The 12-Year Patent

by Dennis Crouch

A patent costs money; and the costs continue even after issuance. The U.S. requires a set of three “maintenance fees” to keep a utility patent in force for its entire 20 year patent term. These are due 3.5, 7.5, and 11.5 years after issuance, but the PTO allows for a 6-month grace period for payment.  The chart below looks at the percentage of utility patents where the fees were paid prior to abandonment. The current schedule requires $2000 for the first fee; $3,760 for the second, and $7,700 for the third.  These fees can be reduced by 50% and 75% for small and micro entities.  There is also a $500 surcharge if you take advantage of that 6-month grace period.

The chart below shows the Percent of Utility Patents Whose Maintenance Fees have been Paid.  The dates along the horizontal axis refer to the 4/8/12 year due date for fees (grouped by months) and the chart reflect number of fee cases were the fees were paid as a percentage of the total number of patents whose fees were due that month.  You’ll note that a bit of a turn-down in early 2013. The USPTO instituted a major fee increase at that time with the 12-year fee up from $4,810 to $7,400.  In addition, some of the 3rd-stage decrease can be explained by the shorter patent term experienced by some patentees under the 20-year-from-filing rule.  This is most commonly experienced by international patent application (PCT) filers. The international application starts the 20-year clock and typically delays patent issuance by 2+ years.

One bottom line: Most utility patents have a post-issuance effective life of 12 years or less.

38 thoughts on “The 12-Year Patent

  1. 10

    As I have mentioned in the past, this type of chart need be evaluated carefully.

    ANY single ‘personal’ decision on a particular application (and whether or not to pay a first or subsequent maintenance fee requires that data points be time-shifted diagonally (wherein the ‘normal’ tendency is to compare the 4, 8, and 12 year lines strictly vertically).

    That being said, looking strictly vertically is a good way to infer NON-patent specific external influences.

    The general lock-step drift upward (across all three maintenance lines) from 2002 to 2009 indicate that (even against the Tr011 narrative) that patent holders were seeing increased value in maintaining their property rights. 2009 of course dips due to the financial crises, with resumption of an upward drift through about 2013, at which point a flex occurs, and the cumulative effect of the AIA and anti-patent case law can be offered as reasons for a downward drift, which would show exacerbated in the 8 and 12 timelines.

  2. 9

    The 12 year fee boost was illogical, unreasonable, and unjustified.

    Anyone recall the Office’s alleged rationale (if any) for it?

    1. 9.1

      Hard to understand why a patent owner, after having a full 12 years after the patent issues to try to license or sell it, still can’t decide even by then if it is not worth a mere $7,700 [or $3,850 if a small entity] to get a full term? [And thus much fairer to charge a higher maintenance fee then than earlier.]

    2. 9.2

      Pretty much every jurisdiction in the world charges more for the renewal fees near the end of the term than near the beginning of the term. That is because this is the economically rational thing to do.

      A firm at the beginning of the effort to market a product needs to build a customer base. They have—at the start of their efforts—essentially no revenue. A high fee at this stage of the product life cycle will essentially serve as an incentive simply to give up and never commercialize.

      By contrast, after the product has been on the market for a few years, a customer base has been built. By this point in the product life cycle, there is a much larger revenue stream. It is less of a hardship and disincentive, therefore, to raise the prices at this stage.

      By making the first set of fees low and the last set high, one is simply acknowledging the commercial realities that the fee-payor faces. That is why this is how nearly every patent system in the world operates.

    3. 9.3

      Back when patent fees were effectively statutory (but with an annual inflation adjustments over the statutory base fee amount) and the maintenance fees were first introduced in December 1980 (35 USC 41(b)), an increase from one payment to the next was already built into the schedule of fees at a 1-2-3 ratio ($400 for the 1st maintenance fee, $800 for the 2nd maintenance fee, $1200 for the 3rd maintenance fee). The House report doesn’t appear to offer any rationale for this ramped fee structure, only the need for maintenance fees in general to help fund the operations of the USPTO. Greg’s explanation in comment 9.2 is as good a reason as any and makes sense to me.

      Since that first fee schedule, the ramp-up has become a bit more steeply skewed toward that 3rd and final payment (currently a 1-1.88-3.85 ratio, albeit somewhat less steep than the last of the effective statutory schedules). Since the 2011 America Invents Act, the setting of fees is largely in the hands of the USPTO itself on an aggregate cost recovery basis (with mandatory input from a Patent Public Advisory Committee).

  3. 8

    Another way of saying what Dennis said is, most patent owners choose an effective term of twelve years. I’m assuming the Office has Moneyballed this to maximize revenue (which, that’s a whole separate debate about public good and the soul/mission of an administrative agency versus solvency/profitability)

    1. 8.1


      You raise some decent enough points, although I would disagree with you that the Office has “Moneyballed this to maximize revenue” – you give more credit than what is due. I would dare say that the OPPOSITE of ‘moneyballing’ is shown with the numbers.

      As to ANY sense of “solvency/profitability,” well, the notion that the Office is not — and cannot be — a profit center need be taken into account.

      As to revenue machinations in view of (limited) powers to set fee levels, I WOULD agree that transparency of such is (and has been) alarmingly lacking.

  4. 7

    Re: “A patent costs money; and the costs continue even after issuance.”
    Just a reminder of the rationales for the U.S. introducing patent maintenance fees for a significant portion of PTO revenue [as other countries have had for much longer]:
    Enabling relatively less expensive obtaining of patents, to encourage invention filings and public disclosures, and
    clearing “deadwood” from the large numbers of issued patents to make new product clearance and right to use studies easier.

  5. 6

    Might be fun to see a breakdown by technology area breakdown. It would seem that there could be a substantial difference between TC1600 and TC3700.

    1. 6.1

      Ben, if you are suggesting that patents with claims more likely to be attacked on Alice or other 101 exception grounds are more likely to get abandoned by non-payment of maintenance fees, I would doubt it for two reasons: First, because I doubt if many corporate patent maintenance review committees would decide on that basis. Secondly, because of the number of such patents being sued on anyway, thus demonstrating value, especially for PAE sales.

      1. 6.1.1

        I can easily envision other reasons than those, Paul, notably, pace of innovation (on direct contrast to Pharma world).

      2. 6.1.2

        Actually, I wasn’t thinking anything along those lines. Instead I was thinking that it seems likely that there’s a discrepancy between the baseline expected value in a patent from TC1600 (chemicals) and TC3700 (mechanical), and that this maintenance payment data would likely reflect that discrepancy.

  6. 5

    Do patents continue on even after all claims are cancelled in a reexamination?

    Some time back, I saw a situation where a patent had expired for failure to pay maintenance fees. I discovered there had been a reexamination that ended with all claims being cancelled because the patent owner did not respond to an Office Action. PAIR explicitly said the patent expired due to non-payment of the maintenance fees. The reexamination certificate issued a bit more than 6 months prior to the official expiration. However, I would have thought that a reexamination certificate cancelling all claims would have immediately resulted in the patent’s “death”. I’m just wondering if this is one of those rare, weird overlaps between the grace period for paying the fee and the issuance of the reexam certificate cancelling all claims, or if a patent can still be “alive” even if it has no claims capable of being enforced.

    1. 5.1

      I don’t think the two areas of the PTO speak with each other. If you hit the deadline without paying the fee, the PTO tags the case as expired due to failure to pay . . . even if the patent had already expired naturally; been terminally disclaimed; or all claims cancelled.

      1. 5.1.1

        “I don’t think the two areas of the PTO speak with each other.”
        I don’t think that comes as a surprise to many patent practitioners. One quickly learns that the Left Hand at the USPTO doesn’t seem to know what the Right Hand at the USPTO is doing though both should technically be in communication with the same Head that should be keeping them informed. I say “technically.”

    2. 5.2

      A patent definitely can be meaningfully “alive” even after all claims have been cancelled in a re-exam. If the patentee believes that there still remains patentable subject matter disclosed in the patent (presumably something narrower than that which had been claimed during the re-exam), the patentee is entitled to apply for a reissue. If, however, the maintenance fees have not been paid, then a reissue is not possible, because a reissue cannot be granted on an expired patent. In re Morgan, 990 F.2d 1230, 1231 (Fed. Cir. 1993). Therefore, there is still some small possibility of value in paying maintenance fees even once all claims are cancelled.

      1. 5.2.1

        Thank you for the case citation though a patent without enforceable claims seems more of a “zombie”; neither truly alive nor truly dead.

  7. 4

    If I understand this blog correctly, a very substantial majority [whatever % it is] of all U.S. issued patents are being abandoned by non-payment of a maintenance fee before or at 12 years from their issue date. That is before the end of their 20 years+ from original filing date patent term assuming that relatively few U.S. patents have 8 years or more of pre-issue pendency [even including deliberately delayed divisionals and continuations] [a determinable statistic?].
    The maintenance-fee-cost-savings for early [pre expiration] abandonment of foreign equivalent patents is even larger.
    Thus, it appears to me that most U.S. applicants are now doing a good job of maintaining cost-reduction programs to review their portfolios for patents that are on now-obsolescent technology, or not protecting their products from infringers, or are not licensable, and are not even sellable to PAEs.

    1. 4.1

      The extent and timing of maintenance-fee-non-payment abandonments is even more significant considering that even after patents are expired they can still be sued-on during the next six years for partial prior infringement back-damages.

      1. 4.1.2

        Reasons for companies continuing to PAY maintenance fees on patents they should have abandoned for cost savings are are even more complex [and unlikely to be seen by many academics].
        Some I have seen are:
        wanting to look more like a high tech company [to goose the stock price, etc.] by being able to advertise having a large number of patents in force;
        strange accounting system tax reasons re the already incurred sunk costs of the prep and prosecution;
        not bothering to set up a patent maintenance review system;
        a patent maintenance review system that does not properly address claim scope and commercial product alternatives;
        managers afraid they might get in trouble for abandoning any patent that someone might want to license later [no matter how unlikely];
        managers unwilling to admit that their costly R&D program which generated such patents was commercially or fiscally unviable;



          The act of abandonment means a relinquishment of any ‘choice of actions.’

          This is directly contrasted with natural expiry, as with such, there is no “choice” from the property holder.



            So what happens in this hypo:

            The patent goes abandoned during an ongoing infringement action.

            Suit dismissed? Automatically by the court? Or must the defendant make a motion for dismissal?

            Does it matter if the abandonment was unintentional?

            Or; assuming the patent owner is successful; is the alleged infringer on the hook for infringement occurring prior to the abandonment?

                1. Correct. And with the PTO providing a petition option to “restore” the patent in such an event . . .

  8. 3

    I suspect that separating out Pharma patents (whose value continues to increase all the way to the end of the term) from other “technologies” may provide two substantially different sets of curves.

    1. 3.1

      Chief Judge Moore wrote an article on this topic back in 2005 titled Worthless Patents. She then extended that work later with a group of other professors in another paper titled Valuable Patents. It would be interesting to me to go back to those original works and see if the same variables continue to be predictive.

      1. 3.1.1

        “Worthless” and “Valuable” are in the eye of the beholder, so I would choose different labels (plus, worth may in some circumstances be of higher value even if for shorter duration).

  9. 2

    I’m intrigued my your statement, Dennis, that “most” patents issued by the USPTO are effective for less than 12 years. This would suggest that “most” issue more than 8 years after their filing date. Does that accord with the experience of readers here?

    I see that a full 40% of all issued patents are renewed right through to the end of their 20 year term. I was expecting it to be much lower. But perhaps it is that the later the issue date the greater the likelihood that the owner will find it worthwhile to pay the relatively expensive 3rd renewal fee.

    1. 2.1

      would suggest that “most” issue more than 8 years after their filing date.

      Not at all — the point being presented as a that with the nominal choice to not pay the third maintenance fee (and to other extents, not paying the second or first such) the effective duration is lessened.

      This is not indicative of any issue delay mechanism.

    2. 2.2

      Hi Max – I always struggle with explaining the maintenance fee data in a clear way. The confusion, I think, comes from all of the different dates at issue.

      Lets look at patents issued in January 2010 (12 years ago).
      * 13% expired in January 2014 for failure to pay the 1st maintenance fee.
      * Another 20% expired in January 2018 for failure to pay the 2nd maintenance fee — bringing the total to 33% expired.
      * Finally, another 24% expired in January 2022 for failure to pay the 3rd maintenance fee — bringing the total to 58% expired (after some rounding).

      The result here is that 58% of the patents issued 12 years ago are no longer in force.

      1. 2.2.1

        OK, Dennis. Thanks. My misunderstanding, a consequence of my mindset that “effective life” means that portion of the 20 year term available after issue, for enforcing the rights granted by the issued patent. I see that for you the expression means that portion of the 20 year term which the patent owner chooses to pay for.

        There is a related misunderstanding of mine over the meaning of the concept of “expiry”. For me, the “expiry date” of a patent is the date at the end of the 20 year term (perhaps longer if there was delay at the USPTO). For me, it is misleading to dub the date when the patent owner lets the patent lapse by failure to pay a renewal fee the date of “expiry” of that patent.

        The main thing though is that the 58% figure is for that set of patents issued 12 years ago, a set that includes a large number of patents granted soon after the PTO filing date of the application on which they are founded. Thanks for the clarification.


          No MaxDrie, there is nothing here to begin making the suggestion that you insert that, “that includes a large number of patents granted soon after the PTO filing date of the application on which they are founded

          You (again) are trying to ‘read in’ a condition not presented.

  10. 1

    I have a question on the data we’re looking at:

    Is the 60% rate for the 2nd maintenance fee based on the total number of originally issued patents, or on the 80% that had paid the 1st maintenance fee? If the latter, then 60% of the 80% amounts to just 48% of the original patents and most patents would have only an 8-year life after issuance.

    Likewise, for the 40% rate for the 3rd maintenance fee. Is it 40% of the originally issued patents, or 40% of those whose 2nd maintenance fee was paid? If the latter, then 40% of the surviving 48% amounts to just 19% of the original patents.

    Please clarify. Thanks.

    1. 1.1

      Thanks Mark — The denominators for all of these are the same — the # of patents issued X years ago.

      Example using the January 2022 data points:

      The top line shows 85% payment at 4-year mark for Jan 2022. This looks at all patents issued in Jan 2018 (4-years prior) and reports the percentage that paid the first maintenance fee.

      The second line down showing 64% payment at 8-year mark. This looks at all patents issued in Jan 2014 (8-years prior) and reports the percentage of those that paid the second maintenance fee.

      The bottom line showing 42% payment at 12-year mark. This looks at all patents issued in Jan 2010 (12-years prior) and reports the percentage of those that paid the third maintenance fee.

Comments are closed.