This Saturday, September 15, 2017, marks the five-year anniversary of the first filing of an inter partes review. We’ve seen nearly 7,000 post-grant reviews filed since then, a Supreme Court case dealing with IPRs, and there are a pair of IPR Supreme Court cases up this term. [Oil States] [SAS]
Over the next few weeks, I’ll tell you about particular stories where inter partes and covered business method patent reviews have curtailed abusive litigation and allowed smaller companies to defend themselves even if they might not have been able to fight a full-fledged patent lawsuit.
But today I just want to step back and look at the effects of the inter partes review system as a whole. One of the reasons IPR was created was to “provid[e] a more efficient system for challenging patents that should not have issued; and reduc[e] unwarranted litigation costs and inconsistent damage awards.” (Page 39-40 of the House Report on the AIA.)
All in all, the IPR system has been incredibly effective at achieving these goals—I estimate that the implementation of inter partes review has helped plaintiffs and defendants avoid at least $2.31 billion in deadweight losses by providing an efficient system for challenging patents.
This benefit is purely based on avoiding deadweight loss from legal fees; it does not account for the benefit of preventing transfers from defendants to plaintiffs based on patents that should have been invalidated. The financial data used in this analysis is based on publicly available data, as well as some data derived from the 2017 AIPLA Economic Survey. Unfortunately, this survey is not publicly available; where possible, I have linked to open summaries of the data contained in the survey.
Amount of Deadweight Loss Due To IPRs
According to the AIPLA survey, the median cost of filing an IPR through to the petition stage is $100,000. This is the total cost—everything from attorneys to expert witnesses to the filing fee paid to the PTO. (IPRs are user-fee funded, meaning that the PTO’s expenses in deciding an IPR are not borne by taxpayers but only by the petitioner.) The median cost of an IPR through obtaining a final written decision is $250,000. (RPX provides similar estimates for IPR cost.)
As of July 31, 2017, there have been 6,390 petitions filed and 1,674 final written decisions. Approximately 92% were inter partes reviews, with the remainder consisting of covered business method reviews and a few post-grant reviews. So we’ve seen approximately 5,879 IPR petitions and 1,540 IPR final written decisions. That means that petitioners have spent approximately $818 million on IPRs. The cost of defending an IPR is generally considered to be roughly equal to the cost of pursuing one, according to the AIPLA survey, so patent owners have spent approximately the same amount, for a total cost of $1.637 billion.
So, would the equivalent litigations have cost more than $1.637 billion?
Amount of Deadweight Loss Due To Litigation
Based on data from RPX, the average amount spent by a defendant in NPE litigation on legal fees is $950,000. NPE litigation is typically less expensive than operating company litigation; however, some of those fees will have been incurred regardless of whether an IPR is filed, as they occur in the earliest portion of litigation. As the RPX data shows, that early portion of the litigation is a relatively low portion of spending on the whole case. A reasonable estimate based on publicly available data for the costs incurred in litigation prior to IPR filing is approximately $200,000. This means that post-IPR litigation spending which an IPR can prevent can be estimated at a rough average of $750,000. Broadly speaking, this is consistent with similar data in the AIPLA survey, which gives a median cost through trial for a medium-sized case of $1.7 million, with approximately half that amount being spent in the discovery phase.
However, not every IPR petition maps to a unique litigation. It has been estimated that 70% of IPR petitions are related to patents that are being litigated. So 30% of IPRs create IPR cost without any corresponding litigations; we include the cost of those IPRs (since they’re part of the IPR system), but do not include any corresponding litigation cost.
Further, while the majority of patents only have one petition filed against them, multiple petitions directed to the same patent do exist. This can either be because a single petitioner filed multiple petitions, or because multiple parties have been sued and filed parallel petitions. In the first case, we shouldn’t estimate additional litigation cost (because both petitions relate to the same litigation), while the second case does represent additional litigation cost (because both defendants will be paying to litigate.) Based on an analysis of petitions that I conducted, there are approximately 80% as many unique patent/petitioner pairings as there are IPR petitions, leaving 20% of petitions as directed to litigations already accounted for. So 20% of our remaining IPR petitions also generate cost without associated litigation savings.
Finally, some IPRs with associated litigations generate lower savings, because not every instituted IPR stays the progress of the associated litigation. Without a stay, the litigation still generates cost. Approximately 80% of cases with co-pending IPRs are stayed. Thus, another 20% of remaining IPR petitions generate cost without associated litigation savings.
Taking the number of IPR petitions filed (5,879) and using our average estimate of overall cost ($750,000), then discounting it by 56% to reflect the effect of non-litigant petitioners, multiple petitions from the same party, and unstayed litigation, I estimate $1.975 billion avoided by defendants that otherwise would have been spent on litigation costs. Approximately the same amount will have been avoided by plaintiffs.
There may be additional savings my estimate does not account for; for example, non-litigant petitions can be filed by licensees who wish to escape a license to a patent they believe is invalid. Instead of filing a lawsuit, they can opt for an IPR. However, as a conservative estimate, it’s fair to say that the litigation costs that IPRs substitute for would have incurred deadweight loss estimated at approximately $3.95 billion.
Adding It Up
That’s $3.95 billion saved by spending $1.637 billion. These rough estimates suggest that, in the not quite five years of IPR data available, IPR has saved plaintiffs and defendants $2.31 billion in deadweight losses, primarily in the form of legal fees. That’s approximately $460 million a year that companies can spend on creating new jobs and researching new technologies, instead of paying lawyers to write motions and argue in court.
To me, those numbers say IPR has been a tremendous success.