Patent Pendency as a Function of Claim Count

This is an updated version of a chart that I made a few years ago — this time using utility patents issued 1/2009 through 11/2011. To create the chart, I grouped patents according to the number of claims in the issued patent and then calculated the average number of claims for each group. As usual, the 20-count claims is an anomaly.

38 thoughts on “Patent Pendency as a Function of Claim Count

  1. 37

    I wonder whether the number of independent claims vs. the number of dependent claims matters. Independent claims represent new inventions. If an independent claim is allowable, then all claims depending from it must also be allowable. So, if it is to be a first office action allowance, a single independent claim with several dependent claims should go through faster.

  2. 36

    Newman, then, may have been more on point if she had argued that in deciding whether there is a “substantial new question” of patentability, the Office must whether the issue asserted was raised and decided in court.

    Agreed. It may be the case that the Office already does this. I’m somewhat surprised the issue hasn’t come up already.

  3. 34

    Newman, then, may have been more on point if she had argued that in deciding whether there is a “substantial new question” of patentability, the Office must whether the issue asserted was raised and decided in court.

    Now, if new art raises a substantially different issue, not decided by a court, I will agree that the PTO not only has the legal jurisdiction to re-examine under the statutes, but they should do so.

  4. 32

    “The drop is because examiner’s subconsciously see 20 claims as the “norm”. See a case with 18 claims and an examiner will find themselves thinking, “look at this arse, he could have just put 15 and it would have been fine, NEXT”. But at 20 you’re just a run of the mill ar se.”

    Good luck trying to follow the “logic” of this.

  5. 31

    I concur that 6 can neither get his thoughts well enough in order or come to a decent conclusion.

  6. 29

    The 20 claim dip could be explained with restriction requirements. Per the graph, total prosecution time for less than 20 claims approximately equals that for 20 claims, plus an extra 2 months or so. A typical restriction/response cycle takes about 2 months or so.

  7. 26

    Dennis – perhaps someone mentioned this above, but it would possibly be more interesting to see the relationshiop using the number of claims INITIALLY examined – esp in light of your hypothesis that 20 claims allowed implies no amd’s were required (a good hypothesis).

  8. 24

    There’s also a factor of delay because examiners typically have a few dozen new applications docketed to them at any one time. When examiners need easy counts, they generally cherry pick their dockets for applications with fewer claims, so time to first action – and therefore likely the time to issuance – is shorter for those cases.

  9. 21

    it is a non-issue

    Agreed. Please let’s stick to the much more fascinating topic of the correlation between the number of issued claims and application pendency. 😉

  10. 20

    Malcolm, true. But the argument she raises is very much like the 7th Amendment, which states, in material part,

    “no fact tried by a jury, shall be otherwise re-examined in any Court of the United States, than according to the rules of the common law.”

    But that’s exactly what I was saying Ned (“one could cobble together some argument that it’s unconstitutional to relitigate the exact issue at the PTO which was previously decided by the court.”).

    And that’s actually a fairly boring proposition.

    The idea that once a court decides a patent is valid in view of prior art B that it’s also valid in view of previously unknown prior art refs M,N,O,P and Q (not to mention a change in the law) until another court says otherwise is, quite frankly, looney-tunes.

  11. 19

    “If the question is squarely presented, do you think that the Supreme Court would limit the 7th Amendment to just the courts, and allow Federal Agencies to “reexamine” facts found by a Court?”

    I don’t know about MM, but I doubt it, although on the other hand, there is no evidence of that having taken place in this case so it is a non-issue as of this moment.

  12. 18

    As much as I might like to agree with Newman in that case, I’m afraid that her colleagues are correct in that if her rule were to be in place then the entirety of reexamination proceedings would get boinked at least in so far as they involve a case that went through courts. That is not the law congress wrote.

    She reminds me of NAL with her res judicata and issue preclusion arguments, indeed, we can tell whose side Newman would be on if it were the finality of Taffas at issue if the office made a new slightly different attempt to make rules. On the other hand, there would be all the administrative lawl lawlyers and probably the rest of the CAFC.

    If it be the will of the congress that the office have procedural rulemaking authority and the ability to continue to propose rules and also that cases should be available for reexamination at any time then so be it.

    As tempting as it is to have courts “finally” decide an issue for all time, it does not seem to me that taking into account all the things that may happen in the future that this would be a wise course of action in some circumstances.

  13. 17

    Malcolm, true. But the argument she raises is very much like the 7th Amendment, which states, in material part,

    “no fact tried by a jury, shall be otherwise re-examined in any Court of the United States, than according to the rules of the common law.”

    If the question is squarely presented, do you think that the Supreme Court would limit the 7th Amendment to just the courts, and allow Federal Agencies to “reexamine” facts found by a Court? But isn’t that what the PTO does in fact when it takes jurisdiction of a reexamination involving art that was before a jury?

    Newman has simply extended the 7th Amendment to the Judicial Power of an Article III court. Once they have decided an issue, it is decided and the executive branch cannot reexamine it and overturn the court.

  14. 16

    Actually that is probably it, I was trying to think along those lines but I couldn’t get my thoughts well enough in order to come to a decent conclusion.

  15. 15

    The drop is because examiner’s subconsciously see 20 claims as the “norm”. See a case with 18 claims and an examiner will find themselves thinking, “look at this arse, he could have just put 15 and it would have been fine, NEXT”. But at 20 you’re just a run of the mill ar se.

    That and it is possible that most of the “good” attorneys more often than not use 20 claims and they also happen to be good at getting that sh t allowed and done.

  16. 14

    She argued that it was unconstitutional for the PTO to reexamine a claim held valid by the court.

    I suppose one could cobble together some argument that it’s unconstitutional to relitigate the exact issue at the PTO which was previously decided by the court.

    But I’m not aware of the CAFC or any other court ever holding that a given patent was “valid forever, regardless of what new prior art is discovered and regardless of any changes in the law.”

  17. 13

    Works well here (Firefox 8.0). Hovering over a data point brings up the actual numbers. Neat feature.

    Interesting slope change going from <20 to >20, somewhat reminiscent of the appeal backlog graph though unrelated. One odd thing about the above graph is that there are so many data points >20 in the first place. Perhaps applicants should be reminded (informed?) that the extra claims fees do not actually buy extra time (“other time”) for the examiner to digest the extra claims, as a rational person might assume. The pxtxnt office is spending that on something else (same goes for the jumbo disclosure fees). An application with 83 claims is not likely to receive any more attention from an examiner than one with 15 claims. Or was that the point all along? A little extra dough up front and a slight increase in pendency, but patent virtually assured after the examiner gets tired of looking at it…

  18. 12

    Guys, we have a blockbuster dissent today by Newman in IN RE CONSTRUCTION EQUIPMENT COMPANY
    link to cafc.uscourts.gov

    She argued that it was unconstitutional for the PTO to reexamine a claim held valid by the court.

    She somewhat argued the particular status of the case where the very defendant that had lost in court had sought reexamination.

    She may, by this argument, have left open whether her logic would extend to third party requesters. Arguably, it might, as the only reason third parties have a right to relitigate issues decided by an earlier court is their own constitutional right to be heard — in court. Nothing grants the government a right to overturn a court decision. Nothing grants a third party a constitutional right to a reexamination.

  19. 11

    Great point. There just might be enough accelerated cases with 20 claims issuing rapidly (<1 year) to drag the average down a bit.

  20. 9

    Might the 20-claim drop also reflect the fact that accelerated applications are capped at 20 claims?

  21. 8

    Is there any way to further refine the data to a count of total pending independent claims? It seems like that information would be more useful. I’d be curious to see if there is a certain number of independent claims after which the average pendency increases significantly (I’m guessing 3 or 4 would be that number).

  22. 7

    I think that the 20-claim artifact is entirely explained by the fact that 20-claims is the most common number of claims in a patent application. Patents that issue with 20 claims are likely to have issued without substantive amendments (since most likely no claims were added or deleted) and that means that they issue more quickly. Patents that issue with 19 claims likely began as 20-claim applications that were then faced with a solid rejection that required amendment and argument.

    To be clear, this is just my educated guess. I’ll need to run some more numbers to see if it is true.

  23. 6

    Uh … my 4:54 comment is addressing a different graph that exists only in my imagination. DC’s graph concerns the claim count at issuance, not the claim count initially presented.

    Never mind …

    /Emily Lutella off

  24. 5

    It is odd, isn’t it? At least, it’s counter-intuitive that patents with 20 claims issue faster, on average, than patents with, e.g., 10 claims.

    I’m certain that you are correct about the relationship to the filing fee. I believe that because of the way that fees for claims are calculated, a greater proportion of cases with 20 claims are filed compared to any other number of claims. And there’s something *else* about that large portion of cases with 20 claims that is causing those claims to issue a bit faster than cases with lower numbers of claims. It could be that most of those 20 claim cases are first-filed cases and therefore the applicants are responding quickly to the Office Actions and the issues are easier to resolve, whereas many cases with fewer claims are continuations where the applicant is trying to find some scrap of subject matter that is less important (and less patent-worthy?) than the original case, causing prosecution to slow down.

    Just a guess. 😉

  25. 4

    Looking good here too.

    I wonder how many of those 20-or-fewer claim cases had election requirements that cut down on claim count.

  26. 3

    Can anyone figure out why there is the drop at 20 claims? Probably has something to do with you get 20 claims included in the filing fee, but I can’t really figure out more than that.

Comments are closed.