Guest post by Heather F. Auyang, Senior Counsel at LTL Trial Attorneys in San Francisco, California. The views and opinions expressed herein are those of the author and do not reflect the views or opinions of LTL Trial Attorneys.
This is the Teva-update to last summer’s Patently-O Guest Post titled “Why Lighting Ballast Won’t Solve Claim Construction” (http://patently.wpengine.com/patent/2013/07/guest-post-why-lighting-ballast-wont-solve-claim-construction.html), which discussed two then-recently published empirical studies – a study of “close cases” (analyzing all post-Markman claim construction cases where the Federal Circuit panel was split) and a study of “reversals” (analyzing all post-Phillips claim construction cases in which the Federal Circuit reversed the district court). Updates to both studies have now been published in a sequel article, which provides further explanation and answers to some of the questions that were raised in this very forum last year. The past year’s data is unsurprising – it is exactly what the models provided by the original studies predicted, and it confirms, yet again, that changing the standard of review will not have a particularly beneficial effect on claim construction predictability or consistency. This conclusion is based on two key facts shown by the data: (1) that Federal Circuit judges remain divided on how to approach the task of claim construction, and (2) that when district courts are reversed, it’s almost always because they have misapplied settled claim construction principles. In fact, giving more deference to district courts will likely exacerbate, rather than fix, problems with claim construction.
The close cases study (of 105 cases) continues to show that each Federal Circuit judge who has participated in more than five close cases falls into a distinct camp. Judges Wallach, Linn, Clevenger, and Moore “go broad” in over 66% of close cases (Judge Rader was also in this group); Judges Bryson, Prost, Mayer, Schall, and Dyk “go broad” in between 47% and 55% of close cases; and Judges Newman and Lourie “go broad” in under 31 percent of close cases.
As pointed out last year, if the goal is to make claim construction more predictable, a good place to start would be to ensure that all the Federal Circuit judges are following the same rules of claim construction. This data clearly shows that they are not. While any given judge in any given opinion can cite the same agreed-upon rules, the rules are not preventing significant differences in approach among the judges. Last year’s post and article posited that some judges are expressly or implicitly following an “actually invented” standard that other judges reject. Whatever the explanation, until these differences are ironed out, tinkering with other issues – like deference to district courts – is unlikely to be productive.
The reversals data (based on 153 cases) is even more relevant to the question of deference than is the close cases data, since it deals directly with the relationship between the Federal Circuit and the district courts. Focusing on reversals makes sense because this dataset, presumably, includes the cases whose results would change if the Supreme Court determines that district courts should get more deference in claim construction. Accordingly, these are the decisions that should be studied before one concludes that the current standard of de novo review should be changed. As it turns out, the reversals data suggests that the high reversal rate is caused by district court error, not Federal Circuit arbitrariness. If that’s the case, why in the world are people advocating more deference to district courts?
The reversals study coded for whether the Federal Circuit decision was in a broadening or a narrowing direction. The result for the 11-month period since the last study was a broadening rate of 87.5%, even higher than the previously-reported overall broadening rate of 72.5%. In other words, in cases where district courts are “getting it wrong,” according to the Federal Circuit, they are systematically interpreting the claims too narrowly. As pointed out last year, these narrowing interpretations typically enable district courts to grant summary judgment of non-infringement (or encourage the parties to enter such a stipulation), and thereby permit the Federal Circuit to review the claim construction issue without going through the trouble and expense of a trial. In the year since then, we have not heard any other plausible explanation. Last year’s post and article also challenged any advocate of deference to provide one or more examples from the “reversals” cases that (1) would have come out differently under a deference regime, and (2) should have come out differently. We haven’t heard anything on that either.
This year’s article – including fancy color charts analyzing Federal Circuit judge proclivities across broad/narrow, less spec/more spec, pro-patent/anti-patent, and pro-affirm/pro- reverse; and pie-charts and tables for the reversals data – is What Reversals and Close Cases Reveal About Claim Construction at the Federal Circuit – The Sequel, 13 J. Marshall Rev. Intell. Prop. L. 525 (2014) (available at http://repository.jmls.edu/ripl/vol13/iss3/3/). The comments here on last year’s article were of a very high caliber; hoping for the same again this time!