By Paul R. Gugliuzza, Professor of Law, Temple University Beasley School of Law
The Federal Circuit decides lots of mandamus petitions on transfer of venue under 28 U.S.C. § 1404(a). Most federal courts of appeals decide one or two per year, at most. The Federal Circuit, in 2021 alone, decided 30.
Yet the Federal Circuit rarely designates those decisions as precedent. Over 85% of the court’s venue mandamus decisions since 2008 have been nonprecedential. And, it’s worth noting, the Federal Circuit’s few precedential decisions are wildly unrepresentative: nearly 80% of them grant mandamus, even though, overall, petitioners succeed less than a third of the time.
In an order issued earlier this month, District Judge Alan Albright—who, despite recent changes in how cases are assigned to judges in his district, still has twice as many patent cases on his docket as any judge in the country—picked up this thread about the proliferation of nonprecedential Federal Circuit transfer decisions.
In Motion Offense v. Google, the patentee, who opposed transfer, argued that Judge Albright should follow Fifth Circuit case law on transfer of venue—and ignore the nonprecedential Federal Circuit decisions Google cited in its transfer motion. Judge Albright rejected the patentee’s argument, noting that he “cannot and does not overrule the reasoning of the Federal Circuit in a patent case.” (Motion Offense slip op. at 7.)
He continued: “Although the Federal Circuit issues unpublished, nonprecedential transfer opinions, the Federal Circuit frequently cites these opinions as though they precedentially interpret Fifth Circuit law.” (Id. (emphasis mine).)
Given how important transfer disputes are to patent litigation—and how often they occur—it’s worth asking whether Judge Albright is correct about how the Federal Circuit treats its nonprecedential transfer decisions.
Since 2008, the Federal Circuit has issued well over 100 nonprecedential transfer decisions, most of them coming out of two district courts in the Fifth Circuit: the Eastern and Western Districts of Texas. Those nonprecedential Federal Circuit decisions technically aren’t binding in future cases. Indeed, under the Federal Circuit’s unusual choice-of-law regime, the venue analysis is ostensibly governed by Fifth Circuit precedent, so even Federal Circuit decisions designated as precedential are binding only to the extent they’re interpreting Fifth Circuit law.
Yet the Federal Circuit does, in fact, often cite its own nonprecedential transfer decisions as if they were binding interpretations of Fifth Circuit law.
* * *
Drawing on a dataset that Jason Rantanen, Jonas Anderson, and I developed for a soon-to-be published article on the Federal Circuit’s mandamus practice, I looked at every Federal Circuit ruling on a mandamus petition challenging a decision by Judge Albright to deny transfer under § 1404(a)—38 rulings in total, all issued since June 2020. I then examined the citations in those rulings to prior Federal Circuit and Fifth Circuit decisions.
The bottom line: nearly a quarter of the citations were to nonprecedential Federal Circuit opinions.
This result holds across both (1) the precedential status of the citing decision and (2) whether the citing decision granted or denied mandamus. The only possible exception is Federal Circuit decisions denying mandamus, which tend to be quite short and light on citations as whole, including citations to nonprecedential Federal Circuit decisions.
* * *
It’s not just the numbers but also the substance of the rulings that reflect the Federal Circuit’s heavy reliance on prior nonprecedential decisions.
For instance, in granting a mandamus petition filed by Apple, the Federal Circuit chided Judge Albright for “erroneously discount[ing]” the location of third-party witnesses in the transfer analysis. The Federal Circuit quoted from its prior nonprecedential opinion in In re HP for the proposition that “there [is] ‘no basis to discount’ third-party entities . . . ‘just because individual employees [i.e., witnesses] were not identified’” in the transfer motion. (Apple slip op. at 6.) The Federal Circuit’s opinion in Apple also relied on and quoted from two prior, nonprecedential decisions involving Google to reject Judge Albright’s treatment of the location of party witnesses in the transfer analysis. (Id. at 7, 10.)
Indeed, one of those Google opinions itself relied on the Federal Circuit’s prior nonprecedential decision in In re TracFone in granting transfer, writing that “[t]his case fits squarely within that line of precedent”—the “line of precedent” consisting of two cases: the Federal Circuit’s precedential opinion in another dispute involving Apple, In re Apple Inc., 979 F.3d 1332 (Fed. Cir. 2020), and the nonprecedential decision in TracFone. (Google slip op. at 10.)
Similarly, in granting mandamus in In re Juniper Networks—a precedential decision—the Federal Circuit cited its prior nonprecedential opinion in In re Hulu five times, concluding that “this case is a very close cousin of our recent decisions in Samsung [precedential] and Hulu [nonprecedential], and the disposition of this case is largely dictated by the disposition of those cases.” (Juniper slip op. at 14 (emphasis mine).)
Perhaps the most stunning treatment of nonprecedential mandamus decisions by the Federal Circuit is In re Pandora Media, which actually contains more citations to nonprecedential Federal Circuit rulings than to precedential Federal Circuit rulings. The Federal Circuit’s determination that Judge Albright erred by disregarding the subpoena power of the proposed transferee court was based entirely on three nonprecedential opinions: a 2014 case involving Apple, the Hulu case mentioned above, and a 2018 case involving HP. (Pandora slip op. at 6.)
The Federal Circuit in Pandora also leaned on the nonprecedential opinions in TracFone and Google (both discussed above) in criticizing how Judge Albright analyzed considerations about convenience for potential witnesses, noting that “we held in TracFone and Google that the litigation should be conducted where more witnesses could testify without leaving their homes or their regular places of business.” (Pandora slip op. at 12 (emphasis mine).)
* * *
It’s tempting to brush off the Federal Circuit’s use of nonprecedential decisions to justify its venue mandamus rulings. Venue mandamus is a niche issue, brought to prominence by a semi-specialized appellate court that is fighting a problem of court competition somewhat unique to patent litigation. And sometimes the Federal Circuit discusses nonprecedential decisions because it was the parties who cited them in the first place, creating a bit of a chicken-or-egg dynamic.
Yet the court’s own rules are clear about how it’s supposed to treat nonprecedential decisions. Federal Circuit Rule 32.1(d) states, in relevant part:
The court may refer to a nonprecedential or unpublished disposition in an opinion or order and may look to a nonprecedential or unpublished disposition for guidance or persuasive reasoning but will not give one of its own nonprecedential dispositions the effect of binding precedent.
This rule makes plain that nonprecedential venue mandamus rulings are not binding law. But the Federal Circuit arguably treated them as such in cases like Google, Pandora, and many others.
Viewed in that light, Judge Albright’s frustration is understandable: how can he make transfer decisions that are consistent with Federal Circuit law when it’s impossible to tell precisely what that law is?
* * *
We live in a time of declining public trust in government institutions and skepticism of the rule of law. Though transfer disputes in patent litigation might not capture the public’s attention like Supreme Court cases about guns, abortion, or voting rights, the Federal Circuit’s unprincipled use of nonprecedential opinions is precisely the sort of everyday lawlessness that, over the long term, can strain trust in the courts.
* * *
A few methodological notes and caveats:
- In calculating the numbers reported above, I counted only citations within the analytical portion of the Federal Circuit’s ruling, and only in the portion of the court’s analysis addressing the § 1404(a) issue specifically.
- I ignored citations in “citing” or “quoting” parentheticals.
- I hand counted the citations, so it’s possible I missed or misclassified a small number of them. But the numbers are sufficiently large (545 relevant citations across 38 decisions) that a few errors, if they exist, wouldn’t affect the overall results much.
- David Hricik wrote for this blog about how, in an earlier version of the Motion Offense order, Judge Albright claimed he’d have denied transfer under Fifth Circuit venue law but was compelled to grant transfer under Federal Circuit law. That claim remains in the amended order mentioned above, though Judge Albright altered the text slightly, writing: “The Court’s decision turns on whether it should apply traditional Fifth Circuit transfer law or apply the Federal Circuit’s interpretations of Fifth Circuit transfer law. This Court cannot ignore or overrule cases from the Federal Circuit. Only because of those Federal Circuit cases, this Court GRANTS Defendant’s Motion to Transfer.” (Motion Offense slip op. at 1. (first emphasis mine))