Argument review: Justices frustrated with posture of case on joint infringement
on May 1, 2014 at 3:20 pm
It was doubtless with some sense of relief Wednesday morning that the Chief Justice called for argument in “our last case this Term,” Limelight Networks v. Akamai Technologies. Fittingly, it was a patent case – the Court has heard more patent cases this Term than in any prior Term since the Patent Act’s enactment two generations ago. Unfortunately, the Court also has had more bad “vehicles” this Term than in any other Term I can recall. So it was all too familiar for the Justices to spend such a large part of the argument debating whether the case properly presents a question worth deciding.
This case involves the vexing problem of two-party infringement: Limelight markets a service under which Limelight and its customers (allegedly) take all the steps to implement the patented method; however, neither party, standing alone, takes all of those steps. The Federal Circuit for a decade has rejected such claims, but reversed course in this case to accept such a claim. The central problem that dominated the argument was the relation between the two relevant subsections of Patent Act § 271, a problem that can be explained only with some doctrinal and procedural background.
On the doctrinal point, simplifying the matter starkly, subsection (a) deals with direct infringement and imposes almost strict liability; all agree that this can cover some cases when the steps of a patented method are taken by more than one person – the provision extends at least to parties in a principal/agent relationship. Subsection (b), by contrast, provides for “induced” infringement, when one party induces another to “infringe,” and results in a somewhat lower standard of liability.
To summarize the relevant problem with the case’s procedural posture, the patent-holder Akamai brought suit against Limelight under both subsections, but the case went to trial under subsection (a). A jury held in favor of Akamai under that provision, but the district court (applying settled Federal Circuit law) granted Limelight judgment notwithstanding the verdict. A panel affirmed, but the Federal Circuit vacated the panel decision and granted rehearing en banc. On rehearing en banc, a splintered Federal Circuit shifted direction and decided the case not as a matter of direct infringement under subsection (a) (the topic of the jury verdict), but rather as a matter of inducing infringement under subsection (b). And considering that provision, the en banc court held Limelight liable for inducing infringement under subsection (b) (the provision that was not at issue at trial).
Limelight understandably sought review of the decision imposing liability for inducing infringement, which the Court granted. Akamai filed a cross-petition challenging the panel’s vacated holding that it could not recover for direct infringement; the Court has been holding that petition since it granted Limelight’s petition. Although Limelight did not address direct infringement in its briefs (understandably, because the Court didn’t grant Akamai’s cross-petition), Akamai defended the judgment below on that basis, and IBM filed a noteworthy amicus brief suggesting that subsection (a) provided the most promising avenue for dealing with the general problem of two-party infringement.
So the Justices came to argument preoccupied with the concern that the question of direct infringement is a necessary predicate for any sensible consideration of the induced infringement question on which it granted review.
Much of the argument was consumed with a discussion of the exact status of the subsection (a) question. And on that question, Limelight and Akamai both agreed that the subsection (a) question remains to be addressed in the court below – that is, if the Court were to reverse Akamai’s en banc victory under subsection (b) and send the case back to the Federal Circuit, that court would remain free to rule in favor of Akamai under subsection (a).
The obvious concern such a scenario raises is that it would render any ruling the Court might make about induced infringement under subsection (b) a “nullity” (in Justice Kagan’s words). As she emphasized: “I can’t think of a way in which our decision on the 271(b) question would be relevant for any case if the Federal Circuit on remand goes the opposite way on the 271(a) question.” Similarly, Justice Alito seemed deeply concerned that “the question before us really has no significance that I can think of . . . [s]o you’re asking us to decide a question . . . that is of no value?”
At least in theory, the Justices could send the case back to the Federal Circuit with instructions to address the direct infringement question, giving the Court the benefit of the Federal Circuit’s views on that question before the Court resolves the induced infringement questions. The Justices, however, didn’t even discuss that possibility, which would involve a rather advisory process – asking the Federal Circuit what it might think about subsection (a) when it hasn’t yet told the Federal Circuit what the Court thinks about subsection (b).
Another possibility would be to proceed to decide the direct infringement question now, because it has been briefed thoroughly by Akamai and, to some degree, by Limelight in its reply. Justice Kagan, at least, thought that a poor choice, because it would effectively be “sandbagging” Limelight. To satisfy her concerns, the most obvious resolution would be to call for briefs on the subsection (a) question and set the case for re-argument in the fall. Seth Waxman (arguing on behalf of Akamai) welcomed the possibility of that approach, which effectively results in a grant of Akamai’s hitherto-languishing cross-petition.
Again, that would make sense only if the Justices thought the questions were sufficiently interrelated that they need both of them to make a sensible decision. As discussed below, Justice Kagan in particular was convinced that the Federal Circuit’s decision rested on an inseverable logical link between the two provisions.
Yet another possibility, suggested in a joking way by Waxman, was that the Court could dismiss the petition as improvidently granted. Although his suggestion drew laughter for its obviously self-serving nature, the frustrations voiced in the argument do suggest it is a very real possibility. The problem with that approach, though, is that it would make it quite difficult for the Court ever to get a case that properly presents the question under both subsections: affirming the Federal Circuit’s finding of liability under subsection (b) would vitiate all pressure to litigate the propriety of liability in these cases under subsection (a). And if the Justices are unsatisfied with what the Federal Circuit has done (as they seem to be), they’ll have to deal with the problem sooner or later.
What little time the Justices spent discussing the merits of the question involved two basic points – one strongly favoring Akamai and one strongly favoring Limelight. So, for example, about the only subject Aaron Panner (arguing on behalf of Limelight) got to address during his time (other than questions about the posture of the direct infringement question) was the concern that his “position makes it pretty easy to get around patent protection,” in the words of the Chief Justice. “All you’ve got to do is find one step in the process and essentially outsource it . . . and you’ve essentially invalidated the patent.”
Panner’s response, as in his brief, was to blame sloppy drafting by Akamai’s patent attorneys. He argued that they easily could have written the patent to direct the claims to behavior by a single actor. But the Chief Justice and Justice Scalia openly scoffed at that argument. First, the Chief Justice: “[W]hen you say it could be written that way [i.e., describing the activity of a single person], it would be by not claiming one step in the method, I assume, which changes the whole patent.” Then, even more pointedly, Justice Scalia:
I don’t understand what you’re saying. You’re saying that you can avoid the problem if the patent is drawn in such a way as what, to require a single person to do all the steps? How does that give you any more protection? You can still violate the patent by not having one person do all the steps; have another person do some of the steps. And it’s just as effective in stealing the idea and yet there would not be a violation of the patent.
To Justice Kagan, the concern that a ruling for Limelight would create a “real problem here in terms of an end run” only underscored the problem with considering the induced infringement question in a vacuum. As she read the opinion of the Federal Circuit, the need to solve the “end run” problem motivated the court below to find some answer in the statute; its conclusion that liability was available under subsection (b) simply reflected the judgment that (b) was a better vehicle for liability than (a).
The same concern dominated the argument of Assistant to the Solicitor General, arguing on behalf of the Patent and Trademark Office. Justice Alito tried repeatedly to force her to identify “any policy reason that could support the conclusion that there was not infringement on the facts here? . . . [I]f you were writing a statute, . . . what policy concerns would support a conclusion that there was no infringement on the facts here?” When Justice Scalia suggested some interest in whether Anders’s attempts to answer involved subsection (a) or (b), Justice Alito reiterated: “I don’t care. (a) or (b), whatever, or (c) or (d) or (z). What policy concern would support the conclusion that there is an infringement on the facts of this case.”
That discussion might have seemed to portend a poor outcome for Limelight, but a largely overlapping group of Justices also seemed convinced that the language of subsection (b) could not possibly be read to justify imposing liability on Limelight. So, for example, when Justice Kagan suggested that the judges of the Federal Circuit “thought they were being very clever by putting it into a 271(b) box and avoiding the strict liability consequences of what they were doing, but also avoiding the possibility of an end run of the patent law,” Justice Scalia interjected that they were “also avoiding the text of the statute.” Justice Kagan immediately agreed: “There is that problem. There is that problem.”
You have to know the Justices are not at rest when the same person can both suggest that a victory for the defendant justifies “stealing” patented inventions while a victory for the plaintiff amounts to “avoiding the text of the statute.” There always is the likelihood that when the Justices return to their chambers they will decide to make the best of what they have. But having said that, it seems quite likely, when the Justices meet to consider the case tomorrow, that they will settle upon some course other than proceeding directly to resolve the induced infringement question on which they granted review. My heart goes out to the unfortunate law clerk who recommended review of the petition in this case (if in fact that was the source of the Court’s action).