The Board of Patent Appeals and Interferences takes into account a number of authorities when deciding appeals. First at the BPAI’s disposal are the higher courts of the Federal Circuit and Supreme Court. Then there are the rulemaking guidelines issued by the PTO. Then of course there’s the MPEP. I’ve been interested in the whether the Federal Circuit affects the BPAI in terms of the number of rejections the Board issues.
Appeals before the Board are initiated after an examiner rejects a claim at least twice. The applicant, rather than continuing to spend time arguing with the examiner, can elect to appeal to the Board where a panel of three judges will hear both arguments. The Board then has three options. It can reverse the rejection, it can affirm the rejection, or it can introduce a new rejection. Since Bilski v. Kappos has been decided, the Board has used § 101 to reverse 95 times, affirm 166 times, and introduce 65 new rejections. Up until today, the percent of new rejections to how many times the Board has acted is about 20%.
To see whether the Board is affected by Federal Circuit cases, I divided the time frames up into intervals roughly 6 months apart when big Federal Circuit cases were decided. Here are the Federal Circuit cases: Research Corp v. Microsoft, Ass’n for Molecular Pathology v. Myriad, and Ultramercial v. Hulu. At times, there were two Federal Circuit cases decided within a short period of time, which I just combined.
A little about the time frames. The first time period is from June 28, 2010 to December 9, 2010, starting with the Supreme Court case Bilski v. Kappos and ending with the first big Federal Circuit decision decided on the merits of § 101. In this time period, the Board introduced new rejections 22.4% of the time, higher than the overall average. This was likely a result of the recent Supreme Court case and the PTO publishing post-Bilski guidelines a month after Bilski. In addition to these factors, two Federal Circuit cases that did not receive that much attention mentioned § 101 subject matter patent-eligibility: Intervet v. Merial Ltd included a dissent with a 101 focus and Astrazeneca v. Apotex was mostly decided on other grounds, but did mention subject matter. Both cases used 101 to poke holes in the patent at issue and the effect may have been felt at the BPAI.
Research Corp v. Microsoft was a subject matter game changer in the sense that it broadened the Bilski analysis by saying that unless the invention was very obviously an abstract idea, the invention should be statutory subject matter. As noted on Ryan Alley’s blog, the case looked as if it would affect the BPAI’s decisions when followed by the BPAI. Prometheus v. Mayo was decided over a week later which was also favorable to patent-eligible subject matter, especially for diagnostic methods. This time period saw new rejections dip to 18.4%.
Next, AMP v. USPTO. The now-famous gene patenting case is familiar to most. While the take-home message from this Federal Circuit case is that the District Court was wrong in concludingi gene patents are not statutory subject matter, the method claims which were affirmed received far less attention. As noted on the Patents4Life blog post, the method claim analysis was very brief and the arguments not very well developed. Very shortly after AMP, the Federal Circuit decided Cybersource v. Retail Decisions. These two cases seemed to propel the BPAI forward with new rejections, cresting to 24.3%.
Ultramercial v. Hulu was another game changer. Like Research Corp, Ultramercial was another Judge Radar decision which seemed to broaden the 101 analysis from the earlier cases of AMP and Cybersource. This decision until the present has seen a dip in the number of new rejections at the BPAI to 16.7%.
In conclusion, it is interesting to see how the patent tides of change can be felt not only in the higher courts, but at the Board and the effect most likely trickles down to each examiner. I found that differences in rejections in the different time frames were interesting and expected, but not all the time frames were statistically significant at an acceptable confidence level.
All data was mined from the website http://www.bpai101.com/decisions. The word “affirmed” was searched in the outcome field for those decisions that were affirmed. The word “reversed” was searched for reversed decisions and the word “new” was searched for new rejections. Some cases performed multiple operations in one decisions, such as affirming one claim, reversing another or introducing a new rejection. In such cases of multiple operations, I included each operation separately. For instance, one decision could potentially yield a reversal, affirmation, and new rejection. Most decisions, however, had only one operation.