State of Innovation

Patents and Innovation Economics

Patent Quality Myth

A number of papers have suggested that the United States Patent and Trademark Office’s (USPTO) quality in issuing patents is substandard.[1] These papers suggest that the USPTO’s allowance rate is significantly higher than the allowance rate in Europe and Japan.  From this they infer that the quality of patent examination, in the U.S. are not as thorough as those in Europe and Japan.

In the case of Jensen et at, Disharmony in International Patent Office Decisions, the population of patents studied were patents that issued in the U.S. and were filed in the EPO (European Patent Office) or the JPO (Japan Patent Office).  Jensen et al found that in this population only 37.7% of this population of patents were also allowed in the EPO or JPO.  Note that based on the population selected by Jensen et al. they could never find that the grant rate was higher for EPO or JPO than in the USPTO (United States Patent and Trademark Office), since the selected population included a 100% grant rate for the U.S.  Of course, the U.S. has never had a 100% grant rate.  The chart below shows the grant rate was never greater than 72% in the U.S. in the last 40 years.

090102USPTO_ALLOWANCE_RATE

Jensen et al. by their methodology biased the outcome so that the USPTO’s grant rate had to be higher than the EPO or JPO.

Ron D. Katznelson’s paper Bad Science in Search of “Bad” Patents[2] explains this error and other errors in Jensen et al.’s methodology.  Comparing grant rates between countries is extremely difficult.  A major reason that patent applications are not allowed is that the inventor or their employer decides that the cost of obtaining a patent is no longer worth the cost.  The longer the time between the filing of a patent application and its examination, the higher the likelihood that the applicant will abandon the case.  Both the EPO and JPO have significantly longer waits for examination than the U.S.  The extended delays for examination in Europe and Japan are not features that we want to imitate.

Another paper that purported to show higher allowance rates in the USPTO than other countries was Continuing Patent Applications and Performance at the U.S. Patent and Trademark Office-Extended[3], by Quillen and Webster.  Quillen and Webster assume that continuations, divisionals, and continuations-in-part are all attempts to claim the same invention.  This is clearly mistaken in the case of divisionals and continuations-in-part.  As pointed out by Lawrence B. Ebert in Patent Grant Rates at the United States Patent and Trademark Office[4] “The use of such an approach, as applied to patent families in which more than one continuing application issues as a patent, allows a grant rate in excess of 100%, — a nonsensical result.”  Quillen and Webster attempted to fix this error in a subsequent paper, but as Ebert points out they were not successful.  Ebert concludes “The patent grant rate studies of Quillen and Webster are flawed and the published grant rate number should not be relied upon in making arguments about patent reform.”

Economist Pat Choate has reviewed the question of numerous junk patents being issued by the USPTO from another point of view.  He reviewed the number of USPTO patents that were held invalid in litigation or in reexamination proceedings.  He compares this against the total population of patents issued in the same period.  He states:

As to the massive numbers of “unworthy patents” argument, the real-world test is how many patents are challenged and the outcome of those challenges. Between 1981 and 2006 the USPTO issued more than 3.1 million patents. In that period, 8,600 were challenged at the Patent Office through inter partes and ex parte reexaminations. The number challenged amounts to less than three-tenths of one percent. Of those challenged, about 74 percent resulted in claims narrowed or cancelled. In addition, almost 60 percent of the relatively few patents challenged in a court trial are sustained.

My point is that the USPTO’s work is certainly not perfect, but the Patent Office is also not pouring out a stream of bad patents.[5]

Since there are around 700 patent lawsuits per year with substantive adjutication, a high end calculation of the number of patents held invalid between 1981 and 2006 would be no more than 1,000 patents and of the 8,600 patents through reexamination cases no more than 6,784 patents were narrowed or held invalid.  This means that out of 3.1 million patents 7,450 patents were actually held invalid or had their claims narrowed.  This means that less than 0.25% of all the patents issued during this time-period were actually found to have problems.

There has been a concerted effort to suggest that there are numerous “bad” patents issued by the USPTO.  This campaign is based on disinformation and has no factual support.  Patents are the most expensive and time consuming process for obtaining title to property in the U.S.  Even large corporations cannot afford to file patents, at a cost of $10k to $20K per patent per country, willy nilly.  It is common for a patent to take three or more years to issue.  Only inventors who are serious are going to file a patent application under these circumstances.  When they consider whether to file a patent application they are likely to err on the side of not filing a patent if it looks questionable whether their invention will obtain a patent.

If it took this amount of time and money to obtaining title to real property, we would assume we were talking about some despotic third world country.  Copyrights are not examined, are inexpensive to file and assumed to be valid.  There is no justification for the U.S.’s poor treatment of inventors compared to authors.

The “bad patents” myth has no basis in fact, denies inventors’ title to their inventions, and hurts U.S. innovation and our economy.


[1] Cecil D. Quillen and Agden D. Webster, Continuing Patent Applications and Performance of the U.S. Patent and Trademark Office, 11 Fed. Cir. B.J. 1 (2001-2002); Paul H. Jensen et al., Disharmony in International Patent Office Decisions, 15 Fed. Cir. B.J.679, 698 (2006).

[2] Katznelson, Ron D., Bad Science in Search of “Bad” Patents, Federal Circuit Bar Journal, Vol. 17, No. 1, pp. 1-30, August 2007.

[3] Cecil D. Quillen and Agden D. Webster, Continuing Patent Applications and Performance of the U.S. Patent and Trademark Office, 11 Fed. Cir. B.J. 1 (2001-2002)

[4] Ebert, Lawrence B., Patent Grant Rates at the United States Patent and Trademark Office, Chicago-Kent Journal of Intellectual Property, 2008, pp. 108-116.

[5] Choate, Pat, Patent ‘Reform’ is Anything But, Manufacturing & Technology News, Vol. 14, No. 12, June 29, 2007.  http://www.manufacturingnews.com/news/07/0629/art2.html

Advertisements

August 18, 2009 - Posted by | -Law, -Philosophy, Patents

9 Comments »

  1. […] ***on patent quality […]

    Pingback by "Is intellectual property law a scam?" « YuvaBlog.in | September 11, 2009 | Reply

  2. […] show that there is a patent quality problem have flawed methodologies.  For more information see Patent Quality Myth .  What we have in this country is a small number of large companies that have made a business out […]

    Pingback by Patent Reform Propoganda « State of Innovation | November 9, 2009 | Reply

  3. “In the case of Jensen et at, Disharmony in International Patent Office Decisions, the population of patents studied were patents that issued in the U.S. and were filed in the EPO (European Patent Office) or the JPO (Japan Patent Office). Jensen et al found that in this population only 37.7% of this population of patents were also allowed in the EPO or JPO. Note that based on the population selected by Jensen et al. they could never find that the grant rate was higher for EPO or JPO than in the USPTO (United States Patent and Trademark Office), since the selected population included a 100% grant rate for the U.S. Of course, the U.S. has never had a 100% grant rate. The chart below shows the grant rate was never greater than 72% in the U.S. in the last 40 years.” You didn’t get the point of Jensen et al.: Of a set of patents granted by the USPTO (so, of course 100%) only 37% got a grant in the EPO or JPO. Jensen et al. does not state that the grant rate of the USPTO is 100%. The procedure of Jansen is obviously correct.

    “Since there are around 700 patent lawsuits per year with substantive adjutication, a high end calculation of the number of patents held invalid between 1981 and 2006 would be no more than 1,000 patents and of the 8,600 patents through reexamination cases no more than 6,784 patents were narrowed or held invalid. This means that out of 3.1 million patents 7,450 patents were actually held invalid or had their claims narrowed. This means that less than 0.25% of all the patents issued during this time-period were actually found to have problems.” But, this does not show that the USPTO does not have low quality!? It only shows the number of patents which were reexamined and found deficient. This number is approximately the same for the EPO. As you know, a reexamination costs money. Not only the USPTO fees but especially the attorney fees. This means that, no matter what the quality of the USPTO, the real incentive to ask for a reexamination are the costs. I agree that it is not straight forward to compare different patent systems. However, it isn’t rocket science either. It suffices to choose the same application, let’s say a PCT filled in the US and follow its outcome in the JPO and in the EPO. After reading Janssen, I was happy to see that this is their next approach. Let’s see then…

    Comment by Miguel | December 7, 2009 | Reply

    • Miguel,

      What you are missing in Jensen is that the incentives, costs, and timelines are different in the JPO, Europe, and the U.S. Even if you “choose the same application, let’s say a PCT filed in the US and follow its outcome in the JPO and in the EPO” you will not receive meaningful results. The costs of filing in the JPO and EPO are significantly higher than in the U.S., so all other things being equal most applicants will be more likely to follow through with the U.S. application and let the other applications drop. In addition, the U.S. is the largest market (for the lowest cost) so applicant’s have more incentive to obtain a patent in the U.S. than in the EPO and JPO. In addition, we know that the longer it takes to examine a patent the more likely that the applicant will abandon the patent application. This is consistent with the fact the most applicants never pay for their last maintenance fee in the U.S. Both the EPO and JPO take significantly longer to examine patent applications, so we would expect fewer allowance in the EPO and JPO. Finally, the maintenance fees in the EPO and JPO are significantly higher than the U.S., further reducing the incentive of obtaining a patent.

      There are far too many variables to make any meaningful comparisons between the allowance rates. Finally, a lower allowance rate does not necessarily mean that higher quality examination is occurring. Any rational definition of quality would include both false rejections of patents and false allowances. If the JPO and EPO have more false rejections than the U.S. is this higher quality? Is this a system we want to imitate? None of the quality studies are objective because they only focus on false allowances and ignore false rejections. Any patent attorney prosecuting patents can tell you that there are many more false rejections, particularly in the last six years, than false allowances. Ron Katznelson has a mathematical proof that shows the societal damage is greater for false rejections than false allowances.

      Comment by dbhalling | December 7, 2009 | Reply

      • Dale,

        If a US filled and granted patent is followed up in the EPO and the JPO, until is either granted or refused by the EPO or the JPO, your argument about incetives, cost, examination time, etc. does not apply. Because by definition, the applicant hasn’t quit (due to the causes you cited) but because the EPO or the JPO refused it. And since this is what the Janssen study shows, I state again that it is a valid comparison. As an example just check any automotive multinational who must file in these 3 offices. They will not usually quit due to causes you mentioned.
        On the other side, for the small/medium sized applicant the causes you cite are significant and relevant. And I do agree that in comparison the EPO and the JPO are not as fast in terms of pure legal procedure. However, I don’t believe its a question of lousy management or service but its more of a reflex of the different legal systems. In Europe (and more extreme even in Japan) it is not common to use the courts to settle patent legal disputes. In Europe the patent system is geared to “discourage” litigation after patent granting. Therefore, more care is taken in examination to guarantee that the inventive step is (as much as possible) beyond dispute. In the US, it is far more common to dispute a grated patent in court. Therefore, the USPTO does not need to have such an exhaustive search and examination because the quality of a patent can still be improved (if challenged) in the courts. However, in this case, the small and medium sized companies are in disadvantage due to these extra litigation costs. So, and jokingly, I would say the USPTO is like McDonalds whereas the EPO and JPO are more like a standard restaurant :-). So if you can compare such a thing…

        Comment by Miguel | December 12, 2009

  4. […] While a number of papers have shown methodolical problems with  these academic papers (see Patent Quality Myth ), the bigger question is whether they selected the correct metric in the first place.  This post […]

    Pingback by Patent Quality Nonsense « State of Innovation | January 7, 2010 | Reply

  5. […] R&D spending per patents, and number of citations per patent.  See Patent Quality Nonsense and The Patent Quality Myth.  It is disappointing that the Economist repeats these diatribes against patents without even a […]

    Pingback by The Economist Weights In on Patent Deform « State of Innovation | March 17, 2011 | Reply

  6. […] By every objective measure: R&D per patent, GDP per patent, and number of citations per patent patent quality is increasing.  See https://hallingblog.com/2010/01/07/patent-quality-nonsense/ and https://hallingblog.com/2009/08/18/patent-quality-myth/. […]

    Pingback by Are Transaction Costs for Patents Too High? « State of Innovation | April 15, 2011 | Reply


Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

%d bloggers like this: