Monday, August 22, 2011

Sterns v. Ticketmaster (9th Cir. - August 22, 2011)

I guess today has a theme.  "It seems so obvious, yet it's not."  The disconnect between intuitive justice and the accomplishment of that goal in practice.  Particularly in light of the procedural and substantive obstacles that the judiciary sometimes puts in the way.

Take this case.  It is, in my view, a pretty simple one.  Ticketmaster sold (and still sells) tickets, and had a pretty good tie-in going with EPI.  Basically, after you bought a ticket, the confirmation screen would say "Get $25 cash back on your next Ticketmaster Purchase.  Click here for details."  Then there's a "Continue" button.  You've probably seen things like that on Travelocity or a whole bunch of other web sites as well.  Who wouldn't want to get a free $25, after all?

Anyone who hit "Continue" would then get taken to a new screen, and if they entered their e-mail address twice and hit "Yes", they were signed up for a program that automatically charged the customer a monthly fee.  The customer didn't have to enter their credit card information or any other data; all this stuff was sent automatically by Ticketmaster.  Was there a tiny portion of the screen that said "We'll be charging you $10 [or whatever] for the rest of your life for this meaningless service that virtually no one ever uses?"  I imagine there was.  But it was carefully hidden.  Which is also why they didn't want you to have to reenter your credit card information.  Because then you'd know that you were actually buying something rather than getting it for free.

Or at least that's what plaintiffs claim.  A claim that seems quite plausible to me, having seen several of these schemes myself.  Do some people purchase these things knowing what they're buying?  Absolutely.  I have no doubt.  Do many others click "Yes" having no idea that they're being charged for it, and are these sites expressly designed to take advantage of this fact?  Completely.  I have absolutely no doubt.

So plaintiffs file a class action, claiming that the defendants have engaged in unfair competition.  That claim seems reasonable.  Maybe the web site's disclosures are accurate, of course.  But maybe they're not.  A judge or jury would presumably apply the relevant law and figure out the scoop.  If the sites are unfair, the people who are injured should get their money back.  If the sites are okay, the defendants get to keep what's fair.  Again, we could figure this out before law school.  That's justice.

But, of course, in practice, it's not nearly that simple.  The trial court refuses to certify the class.  On a plethora of grounds, some of which the Ninth Circuit thinks might be okay.  The district court says that the named representatives may not be typical of the class, and the Ninth Circuit agrees that there might be a class definition problem.  Ultimately the Ninth Circuit affirms the dismissal of a number of the claims but remands to allow the district court to consider a recently-decided California case that interpreted the UCL (Tobacco II) to see whether, in fact, a properly-defined class might be certified.

Both the district court and (to a slightly lesser extent) the Ninth Circuit are profoundly concerned with the fact that some class members might well have gotten what they wanted.  For example, Judge Fernandez's opinion says that it might have been okay (had plaintiffs gotten their act together in time) to certify a class limited to customers who told defendants that they never intended to enroll in the service at the time they cancelled their membership in the program, but that a larger class of all subscribers might be profoundly troubling.  Even a class limited to those who never used the program (despite paying for it), Judge Fernandez argued, might well still fail the typicality requirement, since some people might well have deliberately enrolled in the thing but simply never got around to using it.  Maybe they were lazy.  Maybe they died.  Who knows?  We can't have people like that in the class, and no class definition timely proposed by the plaintiffs gets around this problem.

This seems to me to conflict not only with a common sense approach to litigation, but also with precedent.  A class definition doesn't have to be perfect.  It can include people who might not recover.  Take an extreme case, for example.  Imagine that a company has a web site that says, in its entirety:  "Would you like a candy bar?  If so, give us your credit card number so we can verify your identity and we'll ship you one."  Which then charges your credit card $20 without disclosing the charge.  That's perfect for a class action, right?  A class should (and would) be certified for everyone who got charged the $20.  Are there some people in the class who weren't injured because they would, in fact, have paid $20 even if the charge had been disclosed?  Sure.  Are there some people who aren't injured because they went bankrupt or failed to pay their credit card anyway?  Absolutely.  Are there some people who might have read the web site and assumed that they were being charged $20 -- or even $40 -- for the candy bar?  Possibly.  Are there some people who did not care at all about a tiny little $20 charge (Bill Gates?), or who thought it was going to a good cause anyway so wouldn't complain?  Seems plausible.

But we'd still certify the class.  Just because the class contains people who might not be injured doesn't mean the class fails the typicality prong.  Sure, if a huge number of people in the class got what precisely what they bargained for then we've got a typicality problem.  But that's not the Ticketmaster case, in which there's little doubt that the alleged failures to disclose concerned a material condition of the contract (the relevant state law standard) and might well have led a reasonable consumer to not know what s/he was purchasing.  That's not enough to refuse to certify a class.

It's no different than routinely-certified securities class actions.  Are there some members of the class who purchased stock in individual sales, or without reference to market prices, or who already knew about the (alleged) fraud?  Sure.  And at the damages stage, we can refuse to pay those people.  But we don't refuse to certify the class.

The same is true here.  There might well be people, for example, who used the coupons.  That might be some decent evidence that they in fact intended to purchase the program, and hence weren't deceived.  Those members of the class might not get paid at the end of the day -- though some of them might get paid at least a portion; e.g., if they could establish that, sure, they used the coupons, but only b/c they had already been charged, and then they immediately cancelled the thing, etc.  The point is that it's a factual issue, just as it is in non-consumer class actions.  Not everyone gets paid the same amount, and not every member of the class gets paid at all.  That doesn't defeat the availability of class relief.

Nor should it, by the way.  What's the alternative, after all?  Does the candy bar seller really get to keep the millions he rips off from other people, and only have to refund the $20 to those few people who can afford to spend $300+ to file a $20 lawsuit?  You mean to tell me that because we can't precisely define exactly the subset who got injured the defendant totally gets away with it?  Come on.

Nor is this an exaggeration.  Plaintiffs here, for example, see the handwriting on the wall and attempt (albeit belatedly) to define the class so it only consists of people who (1) cancelled the service, and (2) said at the same time as (1) that they didn't knew they had signed up.  That's an incredibly small number of people -- though, given the facts, probably not a trivial number.  But it means that plenty of people don't get to recover money that was (allegedly) illegally taken from them.  People who simply screamed "Cancel my service you bastards!" don't get to recover.  People who didn't feel like giving a reason.  Lots and lots of people.  Sure, you've defined the class so that virtually everyone was indeed injured.  (Even then, the representative may not be typical -- after all, some people who said they cancelled b/c they didn't know what they were ordering may have in fact known after all.)  But in doing so you've excluded giving compensation to a huge number of people who were in fact injured -- indeed, my guess is that you've excluded most of the people who were in fact injured.  That seems neither required by the typicality rules nor the normative principle we should prefer.

Yes, at some point, we have to separate the damaged from the nondamaged.  We can, and should, do that at the damage stage, at least when, as here, there's ample reason to believe that a material portion of the class -- defined as people who ordered the service -- were in fact damaged.  That some of these people may not be damaged shouldn't mean that no one, or virtually no one, gets to recover.  Otherwise you're creating a rule that has the practical effect of letting defendants successfully scam people with little recourse.

So here's another case where I think the right answer is the intuitive one.  And even though the decision from the Ninth Circuit comes close to being the right one -- and is certainly closer than what the district court did -- I think there's definite room for improvement.

It's a good case.  There's a real problem here.  One the legal system is designed to solve.  And the judiciary is making rules here that make things worse, not better.