The FTC’s PAE Study: Doing More Harm Than Good

US Federal Trade Commission, Washington DC.

Recently, the Federal Trade Commission (FTC) released a report of its study of patent assertion entities (PAEs). The report was long anticipated and could have gone a long way to shining some light on patent licensing firms – who they are, how they operate, and so on. After all, patent licensing firms are misunderstood, partially because so much of their activity is not visible to the public. In theory, because the FTC has the power to obtain this invisible information, the study could have provided the data and insight needed to better understand these firms and improve the policy dialogue surround their behaviors. This PAE study had the potential to do a world of good…but unfortunately, the resulting report is likely to do more harm than good.

To be fully transparent, I was unconvinced from the outset that the FTC’s PAE study would promote any real insight into patent licensing firms because the study was poorly designed. Specifically, I argued that 1) the subjects of the FTC PAE study did not represent a realistic picture of the patent licensing landscape and 2) the questions asked by the FTC were not designed to develop an understanding of these firms and how they operate. I concluded that the FTC’s study had the potential to cause more harm than good, because “inaccurate and incomplete data will go far to bolster the inapt theories that are currently holding the day for those who wish to hinder, if not destroy, patent licensing firms.” Now that the report has been released, my fears have been realized.

To put my remaining remarks in context, here is a quick summary of the relevant portions of the FTC report. The FTC identified two types of business models in its study of PAES. Portfolio PAEs are those firms that license large (and valuable) portfolios of patents, that generally negotiate licenses without filing lawsuits, and that generate significant revenue. Litigation PAEs are firms that instead have small portfolios, that typically sue first then settle, and that generally settle for small amounts of money (less than $300K). Litigation PAEs, according to the report, engage in what could be considered nuisance litigation. Because of the extent of nuisance litigation, the FTC provided four policy recommendations:

• Develop rules and case management practices to address discovery burden and cost asymmetries in PAE litigation
• Amend rules to reach a broader range of non-party interested entities or persons
• Establish procedures encouraging courts to stay a PAE’s infringement action against customers or end-users, where the PAE has also sued the manufacturer of the accused product under the same theory of infringement
• Ensure that patent infringement complaints provide sufficient notice to accused infringers.

Putting aside my concerns with the study’s design mentioned above, I think the FTC report itself is terribly flawed. Here are four primary reasons why the FTC’s report is likely to do more harm than good:

The FTC failed to capitalize on the one shining point of light from the PAE study

The FTC report should be applauded for its recognition that not all patent licensing firms are the same – this is an argument that academics (including myself) have been making for years. The FTC’s definition of PAEs had already identified NPEs as something separate – PAEs acquire patents to monetize them, while NPEs develop and patent technology and then monetize them. Now, in the report, the FTC admits that even PAEs are not a homogenous group. Specifically, the FTC categorized PAEs as either portfolio PAEs or litigation PAEs. Portfolio PAEs are those with large portfolios, they tend to license before they sue, they make lots of money…in short, the implication is that these are “good” PAEs. Litigation PAEs, on the other hand, have smaller portfolios, they like to sue and then settle, they make much less money, and generally they engage in “nuisance” lawsuits. Because the FTC makes policy recommendations to address this “nuisance” litigation, clearly the FTC has classified litigation PAEs as “bad.”

So while I congratulate the FTC for finally appreciating that PAEs are not one big homogenous group, there are two problems that flow from this appreciation.

First, the categories themselves are problematic. The division between the two categories is too coarse a filter to be at all meaningful – patent licensing firms are simply too diverse and too complex to divvy into two overly simplistic classifications. Further, by virtue of how the PAEs were selected, the study was designed to be focused on the litigation, or “bad,” PAEs. There is no real understanding of the patent licensing landscape to be gained from these artificially created categories based on poorly selected study PAEs. In fact, continuing to essentially classify PAEs as either “good” portfolio or “bad” litigation PAEs will simply continue (and likely promote) the unsupported narrative that existed prior to the study.

Second, and this should be made clear to those PAEs who breathed a sigh of relief when they discovered they may fit into the portolio PAE category…it doesn’t matter. Although the FTC took a brave step towards advancing the policy dialogue by recognizing the heterogeneity of patent licensing firms, it then takes a giant leap backwards by promoting policy recommendations that fail to retain any distinction between PAE types. Although the report acknowledged that different patent licensing firms may act differently, may exist for different purposes, etc., when the FTC set forth its recommendations, any differentiation between patent licensing firms failed to exist – the legislative and judicial reforms suggested by the FTC’s report apply to PAEs generally. In my opinion, to say that PAEs differ from each other, in at least some respects, but to then propose changes to the law that apply across the board is more harmful than not acknowledging the diversity of patent licensing firms in the first instance.

The FTC’s policy recommendations are based on incomplete facts

Again, putting aside my concerns with the study design, there are at least two big pieces of evidence that the FTC study is missing in order for its policy recommendations to be supported by facts.

First, the FTC admits it cannot speak to any efficiencies created by patent licensing activity, because the data were just too hard to collect. I’m not terribly convinced by this point – but assuming that the FTC truly was not capable of collecting and understanding these data, how in the world can it make policy recommendations with only part of the information necessary? It could turn out, and people have argued, that patent licensing creates all sorts of advantages for inventors, innovators, the economy, etc. Yet, without any evidence about the extent of these benefits of patent licensing activity, with no assessment about how patent licensing positively affects competition, and with no ability to balance these benefits against the negatives, the FTC made recommendations that could have expansive effects on patent licensing activity.

Second, the FTC claims that at least some of its policy recommendations are designed to address information asymmetries between PAEs and defendants in patent litigation. In order to validate the existence of information asymmetries, however, don’t we need to have some data from the other side…specifically, the defendants? As it stands, the existence of information asymmetries is largely based on anecdote, not evidence, and the extent to which these information asymmetries cause quantifiable harms is even less supported. If, as one would understand from reading the FTC’s report, information asymmetries are one of the biggest problems with PAE activity, this information could have – should have – been gathered through this extraordinary opportunity provided by the FTC’s study.

Basing policy recommendations on no evidence, or at best anecdotal evidence, has great potential to do more harm than good…especially when some of the missing evidence is the other side of the equation – the benefits afforded by patent licensing activity.

The FTC’s policy recommendations are not supported by the facts gathered in the study

In addition to my concern that the facts alleged to support the policy recommendations are incomplete, there is a larger concern that the facts that actually were gathered by the study simply do not support the policy recommendations. There is very little connection between the data gathered by the PAE study and the recommendations made. Perhaps Judge Douglas Ginsburg, speaking at event put on the by Searle Center on Law, Regulation, and Economic Growth at Northwestern Pritzker School of Law, said it best: he remarked the report concludes with “a set of recommendations that is free-floating from the study but worthy of consideration.”

One example of the disconnect between the study results and the recommendations goes back to my first point, above. Although the study demonstrated there are different types of PAEs and they behave in different ways, NONE of the policy recommendations is aimed at one type or another type of PAE…the recommendations are simply aimed at the broad, homogenous category of patent licensing behavior. Another example is the recommendation to increase disclosure of related entities in patent infringement litigation involving PAEs. Although the study may have uncovered that some patent licensing firms have a large number of related entities, and some of these related entities may assert patents and a lot of them do not, this fact alone – the existence of related entities – does not justify a call for increased disclosure. The FTC discovered no evidence that a lack of disclosure of related entities creates any problems for innovation nor does the lack of disclosure, based on the study, create any real hardships for the defendants in these cases. There are other examples of this lack of connection between the FTC’s policy recommendations and the results of the study, but the bottom line is this: regardless of how valid some of these policy recommendations may be (and I’m not taking any position on the validity of these recommendations in this piece), to set forth a set of policy recommendations as part of an extensive study, where the policy recommendations are virtually divorced from anything learned in the study is not just irresponsible, but likely to do more harm than good.

The FTC’s policy recommendations are based on a poor proxy for “nuisance” litigation

A lot of the FTC’s report, starting with its categorization of “bad” litigation PAEs, is based on the assertion that PAEs engage in “nuisance” litigation…and the recommendations are suggested specifically to achieve a balance between protecting patent rights and “nuisance” litigation. Accepting the FTC’s proxy for nuisance litigation will result not only in more harm than good, but will also eviscerate a fundamental right of patent law. The FTC alleges that, especially for the litigation PAE type, patent licensing firms often engage in litigation that results in settlements of less than $300K. This number was not plucked from the air by the FTC; it represents the lower band of costs required to defend a patent suit according to AIPLA (for the relevant time period of the FTC study). The reasoning is that because the lawsuits settle for less than it would cost for the defendant to defend itself in court, this must be a nuisance…implying (or perhaps explicitly stating) that these lawsuits need to be stopped. But <$300K is a terrible proxy for what suits shouldn’t be brought.

First, there is a substantial difference between nuisance and meritless. Nuisance means that the alleged cost-benefit analysis would lead us to believe that bringing the lawsuit is not the best use of resources…but that doesn’t mean there is no infringement and it doesn’t mean the patent is invalid. If we start taking an axe to any case that results in <$300K in damages or settles for less than that, we seem to be permitting, or even encouraging, de minimis or low level infringement. But that is not part of the patent system; in fact, the US Court of Appeals for the Federal Circuit has specifically said that de minimis infringement is still infringement. Setting an artificial cutoff for what cases are “worth” bringing (and determining whether patent licensing activity is “good” or “bad” based on this artificial cutoff) seems to be antithetical to patent law, which provides a right to exclude even a tiny bit of infringing behavior.

Second, the factors surrounding patent litigation and settlement are extensive and case-by-case unique…and cannot be understood or explained simply by looking to a bare number, like $300K. In fact, in an old patent textbook I used to teach from, there was a case that went to verdict and the patentee was awarded $69 and change. You read that right, just $69. The case was NOT included in the textbook as an example of cases that should not be brought; in fact, I loved teaching that case to provide the insight that there are a lot more things going on in patent cases than you can gather from the damages award…and understanding what is going on is critical. By looking at settlement numbers and concluding, ah, that must be a nuisance case, is a vast oversimplification of patent law. Patent law, and specifically patent licensing behavior, is far too important and complex and fact-specific to base recommendations on this type of oversimplification. And while it’s hard to understand what is going on behind the scenes that is precisely what the FTC study was supposed to do for us.

Instead of seizing the opportunity to survey the patent licensing landscape and shed light on behavior that otherwise is invisible to the public, the FTC squandered the chance and instead developed two arbitrary categories of PAEs, determined that one of these categories was not good, and developed a set of policy recommendations because of “nuisance” litigation. By making recommendations without gathering or using the very facts that were supposed to be the public benefit of this PAE study, the FTC’s report is undoubtedly going to do more harm than good.