“While the professors’ conclusions regarding increasing time spent by examiners on patent applications are seemingly innocuous, the entire paper is infected with the ‘bad patents’ premise that has proven to be incredibly detrimental to the U.S. patent system in recent years.”
On Thursday, October 17, a Capitol Hill staff briefing will take place at 3:30 PM in 226 Dirksen Senate Office Building on a proposal to increase the amount of time that patent examiners at the U.S. Patent and Trademark Office have to examine patent applications. At the briefing, Professors Michael Frakes of the Duke University School of Law and Melissa Wasserman of the University of Texas at Austin School of Law will present findings from their paper, Irrational Ignorance at the Patent Office, recently published by Vanderbilt Law Review.
While the professors’ conclusions regarding increasing time spent by examiners on patent applications are seemingly innocuous, the entire paper is infected with the “bad patent” premise that has proven to be incredibly detrimental to the U.S. patent system in recent years. Branded as a work that updates prior research on patent quality produced by law professor Mark Lemley, the Frakes and Wassserman paper concludes that the costs of increasing USPTO resources for weeding out “bad patents” during the patent prosecution process are far outweighed by the costs borne by society in waiting for the courts to invalidate those patents during litigation.
“It gets tiring hearing about bad patents all the time,” said Gene Quinn, patent attorney and founder of IPWatchdog, Inc. “The truth of the matter is it is irrational to expect only perfection from the patent office when they grant patents. Upwards of 90% of patents are never litigated, commercialized or licensed, and tens of thousands of dollars are spent to prepare and prosecute those applications. When they are challenged, upwards of $1 million—sometimes far in excess of $1 million—is spent to invalidate claims. Only an irrational person would believe it possible to achieve perfection for $10,000 to $20,000, given the nature of how patent challenges are rigged to ignore the presumption of validity.”
Authors’ “Hope”: More Examination Time Means Fewer Bad Patents
The main thrust of the Frakes and Wasserman paper is that patent allowances increase as the time to examine a patent application decreases. Pulling from data on 1.4 million patent applications and examiner roster information issued by the USPTO under Freedom of Information Act (FOIA) requests, the authors found that patent examiners that are higher in the general schedule (GS) pay scale, who are given fewer hours to examine patent applications, have a higher allowance rate than examiners at lower GS pay scales. The authors present data showing that patent examiners at GS-level 14 had a 29% higher grant rate than examiners at GS-level 7.
If patent examination times were doubled, however, the authors argue that the USPTO’s overall grant rate would drop by 19 percentage points, resulting in 80,000 fewer patent grants each year. “[W]e would hope that the affected patents would indeed be invalid patents,” the authors wrote. Preventing the issue of invalid patents would prevent societal harms that the authors admittedly can’t quantify, including supracompetitive pricing, stunting follow-on innovation, reducing competition, inhibiting startup access to venture capital and “opportunistically extract[ing] licensing fees from innovators.”
“What is missing from the analysis is any rational explanation as to how or why a patent drafted and prosecuted for two times the present cost—say $20,000 to $40,000—would fare any better in the face of a $1 million challenge,” Quinn explained. “These professors assume at the time of examination the best prior art can and will be found, which is simply erroneous.”
What the authors do end up quantifying is the amount of money saved by reducing litigation through reductions to patent grants. The authors theorize that reducing the number of patents issued by the USPTO would reduce litigation driven by challenges to patent validity, including declaratory judgment actions seeking a district court’s declaration that a patent is invalid. Reduced litigation also includes lawsuits from non-practicing entities (NPEs) “that have acquired a pool of low-quality patents and that seek nuisance settlements.” Analyzing patent lawsuit data from Lex Machina, examiner data from the USPTO’s Patent Application Information Retrieval (PAIR) database and data on filings at the Patent Trial and Appeal Board (PTAB), the authors determine that doubling patent examination times at the USPTO would lead to a 44 percent reduction in patent grants that are eventually asserted in litigation.
Reducing Patents Saves Hundreds of Millions in Litigation Costs
There seems to be little doubt that such a reduction in patents asserted in litigation would necessarily lead to major savings in U.S. litigation costs. Assuming that a 44% reduction would lead to 2,436 fewer patent-lawsuit pairs each year, a patent-lawsuit pair costing an average of $234,761 when dividing the total litigation cost of an average patent lawsuit ($539,949) by the average number of patents included in a patent lawsuit. Frakes and Wasserman posit that doubling examination time would lead to litigation savings of nearly $572 million each year.
“What about foreign patent litigation?” Quinn asks. “Frakes and Waserman seem to believe patent litigation is a zero-sum game, and by further destruction of the U.S. patent system, companies will save money. That ignores what is actually happening in the marketplace. As the United States becomes an increasingly less favorable jurisdiction, patenting activities and litigation are moving to Europe—Germany in particular—and to China. So, too is investment in industries critical for the future, such as artificial intelligence.”
Reducing the number of asserted patents would also lead to obvious savings in PTAB litigation, Frakes and Wasserman claim. The professors state that, of the 430,056 utility patent applications completed each year by USPTO examiners, 378 of those patent applications will be the subject of an instituted PTAB petition. The authors conclude that doubling patent examination times would reduce the number of patent applications that eventually result in an instituted PTAB trial by 272 which, when multiplied by the average litigation cost of PTAB proceedings, leads to another $123 million in reduced litigation cost. Frakes and Wasserman acknowledge that there could be a concern that the effects of the PTAB, which didn’t exist during the first decade of the 2001-17 study period, could cause the data to overstate the degree of litigation savings. However, the authors found that removing data after 2012, the year in which the PTAB was established, didn’t change their analysis and that, “[i]n fact, the introduction of PTAB only reinforces this Article’s conclusion that the savings from increased examination time justify the costs” as reductions in issuances of invalid patents would lead to fewer PTAB proceedings.
“The problem with the professors’ conclusion is it ignores the human element at the PTAB,” Quinn said. “We know that the PTAB does not have enough work for all the adminstrative patent judges that have been hired, which leads to a higher percentage of institutions than should occur. This is only compounded by the fact that the people deciding to institute will also be those deciding the merits of the case. They create work for themselves. Why will that change with fewer patents?”
Overall, Frakes and Wasserman conclude that doubling patent examination times would result in annual savings of $602 million to the patent system. This figure also takes into account a reduction in 0.56 rounds of review per patent application, which the authors found would result from increased patent examination time. The data seems to suggest that examiners at higher GS pay scales, who have less time to examine patent applications, issue more office actions, thus leading to the conclusion that more examination time would reduce the number of office actions. A few pages after stating that increased patent examination times would save $602 million per year, the authors then “heavily discount” that figure based on methodological concerns in arriving at the determination that doubling examination time results in 0.56 fewer rounds of office review.
Benefits of Increased Patent Examination Time Outweigh the Cost
To determine the USPTO’s cost for doubling patent examination times, Frakes and Wasserman analyzed the average number of hours allotted to examiners at each level of the GS pay scale and concluded that it would cost $660 million to achieve that increase in examination time. Because the USPTO doesn’t have the resources necessary to increase patent examination times and achieve the associated societal benefit, Frakes and Wasserman end up finding that, unlike Lemley’s charge that the USPTO was “rationally ignorant” in issuing patent grants that are later invalidated, the agency is instead “irrationally ignorant” because the $660 million cost of doubling patent examination times is far outweighed by the estimated $904 million that would be saved in reduced federal litigation ($491 million in savings), reduced PTAB litigation ($112 million) and reduced patent prosecution costs ($301 million).
The authors propose that patent examiner times should be expanded in a way that reconsiders the manner in which time allotments vary over a given examiner’s career rather than doubling times across the board. How is the agency supposed to cover the costs of increased patent examination times? Frakes and Wasserman suggest that additional personnel expenses could be covered through general tax revenue, but they don’t rule out increasing user fees for patent applicants as an option.
Getting Rid of Patents: A Swiftian Modest Proposal
Any notion that the patent system serves to benefit society seems to be limited to the paper’s footnotes. One such note on page 53 (page 1027 of the Vanderbilt Law Review issue) acknowledges that “the patent system’s primary goal is to promote innovation, and society, not just the patent applicant, benefits from the fruits of such innovation. Without a mechanism such as the patent system to recoup their research and development expenses, innovators may choose not to innovate at all, which would be unquestionably bad for society.” The fact that this basic, fundamental aspect of the impact of patent rights has been relegated to a footnote while fears about supracompetitive pricing and preventing follow-on innovation are used to frame the thesis speaks volumes about this particular study.
There are plenty of concerns that could be unpacked regarding the Frakes and Wasserman paper but perhaps the most basic is, what is a bad patent? An invalid one would be the obvious answer. But invalid by whose standard at which time? Are bad patents the ones that are invalidated under Section 101 after the Supreme Court decisions in Alice and Mayo, which has provided an entirely unworkable legal test as is evidenced by the Federal Circuit’s schism this summer in Athena Diagnostics? Are patents that were invalidated at the PTAB under a different claim construction standard than the district courts’ Phillips standard the bad patents? Are patents simply bad because they result in litigation expenses against supposed infringers, as Frakes and Wasserman’s cost-benefit analysis seems to suggest? Why not just get rid of patents entirely? Isn’t that the Swiftian modest proposal that exists at the end of the “bad patent” narrative once patents have been stripped of any meaningful rights in the never-ending quest for higher patent quality?
In short, patent owners should be very concerned that Hill staffers will be briefed on a paper that furthers a narrative which misrepresents the effects of the patent system. While increasing patent examination times may not seem like a controversial suggestion on its face, it represents another cut in the death by a thousand cuts currently being suffered by the U.S. patent system.
“I have no problem giving patent examiners more time, but I do have a problem allowing policy to be made based on a paper that quite frankly ignores basic, fundamental realities of patent practice and the workings of the USPTO,” Quinn said. “The patent office already has difficulty hiring and retaining enough quality people as it is, so while these professors can believe there is a magic wand that can be waved, no such Harry Potter type powers exist. These recommendations, although perhaps well intended, will only play into the bad patent mythology and unfortunately ignore the economic consequences of further driving innovators out of the U.S. patent system to Europe and China.”
Image Source: Deposit Photos
Image ID: 6198782