Friday, November 16, 2012

Going on a Troll Hunt - The Study

First, unlike many of the "patent system is broken" prognosticators who foresee the death of innovation due to perceived innovation stifling and other nefarious activities of patent trolls, the pejorative term for non-practicing entities, Way Better Patents doesn't see impending invention doom.  What we see is a significant case of the economic impacts of asymmetric information.  Some organizations, practicing and non-practicing, benefit from having an asymmetric information advantage.  They know their way around the patent system.  They tend to sit on the "we're holding all the cards side" of the patent licensing negotiation and IP strategy negotiation while the rest of us don't have their patent picking or enforcement prowess.  This is not a crime but an operational reality in the patentsphere.  Or put another way, trolls are patent smart.

The Leahy-Smith America Invents Act (AIA) (P.L. 112-29) mandated the Government Accountability Office (GAO) to conduct a study on the consequences of patent litigation by non-practicing entities (NPEs) or patent assertion entities.

GAO issued a a Request for Quote for that study back in July 2012.  This looks like part of a larger 20 year study defined in the law.

The definition in play for the study says, "while the term NPE is not defined in the law, it generally refers to a wide spectrum of entities that own and assert patents but do not practice or produce (i.e. design, manufacture or distribute) products using the patented technologies."  In lots of industries these definitions are blurred by many complex business models and structures where practice and produce have many definitions.  Some of the largest NPEs are US and post-doctorate research organizations - they invent and patent but they rarely product the inventions they create.   They don't seem very trollish but we digress.

According to the RFQ GAO is going to buy three specific things:

One — A Patent Litigation Database
Access to a complete database of all patent infringement law suits. GAO wants to collect data on the volume of recent patent infringement litigation in the all 94 federal district courts from 2001 through 2011 (even though there are places in their RFQ that say "and the present" which implies 2012 cases as well. GAO wants access to all of the patent infringement cases filed in the International Trade Commission going back to 2001 as well. The database must ensure that all court filings and documents associated with a particular case are linked to that case and are not counted as separate events (i.e. procedural events, transfers of venue, etc.).

GAO want all suits including patent marking cases, trademark and declaratory judgement cases (cases of non-infringement, invalidity, or unenforceability) as part of the patent litigation cases.  (The presence of trademark cases seems a little odd here but we wanted to make sure that it was clear we didn't make this up.)  GAO wants the capability to slice and dice the cases by federal district, judge, case outcome and patent class. A surprising addition - actually trying to figure out what the technology in the case involves. It will be interesting to see how they define the different areas of the technology. Since many non-practicing entities use the US classification to determine the tranche (domain) for licensing fee calculation.

Oh, and the contractor is supposed to demonstrate to GAO that it has methods for ensuring that data are complete and free from errors. (Good luck with that free of errors one.)

Two — A Random Sample
Then the contractor is supposed to generate random sample of 100 patent infringement cases per year from the database for cases filed each year from 2007-2011. It looks like the vendor needs to extract that set and deliver the sample to GAO who will retain ownership of the sample. Depending on membership in the random sample will directly impact the outcome of the study.

Three — An Analysis
Then comes the analysis. The contractor will then identify the key characteristics of the litigants.  GAO expects, "For the sample of 100 cases per year, contractor shall provide an objective and auditable framework for characterizing the extent to which infringement claimants (patent owners) practice, or make products related to, the patents they are asserting." (Practice what? Plenty of firms that own patents practice licensing too.) It looks like GAO is going to have to have a framework of some kind to deal with the continuum of "practice" because just like Intellectual Ventures and GE sometimes a firm is an inventor and sometimes it's  licensee and at other times it is a NPE.

GAO will approve the framework methodology before making the contract award. (Another sign that GAO knows who will be doing the work since most government contractors won't bid on something unless the scope of work is defined.)

"Contractor shall employ legal analysts to identify key infringement claimant (patent owner) characteristics. Contractor shall analyze court documents, Securities and Exchange Commission filings, corporate websites, and other sources of data as necessary. Contractor must employ expertise in patent infringement litigation to identify evidence that is relevant to objectively characterizing litigants involved in each case. Contractor shall record these characteristics for each case in a Microsoft Excel file. GAO needs to be able to trace all characteristics back to supporting documents.

"Contractor's legal analysts shall compile the characteristics into a Microsoft Excel file and then have at least one other analyst confirm each characterization according to the objective framework." Lots of billable hours on this task.
For each suit in the sample, the contractor shall also report:
  • Patent and Trademark Office patent classifications and patent numbers (Which ones - the USPC origin and cross references, the IPC? or the new Cooperative Patent Classification (CPC) symbols
  • Number of Defendants per case
  • Venue (which federal district court the suit was ultimately heard in)
  • Characteristics of the alleged infringer (defendant) as per the above auditable and objective framework
  • Outcome – Categories to be reported (with quotes from court records):
    • i. whether the case settled and how long it took to settle (e.g. X months after litigation filed)
    • ii. whether attorney fees or costs were awarded to one side;
    • iii. case outcomes
    • iv. whether there were Rule 11 (civil procedure) sanctions
    • v. whether an injunction was granted
    • vi. whether damages were awarded and how much

There are some other interesting gems in the RFQ:

The contractor has to have this work done by December 31, 2012. The expected award date is September 3 according to FedBiz Ops.

GAO wants the access to the data within two weeks of contract award. There were only two questions posed by potential contractors - not a sign that GAO is likely to get a broad range of responses from interested firms. (Either the fix is in or the data is coming from one of your favorite litigation reporting systems.  The RFQ and the very short response time implies that they know who they want to do the study.)

What is also interesting here is what is missing from the mix.

Patent litigation is generally a tool. File the law suit when the potential infringer doesn't want to talk then negotiate and settle when the specter of a multi-million dollar lawsuit looms. How will this be accounted for? How will the outcomes be described when most are covered by Non-Disclosure Agreements?

There doesn't seem to be any attempt to determine how the NPE acquired the patent. Was the patent acquired from a practicing entity who is working with the NPE on monetizing their intellectual property assets. (See the discussion on 'practice' above.) Firms like Acacia Research have a wide range of agreements with practicing entities to maximize the value of the practicing entities intellectual property. These relationships are important in understanding the true nature of the litigation and who the real players are. Just because you don't manufacture stuff doesn't mean you don't work for someone who does. Counting these as solely evil troll litigation seems to be a misrepresentation of the nature of the action and the business models.

How will the contractor determine the "real parties in interest"?

How will GAO determine the universe of NPEs and patent assertion entities? Many firms engaged in this work establish separate entities for each transaction not unlike real estate developers who want all their projects in separate business entities. There are plenty of legitimate reasons for these types of entities — and plenty that are solely for the purpose of obfuscating ownership.

How will GAO evaluate the age of the patents being enforced? The Boston University study on NPEs which has gained urban legend status in the patensphere  cited shows the average age of a patent in NPE litigation is over eight years old.

What about considering how many of these cases required Special Masters of some kind because of the complexity of the underlying inventions and science that form the foundation of the case?

What about licensing pools, such as the MPEG Licensing Authority that enforced the intellectual property rights of its members? It is clearly a non-practicing entity but it represents a very large universe of practicing entities.

How will publicly traded patent aggregators like RPX and firms like Acacia Research be evaluated. Both have different models for creating shareholder value (assuming that creation of shareholder value is still a good thing.)

What about taking into consideration the difference between highly structured industries like pharma and biomedical inventions (the chemical domain) versus the world of software, business methods and all things related to electrical and communication patents?

And finally, how will GAO combat patent urban legend like — Intellectual Ventures (IV) owns 40,000 patents — (at the time of this writing, there is a crowdsourced request for money out there looking to raise $80,000 to develop a comprehensive view of the IV patents; there are over 250,000 patents covering smartphones (Google's general counsel seemed to be counting all the claims in all the patents to come up with the total; that there have been more than 40,000 in troll related litigation; NPEs cost firms billions of dollars each year (see the Boston University analysis on that one) and finally, trolls hinder innovation.

This is clearly a difficult and challenging task. Hopefully the research and analysis will be equally compelling.  According to the RFQ the task is to be complete by December 31, 2012.  GAO will need to create and publish it's report after that.

Let the troll hunt begin.