NJOY, Inc. (NJOY), a maker of e-cigarettes, was alleged to have misrepresented the health risks of its e-cigarettes compared to traditional tobacco cigarettes in its advertising. Plaintiffs also alleged that NJOY failed to disclose the ingredients in its products, as well as the potential related health risks. Plaintiffs claimed that, without this misleading advertising and labelling, they would not have purchased NJOY’s products or would have done so at a lower price.

As part of their first motion for certification, Plaintiffs’ expert proposed using a conjoint survey or a direct method survey to measure the value consumers assign to the so-called “safety claim.” In August 2015, a federal judge denied class certification in the matter, finding that plaintiffs had not demonstrated that damages could be calculated on a class-wide basis using these approaches. The Court ruled that that the correct measure of damages is the difference between the price consumers paid and the price that would have been paid in the market absent the alleged misrepresentation. She concluded that the proposed surveys would only measure consumers’ subjective value from a demand side perspective, ignoring supply side and market factors that would affect both actual and “but-for” prices.

Plaintiffs were allowed to submit a revised damages approach in a second motion. Their expert re-proposed conducting a conjoint/direct survey but added a third “hybrid” option, which was to combine the survey with a statistical technique known as Bayesian hedonic regression. Plaintiffs’ expert claimed that hedonic regression could be used to explain observed e-cigarette prices as a function of their quantitative attributes (e.g., brand, number of puffs, battery life), as well as more qualitative attributes (e.g., taste and the “safety claim”). The more qualitative attributes would be estimated using conjoint analysis. NERA Senior Vice President Dr. Kent Van Liere was retained by NJOY to rebut the proposed conjoint survey and Senior Vice President Dr. Denise Martin was retained to rebut the Bayesian hedonic regression. Dr. Van Liere described to the Court why the proposed conjoint survey was flawed and would generate unreliable results. Dr. Martin explained that the strict market conditions in which hedonic regression was applicable did not exist, and that these deficiencies could not be cured by supplementing the regression with the results of a flawed conjoint analysis.  

A new judge was assigned to the matter and ruled that the new conjoint and direct surveys suffered the same limitation of being demand-side only, and thus were not tied to the market price. Moreover, citing Dr. Martin’s testimony in several places, he concluded that “the proposed Bayesian hedonic regression model is simply not designed to measure only those damages attributable to NJOY’s misrepresentations and/or omissions, and thus does not satisfy Comcast.” Consequently, he denied certification on the grounds that Plaintiffs had not established that they could measure class-wide damages tied to the theory of liability.