The GAO Issues a Report on Patent Litigation Trends -- It Turns Out that the Sky Is Not Falling

By Andrew Williams --

As we have previously reported, the patent system is under attack, and has been for quite some time.  Generally, these attacks either begin with the assumption that the patent system is "broken," or that conclusion is reached after a review of anecdotal accounts of alleged patent abuses.  Rarely are these attacks supported by any actual numbers, and there has been a complete dearth of scientific or statistically based studies to support them.  Perhaps not coincidently, when passing the America Invents Act ("the AIA"), Congress attempted to rectify this situation by tasking the GAO (Governmental Accountability Office) to conduct a study of the consequences of litigation by non-practicing entities or patent monetization entities ("PMEs") (referred to more derogatorily in the mainstream press as "Patent Trolls").  The resulting report, entitled "Intellectual Property: Assessing Factors that Affect Patent Infringement Litigation Could Help Improve Patent Quality" ("GAO Report"), was sent to the congressional committees on August 22, 2013.  The Report itself contains some observations regarding the increase in patent litigation, some possible suggestions explaining these observations, and one recommendation:

We are recommending that the Secretary of Commerce direct the Director of PTO to consider examining trends in patent infringement litigation, including the types of patents and issues in dispute, and to consider linking this information to internal data on patent examination to improve the quality of issued patents and the patent examination process.

GAO Report, p. 46.  Even though this was an ambitious report, the GAO was upfront with the limitations of its research.  Indeed, the Report noted that the Comptroller General sent a letter to the chairs and ranking members of these congressional committees before the passage of the AIA, informing them that reliable data was either not already available or could not be obtained.  See id., p. 4 n.10.  Nevertheless, far from concluding that the patent system was "broken," the conclusions actually reached by the GAO suggest that even though there may have recently been an increase in low-quality patents, particularly related to software, it is not a cause for alarm.  In fact, the GAO noted trends in both the federal courts and the Patent Office that are already seeking to correct these perceived problems.

As background on the Report, section 34 of the AIA charges the Comptroller General to conduct a study that should include the following information:

(1)    The annual volume of litigation described in subsection (a) over the 20-year period ending on the date of the enactment of this act.

(2)    The volume of cases comprising such litigation that are found to be without merit after judicial review.

(3)    The impacts of such litigation on the time required to resolve patent claims.

(4)    The estimated costs, including the estimated cost of defense, associated with such ligation for patent holders, patent licensors, patent licensees, and inventors, and for users of alternate or competing innovations.

(5)    The economic impact of such litigation on the economy of the United States, including the impact on inventors, job creation, employers, employees, and consumers.

(6)    The benefit to commerce, if any, supplied by non-practicing entities or patent assertion entities that prosecute such litigation.

AIA, § 34(b).  Because of the aforementioned concerns regarding the existence (or lack thereof) of reliable data, the GAO reinterpreted the objectives of its study to be:

(1)    What is known about the volume and characteristics of recent patent litigation activity;

(2)    the views of stakeholders knowledgeable in patent litigation on what is known about the key factors that have contributed to recent patent litigation;

(3)    what developments in the judicial system may affect patent litigation; and

(4)    what actions, if any, has PTO recently taken that may affect patent litigation in the future.

GAO Report, p. 4.  This deviation is likely to draw criticisms that the GAO did not actually conduct the study they were tasked to perform.  However, in response to the GAO's previously noted concerns, Senator Leahy stated on the record that any resulting report should include any limitations on the data and methodology used.  See id., p. 4, n.10.  The GAO took this opening to develop its own objectives for the Report, which it believed were consistent with these noted limitations.  Id.

The data that the GAO used to assess recent patent litigation was obtained from Lex Machina, a firm that collects such data.  The Report indicated that Lex Machina used all data for patent infringement lawsuits filed between 2000 and 2011, and also selected a random, generalizable sample of 500 lawsuits -- 100 per year from 2007 to 2011.  This sample size allowed the GAO to make conclusions regarding each of these years with a margin of error of ± 10 percentage points for any particular year, but ± 5 percentage points for all of the years.  The GAO also used data obtained from speaking with 44 stakeholders, which included 10 representatives from operating companies who had been sued in recent years (including representatives from software, computer hardware, retailers, and pharmaceutical industries); 8 representatives from PMEs that had regularly sued others; 14 legal commentators, economists, and consultants; representatives from 2 universities, 2 patent brokers, 4 VCs, and 4 individual inventors.  The GAO clearly attempted to solicit the opinions of a representative cross-section of stakeholders.  Nevertheless, any results stemming from these interviews must be viewed with caution because of the size of the sample was necessarily small, and any data obtained is inherently limited by the biases of the particular stakeholders that were consulted.

Perhaps the most interesting results, and likely the most reliable, came from the GAO's analysis of the Lex Machina data.  The GAO found that patent lawsuits fluctuated slightly between 2000 and 2010, but in 2011 there was a 31% increase.  Contrary to the conventional wisdom, however, the GAO found that it was operating companies, not PMEs, that brought the majority of patent infringement lawsuits in the 2007 and 2011 timeframe.  See GAO Report, p. 17.  In fact, in that timeframe, PMEs and likely PMEs were found to have only brought 19 percent the studied lawsuits.  This number does appear to be increasing, with PMEs being found to be responsible for 24 percent of the lawsuits in 2011, but the Report did note that this increase was not statistically significant.  Perhaps the common perception regarding PMEs stems from the fact that they tended to sue more defendants per suit than operating companies.  For example, between 2007 and 2011, operating companies sued on average 1.9 defendants per suit, whereas for PMEs, the number was closer to 4.1 defendants on average.  In fact, the GAO found that PMEs in this timeframe had sued close to 1/3 of all defendants.  The GAO did attribute the increase in patent litigation to the type of patents, and not the entities brining suit.  In the 2007 to 2011 timeframe, 46% of all lawsuits involved software-related patents, with 64% of all defendants facing such suits.  Unfortunately, the GAO was not able to independently determine a cost for all of this litigation, either to the parties directly or to the industries involved.  Moreover, the Report noted that it was limited to cases that had been litigated, and therefore they were unable to determine the cost of patent assertion that occurred outside of the court system.

As for the remainder of the Report, even though the results of the stakeholder interviews should be viewed with caution, the observations and conclusions are informative.  There were three key factors identified that have likely contributed to the increase in recent patent infringement lawsuits:

(1)    unclear and overly broad patents;

(2)    the potential for disproportionately large damage awards; and

(3)    the increasing recognition that patents are a valuable asset.

GAO Report, p. 28.  The first factor relates to the alleged prevalence of low-quality patents.  This view was predominately related to software-related patents.  The Report identified several reasons why stakeholders believed that these software patents were overly broad.  First, the lack of clear terminology in the computer software industry has resulted in the use of unclear terminology in software patents.  Second, the use of functional language in such patents has allowed the patent owners in these industries to claim that their patents not only cover the small improvement described, but claim that they cover entire technologies or potential future technologies.  Third, even though patent infringement is strict liability (being unaware of a patent is not a defense to infringement), the presence of such broad patents results in more infringers that had no intent to so infringe.  Relatedly, some stakeholders pointed out that the sheer number of patents makes it difficult to search for those that might be relevant to a particular technology they are developing.  Also, even if the relevant patents are identified, it can be difficult to identify the owners, because the Patent Office does not currently require notification of changes in ownership.  Interestingly, even though some of the stakeholders interviewed blamed PMEs for the rise in patent infringement litigation, others did note that it was likely the prevalence of low quality patents that played a bigger factor.

The GAO Report did identify several trends that may have a positive impact on the identified increase in patent litigation.  First, the Report noted that the federal court system is implementing new initiatives to handle patent cases.  For example, in January 2011, Congress established a pilot program in certain districts to encourage expertise among district court judges.  So far, 14 federal districts have been designated to participate in this 10-year pilot program.  In addition, in September 2011, the Advisory Council for the Federal Circuit issued a model order to address e-discovery issues targeted to patent cases.  It is too early to tell, however, what impact these e-discovery rules will have.  Finally, the case law is evolving, and as a result could address some of these problems.  The Report highlighted recent decisions about what constitutes patentable subject matter, or what is required to satisfy the obviousness and definiteness requirements.  The Report noted that such decision could help combat the identified overly-broad and ill-defined patents.

The Report also noted that the Patent Office has taken steps to improve patent quality.  For example, in February 2011, the Patent Office issued supplemental guidelines related to the definiteness requirement.  These guidelines specifically addressed the examination of claims with functional language.  Also, in November 2011, the Patent Office began working with the software industry to address the issue of non-uniform terminology in software related patents.  In addition, the Office launched a new patent classification system in January 2013 called the Cooperative Patent Classification (CPC).  Finally, the Report noted that the Patent Office has been working to solve the issue of patent ownership transparency.  Of course, all of these initiatives are in addition to the Patent Office's implementation of the post-grant review proceedings that are mandated by the AIA.

As indicated at the outset, the only recommendation included in the Report was directed to the Patent Office.  Before the Report was sent to the congressional committees, the GAO sent a copy to the Patent Office to review.  Appendix II of the Report is a letter from Acting Under Secretary and Acting Director Teresa Stanek Rea to the GAO that commends them for their efforts, and concurs with the recommendation.  Specifically, the Patent Office's response to the recommendation was:

The USPTO appreciates GAO's recommendation.  The USPTO currently uses information relating to cases involved in patent litigation, and agrees that it would be appropriate to consider making better use of such information by examining trends in patent infringement litigation.  The USPTO also agrees that as part of its ongoing effort to improve the quality of issued patents and the patent examination process, it would be appropriate to consider linking trends in patent litigation to internal data on patent examination.

GAO Report, App. II.  Perhaps this recommendation, along with all of the identified initiatives, will help improve the quality of software patents, and as a result help to restore some faith in the U.S. patent system.

Comments

[GD]

Very nice post. Important topic.

Yes, patents have been under attack. For the most part (or completely) those attacks were not grounded in facts.

The interesting question is why the hostility? Where is it coming from?

It will be very interesting to see the Comptroller General's report on the economic impact of litigation on the U.S. economy and what, if any, impact on commerce arises from non-practicing entities.

Unfortunately, what appears to be missing is an inquiry into the big questions: What is the net impact of patents as a whole (or broken into industry segments) on the U.S. economy, its balance of trade, its standard of living and etc. Are patents good, bad or neutral? My guess is that they are very good on balance. Solid data to confirm or refute that guess is important but long overdue.

[Skeptical]

Trolls? I was always... (see below)

Now Software? I remain... (see below)

[EG]

Andrew,

As the GAO Report shows, of course the "sky wasn't falling" in terms of the so-called "patent troll" problem. Just lots of "hot air" from you know where to keep so-called PT balloon "inflated.

[Skeptical]

As to the software count, Greg Aharonian through Hal Wegner's email updates shows an up to a 10X bogus factor in the count for 'software' patents (in one particular group - and likely a similar order of magnitude overall).

Could I be less...

[Skeptical]

Oh look, look at who supplied the raw data sets: Lex Machina and RPX.

No bias there, of that I am, well,...