Section 337 Caseload and Win Rate Revert to Norms

Guest Post By Michael G. McManus< ?xml:namespace prefix ="" o />

Traditionally, section 337 complainants at the ITC have enjoyed a greater chance of success than district court plaintiffs.  The historical complainant win rate at the ITC has been calculated as 58%.[1]  This contrasts with a district court patentee success rate of 32% (considering both summary judgment and trial stages).[2]  In 2010 and 2011 however, there was an unprecedented spike in the number of cases filed with the ITC.  These complainants fared poorly with a win rate well below the historical norm.  For section 337 complaints filed in 2012, however, both the number of cases and the win rate have reverted to recent historical norms.

In calendar year 2008, there were 36 section 337 complaints filed.  Of these, 20 were resolved by settlement[3] prior to the issue of an initial determination and 16 were resolved by administrative law judge (ALJ) decision[4] (either summary determination or final initial determination).  These 16 ALJ decisions were evenly split – 8 were resolved in favor of the complainant and 8 in favor of the respondent yielding a win rate of 50% at the ALJ stage.  The following year was generally similar.

In calendar year 2009, there were 34 section 337 complaints filed. Of these 34, 21 were resolved by settlement prior to the issue of an initial determination and 13 were resolved by ALJ decision.  Among those cases decided by ALJ decision, 6 were resolved in favor of the complainant and 7 in favor of the respondent yielding a win rate of 46%.

In calendar year 2010, the number of filings increased markedly to 58.  Of these, 36 were resolved by settlement prior to the issue of an initial determination and 22 were resolved by ALJ decision.  Of these 22 ALJ decisions, 10 were in favor of the complainant and 12 in favor of the respondent yielding a win rate of 45%.

In calendar year 2011, the number of section 337 filings increased further to an all-time high of 68.[5] Of these 68 cases, 49 were resolved by settlement prior to the issue of an initial determination and 19 were resolved by ALJ decision.  Of these 19 ALJ decisions, 5 were in favor of the complainant and 14 in favor of the respondent yielding a rather low win rate of 26%.

In calendar year 2012, by contrast, the number of section 337 filings fell to 40.  We cannot calculate the win rate for all cases filed in 2012 as those filed in the latter part of the year may still be awaiting ID’s.  All of those cases filed in the first half of 2012, however, have either settled or been subject to an initial determination.  Thus, we can use those cases from the first half of 2012 as a proxy.

There were 25 section 337 complaints filed in the first half of calendar year 2012.  Ten of these 25 resulted in ALJ adjudications. Of these 10, 6 were in favor of the complainant and 4 were in favor of the respondent yielding a win rate of 60%.

Put in graphical format, the win rate over time is seen as follows:


Overlaying the section 337 caseload over win rate yields the following:


The graph suggests an inverse relationship between caseload and complainant success rate but there is no obvious explanation for that relationship.  One possibility is that the ITC’s increasing popularity attracted weaker cases. Also, it is possible that the strain of a greatly increased caseload caused the ITC to react in ways that were less hospitable to complainants. In any case, the “great disruption” of 2011 seems to have passed and the Commission’s caseload and win rate have reverted to their historical norms.

Michael McManus is a partner at Duane Morris.

[1] Colleen V. Chien, Patently Protectionist? An Empirical Analysis Of Patent Cases At The International Trade Commission, 50 Wm. & Mary L. Rev. 63 (2008), at 97 n. 174 (overall win rate for the period from January 1995 to June 2007).

[2] PWC 2012 Patent Litigation Study at 17 (overall win rate for the period from 1995 to 2011).

[3] Here “settlement” encompasses all cases resolved by means other than contested adjudication (e.g., withdrawal of complaint, default).

[4] This article will evaluate win rate at the initial determination stage rather than the final determination stage to increase the sample size and to permit an “apples to apples” comparison with cases filed in 2012 (most of which have not yet proceeded to final determination).

[5] As this is the number of cases filed in the calendar year rather than the ITC’s fiscal year, it differs slightly from statistics on the ITC’s web page.

13 thoughts on “Section 337 Caseload and Win Rate Revert to Norms

  1. 4

    This article is somewhat uninformed in that it ignores the true elephant in the room — non-practicing entities (NPEs). The rise in ITC filings from 2010-2012 was due almost entirely to a huge spike in filings by NPEs. (By NPE filings, I’m referring generally to those entities whose only “domestic industry” is “licensing” of the asserted patents.) NPEs flocked to the ITC in droves in 2010 and the years following it, with the expectation that they could get a large number of quick settlements and exert maximum pressure on defendants. In the early days this worked. Naturally this attracted the weaker cases, those that tend to settle to avoid litigation costs where the merits are questionable and the risk of actual liability is low. But weak cases do not fare well before the ITC because the judges (unlike juries) cannot be snowed by trial theatrics, witness personalities and superficial arguments. They are going to dig into the merits of the case at a very detailed level, and if your case is weak, you will end up getting nothing for your effort. Not surprisingly, NPEs lost the vast majority of ITC cases that went to hearing. This effect was exacerbated by the fact that NPEs were doing ITC cases “on the cheap,” and it showed, and the judges noticed. Defendants’ fear of NPE cases at the ITC was reduced, causing more cases to go to a full hearing. Case filings, accordingly, dropped.The ITC “experiment” was over, and while NPEs today still bring ITC cases, the number is way down from 2010-2011 and the cases filed today tend to be considerably stronger.

  2. 2

    Statistically, I would say that you really don’t have a large enough sample size to do anything more than make haphazard generalizations as to why the 2011 success rate was down.There are just far too many variables that impact the outcome of any case to suggest (with any confidence) that the win percentage is based upon the number of complaints in a given year. Also, while a “year” is a convenient period of measure, it is a contrived one.

    1. 2.2

      And to the point of generalizing from the data, isn’t it a view of the Supreme Court that no such bright line generalizations should be attempted (especially with patent cases which are so fact-specific)?

  3. 1

    May I suggest the greater win rate in the ITC vs. District Courts might have something to do with the relative experience of the ITC judges in patent cases vs. District Court judges. It also seem to be a truism that accused infringers desperately want to avoid experienced District Courts as well, often accusing them of a pro-patentee bias. But if an experienced trier of fact tends to result in favorable outcomes for patentees, I would think that patentees should consider not requesting a jury trial if they have a case being handled by an experienced judge. I wonder if we had any stats on who requested jury trials in districts like the ED Texas?

    1. 1.1

      I have trouble understanding why patent cases have juries in the first place … ever. Unless it’s a jury of engineers or the like, there really isn’t much chance of a proper outcome.Michael Feigin, Esq.

      1. 1.1.1

        Actually, they do make good judgments primarily based on credibility rather than understanding the basic science. Trial is a battle of experts. They always disagree. But the jury can see who is lying.


          This viewpoint is somewhat naive. At least when it comes to issues like infringement and validity, the issues are very technical and poorly suited to a jury. Neither side’s expert is “lying.” In most cases, they are both presenting different interpretations of the same basic scientific or technological facts, all of which goes over the head of the jury. Both experts will usually be presenting some plausible theory to support their case, which is why the case even got to trial. The experts are not the inventors of the technology, and the only impeachment is usually the fact that they’re being hired by their respective side (which cancels each other out since that applies to both sides). “Credibility” is much less of an issue in this context, which really takes away the one tool they have to make a rational decision.In criminal cases, securities fraud, malpractice, etc., you have that extra intent element, or an element of reasonableness, that juries are good at assessing. Most patent cases, however, do not turn on any of those issues. That’s because infringement is a strict liability tort except in the case of “willful” infringement, but that only comes into play after infringement is found, and it’s becoming more rare these days to even have willful infringement issues tried to juries.I recently handled a jury trial in a complex technology case, and at the end of the case, interviewed several of the jurors. They all asked me why, with all of the PhDs and smart people testifying, why anyone would think they, the jurors, should be deciding the case? They expressed frustration that the system would even entrust them with a decision over something so obviously outside their realm of experience.


            Well, Mr. Runner, I agree with you that the technology often is way over the heads of the jurors. That means they have to decide the case on the only thing they really understand, credibility, and that is based on demeanor, etc. Intangibles. The kind of thing the appeal court never sees. I think experts who really do not believe in what they are saying communicate that very well to the jury.

      2. 1.1.2

        Should cases regarding tax fraud or securities fraud be tried by juries? How about cases of medical malpractice? Do juries who decide criminal cases have a pre-existing understanding of the forensics used to try the case, or is it explained to them by the attorneys trying the case?

    2. 1.2

      In the old days of patent litigation (e.g., the 1980s and even early 1990s), almost all patent cases were tried as bench trials. Jury trials in patent cases were rare. There have been a number of articles about it, but the stats were crazy — it was something like more than 80% of patent cases were tried to the bench. It was just the tradition back on those days, people didn’t demand jury trials all that often.There is no hard-and-fast rule about “relative experience” of the judges versus the outcomes. It depends on the strength of the case. The stronger your case, the more you want an experienced and knowledgeable trier-of-fact. It is only when you have a weaker case or a weaker defense, that’s when you want a jury trial, because you can more easily distract them from the merits of your case with superficial arguments and theatrics. (Brings to mind that wonderful scene from the movie Chicago where Richard Gere was tap-dancing in court to get around bad facts.)

      1. 1.2.1

        Lode Runner, juries still get it right because they see right through something phony. If the case is weak, they will quickly see it.

Comments are closed.