- within Intellectual Property, Environment and Finance and Banking topic(s)
- with readers working within the Technology industries
One approach frequently used by plaintiffs’ experts to attempt to establish common impact at the class certification stage of a class action antitrust lawsuit is the “two-step econometric method,” sometimes also known as “in-sample prediction analysis.” This approach purports to establish common impact by evaluating whether average prices were higher during the class period compared to a benchmark period free of the alleged conduct and identifying which putative class members paid at least one overcharged transaction.
In their latest article published in ABA Antitrust Magazine, Senior Consultants Veronica Postal and David Terner find that this approach is unsupported by econometric theory and is biased to find class-wide impact even where there is none.
The authors analyze the economic foundations of the two-step econometric method and the conclusions that can be drawn from it. They show that the method treats any difference between actual and predicted prices as statistical evidence of transaction-level impact, even though those differences reflect prediction errors that, even in a well-specified regression model, reflect unexplained variation independent of the alleged conduct.
Dr. Postal and Dr. Terner use a simulation to show that this methodology can generate false positives. In particular, the methodology is biased toward identifying class members as impacted by the alleged conduct even when no overcharge exists and supports a finding of class-wide impact regardless of how many members of the putative class were actually harmed. These results underscore the importance of assessing whether econometric methods used at class certification can reliably distinguish common impact from random variation in the data.
Key Takeaways
-
The two-step econometric method estimates a single average overcharge across all putative class members and adds random noise in the form of prediction errors to evaluate transaction- or individual-level impact.
-
The method assigns meaning to prediction errors that, under standard regression principles, represent unexplained variation rather than evidence of the alleged conduct.
-
Simulation analyses show that the methodology will result in a finding of class-wide impact regardless of the share of the putative class that actually incurred an overcharge and regardless of the existence or magnitude of the simulated overcharge.
How NERA Can Help
NERA combines economic expertise with advanced econometric analysis to help clients address class certification and common-impact issues in antitrust litigation. In class certification matters, where plaintiffs often rely on econometric techniques to argue that common impact can be shown across a putative class, NERA brings the expertise needed to test those claims critically and defensibly. Our work helps clients evaluate whether the methods used by opposing experts are economically sound and capable of identifying which, if any, alleged class members were affected.
Our antitrust and competition teams combine deep econometric expertise with careful judgment to produce analyses that are technically rigorous and persuasive in court. We test—and retest—the assumptions and results of the models and examine whether the opposing expert’s conclusions are sufficiently supported by the data and by the tenets of economic theory.
The result is clear, defensible analysis that translates complex econometric issues into evidence that helps counsel and courts determine if a class of plaintiffs exists.
The content of this article is intended to provide a general guide to the subject matter. Specialist advice should be sought about your specific circumstances.
[View Source]
