Product Warnings Litigation: Fixing What’s Wrong
New York Law Journal
A law review article by professors Aaron D. Twerski and James A. Henderson Jr. is about to be published in the Indiana Law Journal, and it merits serious attention by the bench and litigation bar. Provocatively titled, “Fixing Failure to Warn,” the noted authors once again reveal serious ills in the current system of warnings litigation and provide a logical, simple “fix” that is analogous to a widely accepted construct in product design litigation.1 They suggest that, just as a claimant alleging a defective product design must prove the practicable feasibility of defendant incorporating a safer “reasonable alternative design” (RAD), so too, a claimant urging a product’s inadequate warning should have to prove a safer reasonable alternative warning (RAW).
These noted scholars—Twerski is a professor of Brooklyn Law School and Henderson, is professor emeritus at Cornell Law School—are no strangers to products liability law, and the legal standards for gauging when designs or warnings are to be considered “defective.” Both were Reporters for the Restatement of Torts, Third: Products Liability, the American Law Institute’s prestigious and oft-cited collection of black-letter rules and commentary governing products liability. They also authored the 1990 “Doctrinal Collapse” article about warnings in the New York University Law Review which, 10 years later, was cited as one of the 25 most influential articles published by the NYU Law Review over the previous 75 years.2 There the authors unmasked major shortcomings in how courts handled warning litigation. Prior to surveying some highlights of Twerski and Henderson’s brand new proposal to “fix” the law of warnings, let’s review some of the major problems.
I previously referred to the failure to warn cause of action as a “poor man’s design claim.”3 That reference has nothing to do with the wealth or poverty of the lawyer asserting it. Rather, it reflects a relatively inexpensive, albeit circuitous and indirect, way of attacking a product’s design. The true design claim usually requires competent expert proofs that a reasonable, practicable alternative design existed at the time of manufacture or sale that would have prevented or mitigated the harm. The warnings claim, on the other hand, usually posits that some (or additional) verbiage, text, label, sticker or symbol, if only given, would have obviated the injury.
The superficial appeal of this simplistic notion is devilishly beguiling. Indeed, some courts not only buy into it but help it along with application of a “rebuttable” so-called “heeding presumption,” i.e., that, had the warning alleged been given, the law presumes that the user would have heeded it and thereby would have prevented the injury. Sometimes, this aid is referred to as the “causation presumption.” The “heeding presumption,” however, has been so thoroughly debunked and shown to be so flawed,4 one wonders how some courts can intellectually justify such a skewing of the law of warnings.
In any event, warnings litigation has perhaps graver ills than occasional court allowance of the heeding presumption. One is that the range of possible dangers to be warned against attending product use is virtually limitless. The universe of potential warnings one can conjure up is markedly open-ended. One simply identifies a risk or danger that accompanies a product’s design, its packaging, its parameters of normal use and all varieties of misuse and then alleges that the risk selected should have been warned against.
And if a warning was issued, then one just needs to allege that it could have been made better or safer. The print could have been bolder or larger or more emphatic. Or the text might have been more eye-catching if it had employed different colors or used pictorial symbols. Or it could have been conveyed in the language frequented in the community where the injury occurred, and so on. In effect, says the willing warnings expert, show me the accident and injury, and I will give you the warning or admonition or modified warning that could have helped the plaintiff in this mishap. All too often courts hold such testimony to be enough—without demanding or injecting rigor into the analysis and without subjecting the expert’s opinion or methodology to the kind of Daubert/Frye “reliability” probe courts employ in design or manufacturing defect cases.
Another ill of open-ended warnings litigation is that risk/utility balancing in warnings cases proceeds ineffectively. As Henderson and Twerski in their “Doctrinal Collapse” article noted, proliferative warnings about every conceivable risk, no matter how remote, bombard consumers with “nearly useless warnings about risks that rarely materialize in harm” and that cause consumers “to give up on warnings altogether.”
The courts’ assumption that the risks and costs of additional warnings are low is erroneous. Instead, “overwarning” causes users “to discount or ignore warnings that should be heeded, leading to higher accident costs that, although very real, are not before the court in failure-to-warn litigation,” the authors wrote. Courts simply pile up warning after warning in case after case without looking at the total picture of countervailing costs.
Moreover, unlike design cases, warnings jurisprudence has no “available body of hard science from which to draw the data necessary to run sensible risk-utility analyses,” Henderson and Twerski wrote in their “Doctrinal Collapse” article. The relevant sciences tend to be soft, undeveloped and largely under-utilized. The problems involved are polycentric. Since the capacity to give and receive warnings is limited, when those limits are reached, new warnings inevitably “crowd out existing warnings.”
But a court will usually not know when those limits will be reached because it typically focuses on the precise case before it and solely upon the additional verbiage plaintiff says should have been used. Here the “lack of hard science and dependable data comes back to haunt the courts with a vengeance,” the authors of “Doctrinal Collapse” wrote. These problems are dramatically exacerbated by the “empty shell” that causation analysis then becomes. The claimant offers little more than self-serving testimony and anecdotal evidence to establish the causation case. The incisiveness of the scientific evidence that should be presented suffers. In light of available empirical data, rank speculation thus abounds.5
In their 1990 “Doctrinal Collapse” article, the authors further recounted what considerable empirical evidence shows: (1) many people do not process information in a logical and predictable manner; (2) both the type of information and manner of presentation can have a substantial effect on decision-making; (3) an individual’s prior beliefs and information significantly affect the impact of receiving additional information; (4) the order in which information is presented impacts upon an individual’s perception of new data; (5) suggested alternative or additional warnings cannot be comprehended by the user all in one piece; they must be read sequentially and digested piecemeal; (6) the hierarchy of warnings is important: as one moves from more important warnings to those of lesser urgency, the number of warnings that must be supplied increases exponentially at each descending level; (7) there are many variables attending how a hypothetical new warning should have appeared and how the user would have reacted to them; (8) physical causation problems may complicate the process of determination.6
The foregoing complexities should invite rigorous and probing court scrutiny under Daubert7 and Frye8 reliability standards as to whether an expert’s opinion that this verbiage or that text or the size and placement of some proposed warning would have made a causal difference. I covered that topic in some depth in a column years ago,9 citing, among others, impressive writings by warnings specialist, William H. Hardie. Indeed, a federal court nearly two decades ago correctly observed, “[I]n a warning case, a plaintiff must do more than simply present an expert who espouses a new or different warning. He must establish that warning’s feasibility, adequacy and effectiveness. In this case, plaintiffs’ experts’ testimony fell woefully short of meeting these criteria.”10
In Shepherd v. Michelin Tire Corp.,11 an experienced tire worker died from an accident that occurred when he mounted and inflated a tire on a mismatched-size rim. Plaintiff’s expert, Dr. Kenneth Laughery, opined that the tire did not have an adequate warning and that the accident would have been averted with his alternative warning had it been affixed to the tire. The district court held a Daubert hearing and precluded the proposed testimony. The court observed that warnings research generally concludes that warnings are not particularly effective behavior modifiers. Moreover, males and experienced persons are more likely to ignore warnings than females or novices. The two persons injured in the Shepherd case were experienced males. Further, Laughery conceded that variables such as the cost of compliance, familiarity with the product, hazard perception, ease of instructions, task load, training and gender are all factors affecting a warning’s effectiveness.
A warning overload devalues whatever warnings are on a dangerous product. Laughery conceded that he conducted no field testing to support his theoretical conclusions. He agreed that “no empirical data…supports [his] opinion that any particular sidewall warning will make tires safer because there’s been no testing, jurying, or test marketing of any particular sidewall warnings.” Bottom line, Laughery could not state with any degree of certainty that his proposed warning system would probably have prevented this accident. He could not even state that the probability the accident would not have occurred was 50 percent or more. “In other words, Laughery’s testimony does not meet the preponderance test on proximate causation, even if he were allowed to testify.”
Some more recent decisions also confirm that rigorous Daubert reliability analysis should probe experts’ warning opinions.12 Therefore, given the ubiquity of failure to adequately warn claims, we should expect to see a rather robust body of decisional law involving Daubert gatekeeping analyses of proposed warnings, much as we do for designs in product design cases. But we don’t. Are courts letting warnings opinions too easily slide past the admissibility gate? If so, why? Are reliability challenges to warnings testimony too limp? Is warnings law still in need of a “fix”?
Twerski and Henderson suggest in their new Indiana Law Journal article, available as a forthcoming article on the Social Science Research Network,13 that failure-to-warn “remains a doctrine in distress.” Part of the problem is that, while design claims need proof of a “specifically-identified reasonable alternative design (RAD)” available at the time of marketing, many courts “impose no parallel burdens” on the plaintiff in warning cases. The claimant in some jurisdictions needs only assert “in conclusory fashion that the defendant’s warnings of nonobvious product-related risks were inadequate, without specifying exactly what warning the defendant should have given or proving that a different warning would have done any good.”
Their solution is that plaintiffs asserting warning claims “should be required to specify, by suggesting a RAW [a reasonable alternative warning], exactly how the defendant…should have effectively communicated product-related risks and to prove…how the RAW would have reduced or prevented the plaintiff’s harm.” Part I of their new article establishes the conceptual groundwork for this adjustment in perspective. It explains how, notwithstanding interesting differences, design and warning claims “share very similar foundations.” In Part II, the authors demonstrate that the failure to provide a RAW has resulted in the imposition of liability without proof of causation. Part III discusses cases that actually did require a RAW. In Part IV, the noted scholars set forth their proposal for the RAW “fix.”
Thus, in a warning case plaintiff should be required “to spell out the text, location, size and mode of warning (verbal or pictorial).” Failure to provide the reasonable alternative warning should allow the defendant to prevail on a motion for summary judgment. Once the specific RAW is identified, defendant may argue that the proposed warning would not have made a difference and that plaintiff has not proven specific causation. Or the defendant may argue that, if plaintiff’s proposed RAW is given for this particular scenario, then a multiplicity of warnings of the same risk-level would have to be given as well for other situations, creating sensory overload and rendering the suggested RAW useless.
The authors’ newly proposed “fix” can be viewed as part of a continuum of their scholarly writings on warning law which I have cited in endnote 4 of this column. The new proposal merits serious reflection if not immediate adoption.
From the standpoint of practicalities of litigation, I would add that mere assertion of a RAW is insufficient. It cannot be a mere conclusory opinion exercise. The RAW must be adequately proved and that means, in many cases, providing reliable expert testimony that passes muster under Daubert’s (or Frye’s) reliability criteria. Fortunately, the courts’ gatekeeping task and means to perform it are now very well-defined. All that is necessary is that gatekeeping challenges are properly and incisively joined in warning cases. That will require reliable proof of a RAW.
- Twerski and Henderson, “Fixing Failure to Warn,” 90 Ind. L. J. ___ (vol. 1 Fall 2014).
- Henderson and Twerski, “Doctrinal Collapse in Products Liability: The Empty Shell of Failure to Warn,” 65 N.Y.U. L. Rev. 265 (1990); See “Seventy-Fifth Anniversary Retrospective: Most Influential Articles,” 75 N.Y.U. L. Rev. 1517, 1558 (2000).
- Hoenig, “The Ills of Open-Ended Warnings Litigation,” New York Law Journal, Feb. 14, 2011, p.3.
- See James A. Henderson and Aaron D. Twerski, “Doctrinal Collapse in Products Liability: The Empty Shell of Failure to Warn,” 65 N.Y.U. L. Rev. 265, 267, 289—310 (1990); Twerski and Cohen, “Resolving the Dilemma of Nonjusticiable Causation in Failure-to-Warn Litigation,” 84 So. Cal. L. Rev. 125 (2011); Hoenig, supra n. 3, NYLJ, Feb. 14, 2011, p. 3.
- Henderson and Twerski, “Doctrinal Collapse,” supra n. 2, 65 N.Y.U.L. Rev. at 298—310.
- Id. at 307—309.
- Daubert v. Merrell Dow Pharmaceuticals, 509 U.S. 579 (1993).
- Frye v. United States, 293 F. 1013 (D.C. Cir. 1923).
- Hoenig, “The Influence of ‘Daubert’ on Warnings,” NYLJ, June 8, 1998, p. 3 (also available on Westlaw).
- Meyerhoff v. Michelin Tire Corp., 852 F.Supp. 933, 947 (D. Kan. 1994), aff’d, 70 F.3d 1175 (10th Cir. 1995).
- 6 F.Supp.2d 1307 (N.D. Ala. 1997) (appeal dismissed).
- E.g., Roman v. Sprint Nextel Corp., 2014 U.S. Dist. LEXIS 159707 (S.D.N.Y. Nov. 13, 2014) (burn on breast allegedly caused by cellular phone; plaintiff challenged reliability of defense expert’s warnings opinion but court held the proposed testimony admissible); Calisi v. Abbott Laboratories, 2013 U.S. Dist. LEXIS 139257 (D. Mass. Sept. 27, 2013) (plaintiff’s expert precluded from opining that drug’s labels “failed to provide adequate information to doctors” and that his proposed alternative label would do so); Moore v. P&G-Clairol, 781 F.Supp.2d 694 (N.D. Ill. 2011) (allergic reaction to Clairol hair dye; plaintiff’s expert opined that allergy test instructions were vague and imprecise; opinions regarding warnings precluded as unreliable).