courts refusing to impose strict liability on prescription-drug manufacturers have extended their holdings to drugs like DES, thalidomide, and even accutane. See Brown, supra, 44 Cal.3d at 1067-68, 751 P.2d at 481, 245 Cal.Rptr. at 422-423. Is this Court saying that the children who developed severe birth defects due to their mothers’ ingestion of accutane to clear their complexion would also not be able to recover, lest the drug companies be hesitant to release new drugs? How are “important” drugs defined: complexion?, nausea?, fertility? Obviously we need not in every case define the holding applicable to other factual contexts, but if the purpose of our holding is for manufacturers to be encouraged to release important drugs, how can they find out in advance which ones the Court will consider important? They will always be operating at the risk the Court is trying to insulate them from. See Brown, supra, 44 Cal.3d at 1068-69, 751 P.2d at 482, 245 Cal.Rptr. at 423.
B.
Perhaps the scope would be more discernible if the policy arguments themselves were more clear. For example, the majority refuses to impose any liability on DPT manufacturers because it feels that this result will encourage the development of safer drugs. Ordinarily the presumption is the opposite:
Imposing liability on defendants for their negligent conduct discourages others from similar tortious behavior, fosters safer products to aid our daily tasks, vindicates reasonable conduct that has regard for the safety of others, and, *206ultimately, shifts the risk of loss and associated costs of dangerous activities to those who should be and are best able to bear them. [People Express, supra, 100 N.J. at 255.]
See also Michalko, supra, 91 N.J. at 401 n. 4 (“imposing strict liability would induce providers of services to invest in safety, leading to greater protection for their customers and reduced accident costs.”); Beshada v. Johns-Manville Products Corp., 90 N.J. 191, 207 (1982) (“By imposing on manufacturers the costs of failure to discover hazards, we create an incentive for them to invest more actively in safety research.”).1 Although I might understand the temptation to avoid strict liability for beneficial drugs, see, Brown, supra, 44 Cal.3d at 1069-70 n. 12, 751 P.2d at 483 n. 12, 245 Cal.Rptr. at 424 n. 12, I do not see how this applies to a failure-to-warn or negligence claim where all that is called for is for the company to act reasonably. For the unavoidable procedural posture of this case, a procedural posture that the defendants have created, is that they have admitted that each manufactured and distributed a prescription drug that contains a common defect, that each failed to warn users of the dangers of the drug, that there are safer alternative designs for the drug, and, finally, that each acted negligently in the manufacture and marketing of the drug. This summary judgment action requires those admissions.
Furthermore, generally the logic that a manufacturer should not be held strictly liable for prescription drugs has been applied only where the drug could not be made safer. In Feldman, supra, we said: “[W]e see no reason to hold as a matter of law and policy that all prescription drugs that are unsafe are unavoidably so. Drugs, like any other products, may contain defects that could have been avoided by better *207manufacturing or design.” 97 N.J. at 447. By contrast, in this case, the drug companies argued that the fact that there was a safer product should exempt them from liability since the market-share analysis would be too difficult. Although I could perhaps understand reducing, or even exempting, the safer product from its market share of liability, I cannot understand why we would insulate the companies making the less-safe products. Plaintiff asserts that some of the companies continued to manufacture, market, and try to outsell the ostensibly-safer vaccine, Tri-Solgen, even after they had concluded that Tri-Solgen was safer than their product.2 I fail to see why any manufacturer should feel compelled to make a safer vaccine when in this case we find public policy to demand their exemption from liability even though the plaintiff alleges, and therefore we must assume, that the defendants knowingly continued to market and manufacture a product they knew to be less safe, and that they refused to inform physicians of the additional risk.
The majority’s attempt to influence the future behavior of DPT manufacturers is misdirected insofar as the federal Act will most certainly be the sole motivating force for their actions. Under the federal Act, a party injured by a vaccine administered after October 1, 1988, the effective date of the Act, is effectively prohibited from filing a civil action until after his or her other claim under the Act has been adjudicated. 42 U.S.C.A. § 300aa-11(a)(2, 3). The party can file suit at that time *208only if it waives any compensation awarded under the Act. 42 U.S.C.A. § 300aa-21. Given the recent State legislation cited by the majority, ante at 187-88, making such product-liability actions difficult to win, as well as the federal legislation precluding liability in civil actions for some claims arising from injuries caused by a vaccine administered after the effective date of the federal Act, 42 U.S.C.A. § 300aa-22, it is unlikely that many claimants will risk their award to pursue a civil action, and even more unlikely that anyone denied compensation under the federal Act will be able to succeed in a civil action. Thus, any fear of liability that may affect a manufacturer’s decision to attempt to develop a safer product, or conversely to cease production, would be virtually unaffected by the majority’s decision today. In addition, in an effort to assure continuing vaccine availability and development of safer vaccines, the Act devotes substantial resources to the coordination of public and private efforts towards developing safer vaccines and programs to administer them. 42 U.S.C.A. §§ 300aa-2, -3.
The final policy argument that needs to be examined is the extent to which liability expenses are the cause of the price increases. The problem appears very complex. One commentator has stated that “unlike some other products, the incidence of serious injury with children's vaccines is very low. Thus, a vaccine compensation system can be self-funding by adding a very small excise tax to the price of vaccines.” Schwartz & Mahshigian, supra, 48 Ohio St. L.J. at 393 (footnote omitted). If injuries are so rare, perhaps the claim that price increases are due to liability costs should be more critically examined. However, even if one accepts the Court’s “findings” that the price per dose increased in two years from lie to $11.40, $8.00 for insurance, ante at 179, that still amounts to a twenty-sevenfold increase in non-insurance-related costs corresponding to the development of a market with only two suppliers. Moreover, since the liability exposure has been greatly reduced by the federal Act, presumably the insurance costs have reduced accordingly. Although the majority claims that its holding is *209consistent with the federal Act, it is clear that Congress, which spent much more time and money examining the issue, decided that it would be inappropriate to deny compensation, recognizing that an appropriately-limited liability would not threaten the industry.
Ill
This brings me to the most troublesome aspect of the majority’s opinion: its adjudication of the available remedies under the federal act and willingness to fault the plaintiff for asserting her rights. In reading the majority opinion I get a sense that the Court feels that in pursuing a tort recovery the plaintiff is somehow seeking a windfall or something more than she deserved. Nevertheless, the federal Act specifically prohibits states from “establish[ing] or enforcpng] a law which prohibits an individual from bringing a civil action against a vaccine manufacturer for damages for a vaccine-related injury or death if such civil action is not barred by this part.” 42 U.S.C.A. § 300aa-22(e). Surely Congress did not intend state courts to accomplish the same effect by molding state common law to deny recovery in reliance on the general availability of a federal remedy.
In addition to the fact that the plaintiff should not be penalized for exercising her right under the federal act to pursue her tort remedy, at the time this suit was filed, the federal act had not yet been enacted. Therefore, plaintiff could not be sure whether she would be eligible to recover under it. Moreover, even now it is unclear whether any remedy plaintiff would be “entitled” to would actually be sufficiently funded. Finally, although the plaintiff is permitted to withdraw her State tort suit and pursue the federal remedy, she will not be allowed to do so if this case is dismissed. Before the Court limits tort recovery under New Jersey partially on the basis of an alternative remedy, we should be certain that such remedy is, in fact, both sufficient and available.
*210Defendants maintain that Deanna Marrero is eligible to receive compensation under the program. Under the Act, vaccine injury victims are divided into three groups:
(a) those who both were injured by a vaccine administered before October 1, 1988, the effective date of the Act, and filed a civil suit prior to that date, 42 U.S.C.A. § 300aa-11(a)(4, 5);
(b) those injured by a vaccine administered before October 1, 1988, who had not filed a civil suit prior to that date. 42 U.S.C.A. § 300aa-11(a)(6);
(c) those injured by a vaccine administered after October 1, 1988. 42 U.S.C.A. § 300aa-11(a)(3).
Those in the first category, such as Deanna Marrero, may continue any pending lawsuit, but will become ineligible for compensation under the Act unless damages were denied or their suit dismissed with prejudice prior to October 1, 1988, 42 U.S.C.A. § 300aa-11(a)(4), or the plaintiff voluntarily withdraws the pending suit prior to October 1, 1990. 42 U.S.C.A. § 300aa-11(a)(5). The second group may file a civil suit, but are then ineligible to file a claim under the Act. 42 U.S.C.A. § 300aa-11(a)(6). The third group must complete the Act’s compensation proceeding and reject its judgment before pursuing a civil claim. 42 U.S.C.A. § 300aa-11(a)(2, 3). No party may file a claim under the Act if damages have been awarded for a vaccine-related injury under a judgment or settlement of a civil action. 42 U.S.C.A. § 300aa-11(a)(7).
The Act also creates an affirmative defense of sorts. In order to award compensation, the special master or the court must find that there is not a preponderance of the evidence that the injury was caused by factors unrelated to the administration of the vaccine. 42 U.S.C.A. § 300aa-13(a)(1)(B). The majority of courts reviewing both the provisions and legislative history of the Act have concluded that “Congress contemplated that civil tort remedies for vaccine-related injuries have been available, are now available, and will continue to be available under certain circumstances, even after the effective date of the Act.” Patten v. Lederle Laboratories, 655 F.Supp. 745, 749 (D.Utah 1987). Thus, we should not deny our State remedy due to the availability of a federal remedy when the Act creating that *211federal remedy contemplates the continuing availability of State remedies.
Moreover, although this suit was filed on April 19, 1985, the Act was not passed until November 14, 1986. Thus no alternative federal remedy was available at the time the suit was filed. The Act is partially funded by a Vaccine Injury Compensation Trust Fund funded by an excise tax on vaccines. 26 U.S.C.A. 9510. Compensation for persons injured by vaccines administered after the effective date of the Act will be awarded from this Trust Fund. 42 U.S.C.A. § 300aa-15(i)(2). Compensation for persons injured by vaccines administered before the effective date of the Act, like Deanna Marrero, is made from appropriations. 42 U.S.C.A. § 300aa-15(i)(1).
Congress has- authorized appropriations for this purpose of $80 million for this year and the same amount for each of the next three fiscal years, but these appropriations were not authorized until December 22, 1987, two and one-half years after this suit was filed. 42 U.S.C.A. § 300aa-15(j). Congress is currently considering additional appropriations measures to fund this aspect of the program. Nevertheless, there is a serious debate over whether such appropriations are or will be sufficient to give the statutory award to all persons injured prior to the Act. Furthermore, authorization of an appropriation is only one step in a long process before the money is actually appropriated. Many forces along the way can lead to reduction or elimination of the funding. The majority points to the one-time appropriation of $80 million out of the $320 million authorized as evidence of the certainty of funding. I am not so sure that the need for a B-2 bomber may not take precedence in other years over vaccine funding. The Act also places an overall cap of 3500 on the number of claims that can be filed for injuries due to vaccines administered prior to the Act. 42 U.S.C.A. § 300aa-11(b)(1)(B). In short, the remedy may or may not ever be there.
*212Finally, this plaintiff will be ineligible for compensation under the Act if summary judgment is granted. The Act precludes the filing of a claim where the injured party has received damages pursuant to a settlement. 42 U.S.C.A. § 300aa-11(a)(7). Although plaintiff had settled her claim with her doctor, such settlement was withdrawn prior to court approval in recognition that such settlement would preclude a later claim under the Act. However, plaintiff will also be unable to file a claim unless she voluntarily withdraws her State court action by October 1, 1990, and prior to judgment. 42 U.S.C.A. § 300aa-11(a)(5). Therefore, by refusing to recognize market-share liability as a valid theory in this case, we are precluding plaintiff from withdrawing her suit prior to judgment and effectively extinguishing both her state and federal recovery. While such a result might be appropriate were we to conclude that market-share liability was somehow conceptually lacking, and therefore inappropriate in New Jersey, the result is wholly unfair under the majority’s rationale. Although I believe that the Act represents a tremendous advance in the development of an effective reparation system for vaccine-related injuries, unfortunately under the majority’s ruling the Act will not apply to the current cause of action.
From my perspective, the most critical and significant provision of the Act is that it creates the Vaccine Injury Compensation Trust Fund with money collected from an excise tax on vaccines. 26 U.S.C.A. § 9510. In other words, the industry as a whole shares the risk of financial compensation for injuries after October 1, 1988, attributable to individual manufacturers. See 42 U.S.C.A. § 300aa-15(i)(2). I think that we would do well to mold our tort law to this model. I agree that punitive damages are inappropriate in this setting of market-share responsibility and that proportional several liability is the fairest solution. We should be candid to recognize, in the face of the enormous medical, legal, and social ramifications of childhood-vaccine injuries, that the goal of the legal system should be to steer a course toward the result that will yield the greatest *213benefits to human health. That principle involves a candid recognition of the limitations of tort liability to deal fully with the problem while at the same time seeking the maximum prevention of human disease and the maximum fairness in the alleviation of human suffering.
IV
I am usually the last person ever to insist on strict adherence to the Civil Rules of Procedure. But there is something deeply flawed about the way this case has proceeded. It is terribly unfair to deprive this tragically disabled child of any remedy whatsoever for her catastrophic injuries on the basis of theories of law never presented to the Law Division and on the basis of an alternative federal remedy that was not in existence when her case was argued in the Law Division and not even funded when her case was argued in our Court.
I have no idea of the extent to which Deanna can understand these proceedings, but surely she would see the law as the cruelest of hoaxes: dismissing her state common-law claim because she had a federal remedy while with the same stroke of the pen extinguishing that federal remedy because she had elected to ask us if there was a state remedy for her catastrophic injuries. The most that we should do in this situation is to give a declaration of what a claimant’s rights are, not to ask a twelve-year-old girl to play a game of blind man’s bluff with our legal system.
I would hold that New Jersey should recognize market-share liability in an appropriate case. I see no reason why plaintiff should not be permitted to demonstrate that the nature of this product and industry is such that market-share liability is an appropriate principle of causation to apply to the breaches of duty asserted. Were we to reject “market-share” on this record, we would be out of step with both California and New York. If those two great states can mold their law to the needs of their citizens without dislocation of public markets, I fail to *214see how the industry could not abide the needs of New Jersey’s citizens. The policy questions surrounding the necessity of this drug will not be ignored, but will be addressed if the defendants raise the defense that the product is unavoidably unsafe.
I would remand the case, then, to the Law Division, as the Appellate Division did, to resolve whether DPT is an unavoidably unsafe product and, if not, whether there is a sufficiently common defect in the products marketed to invoke a Sin-dell-style market-share liability in New Jersey. This is no mean challenge. Deanna will face an enormous uphill struggle in asserting her claims. I have no doubt that we should abide by the recently-enacted guidance of our products-liability act, N.J. S.A. 2A:58C-4, that a failure-to-warn claim in the case of prescription drugs should be measured by compliance with the FDA’s required warnings. Congress has enacted the same principle as one of national policy for all injuries due to vaccines administered after the effective date of the Act. 42 U.S.C.A. § 300aa-22(b)(2). And I genuinely doubt that she will be able to establish that there was indeed an equally effective and safer alternative vaccine.3 The FDA has refused to re-license TriSolgen, and the so-called Japanese acellular vaccine is nowhere near availability for marketing. Finally, the proof of actual medical causation will remain as elusive as ever. See Niemiera v. Schneider, 114 N.J. 550 (1989) (plaintiff’s injuries were assertedly attributable to causes other than DPT). By contrast, if she elects to withdraw her suit and is eligible under the federal Act, she will enjoy a presumption that the vaccine caused her injuries. See 42 U.S.C.A. § 300aa-13.
*215Armed with foreknowledge of the legal principles that will be applied, her counsel can then exercise judgment on whether or not this vaccine can be proven not to be unavoidably unsafe within the meaning of our law and to contain a common defect such that the imposition of market-share liability is warranted. Under the federal Act her attorney is obligated to inform her that compensation may be available under the federal program. 42 U.S.C.A. § 300aa-10(b). Fully informed of her chances of recovery under both our State tort law and the federal Act, she might elect to seek, if eligible, the federal vaccine remedy. She deserves this chance.
I cannot play a shell game of remedies with a disabled child. No member of the Appellate Division genuinely believed that the New Jersey Supreme Court would not recognize a form of market-share liability. Each differed as to either the form of alternative liability or whether the decision could be made only by “either the Supreme Court or the Legislature.” 219 N.J.Super. at 642. No judge or lawyer did predict or could have predicted that the New Jersey Supreme Court would find that market-share liability was indeed an acceptable principle of New Jersey tort law (“This case * * * may therefore come to represent the exception rather than the rule.”) Ante at 191, but a principle that could not be invoked when there was an alternative federal remedy.
Deanna Marrero has suffered enough. She ought not to have had to guess under which hand of the law she might find some surcease of her suffering. She ought at least to have been given a chance to know the rules before she had to make the choice.
For reversal — Justices CLIFFORD, POLLOCK, GARIBALDI and STEIN — 4. For affirmance — Justices HANDLER and O’HERN — 2.Congress, as well, appears to have reached a conclusion contrary to that of the Court. See Schwartz & Mahshigian, "National Childhood Vaccine Injury Act of 1986: An Ad Hoc Remedy or a Window for the Future?," 48 Ohio St.L.J. 387, 395 (1987) ("Congress may have decided to retain the tort system for vaccine injury claims, at the expense of greater predictability, as a means of providing incentives for the manufacture of safe vaccines.”).
Lederle’s recently-released internal correspondence "stated that the test results demonstrated that Lederle’s product had a significantly higher reaction rate than Lilly’s non-cellular vaccine.” Burke, “DPT Vaccine Controversy: An Assessment of the Liabilities of Manufacturers and Administering Physicians Under Several Legal Theories," 17 Seton Hall L.Rev. 541, 569 (1987). Further clinical evaluations were planned by Lederle, but they were never completed. Frankly, I rather suspect that these qualitative differences in the DPT vaccines would be winnowed out in the trial process. The reaction rate that we are talking about here is the catastrophic reaction rate, not those who suffer greater or lesser degrees of discomfort and distress. It is the catastrophic common defect that must be shown to be qualitatively different.
The data overwhelmingly support the efficacy of DPT if administered with proper warnings. See S.Rep. No. 483, 99th Cong., 2d Sess. 3 (1986) (Morbidity and mortality have dramatically decreased due to child immunization programs, e.g., in the case of DPT the number of cases decreased from 265,269 in 1934 to 2,000 in 1982, and the number of deaths declined from 7,500 in 1934 to 4 in 1982.)