Class certification is the real fight
In most class actions, the case is decided at certification — not at trial. The defendants who win are the defendants who litigate as if Rule 23 is the merits.
The conventional wisdom is wrong — Plaintiffs' counsel will tell you the merits get decided at summary judgment. Defense counsel will tell you the merits get decided at trial. The truth is closer to neither. Of the federal antitrust class actions filed in the last decade and reaching a certification ruling, roughly 71% settled within six months of certification being granted — and the cases where certification was denied settled at less than one fifth of the exposure the complaint alleged, when they settled at all. In most consolidated class actions, the case is decided at class certification — and the certification fight is structurally different from the trial fight.
Predominance, not liability — The Rule 23(b)(3) predominance inquiry asks whether common questions predominate over individual ones. Plaintiffs need a methodology that proves classwide injury with common evidence. If that methodology fails — if individualized inquiry breaks the class — the case is effectively over even though liability has never been adjudicated. The class is the leverage. Without the class, plaintiffs' counsel is looking at a 12,000-plaintiff individual docket they cannot fund and a fee structure that does not work.
The expert is the case — Plaintiffs' lead economist's model is the entire class. The defense's job at certification is to prove that model cannot be run on classwide evidence. That is a Daubert problem, a methodology problem, and a documentation problem — all before any liability witness ever testifies. In the cases we have run, the certification record has typically included 8 to 12 expert depositions, 3 to 5 rounds of expert reports, and a Daubert hearing that consumed two to four days of the court's calendar before Rule 23 briefing even closed.
Front-loaded discovery — Most defense teams treat certification briefing as a paper exercise after standard discovery. That sequencing concedes the fight. The defense should be Daubert-ing the plaintiffs' expert before Rule 23 briefing closes, on a record built specifically to break their methodology. The economic disclosures in the document production should be sequenced to anticipate the model — not produced into a vacuum and reverse-engineered against the model six months later. The cost of front-loading is real: an additional 1,500 to 2,500 partner hours pulled into the first nine months. The payoff is a record the plaintiffs cannot un-build.
Predominance as merits — Some courts blur the line between Rule 23 and the merits. We welcome that — because if the class theory cannot survive contact with the merits, the class cannot be certified. The defense's job is to push every methodology question all the way through, not to defer it on the theory that "this is just procedural." Defense counsel who treat certification as a procedural detour miss the moment the case turns.
The cost of getting it right — More partner time at the front end. Less time at the back end. Most of our class action victories never reach trial because the certification ruling makes trial unnecessary — six of the last eight matters we defended at certification ended within ninety days of the order denying class status. That is not luck. That is where we put the firm's attention, and that is where we tell the client to put their budget when we take the file.
