Opinion for the Court filed by Circuit Judge D.H. GINSBURG.
Plaintiffs are seven of the many so-called “DES daughters” who have sued various drug companies, alleging that they suffer from malformations of their reproductive tracts because they were exposed, in úte-ro, to DES, which was prescribed for their mothers in order to prevent miscarriages. Appellee is Eli Lilly and Co., the only one of nine original defendants that plaintiffs did not voluntarily dismiss. Plaintiffs, who are residents of Maryland and of the District of Columbia, invoked the diversity jurisdiction of the District Court. That court granted Lilly’s motion for summary judg *420 ment because it determined that the plaintiffs had no competent evidence that the defendant manufactured the DES ingested by their mothers, and under the laws of both the District of Columbia and of Maryland, plaintiffs cannot recover without proof that the defendant proximately caused their injuries. Plaintiffs brought this appeal and thereafter also moved this court to refer controlling questions of law to the highest court in each of those jurisdictions. We deny that motion and affirm the judgment of the District Court.
I. BaCKGRound
In 1938, British researcher Dr. E.C. Dodds developed a synthetic form of the female hormone estrogen, the drug diethyl-stilbestrol, or DES. Estrogen is essential to a female’s sexual development and to her ability to carry a pregnancy to term. Unlike its natural counterpart, however, DES is relatively inexpensive to produce and is effective in oral doses. Dodds’ research was accordingly regarded as a revolutionary breakthrough in biomedical science.
Dr. Dodds did not, however, patent DES. A dozen American pharmaceutical houses, including Lilly, submitted New Drug Applications (NDAs) to the Food and Drug Administration, proposing to market DES for a variety of medical conditions not involving pregnancy.
Because of the large number of NDAs, the FDA ordered all applicants to pool their supporting data in a single file. The applicants then formed a committee (the “Small Committee”), composing representatives of four companies, including Lilly, to collate and submit the various clinical studies. The FDA thereafter analyzed and supplemented the companies’ submission, and determined that DES was safe in the uses for which they proposed to market it. The Small Committee was dissolved. In 1941, the FDA approved the individual companies’ initial NDAs, under which DES has been marketed ever since for a variety of indications not occurring in pregnant women.
By 1947, researchers discovered a link between miscarriages and a deficiency of natural estrogen in pregnant women, which could be corrected either by injection with natural estrogen or by oral administration of DES. In 1947, the FDA approved the applications of several drug manufacturers, including Lilly, to market DES as a miscarriage preventative. After the FDA determined in 1952 that DES was no longer a “new drug” within the meaning of the Food and Drug Act, and thus, did not require further agency approval for any company to market it for use in recognized indications and in recognized dosages, hundreds of manufacturers entered the market.
DES was widely prescribed for use by pregnant women in the 1950’s and 1960’s. In 1971, however, when medical researchers discovered a rare form of vaginal cancer in some young women who had been exposed to DES while in útero, the FDA disapproved its continued marketing for use by pregnant women.
This case was filed in the district court in 1980 by 21 women, each of whom alleged that the various noncancerous injuries from which she suffers are attributable to her mother’s ingestion of DES. Most of the original plaintiffs voluntarily withdrew their complaints, and of the nine major drug companies originally named as defendants, the remaining plaintiffs dismissed all but Lilly. On motion, the district court granted summary judgment in favor of Lilly against eight of the nine remaining plaintiffs who could not produce competent evidence that Lilly manufactured the DES that allegedly caused their injuries. Seven of those eight have pursued this appeal.
II. Plaintiffs’ Theories of Liability and the District Court’s Disposition
The district court did not decide which jurisdiction’s law governed appellants’ claims. The court made no findings indicating the jurisdiction in which either their mothers were resident when they took DES or where the plaintiffs’ injuries occurred. Instead, because it was exercising its diver *421 sity jurisdiction, and because all the plaintiffs are residents of either Maryland or of the District of Columbia, the court analyzed the plaintiffs’ claims under the laws of both jurisdictions. Tidler v. Eli Lilly & Co., C.A. No. 80-2795, Mem.Op. at 19 n. 2 (D.D.C. Aug. 23, 1985). As neither party has argued that the law of a jurisdiction other than the District of Columbia or Maryland applies, we join the district court in assuming that the law of one or both of these jurisdictions is applicable to this case.
The seven appellants concede that, after consulting parents and their physicians and pharmacists, they are unable to identify the firm that manufactured the DES product to which they were allegedly exposed. Plaintiffs, therefore, cannot adduce evidence from which a jury could reasonably conclude that Lilly’s product proximately caused their injuries. Because they lack an essential element of a traditional products liability claim, see generally Prosser, The Fall of the Citadel (Strict Liability to the Consumer), 50 Minn.L.Rev. 791, 840-48 (1966), they urged the district court, and now urge this court, to adopt any one or more of several so-called “non-identification” theories of liability that have been allowed in DES cases in a few states. Alternatively, the plaintiffs argue that liability may be predicated on a “bulk supply” theory, holding Lilly responsible as the manufacturer of DES powder later table-tized by an unidentified intermediary in the chain of production.
Plaintiffs’ first theory of liability is an extension of the “concert of action theory,” under which
those who, in pursuance of a common plan or design to commit a tortious act, actively take part in it, or further it by cooperation or request, or who lend aid or encouragement to the wrongdoer, or ratify and adopt the wrongdoer’s acts done for their benefit, are equally liable.
PROSSER and Keeton on Torts, § 46, at 323 (5th ed. 1984) (footnotes omitted). Of course, the premise for liability under this head is that the defendant, either expressly or tacitly, agreed with another to pursue “a common plan or design to commit a tor-tious act,” see id., which plaintiffs cannot show in this case.
Under the plaintiffs’ extension of this theory, however, recovery could be predicated upon either “conscious parallelism” among manufacturers — for example, in negligently failing to test DES — or upon independent actions of the defendant “halving] the effect of substantially aiding or encouraging” negligent or intentionally tortious activity on the part of others. In
Bichler v. Eli Lilly & Co.,
The district court rejected the theory, holding that
“Bichler
would not be applied in the District of Columbia or Maryland, because these jurisdictions retain the requirement of an agreement ... which the facts surrounding the marketing of DES as a miscarriage preventive [sic] do not support.”
See Stevens v. Hall,
Plaintiffs’ second theory, commonly referred to as “market share” liability, requires only that they prove that their injuries were caused by a defective or unreasonably dangerous product, and that they join enough manufacturers as defendants to ensure that those that produced a “substantial share” of the relevant product market are present before the court. Each defendant is then presumed to be liable for damages in accordance with its respective share of the product market, and in order to avoid judgment must introduce evidence tending to prove that its product could not have caused the plaintiff’s injuries. Thus, plaintiffs are completely relieved of the burden of establishing some causal connection between their injuries and the actions of any particular manufacturer. As stated by the California Supreme Court, which accepted the theory in a DES case, its premise is that “as between an innocent plaintiff and negligent defendants, the latter should bear the cost of the injury.”
Sindell v. Abbott Laboratories,
The district court questioned whether this theory had been properly presented by the plaintiffs. Noting, however, that
Sin-dell
“is inconsistent with traditional principles of comparative negligence, which require that the defendant be causally connected with the actionable allegations, before he is liable,” the court held that
“Sin-dell
will not be applied in the District of Columbia and Maryland, which require proof of causation.” The court here cited
Bowman v. Redding & Co.,
Plaintiffs’ third theory, which they refer to as “alternative market share” liability or “risk contribution” liability, would require only that they prove they were injured by DES in útero, and “that the defendant marketed the type of DES involved”; unless Lilly can prove that it did not manufacture the particular DES ingested by the plaintiffs’ mothers, it would be liable. This approach has been accepted upon the premises that:
Each [manufacturer of DES] contributed to the risk of injury to the public and, consequently, the risk of injury to individual plaintiffs. Thus, each defendant shares in some measure, a degree of culpabality in producing or marketing DES.
and that
justice [requires] that the victims of this tragedy should not be denied compensation because of the impossibility of identifying the individual manufacturer....
Martin v. Abbott Laboratories,
initially presumed to have equal shares of the market and are liable for only the percentage of plaintiff’s judgment that represents their presumptive share of the market. These defendants are entitled to rebut this presumption and thereby reduce their potential liability by establishing their respective market share of DES in the plaintiff’s particular geographic market.... To the extent that ... defendants fail to establish their actual market share, their presumed market share is adjusted so that 100 percent of the market is accounted for.
The district court rejected this burden-switching approach to liability because the relevant authorities show that both the District of Columbia and Maryland “require the plaintiff to prove by a preponderance of the evidence that the defendant’s conduct was the proximate cause of her injuries.” The court also noted certain incentive problems with the theory, and observed that
it appears that this theory ... is not merely an extension of established legal principles, but an act of wholesale judicial legislation abrogating longstanding rules of law_ [I]t is beyond the province of this federal Court, sitting in diversity, to effect such sweeping change in local jurisprudence.
Plaintiffs do not shrink from the district court’s characterization; on appeal they frankly “ask this Court to construct [a] market share approach” by which a diligent plaintiff, unable to identify the particular manufacturer that caused her injury, *423 could recover from defendants representing “in excess of 51 per cent” [sic] of the market who “could [not] prove that their drug does not fit the situation.”
Plaintiffs’ last theory of liability — not strictly a “non-identification” theory — is that because Lilly manufactured 90 to 95 percent of the bulk powder DES sold in the Washington metropolitan area, it may be held liable for their injuries, even though another firm may have tabletized, packaged, and marketed the final product (without altering or adding to the active ingredient). Plaintiffs point to the Restatement (Second) of Torts, section 388, which provides:
One who supplies directly or through a third person a chattel for another to use is subject to liability to those whom the supplier should expect to use the chattel.
Additionally, the plaintiffs rely on section 402A and Comment 1 of the Restatement, which state:
One who sells any product in a defective condition unreasonably dangerous to the user or consumer ... is subject to liability for physical harm thereby caused to the ultimate user or consumer....
# % s¡e * * ¡{c
In order for the rule stated in this section to apply, it is not necessary that the ultimate user or consumer have acquired the product from the seller, although the rule applies equally if he does so. He may have acquired it through one or more intermediate dealers_ The liability stated is one in tort and does not require a contractual relation or privity of contract between the plaintiff and the defendant.
The district court rejected the proposition that Lilly is liable in its capacity as a bulk manufacturer of DES, for two reasons. First, the court held that Lilly was not liable as a matter of law. Quoting
Lyons v. Premo Pharmaceutical Lab, Inc.,
“[I]t is not DES which is defective, ... it is only DES when used by pregnant women.” Accord, Doe v. Eli Lilly & Co., [C.A. No. 82-3515, (D.D.C. Apr. 6, 1984)]. The bulk manufacturer of a drug is not liable if it is in a non-defective condition at the time it leaves its control. Id. Rather, it is the intervening, independent tablet manufacturer that has the duty to warn the consumer of any known or forseeable [sic] harmful effects of the dosage and purpose for which it is sold.
The court also held that the evidence adduced by the plaintiffs purporting to establish that Lilly manufactured 95% of the DES sold in the greater Washington metropolitan area during the relevant period of time was incompetent and insufficient to withstand the defendant’s motion for summary judgment. The evidence consisted of the “expert” testimony of three pharmacists, who had surveyed others by telephone. The court found the survey, which was informal to say the least, “untrustworthy because it is unscientific, it was prepared with an eye towards this litigation, ... and it questions persons who may not have been qualified to answer.”
III. Analysis
Plaintiffs do not dispute the district court’s conclusion that neither Maryland nor the District of Columbia has recognized any of the “non-identification” theories of liability that they have advanced. Nor have they cited cases from these jurisdictions that could, in the aggregate, be said to mark a trend toward a more relaxed treatment of the causation requirement in products liability cases. Indeed, the Maryland Court of Appeals is generally indisposed to create new tort theories without the sanction of the representative branches of government:
[I]n considering whether a long-established common law rule — unchanged by the legislature and thus reflective of this State’s public policy — is unsound in the circumstances of modern life, we have always recognized that declaration of the public policy of Maryland is normally the function of the General Assembly; that body, by Article 5 of the Maryland Declaration of Rights, is expressly empowered *424 to revise the common law of Maryland by legislative enactment_ The Court, therefore has been reluctant to alter a common law rule in the face of indications that to do so would be contrary to the public policy of the State.
Harrison v. Montgomery County Bd. of Educ.,
Plaintiffs nonetheless ask this court “to construct [a] market share approach” for DES daughter cases out of whole cloth, using for a pattern only bits and pieces of the decisions of courts in remote states. Judicial pioneers must no doubt make bold forays into
terra incognita
in order to chart the way to justice, but that is not the office of a federal court exercising diversity jurisdiction. Ours is the more modest challenge, faithfully to apply the law of the state that the courts of the jurisdiction in which we sit, the District of Columbia, would apply had the case been filed with them.
Erie R.R. v. Tompkins,
Thus, we take the law of the appropriate jurisdiction as we find it; and we leave it undisturbed.
See Cantwell v. University of Massachusetts,
Absent some authoritative signal from the legislature or the [state courts], we see no basis for even considering the pros and cons of innovative theories.... We must apply the law of the forum as we infer it presently to be, not as it might come to be_ [W]e see no basis for applying any rule other than the traditional one.
Dayton v. Peck, Stow & Wilcox Co.,
A. Applying the Law of Maryland or the Law of the District of Columbia
Neither the highest court of Maryland nor that of the District of Columbia has explicitly rejected the specific theories advanced by the plaintiffs. Judging from the difficulties DES daughters face in adducing reliable evidence with which to identify the product manufacturer, and the creativity that some state courts have shown in overcoming them, moreover, we think it quite possible that if the courts of either jurisdiction were to discard their settled principles they would choose such a case as this to do so. It remains clear, however, that the theory that plaintiffs would have us “construct” requires that we build on a new foundation, not on the structural underpinnings of the traditional common law of torts.
See, e.g., Jensen v. American Motors Corp.,
We must also reject the plaintiffs’ “bulk supply” theory of liability. Even if such a theory were viable under established law, and even if the evidence adduced by the plaintiffs supported the proposition they would infer from it,
viz.
that Lilly manufactured 95 percent of the bulk DES sold in the Washington area during the relevant period, that evidence would not indicate that Lilly produced the powder consumed by the plaintiffs’ mothers. (In fact, however, their evidence demonstrates, at most, that the defendant had some share of the market for bulk powder DES. Lilly’s actual market share could be quite modest or almost 100 percent as far as one can tell from plaintiffs’ proffer.) The bulk powder theory thus suffers from the same infirmity under Maryland and District of Columbia law as do the plaintiffs’ other non-identification theories; plaintiffs cannot demonstrate that any product manufactured by Eli Lilly caused the injuries of which they complain. They have cited no case from either of the jurisdictions that suggests that a substantial share of the product market, even as much as 95% of that market, is a sufficient basis from which a jury reasonably may conclude that the defendant caused the plaintiff’s injury. Indeed,
Langville v. Glen Burnie Coach Lines, Inc.,
The [plaintiff’s] burden [to establish causation] is not met by proof adduced by the plaintiff to the effect that defendant’s negligence may have caused the injuries, or even that it probably did cause them, if it also appears from plaintiff’s evidence that the injuries may have resulted from some other cause for which the defendant is not responsible.
As the district court stated, “the jury in this case is left with no competent evidence from which it could reasonably conclude that the plaintiffs ... have hailed [sic] the right defendant into court. At best, a verdict on liability would be based on speculation.”
The plight of all DES daughters is a difficult one. Some state courts have therefore devised novel methods to ensure these innocent individuals, who through no shortcoming of their own have not been able to identify the responsible manufacturer, recover nonetheless.
See Collins,
B. Certifying Questions of Law to the Maryland Court of Appeals and the District of Columbia Court of Appeals
On appeal, plaintiffs have for the first time requested that questions of law be certified to the Maryland Court of Appeals and to the District of Columbia Court of Appeals, for those courts to determine after all whether the law of those jurisdictions recognize any of the plaintiffs’ non-identification theories.
Both Maryland and the District of Columbia have enacted legislation providing for such a procedure where there is
*426
presented a determinative question of state law “as to which it appears to the certifying court there is no controlling precedent” in the decisions of the highest court of the jurisdiction. Ann.Code of Md., Courts and Judicial PROCEEDINGS § 12-601 (Michie 1984); D.C.Code ANN. § 11-723 (Michie 1987 Cum.Supp.). The decision of a federal court to certify questions of law pursuant to state-established procedures of this type rests with the sound discretion of the court.
Lehman Brothers v. Schein,
The law of both Maryland and the District of Columbia is not lacking in clarity, as far as is relevant here. Of course, we cannot predict with certainty that either jurisdiction would not view the theories proposed by the plaintiffs as compelling in these circumstances and adopt a markedly different view of liability — and perhaps an equally novel idea of their own role in making public policy — than that which they have heretofore held. As we have explained above, however, that is not the relevant consideration; federal courts, sitting in diversity, are obligated to apply, not to amend, existing state law.
In addition to the condition of existing state law, which is a sufficient ground in itself, there are other reasons for our refusing to exercise our discretion to certify questions in this case. First, the plaintiffs chose to litigate these claims before the district court, fully aware that neither of the relevant jurisdictions recognize the theories they were advancing. Obviously, their choice of forum is not helpful to their request for certification.
See Cantwell,
*427 Second, the record before us is so lacking in concreteness that to certify plaintiffs’ questions of law would merely be to invite our state court colleagues to an exercise in speculation. It has not even been determined whether the law of either of these jurisdictions certainly applies to the case. It would be a wholly pointless waste of judicial (as well as the parties’) resources, for an already overworked state court to consider the complex questions of law propounded by the plaintiffs, if it were only to find in the end that its efforts were irrelevant because the law of another jurisdiction is controlling.
Nor does the record in other respects display the necessary factual clarity to pose a well-framed, concrete question of law.
See
Ann.Code of Md„ Courts and Judicial Proceedings § 12-603(a)(2) (Mi-chie 1984) (“A certification order shall set forth ... [a] statement of all facts relevant to the question certified showing fully the nature of the controversy in which the question arose.”);
Florida ex rel. Shevin,
IV. Conclusion
The district court correctly held that the defendant was entitled to summary judgment, inasmuch as the plaintiffs cannot, by their own admission, prove that the conduct of the defendant proximately caused their injuries, as required by the law of both the District of Columbia and of Maryland. We find it would be inappropriate to prolong this litigation by granting plaintiffs’ afterthought motion to certify questions of law to the Maryland Court of Appeals and the District of Columbia Court of Appeals. Accordingly, the judgment of the district court is
Affirmed.