Ever since this blog started, we’ve made plain that we have no use for the so-called “heeding presumption.” This presumption posits that, because under Restatement §402A, comment j, a defendant providing an adequate warning can presume it will be heeded, a plaintiff should also be able to presume that an adequate warning, had it been granted, would have been heeded. That’s false equivalence if we’ve ever seen it. A defendant to such a warning claim needs no heeding presumption, since it wins on adequacy without ever getting to causation. The comment j discussion really involves design defects (about which more below). Plaintiffs, on the other hand, are getting a burden of proof shift on warning causation that simply has no basis in reality. People disregard adequate warnings all the time.
So we fight the heeding presumption whenever it comes up. Some states have good law on the issue. N.C. G.S.A. §99B-5(a); Wis. Stat. §895.047(1)(e); Ford Motor Co. v. Boomer, 736 S.E.2d 724, 733 (Va. 2013); Rivera v. Philip Morris, Inc., 209 P.3d 271, 274 (Nev. 2009); Leaf v. Goodyear Tire & Rubber Co., 590 N.W.2d 525, 528-29 (Iowa 1999); Riley v. American Honda Motor Co., 856 P.2d 196, 199-200 (Mont. 1993); Deere & Co. v. Grose, 586 So. 2d 196, 198 (Ala. 1991); Huitt v. Southern California Gas Co., 116 Cal. Rptr.3d 453, 467-68 (Cal. App. 2010); Harris v. International Truck & Engine Corp., 912 So. 2d 1101, 1109 (Miss. App. 2005); McPike v. Enciso’s Cocina Mejicana, Inc., 762 P.2d 315, 319 (Or. App. 1988); DeJesus v. Craftsman Machinery Co., 548 A.2d 736 (Conn. App. 1988); Muilenberg v. Upjohn Co., 320 N.W.2d 358, 366 (Mich. App. 1982); Potthoff v. Alms, 583 P.2d 309, 311 (Colo. App. 1978); Payne v. Novartis Pharmaceuticals Corp., 767 F.3d 526 (6th Cir. 2014) (applying Tennessee law); Tuttle v. Lorillard Tobacco Co., 377 F.3d 917, 925 (8th Cir. 2004) (applying Minnesota law); Wilson v. Bradlees of New England, Inc., 250 F.3d 10 (1st Cir. 2001) (applying New Hampshire law); Christopher v. Cutter Laboratories, 53 F.3d 1184, 1192-93 (11th Cir. 1995) (applying Florida law); Odom v. G.D. Searle & Co., 979 F.2d 1001, 1003 (4th Cir. 1992) (applying South Carolina law); Muzichuck v. Forest Laboratories, Inc., 2015 WL 235226, at *13 (N.D.W. Va. Jan. 16, 2015); Luttrell v. Novartis Pharmaceuticals Corp., 894 F. Supp.2d 1324, 1345 n.16 (E.D. Wash. 2012).
Almost as many states are adverse. House v. Armour, Inc., 929 P.2d 340, 347 (Utah 1996); Coffman v. Keene Corp., 628 A.2d 710, 717-19 (N.J. 1993); Eagle-Picher Industries, Inc. v. Balbos, 604 A.2d 445, 468-69 (Md. 1992); Bushong v. Garman Co., 843 S.W.2d 807, 811 (Ark. 1992); Arnold v. Ingersoll-Rand Co., 834 S.W.2d 192, 194 (Mo. 1992); Butz v. Werner, 438 N.W.2d 509, 517 (N.D. 1989); Harlow v. Chin, 545 N.E.2d 602, 606 (Mass. 1989); Bloxom v. Bloxom, 512 So.2d 839, 850 (La. 1987); Payne v. Soft Sheen Products, Inc., 486 A.2d 712, 725 (D.C. 1985); Wooderson v. Ortho Pharmaceutical Corp., 681 P.2d 1038, 1057-58 (Kan. 1984); Seley v. G.D. Searle Co., 423 N.E.2d 831, 838 (Ohio 1981); Menard v. Newhall, 373 A.2d 505, 506 (Vt. 1977); Cunningham v. Charles Pfizer & Co., 532 P.2d 1377, 1382 (Okla. 1974); Dole Food Co. v. North Carolina Foam Industries, Inc., 935 P.2d 876, 883 (Ariz. App. 1996); Ortho Pharmaceutical Corp. v. Chapman, 388 N.E.2d 541, 555 (Ind. App. 1979).
Since the turn of the century, our team has a definite edge (more on why that is so, below, as well).
A few states, like Pennsylvania, and Texas, are somewhere in the middle – allowing a heeding presumption in some product liability situations, but not for cases involving prescription medical products. Compare Coward v. Owens-Corning Fiberglas Corp., 729 A.2d 614, 620-21 (Pa. Super. 1999), with Lineberger v. Wyeth, 894 A.2d 141, 145, 149-50 (Pa. Super. 2006); and Technical Chemical Co. v. Jacobs, 480 S.W.2d 602, 606 (Tex. 1972), with Ackermann v. Wyeth Pharmaceuticals, 526 F.3d 203, 212-13 (5th Cir. 2008) (applying Texas law).
And then there’s New York, where the law can charitably be described as a morass. So how did the law of the Empire State crumble into quicksand?
Well, it’s not a problem with basic warning law. The New York Court of Appeals is fine on the burden of proof in warning cases generally. The black letter law, in New York as in most places, is that “[i]n order to prevail at trial in a negligence case, a plaintiff must establish by a preponderance of the evidence that the defendant’s negligence was a proximate cause of plaintiff’s injuries.” Burgos v. Aqueduct Realty Corp., 706 N.E.2d 1163, 1165 (N.Y. 1998). “The burden of proof to establish causal negligence is upon the plaintiff.” Hirsh v. State, 168 N.E.2d 372, 373 (N.Y. 1960). “It is well settled that a plaintiff must generally show that the defendant’s negligence was a substantial factor in producing the injury to satisfy the burden of proving a prima facie case.” Wild v. Catholic Health System, 991 N.E.2d 704, 706 (N.Y. 2013). What’s more, there’s no difference between negligence and strict liability in New York, insofar as warnings are concerned. “New York views negligence and strict liability claims as equivalent.” Martin v. Hacker, 628 N.E.2d 1308, 1311 n.1 (N.Y. 1993).
In a bunch of cases, the New York intermediate appellate courts have followed this traditional burden of proof in product liability litigation warning cases. “[I]n any products liability case, the plaintiff in an action premised on inadequate warnings must prove causation.” Estrada v. Berkel Inc., 789 N.Y.S.2d 172, 174 (N.Y.A.D. 2005). “[A] plaintiff whose claim is based on inadequate warnings must prove . . . that if adequate warnings had been provided, the product would not have been misused.” Banks v. Makita U.S.A., 641 N.Y.S.2d 875, 877 (N.Y.A.D. 1996). “Plaintiffs had the burden to show that had a different warning been given, this patient would not have used the product that caused her injury.” Mulhall v. Hannafin, 841 N.Y.S.2d 282, 287 (N.Y.A.D. 2007). In Reis v. Volvo Cars, Inc., 901 N.Y.S.2d 10 (N.Y.A.D. 2010), the court held that the plaintiff’s “failure to warn claims should have been dismissed because . . . there is no proof in the record that [plaintiff] would have read and heeded a warning had one been given.” Id. at 13. Reis followed Sosna v. American Home Products, 748 N.Y.S.2d 548 (N.Y.A.D. 2002), a prescription drug case holding that:
Contrary to plaintiff’s argument, in this state, it remains plaintiff’s burden to prove that defendant’s failure to warn was a proximate cause of his injury and this burden includes adducing proof that the user of a product would have read and heeded a warning had one been given.
Id. at 549. Accord Santos v. Ford Motor Co., 893 N.Y.S.2d 537, 538 (N.Y.A.D. 2010) (quoting and following Sosna). See also Menna v. Walmart, 975 N.Y.S.2d 710 (table), 2013 WL 3958247, at *2 (N.Y. Sup. 2013) (plaintiff “must adduce proof that he or she would have read and heeded a warning had one been given”); Zapata v. Ingersoll Rand Co., 959 N.Y.S.2d 93 (table), 2012 WL 3553111, at *11 (N.Y. Sup. 2012) (“under well settled law, to prove proximate cause, a plaintiff has the obligation to adduce proof that had a warning been provided, he or she would have read the warning and heeded it”); Granata v. Sub-Zero Freezer Co., 819 N.Y.S.2d 210 (table), 2006 WL 1358468, at *5 (N.Y. Sup. 2006) (quoting Sosna), aff’d, 841 N.Y.S.2d 469 (N.Y.A.D. 2007); Power v. Crown Controls Corp., 568 N.Y.S.2d 674, 675 (N.Y. Sup. 1990) (“the injured plaintiff must, of course, prove that the inadequacy of the warning was a proximate cause of his injuries”).
Traditional causation also finds support in federal decisions applying New York law. “A plaintiff proceeding under a failure-to-warn theory in New York must demonstrate that the failure to warn adequately of the dangers of a product was a proximate cause of his or her injuries.” Bravman v. Baxter Healthcare Corp., 984 F.2d 71, 75 (2d Cir. 1993).
In a failure-to-warn action, a plaintiff bears the burden to prove that defendant’s failure to warn was a proximate cause of his injury and this burden includes adducing proof that the user of a product would have read and heeded a warning had one been given. In the case of prescription medications, where warnings are directed to prescribing physicians, a plaintiff must demonstrate that had a different, more accurate warnings been given, his physician would not have prescribed the drug in the same manner.
Alston v. Caraco Pharmaceutical, Inc., 670 F. Supp.2d 279, 285 (S.D.N.Y. 2009) (citing Sosna). Accord McDowell v. Eli Lilly &Co., 58 F. Supp.3d 391, 408 (S.D.N.Y. 2014) (“a plaintiff must demonstrate” that “a different, more accurate warning” would have “changed” the decision to use the product); Hayes v. New York, 2013 WL 5278879, at *15 (N.D.N.Y. Sept. 18, 2013) (“Plaintiff has failed to offer any evidence that even if a different warning had been issued, the user of the product . . . would have read and heeded that warning”); Cuntan v. Hitachi KOKI USA, Ltd., 2009 WL 3334364, at *17 (E.D.N.Y. Oct. 15, 2009) (plaintiff’s burden of proving causation in a warning case “is well settled” law); In re Aredia & Zometa Products Liability Litigation, 2009 WL 2496873, at *2 (M.D. Tenn. Aug. 13, 2009) (“[i]t remains Plaintiff’s burden to prove that defendant’s failure to warn was a proximate cause of her injury, and this burden includes adducing proof that the user of a product would have read and heeded a warning”) (applying New York law).
Still, numerous New York cases have gone south on the heeding presumption. Union Carbide Co. v. Affiliated FM Insurance Co., 955 N.Y.S.2d 572, 575 (N.Y.A.D. 2012) (“New York law presumes that users will heed warnings provided with a product”); Power v. Crown Controls Co., 568 N.Y.S.2d 674, 675 (N.Y. Sup. 1990) (“the presumption that a user would have heeded warnings can be rebutted by proof that an adequate warning would have been futile”); Adesina v. Aladan Co., 438 F. Supp.2d 329, 338 (S.D.N.Y., 2006) (“[f]ailure to warn law includes a presumption that ‘a user would have heeded warnings if they had been given, and that the injury would not have occurred’”) (quoting G.E. Capital Co. v. A.O. Smith Co., 2003 WL 21498901 at *5 (S.D.N.Y. 2003)); Henry v Rehab Plus Inc., 404 F. Supp.2d 435, 442 (E.D.N.Y. 2005) (“[i]n New York there is a presumption that a user would have heeded warnings if they had been provided and that the injury would not have occurred”); Santoro v. Donnelly, 340 F. Supp.2d 464, 486 (S.D.N.Y. 2004) (same); Anderson v. Hedstrom Co., 76 F. Supp.2d 422, 441 (S.D.N.Y. 1999) (“New York [follows] a presumption that a user would have heeded warnings if they had been provided, and that the injury would not have occurred”).
What created this split? We think two things did – and one thing did not.
First, the “did not.” We can say with some confidence that, unlike most other states, the culprit in New York was not Restatement §402A, comment j. That’s because New York has never recognized comment j. The New York Court of Appeals has cited comment j a grand total of one time: thirty years ago, in a dissent, on a different issue. See Schumacher v. Richards Shear Co., N.E.2d 195, 202 (N.Y. 1983) (comment j cited concerning a “duty to become and remain aware of . . . technical developments”) (Jasen, J., dissenting). Instead of comment j – or §402A strict liability generally – as we’ve already mentioned, in warning cases, New York law treats negligence and strict liability claims as “equivalent.” Martin, 628 N.E.2d at 1311 n.1. The relevant tests negligent warnings are found in Restatement (Second) of Torts §388 (1965), which (in stark contrast to §402A, comment j) the Court of Appeals has cited repeatedly, if not particularly recently. E.g., Sukljian v. Charles Ross & Son Co., 503 N.E.2d 1358, 1362 (N.Y. 1986); Cover v. Cohen, 473 N.Y.S.2d 378, 385-85 (N.Y. 1984); Robinson v. Reed-Prentice, 403 N.E.2d 440, 446-47 (N.Y. 1980). Notably, neither Restatement §388, nor any of its comments, contains language suggesting a presumption, assumption, or anything of the sort with respect to the reading and heeding of warnings. Also notably, none of the New York cases relying on a heeding presumption have even suggested that it also applies in negligence, which the logic of Martin would require, if such a presumption actually existed.
Nor is the comment j language from which the heeding presumption originally arose even valid any longer. Instead, that part of comment j has been expressly repudiated by the American Law Institute (“ALI”) itself. The ALI, in its more recent Third Restatement of Torts, explicitly rejected the position that products with open and obvious design defects are not “defective” in design because users could be presumed to read and heed adequate warnings of their risks:
The fact that a risk is obvious or generally known often serves the same function as a warning. However, obviousness of risk does not necessarily obviate a duty to provide a safer design. Just as warnings may be ignored, so may obvious or generally known risks be ignored, leaving a residuum of risk great enough to require adopting a safer design.
Restatement (Third) of Torts, Products Liability §2, comment l (1998). It’s no surprise that after the Third Restatement, judicial acceptance of the heeding presumption has essentially ground to a halt outside of those states that had already used comment j to create it.
Comment l’s design-related recognition that “warnings may be ignored” by consumers, as recognized and discussed by the drafters of the Third Restatement, is antithetical to any heeding presumption:
Much of the problem was created by unfortunate language in the Restatement, Second, of Torts § 402A, Comment j. . . . The Comment j presumption embodies the behavioral assumption that “reasonable” users can be expected to receive, correctly interpret, and obey every comprehensible warning accompanying every product they use or encounter. . . . [A]lmost all products present substantial risks if improperly manufactured, designed, or used. . . . People would have to read, understand, remember, and follow innumerable product warnings to protect themselves from all product-related risks they may confront. Moreover, . . . people must devote some of their limited time and attention to many other types of choices. . . . [W]arnings should only be used as a supplement to a design that already embodies reasonable safety and not as a substitute for it.
Comment j of the Restatement, Second, is inconsistent with the judicial abandonment of the patent danger rule and with those cases that take the position that a warning will not absolve the manufacturer from the duty to design against dangers when a reasonable, safer design could have been adopted.
Restatement (Third) of Torts, Products Liability §2, Reporters’ Note to comment l, at p. 101 (1998) (emphasis added). And what do the Third Restatement’s drafters cite for this proposition? New York law – specifically the abolition of the “patent danger” rule in Micallef v. Miehle Co., 348 N.E.2d 571, 578 (N.Y. 1976) (“the patent-danger doctrine should not, in and of itself, prevent a plaintiff from establishing his case”). Thus, a comment-j-based heeding presumption would be uniquely antithetical to New York law, not only as to causation in inadequate warning cases, but also with respect to design defects and patent risks.
Now for the “dids.” In our view the culprits in the creation of the New York heeding presumption morass are, first, the failure of certain courts to appreciate the differences between differing kinds of warning defects; and second one grotesquely distorted prediction of New York law by the Second Circuit – ironically at almost precisely the same time that the Third Restatement was kicking out the comment j jams from the heeding presumption. See Liriano v. Hobart Corp., 170 F.3d 264 (2d Cir. 1999).
Warning claims can be divided (as can many things, such as New York resident presidential candidates) between substance and optics. While the law recognizes claims that turn on the conspicuity of warnings, as opposed to their substance, see Martin, 628 N.E.2d at 1312, this type of claim is uncommon compared to allegations that the content of product warnings was absent or inadequate. Martin, for one, spent one sentence on non-informational warning claims and several paragraphs on informational ones. Id. at 1311-13. In conspicuity cases, warning prominence bears directly on causation, and is a way around the otherwise dispositive defense that the actor didn’t read the warning. Most warning cases, including almost all drug/device warnings (since boxed warning claims are preempted) turn instead on what product warnings actually say.
This distinction was discussed in Johnson v. Johnson Chemical Co., 588 N.Y.S.2d 607 (N.Y.A.D. 1992). The general rule “in any products liability case, [is that] the plaintiff in an action premised on inadequate warnings must prove causation,” and a plaintiff in a product misuse case can recover only if “he proves that, [with] adequate warnings . . . the product in question would not have been misused.” Id. at 611. In a warning “prominence” case, however, that “argument loses its persuasive force”:
A second factor to be considered is the prominence with which such language is displayed. . . . A consumer such as [plaintiff] who, by her own admission, tends to ignore one sort of label, might pay heed to a different, more prominent or more dramatic label.
Id. (factual examples omitted).
One of the cases Johnson cited, id. at 611-12, is Baker v. St. Agnes Hospital, 421 N.Y.S.2d 81 (N.Y.A.D. 1979), another conspicuity case. In Baker, a drug case, the plaintiff’s problem was that the substantively adequate warning was typically removed by pharmacists before prescribing physicians received it – a variant of the physician failure to read defense. The court in Baker let plaintiff get away with an argument that other methods (such as “Dear Doctor” letters) would have been a “more effective means of communicating its warning.” Id. at 86. See also LaPaglia v. Sears Roebuck & Co., 531 N.Y.S.2d 623, 628 (N.Y.A.D. 1988) (ignored warning should have been more “prominent”); Hoffman-Rattet v. Ortho Pharmaceutical Corp., 516 N.Y.S.2d 856, 861 (N.Y. Sup. 1987) (product had “allegedly inadequately conveyed updated warnings”); Monell v. Scooter Store, Ltd., 895 F. Supp. 2d 398, 414 (N.D.N.Y. 2012) (“location of the warning was insufficient”); Derienzo v. Trek Bicycle Corp., 376 F. Supp.2d 537, 568-69 (S.D.N.Y. 2005) (warning allegedly “inconspicuous”); Santoro, 340 F. Supp. 2d at 490 (product should have had an “exterior warning”); Anderson v. Hedstrom Corp., 76 F. Supp.2d 422, 441-43 (S.D.N.Y. 1999) (warning should have been “in bold letters on the [product] itself”). Conspicuity cases, cabined by the language of actually-given warnings, involve none of the rank speculation created by assuming that any allegedly “adequate” language, conveyed at unknown times in unknown ways, would have been heeded. Indeed, most conspicuity cases, like Johnson and Baker, do not purport to shift the burden of proof, and thus aren’t really heeding presumption cases at all.
Second, and particularly in federal courts, cases attempting to apply New York law are bedeviled by ill-chosen language in Liriano. The plaintiff’s case in Liriano lacked any evidence that a warning would have altered the his conduct (sticking his hand in an unguarded meat grinder), but the jury nevertheless returned a plaintiff’s verdict. Second Circuit also shrugged off the absence of warning causation evidence, with the comment that the traditional burden of proof “rested on a false premise.” 170 F.3d at 271. Instead, Liriano discovered a factual “inference” that “shift[ed]” the burden of proof:
[Plaintiff] does not bear that burden. When a defendant’s negligent act is deemed wrongful precisely because it has a strong propensity to cause the type of injury that ensued, that very causal tendency is evidence enough to establish a prima facie case of cause-in-fact. The burden then shifts to the defendant to come forward with evidence that its negligence was not such a but-for cause. [Where] . . . the kind of negligence that the jury attributed to the defendant tends to cause exactly the kind of injury that the plaintiff suffered . . ., rather than requiring the plaintiff to bring in more evidence to demonstrate that his case is of the ordinary kind, the law presumes normality and requires the defendant to adduce evidence that the case is an exception. Accordingly, in a case like this, it is up to the defendant to bring in evidence tending to rebut the strong inference, arising from the accident, that the defendant’s negligence was in fact a but-for cause of the plaintiff’s injury.
Id. at 271 (emphasis added).
Liriano did not mention, let alone discuss, New York Court of Appeals decisions such as Burgos and Hirsh stating that plaintiffs always bears the burden of proving causation in warning cases. Instead, the Second Circuit cited, as supposed precedent for “shifting” the burden of proof on causation in product liability warning cases, Martin v. Herzog, 126 N.E. 814 (N.Y. 1920), an ancient negligence case involving an accident allegedly caused by nonuse of headlights on a horse-drawn buggy being driven at night. Liriano did not mention, however, that Martin was a contributory negligence case. The occupants of the buggy − proceeding without lights after dark − were the plaintiffs, so the burden of proving their negligence rested on the defendant. 126 N.E. at 814. Martin was thus inapposite to a plaintiff’s burden of proof in routine product liability cases. The “inference” mentioned in Martin, 126 N.E. at 816, satisfied the defendant’s pre-existing burden of establishing the affirmative defense of contributory negligence. It did not “shift” that burden in any way. But even if Liriano had been about contributory negligence, the Second Circuit got it wrong. “The rule [in contributory negligence cases] does not entitle the jury to presume plaintiff exercised due care at the time of the accident nor does it shift the burden of proof: plaintiff must still establish a prima facie case.” Sawyer v. Dreis & Krump Manufacturing Co., 502 N.Y.S.2d 696, 699 (N.Y. 1986).
In suggesting that an “inference” could reverse the ordinary burden of proof, Liriano seriously misread New York law. Indeed, the New York Court of Appeals has repeatedly complained about precisely this type of holding. “[T]he indiscriminate use of the terms ‘presumption’ and ‘inference’ [has] caused procedural problems.” Morejon v. Rais Construction Co., 851 N.E.2d 1143, 1146 (N.Y. 2006). “[A]n ‘inference’ and a ‘presumption’ are not identical in scope or effect.” George Foltis, Inc. v. City of New York, 38 N.E.2d 455, 462 (N.Y. 1941).
[A] “presumption” is a rule of law attaching definite probative value to a specific fact, as distinguished from an “inference,” which is a permissive conclusion by a trier of the fact, unaided by any rule or theory of law directly applicable.
People v. Hildebrandt, 126 N.E.2d 377, 378 (N.Y. 1955). Inferences do not shift the burden of proof. “Permissible inferences, based on specified underlying facts . . . must first be proved before the inference may be drawn.” People v. Leyva, 341 N.E.2d 546, 552 n.3 (N.Y. 1975). Most recently, in Zamora v. New York Neurologic Assocs., 970 N.E.2d 823 (N.Y. 2012), that court rejected a supposed causation “presumption” in a workers’ compensation case, holding that the lower court improperly “created a presumption out of an inference” that “would shift the burden of proof,” something for which “[t]here is no precedent in our decisions.” Id. at 826.
Liriano isn’t even consistent with prior Second Circuit precedent, which recognized that what the plaintiff was trying to pass off as a “heeding presumption” was really nothing more than a permissible jury inference:
[Plaintiff] somewhat misstates the matter by asserting that New York recognizes a “heeding presumption,” but she is correct in contending that in some circumstances, New York permits the trier to infer that a warning would have been heeded and thereby to conclude that the absence of a warning that was reasonably required to be given was a proximate cause of an injury. Appellee cites no New York decision that refers to a “presumption” of heeding.
Raney v. Owens-Illinois, Inc., 897 F.2d 94, 95 (2d Cir. 1990) (applying New York law); see Topliff v. Wal-Mart Stores E. LP, 2007 WL 911891, at *43 (N.D.N.Y. March 22, 2007) (given Raney, court was “skeptical” of Plaintiff’s characterization of the failure-to-warn law in New York as providing a “presum[ption] that a user would have heeded the warnings if they had been given”).
All those New York federal cases we cited above, which purport to recognize a general heeding presumption, are ultimately dependent upon the erroneous Liriano decision. They either follow Liriano directly, or else cases such as Anderson, 76 F. Supp.2d at 441-43, that, in turn, relied extensively on Liriano. See Adesina, 438 F. Supp.2d at 338 (citing Liriano); Santoro, 340 F. Supp.2d at 490 nn. 153-54 (citing Anderson); G.E. Capital, 2003 WL 21498901, at *5 (same); cf. Henry, 404 F. Supp.2d 442 (not citing any New York precedent).
Thus, the current muddle in New York law concerning the heeding presumption is a product of a result-oriented Second Circuit opinion trying to stretch New York law in violation of basic Erie principles. See, e.g., Travelers Insurance Co. v. Carpenter, 411 F.3d 323, 329 (2d Cir. 2005). “[E]mbrac[ing] the exhilarating opportunity of anticipating a doctrine which may be in the womb of time, but whose birth is distant” is not the business of federal courts in diversity cases. Spector Motor Service v. Walsh, 139 F.2d 809, 823 (2d Cir. 1943) (Learned Hand dissenting), vacated, 323 U.S. 101 (1944). That has combined with an unfortunate tendency of some courts to ignore the difference between warning cases based on substance and those based on conspicuity. Whatever, the reason, such woolly-headedness does not extend to New York’s highest court, which has firmly stuck to the traditional burden of proof in warning causation cases. Given the size of the mess that the lower courts have created on this topic, it will probably require the New York Court of Appeals to clean it up.