Analysis Group Affiliate Edward A. Snyder – Dean and William S. Beinecke Professor of Economics and Management at Yale School of Management – has provided expert testimony on class certification in numerous depositions and pre-trial proceedings.
Below, Dean Snyder addresses some key questions.
Q: Rules 23(a) and (b) of the Federal Rules of Civil Procedure typically define "class" within the context of numerosity, commonality, typicality, and adequacy of representation. How do economists view these legal thresholds?
Dean Snyder: I approach these issues as follows: I consider what questions need to be asked and what evidence needs to be evaluated to determine if a member of a proposed class has suffered injury, and if so, to what extent. In addition, I take a look at whether conflicts exist among various members of the proposed class.
If the means by which one member’s claim can be evaluated also can be used to evaluate the claims of other members of a proposed class, then my inclination is to treat those and subsequent claims on a common, classwide basis. Conversely, if the methods and results I get for a member of a proposed class do not allow me to assess the fact and potential extent of injury to another member of the class, my inclination is to treat those and subsequent claims on an individualized basis.
When examining whether conflicts exist among various members of a proposed class, I might examine whether two members of the class are interacting with each other in such a way that their behaviors would define them as members of different classes needing different representation. For example, in an indirect purchaser case where one member of the class is selling the product to another member of the class, the issue of upstream and downstream pass-through rates raises questions of conflict among the buyer and the seller. The judge’s decision to deny class certification in Sugai Products, Inc., et al. v. Kona Kai Farms, Inc., et al. (D. Haw. Nov. 19, 1997) illustrates these points most clearly.
Q: What is the role of merits analysis in class certification?
Dean Snyder: In general, issues of injury – fact and commonality – and potential conflict among class members are the focal points of economic expert opinions. Whether in cases of direct or indirect purchasers, the degree to which merits issues are considered at the class stage varies across jurisdictions and therefore affects expert strategies.
Usually, part of any expert’s assignment is to assume that the case in question has merit. So, I begin by assuming that the plaintiffs’ allegations of the defendant’s misconduct are true – this may include assuming that some of the plaintiffs have been injured some of the time by the defendant’s allegedly bad acts. At that point, my job is to determine (a) whether there is common evidence to prove injury on a classwide basis, and if so, (b) how to calculate the extent of any injury.
In the case of an alleged overcharge, for example, I would not assume that all members of the proposed merchant classes were injured by the alleged anticompetitive behavior because that would assume the ultimate conclusion. Instead, I would determine whether a common method of proof exists that would allow me to confirm that most or all plaintiffs had been impacted by the defendant’s bad behavior and if so, to what extent.
Being knowledgeable about industry-specific circumstances is important when determining whether an alleged violation is likely to have had classwide impact. This includes understanding the product and transactions at issue as well as the interaction among relevant parties. The decision defeating class certification in Fagan, et al. v. Honeywell International, Inc., No. 04-4903-BLS2 (Mass. Super. Ct. Feb. 5, 2008) illustrates the importance of understanding the industry and the types of transactions and pricing one might expect. In indirect cases, an analysis of upstream and, if relevant, downstream pass-through also may be helpful in determining whether damages can be analyzed using a common classwide approach.
Q: What elements would you consider in determining whether merits and damages can be assessed at the class level?
Dean Snyder: An expert might consider elements specific to the industry or circumstances that characterize the case. For example, is the product in question a commodity or a differentiated product? How many potential intermediaries might exist between the defendant and the plaintiff class? What is the nature of the interaction between the buyer and the seller? In other words, how do we define the nature of the products, the nature of the transactions, and the types of relationships that exist between defendants and plaintiffs?
Economic analysis of classwide impact often focuses on the question of whether the alleged collusion had common impact as opposed to whether it had any impact. For instance, in the context of a proposed class of direct purchasers, we might examine whether an alleged conspiracy had any impact for at least a portion of the class, even though the conspiracy is assumed.
In another case, this time in the context of an indirect class, the analysis of class-wide impact may require determining the size of any alleged overcharges. Plaintiffs may be in a stronger position to prove impact to downstream purchasers when alleged overcharges are large relative to the value of the final product, while defendants might argue that a small alleged overcharge relative to the value of the final product would be difficult or impossible to detect using classwide tools.
In a matter involving the alleged price fixing of infant formula, I reviewed plaintiffs’ claims of conspiracy, antitrust injury, and damages to determine whether they could be adjudicated on a classwide basis. They could not, due to substantial variations in the degree of alleged antitrust injury, and the severity of alleged damages affecting the individuals involved. No classwide methodology could have accounted for the multitude of factors, including delays in price increases or plateaus in price levels – which varied by consumer – affecting determinations of conspiracy, injury, or damages. Any model would have been either so complex or so speculative in its approach that significant and fundamental errors would have been unavoidable.
The court found that the plaintiff had "failed to establish by a preponderance of the evidence that there [were] questions of law and fact common to the members of the class that predominate over questions of fact affecting only individual members." Holmes v. Abbott Laboratories, Inc., Civ. No. 94-221 (unpublished order: Michigan District Court, Calhoun County). ■
From Analysis Group Forum (Fall/Winter 2008)