This week marks the start of three separate hearings on a draft § 101 bill introduced by Senators Tillis (R-NC) and Coons (D-DE). The draft bill—analyzed last week by Patent Progress—is fundamentally flawed. Fortunately, it’s only a draft bill, and there’s still time for it to be modified to make it less harmful to innovation.
Unfortunately, based on the announced witness list for the first hearing, that might not happen.
Patent Eligibility Hearing 1: Tuesday, June 4
On Tuesday afternoon, the Senate Judiciary Committee’s IP Subcommittee will hold the first hearing. The hearing will feature three panels. While the overall character of the day appears to be former government officials, patent professionals, and academics, a little investigation suggests that several of the government officials and patent professionals might be less neutral than they appear, while some of the academics have presented research that simply doesn’t match up with the facts.
Panel 1: Michel, Dickinson, Kappos, Duan, Lefstin
The first panel features a professor and a patent expert from a think tank, along with three former federal officials. Sounds like an unbiased group who are giving their own opinions, not representing anyone else.
But David Kappos, while listed as a former PTO director, is currently a partner at Cravath, Swaine, and Moore—one of Qualcomm’s go-to firms. . He’s also the former IBM Assistant General Counsel for Intellectual Property. These two firms, Qualcomm and IBM—a current client and a former employer—are essentially the only major tech companies that are in favor of overturning recent changes to § 101. This perspective is not reflected in Kappos’s description as a former PTO official. (Kappos also published some incorrect data on the topic of § 101, as I’ll discuss in relation to Panel 3.)
Panel 2: Armitage, Taylor, Knowles, Moss, Lemley
The second panel? Two professors, a representative from a public interest group, and a pair of IP strategists. And again, there’s some history worth looking at.
Bob Armitage—listed as an IP policy consultant—is better known for his long history at drugmaker Eli Lilly, including a decade as their general counsel. Branded drug companies have been heavily in favor of anything that strengthens their patent monopolies for quite a while, and this § 101 bill is no different. Again, this perspective is not reflected in Armitage’s description as an IP policy consultant.
Professor David Taylor is a patent law professor at Southern Methodist University. Prof. Taylor has done research on patent eligibility, and recently released a paper examining venture capital views on § 101. Unfortunately, there are serious questions about both the methodology and conclusions of this paper. The questions were worded in ways which make it hard to draw conclusions from the answers, assume non-legal audiences are familiar with legal terminology, and frame answers around incorrect assumptions. Beyond that, Prof. Taylor’s paper draws conclusions that his own survey data simply doesn’t support. For example, Prof. Taylor concludes that the eligibility decisions over the past decade have had a significant negative impact on VC investment and attitudes. But his own survey shows that the majority of respondents did not change their investment behavior and that the majority of high tech VC respondents would not change their investment behavior if software patents were completely eliminated, statements at odds with the notion that investors consider the changes to patent eligibility as a major negative factor in their investments.
That brings us to the final panel.
Panel 3: Rosen, Gugliuzza, Sarnoff, Kilbride, Mossoff
Three professors and two representatives from think tanks. Unfortunately, we again have an academic representative who has most salient research on the topic is based on fatally flawed data and poor interpretations.
Prof. Adam Mossoff, of the GMU Center for the Protection of Intellectual Property, wrote a paper a few years ago. That paper claimed to have identified 1,700 applications—from a set of 17,000 applications—that were denied in the U.S. because of § 101 but allowed in other jurisdictions. Setting aside whether the foreign jurisdiction patents were truly equivalent, which is not guaranteed, there’s a few problems with his conclusions.
In particular, many of the applications Mossoff identified didn’t actually illustrate what he claimed they did—that the U.S. patent system, based on § 101, was preventing these applications from being patented. Examining a set of 14 applications called out as exemplary in his paper, only one was abandoned solely because of § 101. But three of those applications overcame a § 101 rejection and were abandoned, either because the prior art prevented patentability or in favor of a continuation. And the remainder had significant rejections, either based on the prior art or based on lack of written description or enablement. Reviewing a larger set of 100 applications in the dataset of 17,000 applications supposedly abandoned because of § 101, the trend holds. 19% had overcome their § 101 rejection when abandoned and 11% never received a patentable subject matter rejection under § 101. Only 14% were clearly abandoned due to § 101.
Some of Mossoff’s errors may derive from the dataset he was provided. (The data was provided by Bob Sachs and David Kappos, who you might remember from the discussion above.) That dataset claims to be of 17,000 applications which were rejected under § 101 and then abandoned. That description implies that the abandonment was based on § 101—but that’s far from the case. Examining the dataset, 529 (3%) of the 17,000 had been allowed at the time they were abandoned—hardly § 101’s fault. And 2,106 (13%) had no § 101 rejection remaining when they were abandoned. And of the remainder, a number had non-subject matter § 101 rejections, such as being rejected for claiming a perpetual motion machine or a cold fusion device. Only 1,685 (11%) were clearly abandoned due to § 101 rejections.
With source data as flawed as this, it shouldn’t be any surprise that Mossoff reached incorrect conclusions. But incorrect or misleading data isn’t the kind of thing Congress should use to decide what is and isn’t patent eligible.