Guest Post: Counting Defendants in Patent Litigation

PatentLawImage024Guest Post by Kyle Jensen of PIPRA.

I recently looked at some historical litigation trends and was surprised to learn that the number of patent lawsuits filed annually was roughly constant over the past decade. This year, of course, there are a large number of qui tam complaints being filed for false marking (largely by a small handful of plaintiffs); however, the number of infringement cases will not be dramatically different than 2009.

The relative constancy of litigation activity is a bit perplexing and certainly incongruent with the patent-system-gone-awry stories so commonplace in the popular media. I thought there was an "NPE problem" spinning out of control. I thought costs were spinning out of control. Well, they might be, but it's not reflected in increased lawsuit filings. I pulled back-data from PACER to produce the chart below, which shows the number of patent lawsuits and defendants in those lawsuits as a percentage of their levels in 1990.

There are some striking observations here. * The total number of patent litigations filed increased about 300% between 1990 and 2002, but has been nearly constant for the rest of this decade. (Many others have shown this previously. See, e.g. Shrestha, 2010) * The number of cases with more than one defendant followed a similar trend (data not shown). * The number of named defendants increased nearly 600% between 1990 and 2010.

In short, the number of cases filed per year is constant, but the number of defendants keeps growing.

I don't know why this is. The most obvious hypothesis is that it's driven by an increase in many-defendant filings by NPEs. That is, maximizing the ratio of defendants named to cases filed will 1) maximize potential revenue and 2) decrease costs for an NPE or contingency counsel. I can't think of many reasons that a practicing entity would behave similarly.

Numerous sources report that the number of NPE lawsuits is growing and now make comprise about 1/5th of all filings. For example, according to PatentFreedom, NPE lawsuits increased from less than 100 per year in 1998 to roughly 500 in 2009. (There were about 2,600 total cases that year). Also, Colleen Chien showed that in the IT sector NPE suits account for 1 in 6 cases. Given these statistics, it would be interesting to if those cases account for a disproportionate fraction of the named defendants. (This is somewhat more complicated to show.)

If the number of defendants is growing, costs are certainly growing as well. Assuming each case costs each defendant $1MM (many studies show it's higher), defendants would be spending an extra $400MM/yr on litigation each year.

Where's that money go? Certainly attorneys capture some of it. Below I've overlaid the number of attorneys assigned to patent litigations over the years. As you can see, the growth in attorneys outstrips both cases and defendants.

(All this data is available in PACER, so you can verify my numbers. Also the raw data are online. There are a few obvious caveats, e.g. I haven't removed cases in which multiple subsidiaries of the same company are named in the same suit. About me: I work for the Rockefeller-funded non-profit PIPRA where I teach scientists and entrepreneurs in developing economies about IP. I also own the website PriorSmart, which publishes a daily patent complaint alert. It's free, and includes PDFs of the pleadings. You can sign up here if you'd like to get it.)

59 thoughts on “Guest Post: Counting Defendants in Patent Litigation

  1. It appears that you’ve put a good amount of effort into your article and I require a lot more of these on the World Wide Web these days. I truly got a kick out of your post. I do not have a bunch to to say in reply, I only wanted to say tremendous work.

  2. It gives me interest on this topic that is why i created a thread on my favorite forums. I will get more response to them will will help me ing my projects. Thanks for this informative post.

  3. Don’t discount the Hatch-Waxman case contribution. A lot of generic challenges that in the past would have had one or two filers are now populated by 6, 7, or more generics especially on the Big Pharma NCE-1 challenges. What percentage of the rise has been in Jersey and Delaware, where most of them are litigated?

  4. “I don’t think that. I think the number is probably higher for NPEs, however, as their patents predominantly appear to descend from the art units where no data is required and you can file on some b.s. that you thought of while trying to hose a fly off the back wall of a urinal.”

    JD’s toiletpaper holder art?

  5. What makes you think that “crxp claims” are more likely to be owned by NPEs, or that NPE claims are more likely to be crxp?

    I don’t think that. I think the number is probably higher for NPEs, however, as their patents predominantly appear to descend from the art units where no data is required and you can file on some b.s. that you thought of while trying to hose a fly off the back wall of a urinal.

  6. They explained that the patent was on the system and that if damages were based upon the system it would literally bankrupt their company.

    Makes you think twice about the real world value of those indemnification clauses, doesn’t it?

  7. @James – I don’t have those data, but could easily get them. Feel free to email if you’d like to chat data stuff. It’s a gmail address and I’m kljensen.

  8. There is also point where a manufacturer of a component of a larger system may be liable both for direct infringement and for contributory infringement, and possibly even for inducement. The value of the component sold is typically small in comparison to the value of the larger system. So the question often becomes whether reasonable royalties should be based upon the value of the component or the value the system.

    I remember a situation where I was in the position of the VAR requesting indemnity from my component supplier. No matter how clear I made it that their obligations to indemnify were unambiguous, they refuse to give in. Finally, I asked why. They explained that the patent was on the system and that if damages were based upon the system it would literally bankrupt their company.

    So it seems that to protect themselves even in a vertical relationship of manufacturer and VAR, patent owners must sue every member of the manufacturing and distribution chain to be fully compensated for the infringement and to have the parties with the deep pockets be held liable.

  9. Those poor non-practicing entities with their crxp claims

    What makes you think that “crxp claims” are more likely to be owned by NPEs, or that NPE claims are more likely to be crxp?

    You see a lot of dubious claims from practicing entities like IBM, Amazon, Myriad, and such.

    Also, once a claim is issued, you can hardly blame the patentee for trying to assert it against infringers, whether he practices it or not. That’s what it’s for.

  10. This factual data lends credence to the position that the use of pejoratives such as “Troll” is merely a part of a smear campaign.

    Boo hoo hooo hoo! Those poor non-practicing entities with their crxp claims just can’t seem to get the love from the public that they deserve. I feel so sorry for them being “smeared” by perfectly descriptive and appropriate name. Is there a place where I can donate some money to help these poor corporations out? Boo hoo hoo hoo hoo!!!

  11. I know what this is all about Ned. The first one was stolen from me, while a Design Patent was filed in Duetch land. And now my other is being argued, and just sits there because of the one that was filed while mine was being hidden and held back claims it’s their idea.

  12. Excellent point there about claim construction-especially if all are using a common reference design such as one finds with standards. The courts should favor, strongly favor, one common construction for all.

  13. That’s pretty much assured when you start extending patents over business methods, software etc etc. though right?

    Absolutely especially with regard to business methods. It’s simply not in the DNA of most financial service company executives to be concerned about patent infringement. Historically it just hasn’t been an issue for them.

  14. Well thanks Guys. As you say Winston:

    “…consider a patentee who believes 10 companies are infringing his patent.”

    Never used to but, these days, it happens all the time.

    You also write:

    “If the patentee believes he only has a 5% chance of losing on invalidity or inequitable conduct, he will have his patent invalidated ~40% of the time by suing each individually,”

    for which I am grateful, for I have to say that I never saw it that way. I was thinking that, if Ethicon can’t invalidate it and, later, Boston Sci can’t invalidate it then, even later, Medtronic’s chances of invalidating it are even less, not as you say, going up towards 40%. what is this “Presumption of Validity” thing anyway?

  15. To Anon: my “skim” comment was an attempt to provoke, to get an answer to the question whether the multiple Defendants in any one action are all “in it together” ie parent and many wholly-owned subsidiary companies, or if the multiple Defendants are unconnected with each other (apart from being all members of the relevant industry standard-setting committee).

    Some of these cases involve a single infringing device incorporated into another device by a second company (which in turn may be incorporated into additional devices by additional companies) and sold through a multitude of other companies. Suing all involved both helps with venue and, through indemnification agreements between the alleged infringers in the supply chain, can put greater financial pressure on the principal infringer.

    Also, particularly in standardized industries, if a patent reads on a standard then anyone practicing that standard infringes. After the first filing the plaintiff is likely to be met with DJ actions and/or ex parte reexams by the whole industry anyway, so the patentee may as well bring every infringer in initially.

    Suing an entire industry in a single suit also makes it harder for defendants to argue for claim constructions that none of the defendants meet. Even in the same court, it is possible for the claim construction against party A to cover party B’s product but not party A’s product and vice versa. Where claim constructions are not consolidated, this presents an obvious problem for plaintiffs. Multiple claim constructions (and multiple trials) also increase the risk of an invalidity finding.

    For argument sake, consider a patentee who believes 10 companies are infringing his patent. If the patentee believes he only has a 5% chance of losing on invalidity or inequitable conduct, he will have his patent invalidated ~40% of the time by suing each individually, but only 5% if he sues all at once. He also only has to manage one appeal, so if he lands in that 5% in trial he gets a new bite on all 10 defendants if he wins his appeal. Even if I were a very confident patentee, I’d probably do the math and sue everyone I thought infringed in the initial suit.

  16. Thanks, MaxDrei.

    I think the question of “all in it together’ is a good one and need not lower into a “NPE provoke” mode, as the question readily applies to more than “NPE.” The data indicates that there is no real “NPE” explosion.

    I do not think that a “relevant industry standard-setting committee is implicated – common industry participants would be enough.

  17. To Anon: my “skim” comment was an attempt to provoke, to get an answer to the question whether the multiple Defendants in any one action are all “in it together” ie parent and many wholly-owned subsidiary companies, or if the multiple Defendants are unconnected with each other (apart from being all members of the relevant industry standard-setting committee).

  18. “How about more infringement?”

    That’s pretty much assured when you start extending patents over business methods, software etc etc. though right?

  19. Overall, I agree with Kyle Jensen in that a mere doubling of litigation in a twenty year period (with the last seven years showing a flat or modestly declining level) is incongruent with popular media’s portrayal of a system out of control.

    This factual data lends credence to the position that the use of pejoratives such as “Troll” is merely a part of a smear campaign. Those involved in setting policy should view this data appropriately and should ensure that such scare tactics fail as such tactics should.

    MaxDrei,

    I do not understand your comment about “NPE skims of an entire industry.” What do you mean by this?

  20. I have noted two explanations so far in this thread, namely 1)tethering to a favourable forum (EDTX, VW) and 2)NPE skims of an entire industry.

    Any others, that anybody else would care to mention?

  21. “Interesting article and more evidence that NPEs in possession of crxp patents that should never have been granted are a growing problem.”

    if the crxp patents you are speaking of are indeed crxp, the parties involved in the litigation would not have wasted so much time & money defending their rights…

  22. more evidence that NPEs in possession of crxp patents that should never have been granted are a growing problem…hence ping’s pathetic attempts

    Nice spin Sunshine – cepts ya gotta announce “it’s opposite day” when ya try to paint black as white as ya try here.

  23. Kyle,

    Great article.

    To answer some of the questions here, do you have the number of defendants broken down by district?

    Thus, a graph comparing the number of defendants in ED Tex cases compared to cases pending in other districts might address whether there are more defendants because of venue in ED Tex

    Thanks,
    James

  24. There is a third reason, of course, for the increase in number of defendants: TS Tech and the other recent Federal Circuit decisions regarding venue transfer. With dozens of defendants in a case, including some with clear ties to, e.g., Marshall, it’s thought to be more likely to keep the case in the chosen forum.

  25. IANAE, in that case, then the text is wrong.

    Yes, it is. It’s the same mistake as the “% change” label on the vertical axis (300% increase rather than 300% of), probably because that’s where the author got his numbers from.

  26. “These graphs all displaced upwards by 100%.

    No, the axis is just labeled wrong. It’s common to scale to 100 in the reference year (look at the CPI, for example) and the resulting graph is much more intuitive than if you’d put the reference year at zero.”

    IANAE, in that case, then the text is wrong. Specifically: “The total number of patent litigations filed increased about 300% between 1990 and 2002″

  27. The big rise in multi-defendant patent litigation starting around 2007 stems from the 5th Circuit’s Volkswagen decision, which allowed mandamus review of denied transfer motions in the Eastern District of Texas. Adding more defendants makes finding a single other forum that is “clearly more convenient” more difficult, so plaintiffs that want to litigate in EDTX add more defendants.
    Posted by: Winston Smith | Oct 27, 2010 at 10:27

    Actually, this is right. The phenomena was fully discussed during a session of of the AIPLA last week. Patent owners want to litigate in rocket dockets. Infringers want to slow it down. This is a phenomenon as old as the hills, and really has nothing in particular to do with NPEs.

    To stay in rocket dockets, POs have no choice but to name not only manufacturers, but OEMs and VARs.

    The panalists also noted that adding numbers of defendants tends to drive up defendant costs as the defendants seek to operate under JDAs. But this requires that every issue be discussed in large conference calls among large numbers of attorneys leading to an expensive mess.

    Most litigators like to talk, and few are concise.

  28. These graphs all displaced upwards by 100%.

    No, the axis is just labeled wrong. It’s common to scale to 100 in the reference year (look at the CPI, for example) and the resulting graph is much more intuitive than if you’d put the reference year at zero.

  29. Interesting article and more evidence that NPEs in possession of crxp patents that should never have been granted are a growing problem.

    But of course the lawyers and the NPEs love it, hence ping’s pathetic attempts to find serious fault in the report.

  30. Winston’s comment is interesting, and I think the spike may also be attributed to E-Bay, which changed the NPE gameplan. They now sue large numbers of defendants and collect nuisance value settlements as opposed to trying to hit the big litigation payday.

  31. ping is absolutely right. It is not possible to have a 100% increase in the reference year.

    To get the % change you take the difference between the number of cases in the reference year, 1990, and the year you are interested in and divide that difference by the number of cases in 1990. If the year you are interested in is 1990 itself, then the difference between the number of cases in 1990 and the number of cases in 1990 is 0. Divide 0 by the number of cases in 1990 and you get 0% increase.

    These graphs all displaced upwards by 100%.

    Someone ought to take the calculators away from lawyers who don’t know what they’re doing.

  32. Nah Cy,

    But you did say I was off my game, when clearly I wasn’t – I wouldn’t necessarily call that defensive (more like correctin your mistake is all).

    And perhaps you be too defensive – after all, I did give your man Kyle props. This be some good data as it really clears the dust cloud from some of the windmill chasers.

  33. Max Rev and Min cost is the be all and end all. It is universal. “Getting there” is of course different, but that aint really what I said, now is it?

    Nope. Which is why I didn’t argue with what you said on this topic – I just added a comment. A little defensive today, aren’t we?

  34. btw Cy – the defender cost quibble is with the word “extra” – in case ya missed that

    I didn’t miss it, but thanks. That “extra” reflects the additional (roughly estimated) defense costs today, compared to Kyle’s reference point, which is before 400 “extra” defendants were sued each year. Kyle estimates that $400M more is being spent today than before, because of the extra defendants. Kyle also presents data showing a large increase in defense attorneys, which generally supports his estimate. Kyle is not making the “somehow-the-first-defender-foots-the-bill-for everyone” mistake. He quite clearly says “If the number of defendants is growing, costs are certainly growing as well.”

    The confusion arises from Kyle’s use of “would be” rather than “are,” which would have been clearer. Kyle is incorrectly using a conditional progressive tense. But astute readers like yourself should be able to get past that.

    Criticize me all you want, but go easy on my man Kyle.

  35. If indeed the activity is “roughly constant over the past decade,” the percentage change would hoover around 0%. A 100% change is a doubling of the activity.

    It’s not a graph of change year over year. It’s a graph of the level relative to 1990. So in 2002 the number of cases was at ~300% of the 1990 level and it stayed about the same through today.

    Anyway, the sharp drop off in the number of attorneys since 2007 is interesting. The early decline (2007-2008) could be explained by the corresponding drop in the number of defendants, which suggests less complex cases, but after that the number of defendants bounces back and the number of attorneys continues to decline sharply.

    I guess it’s more evidence of litigants controlling costs in the recession. Using the number of attorneys as the sole metric (and thus greatly oversimplifying things), this suggests litigants are spending roughly 2/3rds as much per case today as they were in 2007. Litigators, does that sound about right? Or are other factors making up for the drop in the number of attorneys?

  36. There is a spike in ’07, so good ol Winston may have a point about a contribution, but the trend was rising anyways.

    Further, if ya take a look at number of attorneys divided by number of defendants and normalize to 1990, the “attorney efficiency factor” is actually less than one, and has been trending down sharply since 2005

  37. The big rise in multi-defendant patent litigation starting around 2007 stems from the 5th Circuit’s Volkswagen decision, which allowed mandamus review of denied transfer motions in the Eastern District of Texas. Adding more defendants makes finding a single other forum that is “clearly more convenient” more difficult, so plaintiffs that want to litigate in EDTX add more defendants.

  38. Kyle,

    It also might be interesting to see the data normed by actual patents “alive” per year. You might actually see a decline in litigation per patent which would further kick the “popular media” in the teeth.

  39. Cy – Ya think I’m off my game?

    PACER has unlimited capacity.

    PACER’s capacity aint at issue – Ya do know what PACER is, dontcha?

    Max Rev and Min cost is the be all and end all. It is universal. “Getting there” is of course different, but that aint really what I said, now is it?

    Pay attention and re-read: the somehow-the-first-named-defender-foots-the-bill-for-everyone mistake.

    Cy – don’t be sayin someone else is off their game in such slipshod way – ya might draw people ta think that ya be a Malcolm sockpuppet. Cynical people at that.

  40. No. This is an errant conclusion based on the sloppy premise that the “extra” defenders somehow don’t need to spend any additional money defending themselves (the somehow-the-first-named-defender-foots-the-bill-for-everyone mistake.

    Kyle didn’t use that “sloppy premise.” His data shows 400 more defendants a year. He says this suggests $400M more in defense costs.

    You’re off your game this morning, ping.

  41. I think there’s no better reason than maximizing revenue and minimizing costs for everybody than trying to insert some sort of NPE factor when such is just not supported.

    Obviously maximizing financial return from a lawsuit is an important objective, but it’s not the only one. NPEs generally don’t have to work with or against the defendants after the case is over/settled. That’s a key difference that might suggest different behaviors. The notion of “repeated games” in game theory might come into play here.

  42. Anonymous – the easy answer, scratch that – easy observation is “NO.” PACER court filings are impervious to any bottle-neck in the courts. Now court resolutions might be affected by such bottle-necks, but not the filings.

    Uh, ping, that answer might have been too easy. I believe Anonymous was suggesting that the plaintiffs’ decisions might be influenced by (perceived?) overloading of the courts. The answer to his question might still be no, but I don’t think we can explain that answer by simply noting that PACER has unlimited capacity.

  43. 1) maximize potential revenue and
    2) decrease costs for an NPE or contingency counsel.

    I can’t think of many reasons that a practicing entity would behave similarly.

    Dontcha mean “wouldn’t behave similarly”? Aren’t those two reason enough for anybody? It seems a reach to differentiate between practicing and non-practicing on this matter – I think there’s no better reason than maximizing revenue and minimizing costs for everybody than trying to insert some sort of NPE factor when such is just not supported.

    If the number of defendants is growing, costs are certainly growing as well

    In other news, water is still wet. Yes it still takes money to defend yourself in court. No real surprise or “news” here.

    defendants would be spending an extra $400MM/yr on litigation each year.

    No. This is an errant conclusion based on the sloppy premise that the “extra” defenders somehow don’t need to spend any additional money defending themselves (the somehow-the-first-named-defender-foots-the-bill-for-everyone mistake).

    Where’s that money go? Certainly attorneys capture some of it.

    True – and misleading at the same time. As these cases become more complex with multiple named defendants, the need for attorney teams grow. What might be more interesting would be a graph of attorney billings, or net attorney rates (billing per effort).

    Still, nice job there Kyle – Kudos!

  44. So much for the “explosion” in patent litigation, or, for that matter, the “chaos at the patent office” Life goes on.

  45. The graph labels are mismarked.

    If indeed the activity is “roughly constant over the past decade,” the percentage change would hoover around 0%. A 100% change is a doubling of the activity. Unless of course, you meant that teh rate of lawsuit change is fairly constant – but that be quite different than “number of patent lawsuits filed annually.”

    Anonymous – the easy answer, scratch that – easy observation is “NO.” PACER court filings are impervious to any bottle-neck in the courts. Now court resolutions might be affected by such bottle-necks, but not the filings.

  46. This “multi-defendant” stuff. Is it a phenomenon peculiar to the USA? Is it “socially useful”? If so, will we see the same trend in the rest of the world, five years from now? If not, why not?

  47. Do we have a bottle-neck in the courts so that’s why we don’t see an increase in the number of litigations? Instead you put your effort on multi-defendant cases.

Comments are closed.