Collateral estoppel is a resource-saving shortcut. Judges consider it when an issue previously received sufficient judicial attention. And they apply it when the issue was resolved against the party now seeking, in some way, to circumvent that resolution. In the absence of some intervening change in the law or new evidence, resolution likely would be the same as before. Addressing the issue anew would therefore be a waste of everyone’s resources. So, the judge bars that party from re-litigating the issue and applies the consequences of the prior resolution. Collateral estoppel’s logical simplicity is manifest, but its application is sometimes complicated.
Late last year, in Google LLC v. Hammond Development International, Inc., 54 F.4th 1377 (Fed. Cir. 2022), the Federal Circuit dealt with complications it apparently hadn’t been presented before. Google had separately challenged claims in related patents as obvious over combinations of published prior art. Google presented these challenges in separate inter partes review proceedings, each directed to one patent. In one proceeding (the ’483 IPR), the Patent Trial and Appeal Board canceled the claim. In the other (the ’816 IPR), however, the Board refused to cancel a claim that, on appeal, Google said implicated the same patentability issue resolved in the earlier ’483 IPR. Arguably, the Board behaved inconsistently. Collateral estoppel, Google argued, should have applied to prevent that inconsistency, the outcome in the ’483 IPR being correct.
As the party pleading collateral estoppel, Google had to show:
(1) the issue is identical to one decided in the first action; (2) the issue was actually litigated in the first action; (3) resolution of the issue was essential to a final judgment in the first action; and (4) [the party against whom collateral estoppel is being asserted] had a full and fair opportunity to litigate the issue in the first action.
Id. at 1381 (alteration in original, citation omitted). On appeal, the parties only disputed the first of these four showings. Both patent claims—the claim the Board cancelled in the ’483 IPR and the claim it upheld in the ’816 IPR—recite a communication system connecting a cell phone, for example, to remote application servers that execute applications the phone’s user requests. The claims differed only in language describing the number of application servers. The two claims were not therefore identical. But that’s of little import as the court previously said literal identity does not matter (link).
“Collateral estoppel,” the court emphasized, “requires that the issues of patentability be identical.” Id. If differences in the two claims “do not materially alter the question of [patentability],” then estoppel may apply. Whether any difference presents a material alteration is a question of law the court reviews de novo based on the Board’s fact-findings, which the court reviews for substantial evidence. Applying these review standards, the court concluded that the difference—number of application servers—between the two claims in the related patents “does not materially alter the question of patentability,” and further concluded that the two claims “are materially identical for purposes of collateral estoppel.” Id. at 1381–82. The court therefore held the relevant claim in the ’816 IPR unpatentable. Id. at 1382.
One interpretation of the court’s conclusion that the two claims “are materially identical” is that the claims recite obvious variants of each other. In view of the presumption of validity, there is a tendency to think that different claims in separate patents must validly be reciting different inventions. The tendency is akin to initially seeing the horizontal lines in the Müller-Lyer illusion as having different lengths.
But a careful assessment reveals they are the same length. The pairs of diagonal lines in each—forming arrows in one and wings in the other—do not yield different horizontal line lengths, despite the tendency to think otherwise. As in the illusion, different words in comparable claims of related patents are often nothing more than those arrows and wings, both interchangeably found in prior art and predictably combinable with what’s commonly recited in both. The assessment often leads to a conclusion the claims recite obvious variants—double patenting—an assessment and conclusion a terminal disclaimer obviates.
During prosecution, the applicant for the patent involved in the ’816 IPR voluntarily filed a terminal disclaimer over the earlier-issued patent involved in the ’483 IPR. Perhaps the applicant foresaw the potential that the examiner would find no material differences in the claims. But with the terminal disclaimer on file, the examiner never had to make that assessment. Whatever the applicant’s motivations, after an expensive inter partes review and appeal, the Federal Circuit effectively made that assessment.
Decisions like Google are helpful where, as here, the Board lacks an explicit rule authorizing it to apply collateral estoppel. After Google, not only should parties to inter partes proceedings more seriously consider the potential collateral estoppel effect of an earlier Board decision on a later Patent Office proceeding involving materially indistinguishable claims, but so too should the Board—if only to preserve its own resources. In contrast to the Board, Patent Office examiners are empowered by an explicit rule, see 37 CFR § 42.73(d)(3)(i), to apply this same doctrinal hand of fate—although, they rarely do.