Friday, July 13, 2007

More on the 97% grant rate that never was

Further to the recent post on the 97% patent grant rate number popularized by Quillen and Webster, note that footnote 22 of the Lemley and Moore paper in BULR accepts the 85% number of the SECOND paper by Quillen and Webster.


n22. Cecil D. Quillen, Jr. et al., Continuing Patent Applications
and Performance of the U.S. Patent and Trademark Office - Extended, 12 Fed.
Cir. B.J. 35, 38 (2002). Quillen and Webster had originally estimated in
earlier work that the grant rate was 95%. Cecil D. Quillen, Jr. & Ogden H. Webster,
Continuing Patent Applications and Performance of the U.S. Patent and
Trademark Office, 11 Fed. Cir. B.J. 1 (2001). This earlier work was properly
criticized for failing to take account of cases in which multiple patents issue
from a family of continuation applications. On the other hand, some of the
critics made equally unrealistic assumptions - for example, that every continuation
filed results in a separate patent. See Robert A. Clarke, U.S. Continuity Law
and Its Impact on the Comparative Patenting Rates of the U.S., Japan and the
European Patent Office, 85 J. Pat. & Trademark Off. Soc'y 335, 338 (2003)
(erroneously assuming that every continuation resulted in a patent and concluding
that the grant rate was 75%). The 85% number provided in the revised Quillen et
al. study is based on actual data about the applications that issue based on
continuations, and reflects the best estimate we have of how often
applications mature into patents.

[In passing, LBE has written a lot about footnote 22 of 84 B.U.L. Rev. 63]

Separately Doody's paper in IPTLJ mentions the first two papers of Quillen and Webster in footnote 52, which footnote also cites to G.A. Clarke.

Footnote 52 appears after the sentence: "The critics allege it is easier to obtain a patent today than it was 15 years ago."

The next sentence states "The critics have alleged that the approval rate today approaches 90%," citing to footnote 53. Footnote 53 is directly to an editorial in the Wall Street Journal, March 1, 2006. There is also a cite to a link,

That link leads one to the text:

The editorial "Patently Absurd" (A14, March 1, 2006) depicts an out-of-control Patent Office approving almost 90% of submitted applications and a powerless court system constrained by a "clear and convincing evidence" standard. In reality, patent grant rates have been steadily declining since 1999, when the rate was 70.8%; the rate was 62.5% in 2004. Efforts to fashion adjusted patent grant rates, initiated by Quillen and Webster and later relied upon by Jaffe and Lerner, have been shown to be flawed on both numerical and legal bases. If there were indeed a tide of questionable patents, the court system would readily invalidate them over prior work, under any evidentiary standard. Studies by Lunney have shown that invalidation of patent claims by the court system has declined over the last twenty years. In situations wherein there is published prior work, either dead-on to the later work or rendering the later work obvious, the procedure of re-examination is available to invalidate claims on a preponderance of evidence standard. The patents asserted against RIM, Microsoft, and eBay have been placed in the re-examination process. The patent system is about disclosure of inventions that meet the requirements of patent law, which disclosure increases the public knowledge base. It is up to businessmen to innovate, with attention to the disclosed knowledge. People who disregard public disclosures may suffer, but ignoring the work of others should be made perilous so that society can operate efficiently. (end of letter)

Of the Metabolite case, on the matter of "patenting" a law of nature, one notes some background information. First, the patent in question was allowed through the Bayh-Dole Act, and is the work of three university professors, two at Colorado (still alive and represented by a different university professor, from the Stanford Law School, who otherwise advocates patent reform) and one at Columbia (now deceased). It does indeed rely on a correlation, first identified by the professors and not accepted by the scientific community initially, rather than a law of nature. There was no evidence at trial that anyone else had discovered the correlation previously, and the current issue is on the indefiniteness of the claim. Second, the present two corporate litigants were previously in a posture of licensee and sublicensee, so this litigation has the appearance of a business deal gone bad.

One can debate whether this sort of patent claim is of the type that fosters innovation. However, it is becoming increasingly clear that the business community does not want to hear about its role in the problems: the failure to conduct negotiations that, if implemented, would decrease the involvement of the court system in the market and the failure to treat the patent system seriously (the RIM case being a notable example of something that could have turned out differently, but for some bad decisions on the front end).
Although one can certainly point to many sound byte examples of bad patents (which largely have been eliminated through re-exam), it is a sad day when the Wall Street Journal and the eBay brief rely on false figures of patent approval rates to advance their arguments.

Lawrence B. Ebert
March 21, 2006

Grant rate trivia buffs may recall the text about the first Quillen/Webster paper:

The authors attempt to incorporate the missing data on continuing applications to produce a useful measure of the PTO's rigor. Taking account of continuing applications and adding a two-year time lag as a rough estimate of when those applications would have matured, Quillen and Webster examine the period from 1993 to 1998 and arrive at a startling maximum allowance rate of 95 percent. In Europe and Japan, comparable rates were 68 percent and 65 percent, respectively. Grant rates were just as high, reaching a maximum of 97 percent once corrected for the distorting effect of continuations and the attendant abandonments. In Europe and Japan, the corresponding rates were 67 percent and 64 percent, respectively. The authors conclude that “the PTO is by far the least selective of the patent offices analyzed.”

This dullness comes from Mickey Davis by way of researchoninnovation.
LBE tried years ago to make a comment on researchoninnovation, but that was not happening.

The recent IPBiz post was The 97% grant rate that never was

It was related to text on a different blog which stated:

More than 100 years ago, Mark Twain understood the corrosive power of numbers to sway a debate, particularly when those numbers are accepted unquestioningly and are based on faulty (but unspoken) assumptions. One example in patent law is the 97% patenting rate of patent applications set forth in Quillen and Webster, 11 Fed. Cir. B.J. 1 (2001); although quoted extensively (most notably, by Mark Lemley and (now Federal Circuit Judge) Kimberly Moore in their paper condemning continuation application practice), the number has been thoroughly debunked by a variety of authors (see, e.g., Patrick Doody, "The Patent System is Not Broken," 18 Intellectual Property & Technology Law Journal 10 (December 2006)).

IPBiz notes that Quillen and Webster assumed that, to determine grant rate, one could use a formula (aps allowed)/[aps allowed + aps abandoned - continuing applications]. ONE reason this is wrong is in the situation wherein one has an allowed parent and an allowed child. The numbers become (2/(2 + 0 -1)) or 200%, which result should be recognized wrong even by the least gifted. THAT SAID, footnote 17 of the first paper by Quillen and Webster shows that this was NOT an unspoken (or unwritten) assumption. The "big story" of the first Quillen and Webster paper is how ANYONE WITH EVEN A PASSING FAMILIARITY WITH PATENT LAW could have read the first Quillen and Webster paper, including footnote 17, and EVER BELIEVED THE 97% NUMBER COULD HAVE BEEN CORRECT. The 97% number was truly "dead on arrival," if anything can be said to be DOA.


Blogger Lawrence B. Ebert said...

The July 15, 07 issue of the New York Times has a piece by Michael Fitzgerald ( a Boston-area writer on business, technology and culture. E-mail: ) titled A Patent Is Worth Having, Right? Well, Maybe Not which includes text concerning James Bessen, of the researchoninnovation website mentioned above:

“Today, over all, patents don’t work; for the information technology industry especially, they don’t work,” said James Bessen, who became a lecturer at Boston University’s law school after a career in business. In 1983, he created the first computer publishing software with Wysiwyg (an acronym for “what you see is what you get”) printing abilities. He also founded a desktop publishing company, Bestinfo, later acquired by Intergraph. (...)

He and a colleague, Michael J. Meurer, are readying a book on the topic, “Do Patents Work?,” due in 2008. (A synopsis and sample chapters are at

The two researchers have analyzed data from 1976 to 1999, the most recent year with complete data. They found that starting in the late 1990s, publicly traded companies saw patent litigation costs outstrip patent profits. Specifically, they estimate that about $8.4 billion in global profits came directly from patents held by publicly traded United States companies in 1997, rising to about $9.3 billion in 1999, with two-thirds of the profits going to chemical and pharmaceutical companies. Domestic litigation costs alone, meanwhile, soared to $16 billion in 1999 from $8 billion in 1997.

Things have probably become worse since then. For instance, patent litigation is up: there were 2,318 patent-related suits in 1999, and 2,830 in fiscal 2006 (though that’s down from the peak year, 2004, when 3,075 were filed). Mr. Bessen said awards in patent cases also seemed to be up, though he was less confident in that data. Worse, he says, companies doing the most research and development are sued the most.

8:05 AM  
Blogger Lawrence B. Ebert said...

Slashdot has a thread on the Fitzgerald piece. Recall that "peer-to-patent" wanted peer voting like on slashdot. Here's an example of a highly-rated "insightful" comment on slashdot:

Everyone thinks they are special. It's a fundamental human attribute.

How else would you explain the people who play the lottery? Gamble at casinos? Think that out of all the millions of oppressed masses, _they_ are the ones who will live the American dream and become someone?

It makes life more interesting; without that drive there would be little innovation, little hard work and drive, few no obsessively hard workers spending three years of nights in a garage writing software, no interest in going for American Idol... ok, scratch that last one.

In the same way companies, which are only an aggregation of people, will think that they can be the one out of a million who will benefit from patents. Even if you can empirically and theoretically show that they are being taken up the arse by a banana. Human nature. Infuriating, isn't it?

Slashdot also noted:

The authors of the book Do Patents Work? (synopsis and sample chapters), James Bessen and Michael J. Meurer of the Boston University School of Law, have crunched the numbers and say that, especially in the IT industry, patents no longer make economic sense. Their views are less radical than those of a pair of Washington University at St. Louis economists who argue that the patent system should be abolished outright.

PatentHawk noted the following of the Fitzgerald piece:

Two number monkeys got a spreadsheet to tell them that "starting in the late 1990s, publicly traded companies saw patent litigation costs outstrip patent profits." Since they didn't know anything about patents, the statistical statement made sense to them, and Fitzgerald.


Fitzgerald touts his naiveté by warming to the idea of "open[ing] patent applications to public comment, which could help patent examiners find applicable previous inventions." The problem, as we all know, is not that patent examiners can't search the prior art, it's that they don't have time to do so, thanks to management cracking the whip to meet production goals. Adding amateur hour to prior art searching as part of the patenting process doesn't make it into the solution category. The Patent Prospector has covered this ground many times.

4:39 AM  
Blogger Lawrence B. Ebert said...

Fitzgerald made a response of sorts to PatentHawk:

There were also this disappointing blog post from PatentHawk.
I’m sure its author knows a good deal about patents and could have written a nuanced critique of the research. In fact, he starts out with two good comments about how he thinks the research should have been characterized, then unravels. Perhaps the testosterone he references midway through kicked in and got the better of him.

For instance, he seems not to have bothered to read even the short chapter by Bessen and Meurer that the article links to. So he commits the logical fallacy of calling me stupid but accepting my discussion of the research as valid. He pokes fun at me for making what he considers broad and obvious statements about patents, as if the vast majority of NYT readers will know everything he knows about patents.

Of PatentHawk on peer-to-patent, Fitzgerald writes:

He decries what he calls ‘amateur hour,’ the use of open comment systems on patent applications, despite the PTO’s decision to experiment with this approach. He must know that the likelihood is that most of the comments will come from corporate patent lawyers. That may be something to fear, but he prefers to make it sound as though the unwashed masses will flood the system with worthless comments.

IPBiz notes that investigation of the comments made-to-date at peer-to-patent suggests that MOST comments did NOT come from patent lawyers. The fear that IPBiz has expressed is that, if the peer-to-patent system were generally implemented (not in the form of this artificial test of a handful of preselected applications) COMPETITORS (whether or not lawyers) would game the system.

9:04 AM  

Post a Comment

<< Home