Investigating Patent Law’s Presumption of Validity—An Empirical Analysis

Etan Chatlynne reports post-KSR patent validity decisions with a special focus on the presumption of validity and the clear and convincing standard for rebutting that presumption. He finds that the presumption and evidentiary standard did not affect thirty-eight of forty-five (84%) Federal Circuit invalidity determinations made between April 2008 and June 2009. The standards may have affected the remaining seven of forty-five (16%) determinations. [Download Chatlynne.PresumptionofValidity.Final]

Etan Chatlynne, Investigating Patent Law’s Presumption of Validity—An Empirical Analysis, 2010 Patently-O Patent L.J. 37.

53 thoughts on “Investigating Patent Law’s Presumption of Validity—An Empirical Analysis

  1. 52

    He doesn’t let facts or reality get in the way of pushing his snarky anti-patent agenda.

    I’m a big fan of valid patents. I think invalid patents should go die in a fire, even if it turns out the PTO allowed them at some previous time.

    If that makes me anti-patent, call me anti-patent.

    Oook.

  2. 51

    “I’m trying to understand the problem you have here,”

    INANE, er, Mooney’s problem is pretty obvious to most of us around here. He doesn’t let facts or reality get in the way of pushing his snarky anti-patent agenda.

  3. 50

    Even the cited study implies that 15% of invalidity determinations are somewhere between “too close to call but leaning toward defendant” and “the defendant has a pretty convincing case”.

    Read it again. The study simply does not say this. If 45 people die today, and I can only say for certain that 38 were not run over by stampeding elephants, that does not mean that 7 poeple were killed by stampeding elephants. Here, the 7 cases you’re clinging to as some sort of error may or may not have depended on a clear and convincing standard, and 6 of those 7 were jury cases where any weighing of facts is opaque.

    I’m trying to understand the problem you have here, but you still haven’t provided a hypothetical where the clear and convincing standard would render a claim valid but a prepondreance standard would render a claim anticipated. Without some substance to your concern, I really can’t agree that there’s a problem.

    Yes, any case where one side thinks it has no chance should settle. But that depends on the side that has no chance actually thinking that it has no chance, and also there are sometimes business or emotional reasons to pursue a case even when you have no legal chance. Besides which, you never really have no chance, and you’ll never get a litigation opinion that says you have no chance.

    Please forgive my simplification. The vast majority of patent cases (and parties in multi-party patent cases) settle prior to dispositive motions or trial. Over the last 10 years, 75% of cases terminated by all parties settling or accepting a consent judgment while only 14% terminated with a decision on the merits (and of those 14%, many parties settled prior to such a decision terminating the case for their codefendants). This only takes into account those cases where litigation was even necessary. There are, no doubt, an irrational few who throw away millions of dollars hoping for a sympathetic jury, but such bad gambles are the exception rather than the rule.

  4. 49

    “too close to call but leaning toward defendant” and “the defendant has a pretty convincing case”.

    This is the most accurate summary you’ve made so far regarding the study. Still, your position seems to be that all 7 of the cases where the presumption matters should have been won by the defendant. The study simply does not analyze the cases to that level of detail.

  5. 48

    As for being close cases, any case where one side thinks it has no chance should settle. The cases we see in decisions are those where both sides thought they could win based on different interpretations of fact or law. To the extent there is a “right” answer in such cases, we should be pleased with a judiciary that succeeds more than 50% of the time.

    Yes, any case where one side thinks it has no chance should settle. But that depends on the side that has no chance actually thinking that it has no chance, and also there are sometimes business or emotional reasons to pursue a case even when you have no legal chance. Besides which, you never really have no chance, and you’ll never get a litigation opinion that says you have no chance. You always have at least the chance of being the more sympathetic party on the day, and convincing the judge/jury that you ought to win.

    I also take issue with your implication that cases are either clear-cut (one party should settle) or too close to call. Courts are called upon to decide cases all the time that fall elsewhere along that spectrum. Even the cited study implies that 15% of invalidity determinations are somewhere between “too close to call but leaning toward defendant” and “the defendant has a pretty convincing case”.

    Finally, we should certainly not settle for a judiciary that “succeeds” barely more than 50% of the time. Most cases do lean one way even when they’re not clear-cut, and the courts should strive to be as predictable as possible. That way, more people will know when they should settle.

  6. 47

    This isn’t an 85% success rate on close cases. This is a statement that 15% of all cases are close cases that we’re getting wrong. 15% are cases where patents were upheld when they arguably would have been invalidated according to the usual civil standard of “the plaintiff probably has no valid right to enforce”.

    Reread the article. It’s not saying that. It’s saying that 85% of cases are not decided on the clear and convincing standard, and that the remaining 15% may or may not have been decided on the clear and convincing standard. The review found no cases that were definitely decided on the clear and convincing standard.

    I’d still like to know a fact pattern where the standard matters, though. The defense will ultimately present adequate prior art to anticipate the claims or it won’t.

    As for being close cases, any case where one side thinks it has no chance should settle. The cases we see in decisions are those where both sides thought they could win based on different interpretations of fact or law. To the extent there is a “right” answer in such cases, we should be pleased with a judiciary that succeeds more than 50% of the time. An 85% success rate would be phenomenal, though, again, this article isn’t claiming that.

  7. 46

    “Yes, and your business is the most important thing ever. We must never forget that.”

    Right. I supposed we could all abandon our business pursuits, walk around in Birkenstocks and live in mud huts.

    Now THAT sounds like progress. You really shed some light on your true philosophies Mooney. You really put the “progress” in “progressive” for me.

  8. 45

    “Because the royalties were lower than the expected costs of litigation …?”

    Well I guess we can change “probably invalid” to “that I wish was invalid” then, otherwise sounds like a sound business decision to me.

  9. 44

    I don’t understand the “Sigh.”

    Is my question redundant? Have you answered this before?

  10. 43

    Cases that actually reach this point are generally close regardless, and an 85% “success” rate on close cases would be phenomenal.

    This isn’t an 85% success rate on close cases. This is a statement that 15% of all cases are close cases that we’re getting wrong. 15% are cases where patents were upheld when they arguably would have been invalidated according to the usual civil standard of “the plaintiff probably has no valid right to enforce”.

    Granted, it’s a small data set, but it’s not at all a reassuring percentage.

    IANAE, are you in academia?

    Sigh. Here we go again.

  11. 41

    PB: Having my hard earned patent subject to reduced burden challenges does not sound like a great thing for my business.

    Yes, and your business is the most important thing ever. We must never forget that.

  12. 40

    Why would anyone in their right mind pay royalties on a patent that was “probably” invalid?

    Because the royalties were lower than the expected costs of litigation …?

  13. 39

    ProBusiness: As the article suggests, the standard of proof makes an outcome determinative difference only around 15% of the time.

    That’s a misreading of the article. The 7 of 45 cases referred to are the cases where they could not exclude the standard as a factor in the determination of invalidity, and 6 of those 7 were jury cases. The article identified no cases where the standard was certainly a factor in the invalidity determination.

    IANAE: One patent out of every six being upheld when it should be invalidated is a shockingly large proportion.

    Cases that actually reach this point are generally close regardless, and an 85% “success” rate on close cases would be phenomenal.

    That being said, can you produce a hypothetical fact pattern where a piece of prior art would anticipate a patent under the preponderance standard but not under the clear and convincing standard? I’m at a loss as to how a jury or finder of fact would make different interpretations under the different standards on an anticipation issue. I suspect, as the linked study shows, that the standard is not terribly important.

  14. 38

    A tautology nobody can disagree with.

    Nobody can disagree with it, except a judge. Because that tautology is inconsistent with the standard of proof for invalidating an issued patent.

  15. 37

    “As the article suggests, the standard of proof makes an outcome determinative difference only around 15% of the time.”

    “One patent out of every six being upheld when it should be invalidated is a shockingly large proportion.”

    Neither of these conclusions can be drawn from the article which merely indicates where the presumption may have affected the result. Surely the “right” result was reached in some of those cases.

    “If your hard-earned patent is anticipated or obvious, it should be invalidated. ”

    A tautology nobody can disagree with. We might also say, courts should not invalidate patentable claims or that claims should not be invalidated because they might be obvious.

  16. 36

    Why would anyone in their right mind pay royalties on a patent that was “probably” invalid?

    Because a judge tells you to. Because “probably invalid” isn’t enough evidence to invalidate a patent in court. That’s what “clear and convincing” means.

    Or didn’t you pay attention in evidence class?

    As the article suggests, the standard of proof makes an outcome determinative difference only around 15% of the time.

    Even ignoring the fundamental flaw in the article that Paul Morgan identified in the very first post, One patent out of every six being upheld when it should be invalidated is a shockingly large proportion.

    Having my hard earned patent subject to reduced burden challenges does not sound like a great thing for my business.

    If your hard-earned patent is anticipated or obvious, it should be invalidated. That’s a great thing for everybody’s business.

  17. 35

    “What I don’t understand is why you think it’s a good idea to have probably-invalid patents commanding royalties and injunctions against people who are practicing the prior art.”

    Why would anyone in their right mind pay royalties on a patent that was “probably” invalid? They would do many things before they did that. As the article suggests, the standard of proof makes an outcome determinative difference only around 15% of the time.

    Sounds like the presumption is working just fine to me…

    Having my hard earned patent subject to reduced burden challenges does not sound like a great thing for my business.

    I’m guess if you’re somehow for this, you probably also fancy taking two or three of the lug nuts off your car’s wheels just for sport.

  18. 33

    the rest of us who paid attention in Admin Law and understand why there is a higher presumption accorded to agency decision making

    I understand the concept that the agency decision is supposed to have a certain degree of finality, because that’s the whole point of delegating the function to someone other than the court. If you’ve been following some of the other threads, you already know my opinion on free do-overs in court.

    What I don’t understand is why you think it’s a good idea to have probably-invalid patents commanding royalties and injunctions against people who are practicing the prior art.

    I also don’t understand why you’re terrified of even the tiniest change to the patent system, when that change would be clearly pro-business.

    Oook.

  19. 32

    Chuckle away Mooney,

    Meanwhile the rest of us who paid attention in Admin Law and understand why there is a higher presumption accorded to agency decision making will watch you blabber on as usual – like a smelly baboon.

  20. 31

    But I guess you wouldn’t understand something like “solid principles.”

    I understand that you’re working very hard to avoid discussing this particular principle. Patent principles don’t get much more solid than “people shouldn’t get patents on things that already existed”.

    If you think the patent system is so tenuous that subjecting it to the same standard of proof as every other civil litigation in the country will bring the whole thing crashing down, come out and say it so we can all have a good chuckle.

  21. 30

    “It means “oppose a reasonable change by scaring people about unlikely future unreasonable changes”, right?”

    Sounds like something Mooney would say…

    What it means is, when you compromise on solid principles, sooner or later everything comes tumbling down. But I guess you wouldn’t understand something like “solid principles.”

  22. 29

    INANE never learned the meaning of “slippery slope”

    It means “oppose a reasonable change by scaring people about unlikely future unreasonable changes”, right?

    Seriously, explain why I should have to pay royalties on a patent that is more likely than not to be anticipated. Explain how it aids in the progress of innovation to allow people to patent the prior art. Because there’s a slippery slope in that direction too.

  23. 28

    “This is a real benefit of the patent system that aids in the progress of innovation.”

    It signals to companies what areas NOT to enter and, in a strong regime, signals what areas may emerge as areas that provide support for primary areas of innovation.

    Example: a company reveals, through its portfolio, that it is building up in the area of television display technology would signal entrants to select something like, say, advanced tuners or antennas.

  24. 26

    One of the great benefits of patents that is missed by all the academics that I know of is that the patents are actually used to understand what a company is doing by other companies.

    This is a real benefit of the patent system that aids in the progress of innovation.

  25. 25

    Beware: attempting to kill the presumption of validity is what the anti-patent crowd views as a key to deconstructing our strong system.

    Considering that it’s logically impossible to prove a patent valid, I don’t see how anyone could reasonably want to “kill the presumption of validity”.

    What people are saying is that the evidentiary standard is too high. So high, in fact, that a patentee can obtain damages and an injunction for a patent that is probably invalid. Does the patent system need to be that strong?

  26. 24

    Here we go, more anti-patent subterfuge.

    Beware: attempting to kill the presumption of validity is what the anti-patent crowd views as a key to deconstructing our strong system.

  27. 23

    ESC: When I first decided to perform an empirical analysis, I wanted to collect jury instructions and correlate them with invalidity determinations. Comparing pre-1982 jury determinations (when the regional circuits permitted preponderance of evidence instructions) to more recent determinations (clear and convincing only) would have been very useful. Alas, my meager student budget does not allow for docket runners.

    Lex Machina (www.lexmachina.com) is a great source of data over the last 10 years, and you should be able to get free access as an academic.

  28. 21

    For a valid study of the effect of the CAFC “clear and convincing evidence” standard one would also have to exclude all cases since the Sup. Ct. KSR decision, which I would think has had far more effect on 103 invalidity decisions, both in district court summary judgements and in Fed. Cir. appeals?
    I also wonder if the “clear and convincing evidence” requirement might be behind several CAFC decisions that seem to have taken extreme non-infringment positions [making that other law more unpredictable] in cases of particularly obvious looking claims on appeal, where the CAFC could not, or did not want to, review the [usual] jury verdict against invalidity?

  29. 19

    That reference above, to finding facts as they see fit, reminded me of the Comaper case (see Patent Prospector today) in which the jury did not find claim 1 old or obvious, yet did find obvious the narrower claim that depended from it. Such nuances as “preponderance” or “clear and convincing” are (apparently) altogether too much for most jurymen, when asked to second guess the experts at the USPTO, and declare that a claim embraces within its ambit some product, process or apparatus that would have been obvious, to an imaginary hypothetical person, ten years earlier.

  30. 18

    judesf: it assumes that failure of a decision to discuss the clear and convincing standard explicitly is taken as an indication that it played no role.

    The frequency with which the court discussed the evidentiary standard was not a driving factor in determining whether an invalidity determination may have been different with a preponderance of evidence standard. Rather, I collected this data based on Professors Lichtman’s and Lemley’s suggestion that finders of fact sometimes find facts as they see fit, regardless of the presumption of validity. For more discussion on this, please refer to my Note at p. 316-318 (which includes a discussion of a Federal Circuit case where this might have happened). link to cardozolawreview.com.

    judef: the standard is likely more of an issue with jury trials

    When I first decided to perform an empirical analysis, I wanted to collect jury instructions and correlate them with invalidity determinations. Comparing pre-1982 jury determinations (when the regional circuits permitted preponderance of evidence instructions) to more recent determinations (clear and convincing only) would have been very useful. Alas, my meager student budget does not allow for docket runners. So, instead of analyzing the direct affect of the jury instruction on the invalidity determination, I investigated the degree of change–across all cases–we might expect should the evidentiary standard be lowered to a preponderance of evidence. Surely, a given defendant that wants a lower standard will not be satisfied with this. But it might be useful for balancing policy decisions. For example, a clear and convincing standard may encourage use of reexamination and possibly forthcoming post-grant review; a preponderance standard may discourage using these alternatives to litigation.

    Microsoft had asked the Federal Circuit to allow preponderance of evidence instructions in its en banc petition last November. It did not include this question in its recent cert petition. I wonder why. I think it had a good shot at winning on this issue. If and when the standard is lowered, the opportunity will present itself to “inexpensively” compare jury verdicts reached based on both evidentiary standards. And at that point, if all goes according to plan, I won’t be on a student budget anymore either.

    Thank you for reading my essay.

  31. 17

    Agreeing with “juddef,” just because a CAFC panel does no repeat the words “clear and convincing evidence” in its decision does not mean that they do not know they are stuck with it by stare decisis and should apply it.

  32. 16

    One beef with this article’s analysis is that it assumes that failure of a decision to discuss the clear and convincing standard explicitly is taken as an indication that it played no role. While that might be true, I can see where courts that are familiar with the standard might not discuss the standard, yet take it into account in favoring the patent’s validity. Also, I agree with the earlier comment that the standard is likely more of an issue with jury trials. So, it’s an interesting analysis, but I’m not convinced they’re right in their conclusions.

  33. 15

    If done modestly well, it reduces the number of crxppy patents.

    Also, reduces the number of lawsuits to assert patents that have been drafted and registered without even glancing at the prior art. There are enough patent applications out there that it’s much more efficient for them to get examined by a dedicated (in that it’s their only task) corps of civil servants than by defendants.

  34. 14

    As I tried to say earlier, if the fault is with an insufficient examination system, then you either need a “sufficient” one that supports the presumption and heightened standard, or you drop the figleaf entirely, and just register the patent.

    Why throw away examination entirely, just because it’s not (and may never be) thorough enough to support a clear-and-convincing standard?

    Even as it currently stands, examination is good enough to support a bare presumption of validity. The patentee needs such a presumption anyway, because it’s impossible for him to prove his patent valid.

    Examination also gives you a chance to see the examiner’s search results and adapt your claims to the cited art before they issue and you’re stuck with them.

    Let’s try make the examination system work as well as it can, and quit holding it to the unrealistic standard of underpinning a system based on clear-and-convincing.

  35. 13

    LH As I tried to say earlier, if the fault is with an insufficient examination system, then you either need a “sufficient” one that supports the presumption and heightened standard, or you drop the figleaf entirely, and just register the patent.

    I dunno, seems like there are more options than that.

    what’s the point of the time and expense of an examination system

    If done modestly well, it reduces the number of crxppy patents.

    James: Malcolm, one of Kieff’s primary papers on the subject of a registration system is “The Case for Registering Patents and the Law and Economics of Present Patent-Obtaining Rules,” which is available on SSRN here

    Thanks. That might go down well tomorrow am with some coffee.

  36. 12

    As I tried to say earlier, if the fault is with an insufficient examination system, then you either need a “sufficient” one that supports the presumption and heightened standard, or you drop the figleaf entirely, and just register the patent. In the present system, there are few, if any issues, that are not entirely contested from scratch in litigation (in addition to inter parte reexamination). If you’re going to do a better prior art search when the stakes are high, and you’re going to have to argue inequitable conduct (which will not arise in a registration system) and the conclusions of the PTO do not bind the Court (not even those of and appeals board) then what’s the point of the time and expense of an examination system? Just have the applicant submit an application that meets all statutory requirements, collect a fee and publish it ASAP. Save the fighting for when it really matters — litigation over infringement. Want to really make it work? Enact the British rule: loser pays.

  37. 11

    Malcolm, one of Kieff’s primary papers on the subject of a registration system is “The Case for Registering Patents and the Law and Economics of Present Patent-Obtaining Rules,” which is available on SSRN here

    link to papers.ssrn.com

  38. 10

    DC (1) In the past, Scott Kieff (GWU) has proposed switching to a registration system.

    Thanks, I’ll have to check that out.

    (2) I actually believe that a *design patent* registration system would represent an improvement over the status quo.

    I sympathize with that view, mainly because the status quo is a de facto registration system in cheap clothing.

  39. 9

    one seemingly can’t presume that the Office performed a thorough search.

    Presumptions don’t have to be 100% accurate, they’re just the starting point when you get to court. If a patent is to be presumed valid, and I can’t imagine the system working otherwise, we have to assume that the search was adequate to justify that presumption. If there’s a systemic problem with examiner searches, that’s where the problem needs to be fixed.

    Also, even a thorough search might not find everything, which is why we even allow the infringer to challenge the issued patent under 102/103, and why it’s possible to get a do-over from the PTO on an already-issued patent. You always find more art on a litigation budget than on a prosecution budget.

    Art not on the cover might have been found in the examiner’s search but (wrongly) dismissed as irrelevant. You can’t be sure that the PTO never considered a reference. You can be sure the PTO never properly considered the reference if it’s not on the cover and it’s clearly relevant, but that’s a whole other issue.

    My real concern is whether the clear and convincing standard is proper at all. There’s nothing wrong with having both a presumption and the normal civil standard of proof. It would also mean that judges couldn’t come up with rulings like “the patent is probably invalid over the art, but not probably enough, so you have to pay”. Since when do we stand for civil liability in cases where the plaintiff more likely than not has no valid legal right to assert?

  40. 8

    Why not a non-exam but non-register system? You submit the application, the Searcher (Examiner) searches the art and provides you with references. You can choose to get your patent, abandon, or amend and take the amended patent. The references are listed on the face of the patent as today.

    It’s not a simple registration system, and it’s not a strong presumption of validity from a flawed examination process.

  41. 7

    So I just received an Office Action rejecting, over a newly-discovered reference, a set of claims that the Examiner had previously indicated were free of the art. In my response to the previous Office Action, I had canceled all other claims, and expected an allowance (no other bases of rejection). I did not amend the claims in the response.

    Apparently, the Examiner re-serched the claims and found the new reference. Unfair, because the Examiner should have found and considered that reference the first time around, but so it goes.

    However, what if the Examiner had simply allowed the claims? That piece of art would be out there for a potential litigant to use. I can’t imagine that my application is the only one to have been so handled, which seems to me to be a good argument for NOT affording a patent a clear and convincing evidentiary standard for art not considered. After all, the Office wouldn’t have considered the art, and one seemingly can’t presume that the Office performed a thorough search.

  42. 6

    The US tried a registration system for awhile and decided it was a complete disaster.

  43. 5

    Malcom Mooney: Are you aware of any person who actually (or apparently) believes that a utility patent registration system in the US would represent an improvement over the status quo?

    (1) In the past, Scott Kieff (GWU) has proposed switching to a registration system.
    (2) I actually believe that a *design patent* registration system would represent an improvement over the status quo.

  44. 4

    LH: If, as proponents of the change in standard say, “patents are insufficiently examined” the cure is to just move to a registration system since all matters are challenged de novo in civil litigation anyway.

    Yes, killing the patient will cure the disease. Great idea, Larry!

    Are you aware of any person who actually (or apparently) believes that a utility patent registration system in the US would represent an improvement over the status quo?

  45. 3

    the cure proposed by Lawrence Husick is akin to bury the patient because he is sick.

    No thank you.

  46. 2

    If, as proponents of the change in standard say, “patents are insufficiently examined” the cure is to just move to a registration system since all matters are challenged de novo in civil litigation anyway.

  47. 1

    The CAFC is not where the CAFC-added “clear and convincing evidence” requirement is a major problem. The real issue is the influence on JURIES, in jury instructions and attorney remarks, of that high standard imposed on top of the statutory standard of presumed validity.

Comments are closed.