Anderson & Menell: Informal Deference

Is it finally time to adopt a deferential standard of review in patent claim construction appeals? Professors Jonas Anderson and Peter Menell respond emphatically in the affirmative in their recent patent law scholarship, From De Novo Review to Informal Deference: An Historical, Empirical, and Normative Analysis of the Standard of Appellate Review for Patent Claim Construction. In the article, Anderson and Menell argue that the time is ripe to abandon the de novo review standard in favor of a deferential standard that comports with the inherent and unavoidable factual aspects of patent claim construction. The article first presents a comprehensive and insightful historical analysis of patent claim construction. It then reveals enlightening results of the authors' empirical study of Federal Circuit jurisprudence from 2000 through 2010, and provides a normative analysis to explain these findings. Finally, it proposes a pragmatic framework for a dual standard for reviewing claim construction rulings, a framework that promotes transparency and uniformity moving forward. The result is a must read for anyone interest in patent litigation, and more specifically in the appellate review of patent claim construction at the Federal Circuit.

Part I. Historical Analysis

Following Markman, there has remained much division among Federal Circuit judges regarding the appropriate standard of appellate review. The Federal Circuit's en banc decision in Cybor Corporation v. FAS Technologies, Inc. in 1998 resulted in a divided decision holding that claim construction is a purely legal issue subject to de novo review. Since 1998, the reversal rate in patent cases has remained high, however, which has fueled criticism of the patent litigation system. In 2004, this induced the Federal Circuit to grant en banc review of claim its construction jurisprudence in Phillips v. AWH Corp. Given the court's notice announcing questions it would consider at the oral argument, it appeared the court would attempt to address dissatisfaction among district judges and practitioners with the de novo standard of review of claim construction decisions.

In the Phillips decision in 2005, the majority emphasized the role of intrinsic evidence and the specification as being typically dispositive, but still authorized trial courts to consider extrinsic evidence. The majority also back away from a presumption that the dictionary meaning of claim terms would apply, shifting the focus of construing claim terms to the context of the patent. The court, however, did not revisit the question of appellate standard of review, leaving the de novo standard in place. Division remains among Federal Circuit judges on this issue, prompting several calls for more deference to lower court claim construction determinations and for overturning Cybor.

Part II. Empirical Results

The authors complied a database for appellate claim construction decisions from 2000 through 2010, with the purpose of taking a comprehensive look at the impact of the Phillips decision on claim construction. The database includes all precedential, non-precedential, and summary affirmance (Rule 36) opinions, resulting in a total of 1,039 cases and 1,895 individual claim terms. This means that, on average, each case contains 1.8 appealed claim terms. The authors coded all cases and claim terms across four categories: case data, claim term data, construction evidence, and appellate reasoning. The authors stress that the data reflects appellate behavior, and is not designed to answer any questions about district court accuracy.

The collected data indicate that the reversal rate of claim construction appeals at the Federal Circuit has dropped substantially since Phillips (see "Table 2: Pre- and Post-Phillips Reversal Rates" in the article). The authors note that, since Phillips, the reversal rate on claim construction issues is much closer to that of other patent-related issues, demonstrating that claim construction is no longer an extreme outlier at the Federal Circuit. According to the authors, the data further showed that, in accordance with the Phillips opinion, the Federal Circuit relies heavily on the specification, with decreasing reliance on prosecution history in construing claim terms. Additionally, the authors point out that the Federal Circuit has decrease its reliance on extrinsic evidence, from 32.3% to 24.5% of terms construed since Phillips. (This drop in reliance on extrinsic evidence particularly reflects decreased use of dictionaries, but notably use of expert evidence appears to be relatively unaffected). Finally, there was a significant increase in the number of claim construction appeals decided by Rule 36 summary affirmances, jumping from 16.1% to 28.9% of cases decided since Phillips.

In interpreting the empirical results, the authors considered several potential theories to explain the reduction of the reversal rate following Phillips. They do not believe that the clarification of law that occurred after Phillips can completely explain the significant, immediate, and lasting drop in reversal rates. Such a theory would expect to see gradual reductions in the reversal rate, as appeals began coming in from the lower courts. This study, however, found an immediate and drastic decline in the reversal rate. The authors conclude that the most likely explanation is that the Federal Circuit judges backed away from the de novo standard of review following Phillips, marking the beginning of the era of "informal deference" for claim construction appeals.

According to the authors, the immediate drop in reversal rates following Phillips, further supported by anecdotal evidence and increased summary affirmances, demonstrates a more deferential Federal Circuit than previously thought, providing more predictability in patent claim construction.

Part III. Toward a Coherent Standard of Appellate Review

While informal deference has improved the predictability of patent claim construction, Federal Circuit judges remain divided over the proper formal standard of review and the degree of deference that should be conferred to district judge rulings. The authors point out that the de novo standard discourages trial judges from hearing expert testimony debating how the PHOSITA (person of ordinary skill in the art) would interpret claim terms as they are customarily used in drafting. Furthermore, the authors identify many different private and social costs of the de novo standard of review, which cumulatively raise the cost of patent litigation without any discernible benefits.

Because the intrinsic evidence in a patent case is viewed from the standpoint of a skilled artisan, the authors reason that the proper standard of review must integrate deferential review of factual predicates with something approaching de novo review of documentary sources to achieve the ultimate appellate determination. This proposed "mongrel" appellate standard would afford great weight to district court rulings on industry terms particularly necessitating a construction reflecting the understanding of a PHOSITA, while applying a lesser yet still heightened standard for reversal where the intrinsic record controls. Thus, according to the authors, "trial courts' claim construction rulings should be upheld if not clearly erroneous or clearly contradicted by the specification or prosecution history."

In the end, Professors Anderson and Menell assert that the more deferential standard of review they advocate would invigorate development of the factual record in trial court Markman hearings, and in turn would further the "overarching notice goal of the patent system - that patent claims should be understood objectively from the standpoint of skilled artisans unless the patent affords unambiguous scope."

Anderson and Menell's Contribution to the Ongoing Debate

This article instills optimism that the patent machine is lumbering in the right direction, and that the courts may soon acknowledge the "failure of the de novo experiment." This optimism appeared to experience a setback this past November, as the U.S. Solicitor General's brief to the Supreme Court recommended denial of the writ of certiorari in Retractable Technologies, Inc. v. Becton, Dickinson & Co. As pointed out at Patently-O, the Solicitor General's recommendations have been followed 80% of the time in recent years, and the Supreme Court has in fact now denied the petition. The rationale provided for denying certiorari on the question of the appropriate standard of review of patent claim construction rulings, however, was simply that Retractable Technologies. was not the appropriate case to consider the question. The Solicitor General explained that "[b]ecause the district court's claim-construction ruling did not depend on the resolution of any questions of fact, this case does not present the question whether a claim-construction ruling that is predicated on factual determinations should nevertheless be subject entirely to de novo appellate review." This can certainly be interpreted as revealing that, when an appropriate case does come along, the Solicitor General will recommend a grant of certiorari. And if the Supreme Court finally considers the issue, I think the Court would do well to adopt the approach proposed by Anderson and Menell.

Also noteworthy are these recent postings at Patently-O discussing the current number of judges on the Federal Circuit and the opportunity it presents for a potential en banc review of Cybor.

Drafted by Alex Young (AlexY@smu.edu), research assistant to Professor Sarah Tran, and a 2015 Juris Doctor candidate at SMU Dedman School of Law. While Professor Tran is on leave, Professor David Taylor is supervising her research assistants. Alex received his B.S. in Mechanical Engineering from the University of Virginia. Prior to law school, he worked as a patent examiner for the USPTO.