concurring.
I join in Parts I, II and IV of Judge Weis’s opinion and concur wholly in the result. I also join in portions of Part III, in which Judge Weis explains why he finds a rational basis for New Jersey’s distinguishing asbestos cases from prescription drug cases in terms of the state-of-the-art defense.1 However, I do not believe that Judge Weis has identified with sufficient precision the New Jersey Supreme Court’s reasons for making the distinction under review, a distinction I believe to be supported by a valid government objective and rational within our equal protection jurisprudence. Specifically, I believe that, on the basis of adjudicative facts determined in cases that had the full panoply of procedural protections, the New Jersey Supreme Court has determined a legislative fact— that the hazards of asbestos exposure were knowable to the industry at all relevant times. The subject of legislative factfinding is rarely discussed in the jurisprudence, and I write separately to explain why I think it validates the New Jersey Supreme Court’s distinction.
Appellants complain that the New Jersey Supreme Court deprives manufacturers and distributors of asbestos-containing products of the state-of-the-art defense in products liability/failure to warn actions, which other manufacturers and distributors are entitled to assert in the same kind of lawsuits. Appellants urge that no rational theoretical basis exists for differentiating between these classes of litigants. Judge Weis in his opinion demonstrates that is not clear that all other manufacturers may take advantage of the state-of-the-art defense, although he agrees that New Jersey “does treat asbestos cases differently than other product liability cases.” Weis Op., at 1241. I believe that we need not find constitutional infirmity even if we assume for the sake of argument appellants’ worst case scenario, i.e., that New Jersey singles out the asbestos industry.
I.
As I read the New Jersey Supreme Court’s cases, the court does not deny asbestos defendants the state-of-the-art defense on theoretical grounds. Instead, I believe that the New Jersey court, via the Beshada-Feldman-Fischer trilogy, has determined a legislative fact — that, at all relevant times, asbestosis harms were knowable to the industry. That being the case, the New Jersey Supreme Court has reasonably decided to preclude endless relitigation of what was “knowable” to the asbestos industry.2
Believing that “Feldman provides yet another underlying reason for precluding the state-of-the-art defense in the asbestos setting,” Judge Weis observes that “the [Feldman] opinion suggests that these manufacturers knew the dangers of asbestos and, consequently, the state-of-the-art defense could not be sustained.” Maj. ap. at 1241 (footnote omitted). Although I agree with Judge Weis that the New Jersey Supreme Court has determined that the harms of asbestos were actually known to at least one asbestos manufacturer, I do not believe that it found in Feldman that the harms of asbestos were actually known to the industry. Rather, on the basis of what could have been known, that opinion determines constructive knowledge sufficient to defeat the state-of-the-art defense. *1246It was not until Fischer v. Johns-Manville Corp., 103 N.J. 643, 512 A.2d 466 (1986), that the court found that at least one asbestos manufacturer actually knew of the harms.
The Beshada court did have before it the factual dispute concerning what was known. It found, however, that it need not resolve the “substantial factual dispute about what defendants knew and when they knew it,” 447 A.2d at 542, referring both to a finding that “[k]nowledge of the danger [of asbestos] can be attributed to the industry as early as the mid-1930’s____” Id. 447 A.2d at 542, (quoting Hardy v. Johns-Manville Sales Corp., 509 F.Supp. 1353, 1355 (E.D. Texas 1981)) and the contrary assertions of the asbestos industry.3 Pertinent in Feldman is the exact manner that the court restricted the application of Beshada when it differentiated between asbestos products and other products in the context of the state-of-the-art defense:
We do not overrule Beshada, but restrict Beshada to the circumstances giving rise to its holding. See, e.g., Friedman v. Podell, 21 N.J. 100, 105, 121 A.2d 17 (1956); Konrad v. Anheuser-Busch, Inc., 48 N.J.Super. 386, 388 [137 A.2d 633] (Law Div.1958) (“Cases state principles but decide facts, and it is only the decision on the facts that is binding precedent.”). We note in passing, that, although not argued and determined in Beshada, there were or may have been data and other information generally available, aside from scientific knowledge, that arguably could have alerted the manufacturer at an early stage in the distribution of its product to the dangers associated with its use.
97 N.J. at 455, 479 A.2d 374. Thus, while the New Jersey Supreme Court in Beshada expressly did not decide “what defendants knew and when they knew it,” 447 A.2d at 542, it also refrained from deciding that question in Feldman. In Feldman, it stated only that the manufacturers “arguably could have [been] alerted ... at an early stage in the distribution of its product to the dangers associated with its use,” 97 N.J. at 455, 479 A.2d 374 (emphasis supplied). Rather than finding actual knowledge, the Feldman opinion thus intimated its views on constructive knowledge, which is sufficient to defeat the state-of-the-art defense, based on facts brought to its attention in Beshada.
In Fischer the New Jersey Supreme Court determined not only that asbestosis harms were knowable by the asbestos industry, but that those harms were actually known by at least one manufacturer. In that case, the court recited two single-spaced pages of facts that determined to its satisfaction that at least one company “did in fact have knowledge of the hazards of asbestos ... as early as the 1930’s.” 103 N.J. at 649, 512 A.2d 466 (quoting Fischer v. Johns-Manville Corporation, Bell Asbestos Mines, 193 NJ.Super. 113, at 117, 472 A.2d 577). For example, the court referred to “eleven scientific articles published between 1936 and 1941 documenting the grave pulmonary hazards of exposure to asbestos and discussing the measures which could be taken to protect workers.” Id. at 650, 512 A.2d 466 (quoting 193 N.J. Super, at 118, 472 A.2d 577). On the more individualized level of worker exposure, a doctor testified that “from the beginning of his employment [in 1944] he saw persons with asbestosis ‘on a regular and frequent basis’ and frequently made recommendations that such employees receive job re-classifications which would remove them from continued exposure to asbestos dust.” Id. at 651, 512 A.2d 466 (quoting 193 N.J. Super, at 120, 472 A.2d 577). As early as 1933, noted the court, asbestos workers were filing claims against at least one asbestos manufacturer. See id. at 650, 512 A.2d 466 (quoting 193 N.J.Super. at 118).
The state-of-the-art defense decides not what the defendant or another party knew — a fact relating to a particular party — but what was knowable — a fact about the state of the world. In Feldman, the court adverted to what was knowable, but *1247in Fischer it made a concrete determination. By deciding that even one company in the industry — Johns-Manville — knew of asbestosis, the Supreme Court of New Jersey, culminating the trilogy, found that the harms were knowable to the industry as a whole.
To my mind, it was not inappropriate for the court to rely on this determination — or even a similar earlier determination, see, e.g., Hardy, cited in Beshada, 447 A.2d at 542 — in finding legislative facts4 when it determined that the rule of Beshada should continue to govern the asbestos industry. As Justice Holmes recognized long ago, “the court may ascertain as it sees fit any fact that is merely a ground for laying down a rule of law.” Chastleton Corp. v. Sinclair, 264 U.S. 543, 548-49, 44 S.Ct. 405, 406, 68 L.Ed. 841 (1924). To forbid such recognition would force courts to fashion laws without reference to reality.
Judge Hunter, in dissent, acknowledges that one manufacturer was found to have known of asbestos’ harms in Fischer, but he finds that case unhelpful for judging the rationality of the challenged differentiation between asbestos and non-asbestos defendants because it was decided after the differentiation was first drawn. See Dissent, at 1259, n. 5. I disagree. I believe that the New Jersey Supreme Court rested the differentiation on the views it intimated in Feldman, i.e., the legislative fact that asbestos’ harms were knowable. This avenue was open to New Jersey’s highest court despite the putative protestations of a federal court, see Hardy, or the lack of “a certain degree of consensus,” id. As Professor Davis has opined,
judge-made law would stop growing if judges, in thinking about questions of law and policy, were forbidden to take into account the facts they believe, as distinguished from facts which are ‘clearly * * * within the domain of the indisputable.’ Facts most needed in thinking about problems of law and policy have a way of being outside the domain of the clearly indisputable.
Fed.R.Evid. 201 advisory committee’s note quoting Davis, A System of Judicial Notice Based on Fairness and Convenience, in Perspectives of Law 69, 82 (1964).
Leaving aside the argument that the court’s pronouncement in Feldman amounted to such a legislative fact-finding, and assuming that that fact was not found until Fischer, I still do not believe that we should today pronounce it irrational. La-belling the differentiation irrational would be contrary New Jersey’s current position regarding manufacturers’ ability to have known. Moreover, it would involve this court in a meaningless ritual of striking down a differentiation that we know to be supportable, only for the state court to revive it as supported on the basis we have already identified. For similar reasons, Justice Stevens has warned about unduly emphasizing actual motivation for the differentiation under review in equal protection challenges of legislative enactments.
Actual purpose is sometimes unknown. Moreover, undue emphasis on actual motivation may result in identically worded statutes being held valid in one State and invalid in a neighboring State. I therefore believe that we must discover a correlation between the classification and either the actual purpose of the statute or a legitimate purpose that we may reasonably presume to have motivated an impartial legislature.
United States Railroad Retirement Bd. v. Fritz, 449 U.S. 453, 180-81, 101 S.Ct. 453, 462, 66 L.Ed.2d 368 (1980) (Stevens, J., concurring in the judgment). Because the legislative fact-findings of Fischer may reasonably be presumed to have motivated a hypothetical impartial court, we would have to affirm even had Feldman not preceded that case.
The determination of knowability implicates the standard to which the state is willing to hold a manufacturer of a product: not only must a manufacturer stay *1248abreast of what has already been discovered about his product, but he must also diligently pursue information about its possible dangers before he introduces it for distribution throughout the marketplace. See supra at 1245 n. 2. As such, under the aegis of “constructive knowledge,” the New Jersey Supreme Court has made a policy judgment concerning the diligence with which the manufacturers should have undertaken additional investigation. See Feldman, 479 A.2d at 386-87. Thus, regardless of whether a given manufacturer is found to have actually known of the harms, the New Jersey Supreme Court could find that the industry was chargeable with the knowledge that was attainable had the manufacturers undertaken the task of discovery. As Judge Weis indicates, the court in Feldman made clear that “a reasonably prudent manufacturer will be deemed to know of reliable information generally available or reasonably obtainable in the industry.” 479 A.2d at 387 (emphasis supplied); see Weis Op., at 1242. Such a determination cannot be made without reference to facts concerning the availability of information to the asbestos industry as a whole. However, once this factual assessment had been made, the court was also justified in precluding the relitigation of the factual basis of the state-of-the-art defense.5
II.
Judge Hunter in dissent argues that, because the use of legislative facts concerning the knowability of asbestos harms does not satisfy the requirements of the collateral estoppel doctrine, it violates due process. I disagree. The above discussion demonstrates that, in choosing to allow the state-of-the-art defense for other industries, the New Jersey Supreme Court did not have to turn a blind eye to its belief that the harms of asbestos were knowable to the asbestos indutry as a whole. Common law courts could not fashion rules grounded in reality if they were obliged to proceed without aid of legislative facts. As is evident from other cases that have found legislative facts, legislative fact-finding by such a court need not conform to the requirements of collateral estoppel to pass due process muster. I therefore do not believe that the New Jersey Supreme Court can be held to have acted unconstitutionally in finding the legislative facts it did when fashioning the law at issue in this case.
In the New Jersey Supreme Court cases discussed above, the factual determination concerning the knowability of asbestos harms is a legislative fact. “Adjudicative facts are simply the facts of the particular case. Legislative facts, on the other hand, are those which have relevance to legal reasoning and the lawmaking process, whether in the formulation of a legal principle or ruling by a judge or court or in the enactment of a legislative body.” Fed.R. Evid. 201(a) advisory committee’s note. Under this dichotomy, the pertinent facts, which concern the nature of the asbestos industry as a whole and what it could have known, were properly treated by the court as legislative. As one commentator has noted,
An industry-wide question calls for facts about the industry; facts about each company may be unhelpful and may even get in the way. If a court ... is making law to govern an industry of 100 companies, the useful facts are about the 100 companies as a group, not about each company. The facts about the group are legislative, even though they are the sum of adjudicative facts about each company.
3 K. Davis, Administrative Law Treatise § 15:5, at 152-53 (2d ed. 1980) (emphasis in original). The issue of the knowability of asbestos harms concerns not what a particular litigant knew, but rather what knowledge was in the realm of the possible for the industry as a whole. Because the knowability of the harms of asbestos may thus be a legislative fact upon which a court can fashion a rule of law, it was not inappropriate for the New Jersey Supreme *1249Court to base a legal rule upon these legislative facts. See generally FecLR.Evid. 201(a) advisory committee’s note; Morgan, Judicial Notice, 57 Harv.L.Rev. 269, 270-71 (1944).6
Many courts and commentators have argued that a court should not rely on legislative facts without giving the parties an opportunity for comment upon them. See, e.g., Bulova Watch Co. v. K. Hattori & Co., 508 F.Supp. 1322, 1328-29 (E.D.N.Y. 1981); 3 K. Davis, Administrative Law Treatise § 15:9; S. Saltzburg & K. Redden, Federal Rules of Evidence Manual 60 (4th ed. 1986) (“We do not claim Judges cannot rely on a broad range of facts to force the law forward. We suggest only that the parties should be permitted to participate in the march.”). It is therefore noteworthy that the Beshada case was itself concerned with developing facts pertinent to the knowability of asbestos harms. Additionally, the Hardy and Borel cases, on which Beshada and Feldman appear to be predicated, as well as the Fischer case, were litigated with the full panoply of procedural protections. Moreover, the court in Beshada did not hear from only one member of the industry; rather, a broad cross-section of the industry was represented. See 447 A.2d at 541-42. Finally, the members of the asbestos industry had the opportunity to advocate a contrary conclusion in In re Asbestos Litigation Venued in Middlesex County, 99 N.J. 201, 491 A.2d 700 (1984), and again in Fischer. Hence, the affected industry has had relevant opportunities to respond to the New Jersey Supreme Court’s determination of the pertinent legislative facts.
It is unclear whether Judge Hunter would find relevant to his due process point the legislative nature of the facts relied on by the New Jersey Supreme Court. Be*1250cause a legislative fact is not an individualized fact, however, the Constitution does not mandate that it be found through a process of individualized fact-finding.7 Because legislative facts are the basis for a rule of law, they need not be relitigated in each succeeding case that invokes the rule and thereby indirectly relies on the legislative fact. It does not offend due process to craft the rule without allowing individualized process for redetermination of the legislative facts on which the rule is based. See, e.g., New York Times v. Sullivan, 376 U.S. 254, 278, 84 S.Ct. 710, 725, 11 L.Ed.2d 686 (1964) (finding state libel law unconstitutional because of “the pall of fear and timidity imposed upon those who would give voice to public criticism”); Hawkins v. United States, 358 U.S. 74, 78, 79 S.Ct. 136, 138, 3 L.Ed.2d 125 (1958) (“Adverse testimony given in criminal proceedings would, we think, be likely to destroy almost any marriage.”).
Judge Hunter would apparently require a state Supreme Court to adhere to the strictures of collateral estoppel doctrine whenever it rests a rule of law upon a legislative fact that had been previously determined as an adjudicative fact in a prior court case. Such a rule would deprive courts of the ability to fashion rules with reference to the reason borne of experience. “The history of the common law shows a constant pattern of questions once treated as fact growing into matters of law after the courts have gained knowledge and experience concerning them.” Korn, Law, Fact, and Science in the Courts, 66 Colum.L.Rev. 1080, 1105 (1966). Thus, for example, the decision to admit into evidence novel scientific testimony is first tested by individual adjudications before judicial recognition eliminates the need for a preliminary foundation. See United States v. Downing, 753 F.2d 1224, 1234 (3d Cir.1985).8 Courts have even elevated the fact finding of a single jury verdict to the position of legislative fact on which to base a rule of law.9 In all instances of adjudicative facts’ elevation to legislative facts, the prior adjudications, with their full panoply of procedural protections, influence the court’s view of reality. The legislative facts, in turn, influence the rule that is fashioned, and the due process clause does not require individualized determination or reconsideration of the legislative facts in all subsequent cases.
Finally, Judge Hunter goes beyond testing the New Jersey Supreme Court’s determination against the due process clause of the fourteenth amendment. He also finds that, “[i]f judicial notice can be taken of such ultimate facts at all, it can only be within the strictures of Fed.R.Evid. 201.” Dissent, at 1259; see also id. At 1260 (finding “does not satisfy the requirements of Fed.R.Evid. 201”). While I find the advisory committee’s comments illuminating on the distinction between adjudicative and legislative facts, I do not believe that Rule 201 binds any state court. More important, Judge Hunter’s point seems inapposite to the central issue. The question is not whether New Jersey wisely found legislative facts — I do not endorse the New Jersey Supreme Court’s finding, for I do not pass on the wisdom of what it has done. Rather, for our purposes it is sufficient *1251that New Jersey behaved rationally.10 Because legislative facts underlie the New Jersey Supreme Court’s decision to continue to withhold the state-of-the-art defense from asbestos manufacturers, the differentiation under review cannot be deemed irrational.
III.
I have not joined the balance of Part III of Judge Weis’ opinion because I do not find a rational basis for New Jersey’s distinction in any of the other justifications he advances. I briefly note my differences with the remainder of Judge Weis’ Part III.
For the most part, the other rationales advanced by Judge Weis do not explain the New Jersey Supreme Court’s distinction; rather, they imply that, because of the New Jersey Supreme Court’s conceded hegemony over the development of that state’s tort law, we must defer to whatever distinction the court draws. For example, he appears to regard the differentiation in and of itself as a matter of state policy in a matter of state expertise, which is said to deserve our deference. See, e.g., Weis Op., at 1243 (“Because the court’s reasoning may be applicable to other defendants in similar circumstances, the justification advanced both directly and indirectly by the New Jersey court may be regarded as weak and ill-advised. We cannot, however, conclude that the state court’s position is irrational.”). Additionally, Judge Weis draws attention to “the federalism concerns that color this controversy” — concerns, I must note, that are ever present in Fourteenth Amendment challenges. Id., At 1243.
To determine the issue on the basis of such deference, however, assumes the answer to the equal protection inquiry. If we were to defer to the extent suggested by Judge Weis in all equal protection cases, no differentiation could be found irrational. Moreover, this extra dose of deference is duplicative. As Judge Weis so artfully demonstrates in Part II of his opinion, deference to legislative and state decisionmakers is part of the rationale behind the rational relation standard itself. To add greater deference would totally eviscerate that standard. This is clearly not the intended result of rational relation scrutiny. See, e.g., Cleburne v. Cleburne Living Center, 473 U.S. 432, 105 S.Ct. 3249, 87 L.Ed.2d 313 (1985) (finding city ordinance unconstitutional under rational relation equal protection scrutiny). Rather, the rational relation standard requires at least one justification for the challenged differentiation. It may be that the justification is to be judged by the lax standard enunciated in Vance v. Bradley, 440 U.S. 93, 111, 99 S.Ct. 939, 949, 59 L.Ed.2d 171 (1979) (reviewing court must be convinced “that the legislative facts on which the classification is apparently based could not reasonably be conceived to be true by the governmental decisionmakers”), but one such justification must nonetheless exist.
Additionally, although Judge Weis does not rely on administrative convenience simpliciter, he justifies the differentiation at least partially on that basis. I agree with Judge Weis to the extent that he holds that administrative convenience may play a role in prompting differentiation when an independent reason also supports it. As I have explained supra, I find that the New Jersey Supreme Court has reached the conclusion that that the state-of-the-art defense should not be available to the asbestos manufacturers because the harms of asbestos were knowable to the industry. Administrative convenience is thus a justification for the differentiation because state courts are not proscribed by the Equal Protection Clause from refusing to hear a defense *1252that, as a matter of law, is doomed to fail. However, beyond such considerations, administrative convenience fails as a justification.
IV.
While states must be allowed to make their own policy judgments in matters such as tort law, they cannot use that discretion to arbitrarily discriminate against a class of litigants. Where a factual basis supports a differentiation, however, both the policy decision and the underlying factual determination deserves our deference.11 In an equal protection case, those challenging the state law “must convince the court that the legislative facts on which the classification is apparently based could not reasonably be conceived to be true by the governmental decisionmaker.” Vance v. Bradley, 440 U.S. 93, 111, 99 S.Ct. 939, 949, 59 L.Ed.2d 171 (1979), quoted in Minnesota v. Clover Leaf Creamery Co., 449 U.S. 456, 464, 101 S.Ct. 715, 724, 66 L.Ed.2d 659 (1981) (citing other cases). Because the court has found that at least one asbestos manufacturer had actual knowledge of the harms of asbestos at all relevant times, I do not believe that its determination of knowability “could not reasonably be conceived to be true by the governmental decisionmaker.” I therefore concur in this aspect of Part III of Judge Weis’s opinion.12
JAMES HUNTER, III, Circuit Judge, dissenting:
PRELIMINARY STATEMENT
In these diversity cases, we look to the law of New Jersey. New Jersey has embraced the full panoply of products liability law. See, e.g., Suter v. San Angelo Foundry & Mach. Co., 81 N.J. 150, 406 A.2d 140 (1979); Henningsen v. Bloomfield Motors, 32 N.J. 358, 161 A.2d 69 (1960). The “state-of-the-art” defense is normally available to defendants in products liability actions. This defense precludes liability where the manufacturer can prove that it did its work properly and produced the product in accordance with the practices and procedures appropriate to the product’s known dangers. Today this court has ruled that the manufacturers of one product may not use the state-of-the-art defense. That product is asbestos. The court has said to asbestos manufacturers: there are too many asbestos cases, these cases have clogged up the court calendars, schedules and statistics; the proof of “state-of-the-art” is too time-consuming and concerned with too many variables; and, in any event, we do not think you could prove the defense even if we gave you the chance. Thus, one narrow class of defendants is deprived of a potentially exculpatory defense in the interest of expediency and calendar control. The manufacturers of all other products — including Agent Orange, the Daikon Shield and DES — may use the defense, even if they are also clogging up the court calendar and causing statistical chaos. Only the asbestos industry is treated differently. This is just plain wrong and I dissent.
I.
It is beyond dispute that a classification that neither discriminates against a suspect *1253class nor impinges upon a fundamental right does not violate the Equal Protection Clause if it is rationally related to a legitimate governmental purpose. See United States Railroad Retirement Bd. v. Fritz, 449 U.S. 166, 175, 101 S.Ct. 453, 459, 66 L.Ed.2d 368 (1980). While it is beyond our authority to strike down laws simply because we conclude that they are unwise or inartfully drawn, id., neither are we required to give our stamp of approval to classifications that are arbitrary or wholly insubstantial. See Delaware River Basin Commission v. Bucks County Water & Sewer Authority, 641 F.2d 1087, 1097 (3d Cir.1981) (“the rationality standard is not ‘toothless’ ”). Whether or not the lawmakers’ “governmental purpose” under review must be clearly articulated by the lawmakers in order to be deemed legitimate cannot be definitively answered by reference to Supreme Court precedent. Compare Fritz, 449 U.S. at 170,101 S.Ct. at 456 with Fritz, 449 U.S. at 187, 101 S.Ct. at 466 (Brennan, J., dissenting) (citing Weinberger v. Wiesenfeld, 420 U.S. 636, 648 n. 16, 95 S.Ct. 1225, 1233 n. 16, 43 L.Ed.2d 514 (1975)). However, our own precedents teach us that “[s]o long as we are careful not to attribute to the legislature purposes which it cannot reasonably be understood to have entertained, we find that in examining the challenged provisions we may consider purposes advanced by counsel ... or suggested initially by ourselves.” Delaware River Basin, 641 F.2d at 1092. Our job then, is to determine whether the common law doctrine challenged in this case is in furtherance of a legitimate state purpose put forward by the New Jersey Supreme Court, the appellees, the United States District Court for the District of New Jersey, or ourselves, and whether the common law rule fits closely enough with any of those purposes that it can be said to be rationally related to them. I conclude that New Jersey’s common law rule creates a classification that is not rationally related to a legitimate governmental purpose, and therefore deprives asbestos manufacturers of the equal protection of the laws in violation of the Constitution of the United States. Before engaging in this constitutional analysis, I will review the line of cases in question.
II.
In Freund v. Cellofilm Properties, Inc., 87 N.J. 229, 432 A.2d 925 (1981), the New Jersey Supreme Court explained the difference between a failure to warn claim based on strict liability and a failure to warn claim based on negligence. Under either rubric, plaintiff will recover damages from the defendant if, given the dangerousness of the product, the manufacturer’s failure to provide warnings was unreasonable. The failure to provide warnings will be found unreasonable if warnings would have made the product safer. Negligence and strict liability actions differ in the manner in which the manufacturer’s knowledge of the product’s dangerousness is determined. Where a negligence plaintiff must prove that the defendant knew or should have known of the dangerousness of the product, the defendant’s knowledge of the product’s dangers is presumed as a matter of law in the strict liability action. Thus, “negligence is conduct-oriented, asking whether defendant’s actions were reasonable; strict liability is product-oriented, asking whether the product was reasonably safe for its foreseeable purposes.” Beshada v. Johns-Manville Corp., 90 N.J. 191, 200, 447 A.2d 539, 544 (1982) (citing Freund, 87 N.J. at 238, 432 A.2d at 929).
In Beshada v. Johns-Manville Corp., the New Jersey Supreme Court decided that the state-of-the-art defense is inapplicable to a strict liability failure-to-warn claim. Through the state-of-the-art defense, defendants are able to introduce evidence of the technology available at the time the product was manufactured and distributed in order to prove that the state of scientific knowledge was such that they could not have known of the product’s dangers when it was put into the stream of commerce. Id. 90 N.J. at 202, 447 A.2d at 545. The state-of-the-art defense is thus logically incompatible with Freund, which imputes knowledge to defendants as a matter of law. Moreover, because state-of-the-art is essentially a negligence defense, in that it *1254seeks to explain why defendants are not culpable, it seems inappropriate in a strict liability case, in which culpability is irrelevant. Beshada, 90 N.J. at 204, 447 A.2d at 546. The court admitted that “the phrase ‘duty to warn’ is misleading, [because i]t implies negligence concepts with their attendant focus on the reasonableness of defendant’s behavior.” Id. Nevertheless, the court committed itself to strict liability in failure-to-wam cases, and firmly rejected the state-of-the-art defense.
In response to a torrent of criticism, the New Jersey Supreme Court departed from its stand against the state-of-the-art defense in Feldman v. Lederle Laboratories, 97 N.J. 429, 479 A.2d 374 (1984). In Feldman, the court concluded that strict liability for failure to warn was inherently illogical. “A warning that a product may have an unknowable danger warns one of nothing.” Id. at 454, 479 A.2d at 387. The court did not eliminate the strict liability failure to warn cause of action, but chose instead to transform it. Where, in Freund and Beshada, strict liability actions were characterized by the imputation to defendant of knowledge of the product’s dangerousness, in Feldman, they became characterized by the applicable burden of proof. “In strict liability warning cases ... the defendant should properly bear the burden of proving that the information was not reasonably available or obtainable and that it therefore lacked actual or constructive knowledge of the defect.” Id. at 455-56, 479 A.2d at 388. The court refused to overrule Beshada, but did “restrict Beshada to the circumstances giving rise to its holding.” Id.1 It is in the context of this murky state of New Jersey “law” that we must address our problem.
III.
By “restrictpng] Beshada to the circumstances giving rise to its holding,” the New Jersey Supreme Court created the classification challenged here. My review of the Beshada-Feldman line of cases leads me to conclude that the challenged classification represents nothing more than an unprincipled, expedient, and ineffective response to widespread criticism of the Beshada doctrine combined with an unwillingness to give up the application of Beshada to asbestos manufacturers. If so, the Beshada-Feldman classification is undeniably arbitrary. Worse yet, I fear that the New Jersey Supreme Court’s sole purpose may have been to inflict a special punishment on asbestos manufacturers. The Constitution does not permit the New Jersey courts to level either an arbitrary or a punitive sanction against asbestos manufacturers (or the manufacturer of any other product, for that matter). Proper analysis of this classification under the rational basis test should disclose to us whether the classification is impermissibly arbitrary or punitive. I will address each of the justifications for the common law rule advanced by Judge Weis and Judge Becker: the asbestos manufacturers’ knowledge of the dangers of asbestos; case-management; and jury confusion.2
A.
Judge Weis and Judge Becker agree that the principal “legitimate state purpose” underlying the Beshada-Feldman doctrine is “the New Jersey Supreme Court[’s reasonable decision] to preclude [the] endless re-litigation of what was ‘knowable’ to the asbestos industry.” Becker at 1245-46. According to Judge Weis, “[t]he [Feld*1255man ] opinion suggests that [asbestos] manufacturers knew the dangers of asbestos and, consequently, the state-of-the-art defense [cannot] be sustained” by those manufacturers. Weis at 1241. Specifically, the New Jersey Supreme Court concluded in Feldman that “there were or may have been data or other information generally available, aside from scientific knowledge, that arguably could have alerted the manufacturer at an early stage ... to the dangers [of asbestos].” 97 N.J. at 456, 479 A.2d at 388. Thus, since the majority of this panel and the New Jersey Supreme Court claim to know for a certainty that all asbestos manufacturers either acted with full knowledge of the dangers of their product or could and should have obtained such knowledge, those manufacturers should not be permitted to litigate the knowledge question any further. I have trouble accepting this aim as a “legitimate state purpose” under the Equal Protection Clause, because I believe it violates the Due Process Clause of the Fourteenth Amendment. See U.S. Const, amend. XIV, Sec. 1 cl. 3.
This so-called legitimate state purpose can best be described as a de facto exercise of collateral estoppel without benefit of the procedural niceties. On the basis of the New Jersey Supreme Court’s declaratory dicta in Feldman that asbestos manufacturers did indeed possess culpable knowledge at all relevant times, those manufacturers must be collaterally estopped from litigating the question of their knowledge ever again. Judge Weis feels that this “legitimate state purpose” is a creature somehow related to collateral estoppel, as his discussion of a series of fifth circuit opinions dealing with the procedural complexity of collateral estoppel in massive asbestos litigation demonstrates. His review of the line of cases culminating in the fifth circuit’s opinion in Hardy v. Johns-Manville Sales Corp., (“Hardy II”), 681 F.2d 334 (5th Cir.1982), leads him to conclude that “Beshada’s broad language, when applied to the concrete facts of asbestos litigation and read together with Fischer [v. Johns-Manville Corp., 103 N.J. 643, 512 A.2d 466 (1986)], is not completely divorced from reality.” Weis at 1242. I am not sure what this statement means, but it is clear that Judge Weis has concluded that, since plaintiffs have experienced difficulties in prevailing on their collateral estoppel motions heretofore, see, e.g., Hardy II, the courts may constitutionally grant all plaintiffs the benefit of collateral estoppel through application of their own wisdom rather than the traditional channels of legal process. I believe that Hardy II teaches a much different lesson than this. I read Hardy II as reiterating the well-worn legal principle that defendants will be collaterally estopped from relitigating factual issues where
The party asserting the estoppel [can] show that: (1) the issue to be concluded is identical to that involved in the prior action; (2) in the prior action the issue was “actually litigated”; and (3) the determination made of the issue in the pri- or action [was] necessary and essential to the resulting judgment.
681 F.2d at 341. The grant of a collateral estoppel motion where these conditions are not met violates a party’s right to due process. See Blonder-Tongue Laboratories, Inc. v. University of Illinois Found., 402 U.S. 313, 329, 91 S.Ct. 1434, 1443, 28 L.Ed.2d 788 (1971). A fortiori, Judge Weis’ sua sponte grant of collateral estoppel to all future asbestos plaintiffs violates the Due Process Clause and thus cannot provide a legitimate state purpose under the rationality test. Indeed, “to use collateral estoppel in this context would be to ‘elevate judicial expedience over considerations of justice and fair play.’ ” Weis at 1242 (quoting Hardy II, 681 F.2d at 348).
Judge Weis points out that, after its decision in Feldman, the New Jersey Supreme Court found that at least one asbestos manufacturer — Johns-Manville — did indeed have actual knowledge of the dangers of asbestos as early as the 1930’s. See Fischer v. Johns-Manville Corp., 103 N.J. 643, 512 A.2d 466 (1986). I cannot understand how the subsequent adjudication of an unrelated party’s culpable knowledge can render the Feldman court’s attribution of knowability to all asbestos manufacturers *1256by judicial fiat constitutional. This conclusion very simply does not satisfy the requirements of the collateral estoppel doctrine. It is a fundamental tenet of our procedural law that the court’s factual findings in the Fischer case can only be used against Johns-Manville. “The requirement that a person against whom the conclusive effect of a judgment is invoked must be a party or a privy to the prior judgment ... has been repeatedly affirmed[.]” Hardy II, 681 F.2d at 338 (citations omitted). Accord Parklane Hosiery Co. v. Shore, 439 U.S. 322, 327 n. 7, 99 S.Ct. 645, 649 n. 7, 58 L.Ed.2d 552 (1979) (“It is a violation of due process for a judgment to be binding on a litigant who was not a party or a privy and therefore has never had an opportunity to be heard.”) (citations omitted). Thus, with or without the Fischer court’s adjudication of Johns-Manville’s actual knowledge, the principle justification for the challenged classification asserted by Judge Weis is very simply not a legitimate state purpose for the New Jersey Supreme Court’s discriminatory treatment of asbestos manufacturers.
B.
Judge Becker attempts to circumvent the collateral estoppel problem by characterizing the New Jersey Supreme Court’s conclusion in Feldman “that the hazards of asbestos exposure were knowable to the industry at all relevant times,” Becker at 1245, as a “legislative fact” of which the court may freely take judicial notice. I believe that Judge Becker’s argument fails because it relies on a misapprehension of the nature of the legislative facts that may be judicially noticed under the common law of evidence. There are two kinds of fact of which courts may take judicial notice: “adjudicative” and “legislative.”3 “Adjudicative facts are simply the facts of the particular case.” Fed.R.Evid. 201 advisory committee’s note. “ ‘Adjudicative facts,’ ... are the ultimate facts in the case, plus those evidential facts that are sufficiently central to the controversy that they should be left to the jury unless clearly indisputable.” 21 C. Wright & K. Graham, Federal Practice and Procedure § 5103, at 478 (1977). “Legislative facts, on the other hand, are those which have relevance to legal reasoning and the lawmaking process, whether in the formulation of a legal principle or ruling by a judge or court or in the enactment of a legislative body.” Fed.R. Evid. 201 advisory committee’s note.
My understanding of what constitutes a “legislative fact” seems to be at odds with Judge Becker’s. In my view, legislative facts are those social, economic and philosophical facts upon which we rely to fashion just, appropriate and suitable rules of law. We derive these legislative facts from a common-sense assessment of how modern-day Americans live and view their world. Legislative facts are broad conceptions or beliefs about the interaction of law and society that inform judicial policy-making. As such, these legislative facts may be truisms, or may be wholly incapable of proof or disproof. The United States Supreme Court regularly takes judicial notice of legislative facts to resolve cases involving major policy determinations in the areas of constitutional law and criminal procedure. See, e.g., Church of Latter-Day Saints v. Amos, — U.S. -, 107 S.Ct. 2862, 2875, 97 L.Ed.2d 273 (1987) (O’Con-nor, J., concurring) (“The determination of whether the objective observer will perceive an endorsement of religion ‘is not a question of simple historical fact. Although evidentiary submissions may help answer it, the question is, like the question whether racial or sex-based classifications communicate an invidious message, in large part a legal question to be answered on the basis of judicial interpretation of social facts.’ ”) (quoting Lynch v. Donnelly, 465 U.S. 668, 693-94, 104 S.Ct. 1355, 1369-70, 79 L.Ed.2d 604 (1984) (O’Connor, J., concurring)); United States v. Leon, 468 U.S. 897, 918-21, 104 S.Ct. 3405, 3418-19, 82 L.Ed.2d 677 (1984) (discussing the probable deterrent effect of the exclusion*1257ary rule on police officers); Detroit Edison Co. v. NLRB, 440 U.S. 301, 318, 99 S.Ct. 1123, 1132, 59 L.Ed.2d 333 (1979) (“The sensitivity of any human being to disclosure of information that may be taken to bear on his or her competence is sufficiently well known to be an appropriate subject of judicial notice.” (footnote omitted)); Hawkins v. United States, 358 U.S. 74, 77-78, 79 S.Ct. 136, 138, 3 L.Ed.2d 125 (1958) (“The basic reason the law has refused to pit wife against husband or husband against wife in a trial where life or liberty is at stake was a belief that such a policy was necessary to foster family peace, not only for the benefit of husband, wife and children, but for the benefit of the public as well____Adverse testimony given in criminal proceedings would, we think, be likely to destroy any marriage.”); Brown v. Board of Education, 347 U.S. 483, 494, 74 S.Ct. 686, 691, 98 L.Ed. 873 (1954) (“To separate [Negro children] from others of similar age and qualifications solely because of their race generates a feeling of inferiority as to their status in the community that may affect their hearts and minds in a way unlikely ever to be undone.”).
I realize, of course, that not all judicially noticed legislative facts involve the kinds of fundamental and glamorous policy questions central to the cases cited in the previous paragraph. Indeed, the observations made through the medium of judicial notice are just as likely to be obvious and mundane. See e.g., Kessler Inst, for Rehabilitation v. NLRB, 669 F.2d 138, 141 (3d Cir.1982) (court takes judicial notice of “the delays in the postal system which have been increasing over the years”); Neeld v. National Hockey League, 594 F.2d 1297, 1300 (9th Cir.1979) (court takes judicial notice of dangers posed by one-eyed hockey players to uphold rule barring such players from professional hockey teams). Furthermore, courts may take judicial notice of legislative facts for reasons totally unrelated to the development of policy. For example, courts regularly take judicial notice of legislative facts in order to determine whether or not particular activities come within specified statutory prohibitions. See, e.g., United States v. Gould, 536 F.2d 216, 220-21 (8th Cir.1976) (court takes judicial notice of fact that cocaine hydrochloride is derivative of the cocoa leaf in order to determine whether or not it is a Schedule II controlled substance). However, I have never heard of any case in which the resolution of an ultimate fact— e.g., the innocence or culpability of a products liability defendant — was reached through the judicial notice of legislative facts.
The asbestos manufacturers’ actual or constructive knowledge of the potential harms of asbestos does not bear any resemblance to the legislative facts judicially noticed in the cases cited above or in any other case I have found. The knowledge question does not rest upon a generalized assumption or conclusion about the state of things in order to come to a rational policy decision; rather, it demands the resolution of a hard, cold, specific factual dispute that is central to the products liability litigation between the injured plaintiffs and defendant asbestos manufacturers before the court in each case. The question posed involves a determination of who knew what, when they knew it, and when they should have known it. To my way of thinking, this has every indicia of an adjudicative fact.
When a court ... finds facts concerning the immediate parties — who did what, where, when, how, and with what motive or intent — the court ... is performing an adjudicative function, and the facts are conveniently called adjudicative facts____
Stated in other terms, the adjudicative facts are those to which the law is applied in the process of adjudication. They are the facts that normally go to the jury in a jury case. They relate to the parties, their activities, their properties, their businesses.
2 K. Davis, Administrative Law Treatise § 15.03, at 353 (1958) (quoted in Gould, 536 F.2d at 219).
Nor am I persuaded by Judge Becker’s assertion in footnote 6 of the concurring opinion that “if the knowability of asbestos’ harms were purely an adjudicative fact, it could not be the subject of a legisla*1258tive fact-finding.” Not every fact found by a legislature is necessarily legislative. Legislatures can and often do find facts that would undeniably be classified as “adjudicative” if they had been found in a judicial proceeding. For instance, in 1979 the House Select Committee on Assassinations found that
A. Lee Harvey Oswald fired three shots at President John F. Kennedy. The second and third shots he fired struck the President. The third shot he fired killed the President.
1. President Kennedy was struck by two rifle shots fired from behind him.
2. The shots that struck President Kennedy from behind him were fired from the sixth floor window of the southeast corner of the Texas School Book Depository building.
3. Lee Harvey Oswald owned the rifle that was used to fire the shots from the sixth floor window of the southeast corner of the Texas School Book Depository building.
4. Lee Harvey Oswald, shortly before the assassination, had access to and was present on the sixth floor of the Texas School Book Depository building.
5. Lee Harvey Oswald’s other actions tend to support the conclusion that he assassinated President Kennedy____
H.R.Rep. No. 1828, 95th Cong., 2d Sess. 1 (1979). Since these, facts were found by a legislature, Judge Becker would characterize them as legislative facts. Thus, if Judge Becker’s definition of “legislative fact” is correct, then we would have to reach the extraordinary conclusion that a court trying Lee Harvey Oswald for the murder of President Kennedy could take judicial notice of the “legislative fact” that Lee Harvey Oswald killed President Kennedy, and thereby the court could avoid the inconvenience of requiring the government to prove the ultimate fact in the case. It is beyond peradventure that such a use of judicial notice of “legislative facts” would constitute not only a denial of due process but also a violation of the bill of attainder clause. U.S. Const, art. I, § 10, cl. 1. Notwithstanding any legislative pronouncement on the matter and notwithstanding the firmness of the trial judge’s belief of Oswald’s guilt, the court could not deprive Oswald of his day in court simply because “he doesn’t have a chance of winning, anyway.” However, this is precisely the way that the Beskada-Feldman rule treats asbestos defendants. I believe that Judge Becker’s attempt to characterize the know-ability of asbestos harms as a “legislative fact” is misguided and does not justify our countenancing this pernicious rule.
My conclusion that the knowledge question requires the determination of an adjudicative rather than a legislative fact is buoyed by analogous decisions reached in two other asbestos cases. See Hardy II, 681 F.2d at 347-48; Laster v. Celotex Cory., 587 F.Supp. 542 (S.D.Ohio 1984). In Hardy and Laster, the courts were asked to take judicial notice of a fact closely related to the one before us — whether or not exposure to asbestos causes cancer, pleural mesothelioma or asbestosis. Without hesitation, both courts classified the fact to be noticed as an adjudicative fact. See Laster, 587 F.Supp. at 543 (“Clearly, the facts pertaining to whether asbestosis and mesothelioma are caused by exposure to asbestos are ‘adjudicative facts’ under Rule 201.”). The causation issue clearly requires the resolution of an adjudicative (rather than a legislative) fact for the same reasons that the knowledge or knowability issue does. First, like the question of knowledge, the question of asbestos’ role in the development of certain diseases “relate[s] to the [immediate] parties, their activities, [and] their businesses,” and thus must be classified as an adjudicative fact. More importantly, disease causation and culpable knowledge are both ultimate facts in products liability litigation,4 and cannot *1259be properly resolved by judicial notice of legislative facts.5 See Korematsu v. United States, 584 F. Supp. 1406, 1415 (N.D. Cal.1984). If judicial notice can be taken of such ultimate facts at all, it can only be within the strictures of Fed.R.Evid. 201.6
A court may only take judicial notice of an adjudicative fact where that fact is “generally known within the territorial jurisdiction of the trial court[; and] ... capable of accurate and ready determination by resort to sources whose accuracy cannot reasonably be questioned.” Fed.R.Evid. 201. Even assuming that some potential hazards of asbestos were known in the 1930’s, it is by no means “generally known” or indisputable that the various specific uses of asbestos (e.g., the use of asbestos to insulate automobile brake linings) were all known to be hazardous to *1260workers or consumers as far back as the 1930’s. The fifth circuit and the Southern District of Ohio were not even willing to take judicial notice of the fact that asbestos necessarily causes the various diseases with which it has been associated. According to the fifth circuit in Hardy II, “[t]he proposition that asbestos causes cancer, because it is inextricably linked to a host of disputed issues— ... [including whether or not] this manufacturer [was] reasonably unaware of the asbestos hazards in 1964— is not at present so self-evident a proposition as to be subject to judicial notice.” Hardy II, 681 F.2d at 347-48. Similarly, the issue of knowledge or knowability is “not at present so self-evident a proposition to be subject to judicial notice.” Therefore, the issue of knowledge or knowability cannot properly be judicially noticed as an adjudicative fact under Fed.R.Evid. 201.
Fed.R.Evid. 201 also provides that “[a] party is entitled upon timely request to an opportunity to be heard as to the propriety of taking judicial notice and the tenor of the matter noticed.” The parties in this case were not given such an opportunity. Judge Becker argues that “the affected industry [i.e., asbestos manufacturers] had relevant opportunities to respond to the New Jersey Supreme Court’s determination of the pertinent legislative facts.” Becker at 1249. The significant opportunities cited by Judge Becker include the industry’s participation in two fifth circuit cases, Hardy and Borel, and in two cases decided by the New Jersey Supreme Court after the Beshada-Feldman classification had been developed, Fischer and In re Asbestos Litigation Venued in Middlesex County, 99 N.J. 201, 491 A.2d 700 (1984). Judge Becker’s conclusion is unprecedented. I have never heard of any case that has permitted a party’s “opportunity to be heard” to be satisfied in subsequent litigation before the same court, or — even more incredibly — in prior litigation before an unrelated tribunal. Furthermore, even if Judge Becker’s remarkable conclusion were to be accepted, it does not satisfy the requirements of Fed. R.Evid. 201, which requires the court to provide “an opportunity to be heard as to the propriety of taking judicial notice,” Fed.R.Evid. 201 (emphasis added), not simply an opportunity to be heard on the merits. I find the New Jersey Supreme Court’s failure to give the asbestos manufacturers an opportunity to be heard as to the propriety of taking judicial notice understandable, in light of the fact that they probably did not realize they were taking judicial notice (of either a legislative or an adjudicative fact) until so informed by Judge Becker.
In sum, Judge Becker has grasped the notion of legislative facts — a notion that this court has viewed with the greatest of skepticism, City of New Brunswick, 686 F.2d at 131 — in an attempt to fix the broken shell of the Beshada decision and to patch-over New Jersey’s violation of the Equal Protection and Due Process clauses. “All the King’s horses ...”
C.
Judge Weis observes that although “[administrative convenience standing alone is not an adequate ground for the elimination of a substantive defense ... [the court] cannot help but be conscious of the extraordinary size of the asbestos personal injury litigation.” Weis at 1243. While Judge Weis has rejected explicit reliance on the case management rationale as an independent justification for the Beshada-Feldman classification, he continues to believe that it buttresses the other justifications he has put forward. Because I find none of the other arguments advanced by Judge Weis or Judge Becker persuasive, I will examine the administrative convenience argument separately to determine whether it provides any support for the challenged classification. I find that it does not. The simple fact that elimination of a defense for one group saves time is not enough to justify discriminating between that group and another that is similarly situated. The case management rationale simply cannot provide a legitimate basis for depriving this one class of manufacturers of an exculpatory defense, particularly in light of the massive expenditure of court time and judicial energy required by DES, Daikon Shield, Agent Or*1261ange and miscellaneous pharmaceutical and environmental disaster litigation. I do not mean to imply that the Fourteenth Amendment bars any effort to address a perceived crisis in a particular kind of litigation; legislatures all over the country have demonstrated that such crises can be dealt with constitutionally by developing means that are rationally related to the aim of averting them.
Judge Weis looks to the nationwide enactment of no-fault automobile insurance statutes to provide support by analogy for the Beshada-Feldman classification. Due to the extreme diversity of these statutes, I have limited myself to an examination of the laws of New Jersey and Pennsylvania to determine whether Judge Weis’ reliance is justified. The New Jersey Automotive Reparation Act, N.J. Stat. Ann. 39:6A-8 (West 1973 & Supp.1986), “bars suit for bodily injury ... unless the injury either is permanent as opposed to 'soft-tissue,’ or requires treatment with a cost or equivalent value of $200.” Rybeck v. Rybeck, 141 N.J. Super. 481, 488, 358 A.2d 828, 832 (Law Div.1976). Similarly, the Pennsylvania No-fault Motor Vehicle Insurance Act, 40 Pa.Stat.Ann. § 1009.301(a)(5)(B) (repealed 1984) (Purdon Supp.1987), permitted only those victims whose medical expenses exceed $750 to sue in tort. Singer v. Sheppard, 464 Pa. 387, 403, 346 A.2d 897, 905 (1975). Both statutes were upheld in the face of numerous constitutional challenges, including the discriminatory deprivation of tort recovery for victims of lesser injuries. The state courts determined that such a distinction served a legitimate state purpose:
[A]uto accident injury claims were reasonably seen by the Legislature to present a special problem requiring reform. It was that class of claims that created the calendar congestion, that ate up the investigative and administrative insurance dollar, that represented justice worst delayed, and that most invited compensation without fault determination.
Rybeck, 141 N.J. Super, at 498, 358 A.2d at 837. Further, the state courts were convinced that the discrimination against victims of minor injuries created a classification that was rationally related to the achievement of the legitimate state purpose. See Singer, 464 Pa. at 403, 346 A.2d at 905. So am I. Wresting the adjudication of minor injuries from state courts and assuring those victims of a certain and sure recovery through the insurance system is rationally related to the state’s purpose of relieving an overburdened court docket. This method of case-management bears no resemblance to one that simply deprives one class of litigants of a potentially exculpatory defense. Any victim whose injuries reach the threshhold amount is entitled to go to court and present his negligence claim like any other tort plaintiff; and the defendant is entitled to use all common law defenses available. Thus, dockets are streamlined without either party losing any legal right.
According to Judge Weis, workers’ compensation programs share with no-fault auto insurance plans “[the common goal] of simplifying the fact-finding process and thus making it easier for victims to recover.” Weis at 1243. This is not my understanding of the goal of workers’ compensation statutes. I had always learned that they were developed as a kind of social insurance based on the policy decision that the compensation of workers’ injuries should be absorbed by the employer as a cost of doing business. See New York Cent. R. Co. v. White, 243 U.S. 188, 205, 37 S.Ct. 247, 253, 61 L.Ed. 667 (1917). Furthermore, I do not think that the classification created by the workers’ compensation statutes even represents a useful analogy to the Beshada-Feldman classification. Workers’ compensation statutes create a classification based on the ongoing relationship between the employer and the employee, not on the personal (or corporate) identities of those parties outside of that relationship. Manufacturers of asbestos owe no greater duty to their employees under these statutes than do purveyors of paper clips. Workers’ compensation statutes are analogous to inter-familial tort immunity and guest-host statutes, but not *1262to the discriminatory treatment of asbestos manufacturers challenged here.7
D.
The final “legitimate state purpose” advanced by Judge Weis is the reduction of jury confusion through the complete elimination of the state-of-the-art defense in asbestos cases. The Beshada court had observed that “vast confusion ... is virtually certain to arise from any attempt to deal in a trial setting with the concept of scientific knowability,” Beshada, 90 N.J. at 207, 447 A.2d at 548, and that “discussion of state-of-the-art could easily confuse juries into believing that blame-worthiness is at issue.” Id. This confusion does not stem from the nature of asbestos litigation, however, but from the interplay of the strict liability failure-to-warn action with the state-of-the-art defense. See Beshada, 90 N.J. at 204, 447 A.2d at 546. The likelihood of jury confusion provides a justification for one of two paths not taken by the New Jersey Supreme Court: (a) the elimination of a strict liability cause of action requiring the defendant to warn of the unknown and unknowable through the reversal of Freund and Beshada; or (b) the continued unavailability of the state-of-the-art defense to any defendant in a strict liability failure-to-warn action. The New Jersey Supreme Court balked and refused to make this difficult choice. The jury confusion rationale is, in my opinion, totally undermined by the New Jersey Supreme Court’s recent decision in Fischer v. Johns-Manville Corp., 103 N.J. 643, 512 A.2d 466 (1986), holding that punitive damages were available to plaintiffs against asbestos defendants in strict liability failure-to-wam cases. The evidence introduced to determine a punitive damages award is the very same evidence — the state of scientific knowledge at the time of manufacture— that is relied upon in a state-of-the-art defense. Thus, if jury confusion is to be minimized by the exclusion of such evidence, the Fischer rule is sure to restore that confusion by the introduction of such evidence. I think the Fischer holding is fatal to the jury confusion rationale.
CONCLUSION
My position by no means constitutes the establishment of tort law for New Jersey. It simply means that since New Jersey common law is subject to the operation of the Equal Protection Clause, that clause quite plainly works to bar New Jersey from depriving asbestos manufacturers alone of the state-of-the-art defense, a defense that is available to all other manufacturers in products liability cases. The Supreme Court of New Jersey simply cannot establish a law that violates the Federal Constitution. And this court cannot look the other way and ignore its duty to defend that Constitution.
My answer to the question presented is that the judicially imposed denial of the state-of-the-art defense to manufacturers of asbestos-containing products does constitute a violation of the Equal Protection Clause of the Constitution.
. For the reasons I have not joined in the balance of Part III, see infra, at 1251-52.
. The state-of-the-art defense, as the name implies, does not merely rest on an inquiry concerning what a manufacturer knew. Rather, the defense is intended to limit liability to those dangers of which the seller "has knowledge, or by application of reasonable, developed human skill and foresight should have knowledge." Restatement (Second) of Torts, § 402A comment j. Implicit in the defense, therefore, is the obligation to investigate and keep abreast of recent developments — an obligation imposed under the aegis of "constructive knowledge!, which] may encompass virtually all information that is in the public domain.” Green, The Inability of Offensive Collateral Estoppel to Fulfill Its Promise: An Examination of Estoppel in Asbestos Litigation, 70 Iowa L. Rev. 141, 190 n. 276. The relevant determination is thus not what was known, but what was knowable. See also infra at 1247.
. Hardy, in turn, rests to a significant degree on Borel v. Fibreboard Paper Products Corp., 493 F.2d 1076 (5th Cir.1973), cert, denied, 419 U.S. 869, 95 S.Ct. 127, 42 L.Ed.2d 107 (1974).
. For a definition of legislative facts and demonstration that knowability is a legislative fact, see infra at Part I. B.
. See, e.g., Forte Towers, Inc. v. City of Miami Beach, 360 So.2d 81, 82 (Fla.App.1978) (precluding relitigation of factual predicate).
. Judge Hunter believes that the state-of-the-art determination concerning the knowability of asbestos’ harms is an adjudicative fact. Arguing by analogy, he finds the issue similar to whether asbestosis and mesothelioma were caused by asbestos exposure in a particular case, held to be a matter of adjudicative fact in Laster v. Celotex, 587 F.Supp. 542 (S.D.Ohio 1984). The analogy fails, however, because the development of these diseases relates to the peculiar conditions under which a person is exposed— for example, whether the person smokes and the ambient air concentration of asbestos dust. See Laster, 587 F.Supp. at 543. In the context of what was knowable by the asbestos industry, however, the peculiarities of a specific manufacturer are irrelevant. See supra at 1245 n. 2,1247-1248. What may have been known is the heart of the state-of-the-art inquiry; in the causation inquiry, by contrast, "judicial notice that the inhalation of asbestos may cause asbestosis under certain conditions would have no appreciable impact.” Laster, 587 F.Supp. at 544 (emphasis in original).
In any event, if the knowability of asbestos’ harms were purely an adjudicative fact, it could not be the subject of a legislative fact-finding. The contrary is true, however, for the Senate Committee on Labor and Public Welfare made a specific legislative finding concerning the know-ability of asbestos' harms when deciding to enact the Occupational Safety and Health Act.
Asbestos is another material which continues to destroy the lives of workers. For 40 years it has been known that exposure to asbestos caused the severe lung scarring called asbestosis. S.Rep. No. 1282, 91st Cong., 2d Sess., reprinted in 1970 U.S. Code Cong, and Admin. News, 91st Cong., 2d Sess. 5177, 5178.
Judge Hunter also argues that "culpable knowledge,” like disease causation, is an ultimate fact that cannot be resolved on the basis of legislative facts. First, I am perplexed by the phrase "culpable knowledge,” for the state-of-the-art inquiry does not look at what was known by the individual defendant, but what could have been known in the industry by the application of reasonable diligence. See supra at 1245 n. 2, 1247-1248. Second, I fail to understand why legislative fact-finding cannot resolve an ultimate issue; indeed, in the Supreme Court cases Judge Hunter cites as finding legislative facts, those facts seem to relate only to ultimate issues. See Dissent at 1257; Church of Latter-Day Saints v. Amos, — U.S. -, 107 S.Ct. 2862, 2875, 97 L.Ed.2d 273 (1987) (O’Connor, J., concurring) (perception of endorsement of religion in Establishment Clause case); United States v. Leon, 468 U.S. 897, 918-21, 104 S.Ct. 3405, 3418-19, 82 L.Ed.2d 677 (1984) (deterrent effect of exclusionary rule); compare Hawkins, 358 U.S. at 77-78, 79 S.Ct. at 138 (allowing adverse spousal privilege because adverse testimony found to hurt marriage) with Trammel v. United States, 445 U.S. 40, 52, 100 S.Ct. 906, 913, 63 L.Ed.2d 186 (overruling Hawkins because "[wjhen one spouse is willing to testify against the other in a criminal proceeding ... their relationship is almost certainly in disrepair”). Finally, I disagree with the characterization of the state-of-the-art defense as the ultimate issue in a products liability case. It is no more or less ultimate a determinant of liability than, to take Judge Hunter’s other example, causation.
. Thus the Supreme Court long ago found a critical difference between Londoner v. Denver, 210 U.S. 373, 28 S.Ct. 708, 52 L.Ed. 1103 (1908), in which a small number of people were exceptionally affected upon individual grounds by a tax increase and therefore should have been afforded individual due process, and Bi-Metallic Inv. Co. v. State Bd. of Equalization, 239 U.S. 441, 36 S.Ct. 141, 60 L.Ed. 372 (1915), in which the court found that a general tax increase could be predicated without individualized fact-finding.
. Specifically, although originally a subject to be re-established by expert testimony in every case, courts long ago came to give blood-grouping tests conclusive effect in paternity suits. See, e.g., Jordan v. Mace, 144 Me. 351, 69 A.2d 670 (1949); see generally Ross, The Value of Blood Tests as Evidence in Paternity Cases, 71 Harv. L.Rev. 466 (1958).
. See, e.g., Commonwealth v. Sullivan, 146 Mass. 142, 15 N.E. 491 (1888) (finding on the basis of a prior jury decision that, as a matter of law, a certain game was a regulated "lottery”), cited in Korn, Law, Fact, and Science in the Courts, 66 Colum. L.Rev. 1080, 1104 (1966).
. As one commentator has noted in the context of legislative fact-finding by a legislative body:
Given the bent to test due process according to the information available to the legislature, the truth-content of the data is not directly relevant. The question is whether sufficient data exists which could influence a reasonable legislature to act, not whether ultimately this data is true.
E. Cleary, McCormick on Evidence § 331, at 768 (2d ed. 1972) (footnote ommitted); see also Bums Baking Co. v. Bryan, 264 U.S. 504, 517, 44 S.Ct. 412, 415, 68 I-.Ed. 813 (1924) (Brandeis, J., dissenting).
. Because the determination of legislative facts is thus a component of fashioning a rule of law, the clearly erroneous standard of Rule 52(a) does not apply to review of a federal court’s findings concerning legislative facts. See Lock-hart v. McCree, 476 U.S. 162, 106 S.Ct. 1758, 1762 n. 3, 90 L.Ed.2d 137 (1986); see generally Dunagin v. City of Oxford, Miss., 718 F.2d 738, 748-49 n. 8 (5th Cir.1983) (en banc) (plurality opinion of Reavley, J.).
. Another reason exists for deferring to a state supreme court's determination of legislative facts. Because a state court must make a decision concerning legislative facts despite imperfect information, its decision deserves deference. As Justice Blackmun's concurrence in United States v. Leon, 468 U.S. 897, 927, 104 S.Ct. 3405, 3423, 82 L.Ed.2d 677 (1984) (Black-mun, J., concurring), noted concerning the issues in that case,
I see no way to avoid making an empirical judgment of this sort and I am satisfied that the Court has made the correct one on the information before it. Like all courts, we face institutional limitations on our ability to gather information about "legislative facts”____ Nonetheless, we cannot escape the responsibility to decide the question before us, however imperfect our information may be, and I am prepared to join the Court on the information now at hand.
. The meaning of this sentence is by no means clear on its face. Subsequent case law was has shown that “restricting Beshada to the circumstances giving rise to its holding" means that asbestos defendants are not permitted to use the state-of-the-art defense. See In re Asbestos Litigation Ventted in Middlesex County, 99 N.J. 201, 491 A.2d 700 (1984). The restriction has not yet been applied to any other manufacturer, and there is no indication that it will be.
. In Beshada, the court articulated its reason for precluding the use of the state-of-the-art defense in all strict liability failure-to-warn cases. The court justified its ruling on the basis of traditional strict liability policy concerns: deep pockets, risk spreading and accident avoidance incentives. The articulation of purposes underlying the Beshada rule does not, however, provide any basis for the Feldman distinction, which limits the Beshada rule to one class of defendants alone.
. The adjudicative fact/legislative fact dichotomy was first articulated by Professor Kenneth Culp Davis in 1942. See Davis, An Approach to Problems of Evidence in the Administrative Process, 55 Harv. L. Rev. 364, 404-07 (1942).
. In footnote 6 of the concurring opinion, Judge Becker urges that the knowability vel non of the harms of a particular product is not an ultimate fact in a strict liability — failure to warn case. I respectfully disagree. " ‘Ultimate’ facts are those which the law makes the occasion for imposing its sanctions.” The Evergreens v. Nunan, 141 F.2d 927, 928 (2d Cir.1944) (Hand, J.) "Ultimate facts" describe those specific historical or narrative facts which are decisive to the *1259outcome of a particular case because their proof is absolutely necessary to establish the elements of a claim or defense. See Winters v. Lavine, 574 F.2d 46, 57-58 n. 12 (2d Cir.1978). Clearly, the unknowability of harms associated with a particular product is the ultimate narrative or historic fact that a producer of that product must prove in order to establish the state-of-the-art defense.
. Judge Becker argues that adjudicative facts may be “elevated to legislative facts" and relied upon to shape judicial decisions in later litigation. See Becker at 1250. I express no opinion on this statement as a general matter. However, I find the proposition faulty as applied to the instant case for two reasons. First, prior to its development of the Beshada-Feldman classification, the New Jersey Supreme Court had never found as an adjudicative fact that any asbestos manufacturer had knowledge of the potential harms of asbestos. Admittedly, such a finding was made in Fischer, but that case was decided two years after Feldman, and could not have provided a basis for the Beshada-Feldman classification. To the extent that the New Jersey Supreme Court may have relied on the Borel-Hardy line of cases, such reliance is undercut by the fifth circuit’s rejection in Hardy II of a lower court's effort to rely on facts adjudicated in previous cases by means of collateral estoppel or judicial notice. Furthermore, I do not think adjudicative facts can possibly be elevated to the level of legislative facts unless a certain degree of consensus has been reached in the earlier litigation. Otherwise, judges may be tempted to reach into an uncertain and confusing mass of conflicting decisions and — shunning the restraints imposed by the laws of evidence and collateral estoppel — pull out the result most to their liking. Such procedures run afoul of the proper presentation of proof in our adversary system. I hasten to remind Judge Becker that "[t]he doctrines of ‘legislative facts’ and 'judicial notice’ are not talismans by which gaps in a litigant’s evidentiary presentation may be repaired on appeal.” City of New Brunswick v. Borough of Milltown, 686 F.2d 120,131 n. 15 (3d Cir.1982), cert, denied, 459 U.S. 1201, 103 S.Ct. 1184, 75 L.Ed.2d 431 (1983).
For similar reasons, I think that the statement by Professor Davis quoted by Judge Becker on page 1247 of the concurrence does not support his judicial notice thesis. First of all, I do not believe that Professor Davis was thinking about the propriety of applying a fact adjudicated in one or a handful of products liability cases against an entire industry. I think his concern was with the ability of judicial and administrative tribunals to fashion regulatory rules on the basis of generalized observations about a specific industry. Second, I cannot believe that Professor Davis would approve of the development of an industry-wide "legislative fact” on the basis of a fact adjudicated about only one company in the industry. I am certain that he would require some broader consensus about industry-wide knowledge before elevating an adjudicative fact about Johns-Manville to a legislative fact about all asbestos manufacturers.
. Judge Becker takes issue with my conclusion that Fed.R.Evid. 201 applies to the cases sub judice. Becker at 1248. Assuming (as Judge Becker does) that the Beshada-Feldman rule involves the taking of judicial notice of the know-ability of asbestos harms, and assuming (as I have demonstrated) that the knowability of asbestos harms is an adjudicative fact, then it seems clear that Fed.R.Evid. 201 applies to the cases sub judice. I note that it is at least arguable that Fed.R.Evid. 201 defines the minimum process due a litigant, under the Vth and XIVth amendments, before a court can take judicial notice of an adjudicative fact; see Eidt v. City of Natchez, 421 So.2d 1225 (Miss.1982); however, it is not necessary here to resolve whether Fed. R.Evid. 201 describes minimal constitutional protections. All of the cases sub judice are in federal district court, and, therefore, the Federal Rules of Evidence are applicable. Gallup v. Caldwell, 120 F.2d 90, 94 (3d Cir.1941) (“This rule as to judicial notice is not affected by Erie R.R. v. Tompkins, 304 U.S. 64, 58 S.Ct. 817, 82 L.Ed. 1188 (1938).’’); Laster, 587 F.Supp at 542; 19 C. Wright, A. Miller & E. Cooper, Federal Practice and Procedure § 4512 at 190 (1982) ("Of all the procedural and quasi-procedural rules that are applied in federal courts, the Federal Rules of Evidence are the least affected by the Erie doctrine."). Thus, whether or not "Rule 201 binds any state court,” Becker at 1250, it binds the federal court in which these cases await adjudication. If (as Judge Becker contends) the doctrine of judicial notice lies at the bottom of this controversy, then Rule 201 supplies the rule of decision.
. Judge Weis’ observation that "the doctrine of strict liability ... is in itself a classification that imposes discriminatory liability on a particular group of defendants," Weis at 1243, is similarly misleading. Restatement (Second) of Torts, § 402A (1965), sets out a special standard of strict liability to be used by consumers against retailers and manufacturers. This special standard is not discriminatorily applied to such defendants because of who they are, however, but because of what they do, i.e., put products into the stream of commerce. It is the activity, not the particular defendant, that triggers the application of strict liability. If the same defendant injures an individual through a means other than product manufacture and distribution, in a mundane slip and fall action, for example, that defendant will be liable for negligence only. Differential regulation of different kinds of activities pervades tort law and all other areas of public and private law, from criminal law to conflict of interests. Furthermore, strict liability is not exclusively reserved for use against miscreant manufacturers and retailers, but is also available for use against participants in “abnormally dangerous activities." See Restatement (Second) of Torts § 520 (1977).