Monday, December 12, 2016

Wright and Ginsburg on the FTC's PAE Study

As previously noted on this blog, in early October the U.S. Federal Trade Commission released a document titled Patent Assertion Entity Activity:  An FTC Study.  The study is based on information the agency obtained under its authority to collect confidential business information from private entities, and its methodology is spelled out in detail at study pages 2, 38-40, and appendix F (pages F-2 to -4 in particular).  Altogether the study analyzed information from 22 Responding PAEs with over 2,500 “Affiliates and other related entities,” of which 327 had engaged in patent assertion by sending demand letters, litigating, or licensing.  (“Affiliates” are defined for purposes of the study as including parent companies, subsidiaries, divisions, and any other person or persons over which a responding firm exercised supervision or control.  The agency also conducted a separate Wireless Case Study involving analysis of information from six wireless manufacturers and five NPEs.)  The agency found that the responding PAEs could be grouped into two categories:  Litigation PAEs, which filed 96% of the cases in the study and accounted for 91% of the reported licenses, but only 20% of the reported revenue, or approximately $800 million,” id. at 2-4; and Portfolio PAEs, which generated approximately 80% of the revenue reported by responding PAEs, even though they accounted for only 9% of the reported licenses in the study. 3-4, 45.  The study reports a number of interesting findings--among them that 77% of the assertions by responding Litigation PAEs settled for less than $300,000, that is, less than the estimated median cost of litigating even a small-stakes patent infringement suit through the end of discovery, according to the AIPLA's Annual Report of the Economic Survey—and that over 30% settled for less than $50,000.  See id. at 7, 49, 90.

Anyway, Professor Joshua Wright and Judge Douglas Ginsburg have posted on ssrn a short paper titled The FTC PAE Study: A Cautionary Tale About Making Unsupported Policy RecommendationsHere's a link to the paper, and here is the abstract: 
In October 2016, the Federal Trade Commission released its long awaited case study examining the business practices of 22 Patent Assertion Entities (PAEs). One clear policy implication is that PAEs do not present an antitrust problem. While the study makes a number of interesting and potentially important factual findings, it makes four policy recommendations that simply are not substantiated by anything in the study. The FTC acknowledged that the limitations of its sample rendered it inappropriate to extrapolate its findings to PAEs as a whole. Curiously, however, in spite of that acknowledgement, FTC went on to make recommendations applicable to the entire population of PAEs. Because it is unclear whether the policy recommendations in the FTC’s PAE study would survive a cost-benefit test, and because they certainly cannot be substantiated based upon the PAE Study alone, we conclude the policy recommendations should be afforded little weight.
In my view, the authors are a bit too critical of the recommendations, which include discovery reforms similar to those that Congress has considered enacting in recent years; amending Federal Rule of Civil Procedure 7.1 so as to require more transparency with regard to the ultimate parent company and affiliates of LLCs (the corporate form used by many Litigation PAEs); enacting legislation to codify the "customer stay" exception (see previous discussion on this blog here); and requiring greater specificity in patent pleadings.  Their principal critique is as follows (pp. 2-3):
As the FTC correctly acknowledges, because the full population of PAEs is not identified in any publicly available data set, its results are based upon a potentially unrepresentative sample, making it inappropriate to extrapolate its findings to PAEs as a whole.4/ Curiously, however, the FTC goes on to do just that. Specifically, “based on the overall findings of this study,” . . . the FTC offers four policy recommendations to remedy “nuisance infringement litigation” that would necessarily apply to the entire population of PAEs—and beyond. . . .
4/  PAE Study, supra note 1, at F-2. See also Id. at F-12 (“this is a case study, and as such, it is not statistically valid to extrapolate the findings from the case study to the population of PAEs, manufacturers, or NPEs. Instead, the findings of the case study should be viewed as descriptive and probative for future studies seeking to explore the relationships between organizational form and assertion behavior.”).
The authors' quotations are accurate, in that the study doesn't purport to be "generalizable to the population" of all PAEs (F-2).  However, the study also states:
Making the best use of available data, the FTC has designed a subject-selection procedure that will simultaneously be more likely to include more economically important firms (that account for a larger proportion of PAE behavior) while including firms of different sizes (to ensure that firms operating a variety of business models are included). . . . 
. . . Hence, the results of this study will not be extrapolated to the population of all PAEs. Instead, the study’s results should be interpreted as a detailed case study of the PAE industry where the study subjects have been selected to disproportionately include firms with more patents and litigation activity, while still including small and medium-sized firms (F-2, -3).
So, the findings may not be representative of all PAEs--but they do appear to give us an unprecedented window into the characteristics of the ones that, at least individually, have the most economic impact.  That seems pretty important to me.  

Wright and Ginsburg also note that responding "Litigation PAE patents had 80% more citations than the average patent, implying that PAEs are not litigating lower than average quality patents" (p.2), a finding that is consistent with some other research that predated the FTC study.  Given that PAEs also have a lower "win rate" than non-PAEs in cases that go to trial, however--both according to the FTC Study, and to John R. Allison, Mark A. Lemley & David L. Schwartz, How Often Do Patent Assertion Entities Win Patent Suits?, 32 Berkeley Tech. L.J. __, __ (forthcoming 2017)--perhaps the better inference to draw is that patent citations aren't really as indicative of value as people sometimes think they are (a point that Allison et al. have raised as well in other work). The authors also quarrel with the agency's description of the cases that settle for less than the median cost of discovery as "nuisance litigation," stating that "PAEs are not the first plaintiffs to take advantage of our legal system by bringing or threatening to bring cases purely for their settlement value" (p.7).  Aside from the unflattering term "nuisance litigation," though, I'm not sure what the criticism is; if a large percentage of Litigation PAEs are settling for lawsuits for, basically, the avoided costs of discovery--and especially if they aren't transferring any technology in return, as would often appear to be the case if as is commonly believed inadvertent infringement is the norm, not the exception--it's a bit hard to see how this behavior serves the public interest.  

The authors also believe that the FTC should have conducted an actual cost-benefit analysis before making any specific recommendations (pp. 4, 9), and note that much of the data the agency used pre-date the Supreme Court's recent decisions on attorney's fees and on software patents, which might mitigate some of the concerns raised by PAE critics without the need for further legislation. 

By the way, readers of this blog also may be interested in a report issued by the Joint Research Centre, which describes itself as "the European Commission’s science and knowledge service," titled Patent Assertion Entities in Europe. Their impact on innovation and knowledge transfer in ICT markets, available here.  (I only learned about this report recently, and haven't read it yet.  I'll probably have something to say about it on this blog before too long.)  Here's the abstract:
Patent assertion has become a common practice in shaping the balance between technology creation and technology dissemination in the Information and Communication Industry (ICT). The importance of this practice for the functioning of ICT markets has given rise to new entities that enforce patents but do not utilise the patented technology, commonly referred to as patent assertion entities (PAEs). This study provides an overview of patent assertion practices and of PAEs in Europe, taking into consideration their impact on innovation and technology transfer in European ICT markets.

No comments:

Post a Comment