Is there a tougher federal district judge than Stephen Wilson in Los Angeles?  By “tougher,” we do not mean in the usual sense when applied to judges: handing out long criminal sentences – though that certainly does apply to Judge Wilson.   No, we mean in the sense of not suffering fools gladly.  (To borrow a phrase from the movie Rounders, if you are in court for a case for more than a few minutes and you cannot spot the fool, odds are that it is you.)  Judge Wilson, like Judge Rakoff in SDNY or Judge Posner on the 7th Circuit or Judge Kozinski on the 9th Circuit, has a sharp mind and sharp pen (or keyboard).  If he thinks your theory is wrong, he will say so.  If he thinks you are dumb, he will say so.  Sometimes Judge Wilson can come across as a wee bit impatient.  When we were a prosecutor in C.D. Cal. (as Judge Wilson had been several years before us), Judge Wilson was the first judge we encountered who imposed strict time limits during trial.  If he decided that a witness examination had gone on long enough, he would halt the examination on the spot.  We remember a guilty plea in front of Judge Wilson that took place the day before trial was scheduled.  Guilty pleas can be messy affairs.  Some people, no matter how overwhelming the evidence, have a hard time admitting that they did the crime.  And yet, such admission is an essential part of the guilty plea.  Often, the defendant stumbles over the admission.  It can take some goading, some reassuring whispers from defense counsel, and simply some time before the defendant can bring him or herself to utter the magic words.  In our case, the defendant hemmed and hawed a little too much.  Judge Wilson said that there was not enough there for a valid guilty plea, he rejected it, ordered the parties to show up the next day for jury selection, and then called the next case.  It seemed to happen in an instant.  There was panic.  We had already told our witnesses that the case would plead out and that they did not need to come to the federal courthouse the next day.  For all we knew, our case agent was already well outside the jurisdiction, on a protection detail or going undercover.  On the other side of the ledger, the defendant knew what the result at trial would be, and wanted at least to get a couple of points off the sentencing guidelines for acceptance of responsibility.  Somehow the defendant and his lawyer got their act together and returned to court near the end of the day and performed the requisite guilty plea allocution, and all was right with the world.   Judge Wilson’s demonstration of impatience had worked.

That being said, Judge Wilson seems to have displayed quite a lot of patience in dealing with the putative Cymbalta class action.  The plaintiffs alleged that the withdrawal effects of Cymbalta had been understated.  They initially moved to certify a class of consumers in California, Missouri, New York, and Massachusetts who had “received a product that had less value than the value of the product as class members expected to receive it.”  They said they would prove this harm via a “conjoint analysis.”   A conjoint analysis is a statistical technique used in market research to determine how people value different attributes (feature, function, benefits) that make up a product or service.  If you have ever answered a survey where you were asked to assign point values to different attributes of a product or service, you were probably part of a conjoint analysis.  We have dealt with conjoint analyses in cases just enough to be dangerous.  We know this much:  conjoint analysis is complex.  It can be done with rigor.  It can also be outcome-driven hocus pocus.  In any event, Judge Wilson in December 2014 rejected the proposed damages model because it looked only to the demand side of the equation, whereas the prescription drug market is such a hodge-podge on the supply and insurance sides that the relationship between price and consumer value is tenuous at best.  In fact, in Judge Wilson's view, it is "severed."  Different consumers pay different prices with different copays, etc.  Causation and injury cannot possibly be common issues.  Class certification, consequently, would make no sense. 

The plaintiffs decided to try again.  They did not succeed, and Judge Wilson’s order denying class certification in Saavedra v.Eli Lilly & Co., No. 2:12-cv-9366-SVW (C.D. Cal. July 21, 2015), is exactly what we would expect:  succinct and tough-minded.  The plaintiffs moved to certify New York and Missouri subclasses that would seek only the statutory minimum damages ($25 per transaction for Massachusetts, and $50 for New York).  By this technique, the plaintiffs hoped to dodge the need for individualized damages calculations.  But the plaintiffs would still need to establish injury of some sort.  Again, the theory was that the alleged misrepresentations caused the plaintiffs to pay a price premium.   This time, instead of relying on a conjoint analysis, the plaintiffs claimed they could make their point via the defendant’s own internal documents.  But Judge Wilson did not buy what the plaintiffs were selling.  He assigned exactly the right value to the plaintiffs’ argument:  zero.  The plaintiffs admitted that the alleged “premium pricing affected class members to varying degrees because of the existence of health insurance and the varying terms of pharmaceutical benefits that can accompany it.”  Even if the plaintiffs disclaimed the need to quantify the damages, since they were content with statutory damages (at least their lawyers were), they could not use common proof to show the fact of injury.  Whatever the internal documents might show, they could not show (1) that a price premium was actually charged, (2) that the price premium was a premium over the medicine’s true value rather than a premium over competing antidepressants, or (3) that any alleged price premium was passed along to the consumers (who probably handed over a copay).  Judge Wilson denied class certification again.  Tough, but fair.