Last week, the Supreme Court granted certiorari in Google v. Oracle, preparing to review the Federal Circuit’s decision that application programming interfaces (APIs) are copyrightable and that the replication of an API to implement compatibility is not a fair use. There are numerous legal flaws with the Federal Circuit’s decisions, flaws addressed at length in the Disruptive Competition Project’s comprehensive coverage of the case, but one thing that isn’t in the case is a patent—it’s all about copyright.
So why did the Federal Circuit, the appellate court whose jurisdiction centers largely around hearing patent cases, decide key issues of copyright?
28 U.S.C. § 1295
28 U.S.C. § 1295 is the statute that sets the Federal Circuit’s appellate jurisdiction. And, in particular, § 1295(a) gives the Federal Circuit jurisdiction over appeals of decisions regarding civil actions “arising under … any Act of Congress relating to patents.”
But, like I said, there’s no patent issue in the Oracle v. Google decisions by the Federal Circuit. So why is it that the appeal went to them, instead of to the regional circuit that would ordinarily handle appeals of copyright issues?
Well, when the case was originally filed, Oracle alleged patent infringement. And because of that, the district court action is considered to “arise under” patent law—even though the jury ruled that Oracle’s patents weren’t infringed and that decision was not appealed. In other words, by inserting a patent claim into the case early on, a plaintiff can make sure that the Federal Circuit hears the case rather than the regional circuit.
In this case, that meant that the Federal Circuit would decide the issue of copyrightability. It was supposed to do so according to 9th Circuit precedent, which has clearly held that aspects of programs necessary for interoperability are not copyrightable—aspects such as “interface specifications” or “interface procedures”, which are similar to an API. But instead, the Federal Circuit contradicted precedent and ruled in favor of the copyright owner. The Federal Circuit is often criticized for its favoritism to patent owners . It’s not surprising that that bias might extend into copyright, leading to a decision that has been widely criticized as incompatible with the precedent it supposedly relies on.
Federal Circuit Jurisdiction
The current state of law allows anyone to ensure the Federal Circuit hears their case so long as they plead a patent claim at the start of the case—even if that claim is long gone from the case when it’s appealed. And that, due to perceived Federal Circuit bias, is likely to lead to more instances in which copyright plaintiffs do exactly that.
It’s unclear why the Federal Circuit should hear cases when no patent law issue is involved and the case would not otherwise arise under its jurisdiction. It doesn’t seem to be what Congress intended—in fact, when creating the Federal Circuit, the Senate Judiciary Committee explicitly stated that its jurisdiction “is not intended to create forum shopping opportunities between the Federal Circuit and the regional courts of appeals on other claims” and that the manipulative inclusion of a patent claim in a non-patent case should result in the case being heard by the regional circuit.
Fixing the Situation
It’s not impossible to fix, though. Several alternatives to the present situation come to mind.
First, the Federal Circuit could lose jurisdiction over a case if no patent issue existed when the case was appealed. However, simply changing the law so that they would not hear such cases would incentivize plaintiffs to leave patent claims in a case during the appeal as a means of forum shopping, a counterproductive result.
Second, the Federal Circuit could lose its exclusive jurisdiction over patent matters. This would result in some appeals being directed to the regional circuits. However, since the party that appeals would effectively be able to pick its appellate forum, we’d likely see forum shopping and races to appeal that would produce more expensive and less predictable appellate litigation. It also runs the risk of reducing the uniformity of patent law as regional circuits apply Federal Circuit law—though the alternative is having the Federal Circuit apply regional circuit law. And that uniformity has been more elusive than hoped, given the significant differences between district courts reflected in the forum shopping that went on prior to the TC Heartland decision.
Third, we could implement a predominance test. If the predominance of the issues in a given appellate case are governed by regional law, the regional circuit would take the case; if patent law predominates, the Federal Circuit would. This would be closest to the Senate Judiciary Committee’s vision of cases not being forum-shopped to the Federal Circuit if a patent claim is added in order to manipulate jurisdiction. At the same time, it would add one more issue to be litigated as part of appeals, prolonging cases, and the question of who decides predominance in contested cases would need to be resolved.
Fourth, we could codify bifurcation of appeals. If a case has both patent and non-patent components, the Federal Circuit would hear the appeal of the patent issues and the regional circuit would take on the non-patent issues. This would increase litigation cost and complexity due to the need to have separate appeals, and potentially raises complex questions as to how to bifurcate the case when the regional and patent law questions are interdependent. It would avoid non-uniformity of patent law and misapplication of regional circuit law by the Federal Circuit. (An alternative, but similar, approach would be a certification procedure similar to that used between federal courts and some state courts, where the Federal Circuit takes the case and asks the regional circuit to certify what the appropriate rule is on an issue of state law. Despite its attraction, that approach might have Constitutional infirmities.)
Finally, the Federal Circuit might simply be a failed experiment. Earlier commissions reviewing the question of whether to create a specialized patent court worried that specialized judges would succumb to “tunnel vision”, making the law in its area even more esoteric and arbitrary and even allowing the judges to “impose their own views of policy” rather than complying with the law. There’s some evidence that that may have happened with the Federal Circuit, resulting in its capture by pro-patent interests. Maybe it’s time to end the specialized court experiment and return patent law to generalist judges.