Of counterfeit statistics: knocking off the inflated patent grant rate numbers
An article in IPL&B entitled Fake-Onomics; The bogus nature of counterfeiting statistics revealed quotes Mark Lemley: "Counterfeiting statistics are counterfeit." Of the use of the bad statistics, Lemley says: "It may prompt Congress to act on the assumption that the sky really is going to fall if they don't."
Flashback merely to February 2004 when Lemley and Moore wrote:
The effect of these continuations is substantial. Recent work by
Cecil Quillen and others shows that when continuations are taken into account, the PTO issues patents on over 85% of the application chains that are filed.
and cited to footnote 22 which said in part:
On the other hand, some of the critics [of Quillen/Webster] 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.
[from ENDING ABUSE OF PATENT CONTINUATIONS, 84 B.U.L. Rev. 63 (Feb. 2004)]
There are several difficulties with what Lemley and Moore said. Clarke never said (or even implied) that every continuation filed results in a separate patent. The 85% number arises from subtracting continuing applications (not just continuation applications) for which there is no antecedent abandonment, and can yield grant rate numbers in excess of 100%. The 85% number is not the best estimate of how often applications mature into patents, and is in fact the kind of counterfeit statistic bemoaned by Lemley in the IPL&B article.
***
At least, Lemley and Moore rejected the 95% number of the first Quillen/Webster article. From their footnote 22:
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, Polk Wagner's November 2005 article, written more than a year and half after Lemley/Moore used the 95% number!
Within Gideon Parchomovsky, R. Polk Wagner, 154 U. Pa. L. Rev. 1 (Nov. 2005), one finds, surprisingly, that footnote 76 is to the first Quillen/Webster paper:
See, e.g., 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, 3 (2001) (indicating that once continuing applications are included, the patent approval rate is 95%). Quillen and Webster conclude that the PTO might ultimately approve as many as 97% of all patent applications. Id. at 13.
[Note: the number in the first Quillen/Webster paper is as high as 97%!]
The predicate to footnote 76 was the text:
Furthermore, patent applications convey little information about the
potential commercial value of the invention. For example, patentees do
a notoriously poor job of referencing prior inventions in their patent
applications. n75 Without information about competing technologies and
blocking patents, third parties cannot possibly determine the value of the patented invention. Finally, the potential signaling value of the patent application is further weakened by the lax and "patent-friendly" review given by the PTO (which approves nearly all of the applications that it receives). n76 Moreover, after passing this diminished level of scrutiny, a significant percentage of issued patents are declared invalid when challenged in court; hence, third parties cannot rely too heavily on the validity of issued patents that have not been exposed to litigation. Indeed, even Long acknowledges that in many
circumstances patent signals may be ambiguous, in which case their
value becomes suspect. n77
The problem with Wagner's text is both that it cites to false information (i.e., the first Quillen/Webster paper when the second Quillen/Webster acknowledged that the assumptions of the first were in error) but that it neglects to cite to other material information (e.g., it talks about a "significant" portion of patents found invalid, even though Lumney's paper shows that invalidity determinations by the courts are down).
Thus, ironically, Wagner's paper illustrates Wagner's own text:
For example, [law review professors] do a notoriously poor job of referencing prior [law review articles] in their [own law review articles].
Merely fyi, footnote 75 is to Robert P. Merges, As Many as Six Impossible Patents Before Breakfast: Property Rights for Business Concepts and Patent System Reform, 14 Berkeley Tech. L.J. 577, 589-90 (1999) (discussing the poor quality of patent applications in terms of the number and nature of prior art references).
Merges actually cited one of my papers on the patent grant issue, noting that Clarke's number was indeed preferred over the 85% number of the second Quillen/Webster paper. Thus, Polk Wagner and the University of Pennsylvania Law Review are way behind the times.
***
The paper "Probabilistic Patents" by Mark Lemley and Carl Shapiro cites to both the second Quillen/Webster paper and the Parchomovsky/Wagner but fails to note the disparity (i.e., 95% vs. 85%). The Lemley/Shapiro paper mentions the 85% number:
the overwhelming majority of patent applications, perhaps 85 percent, ultimately result in an issued patent – far more than in Europe and Japan (Quillen et al., 2003; NAS, 2004). [Note that the cite to the NAS report is merely an imbedded cite to QWII; the NAS did not independently generate the 85% number.]
Flashback merely to February 2004 when Lemley and Moore wrote:
The effect of these continuations is substantial. Recent work by
Cecil Quillen and others shows that when continuations are taken into account, the PTO issues patents on over 85% of the application chains that are filed.
and cited to footnote 22 which said in part:
On the other hand, some of the critics [of Quillen/Webster] 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.
[from ENDING ABUSE OF PATENT CONTINUATIONS, 84 B.U.L. Rev. 63 (Feb. 2004)]
There are several difficulties with what Lemley and Moore said. Clarke never said (or even implied) that every continuation filed results in a separate patent. The 85% number arises from subtracting continuing applications (not just continuation applications) for which there is no antecedent abandonment, and can yield grant rate numbers in excess of 100%. The 85% number is not the best estimate of how often applications mature into patents, and is in fact the kind of counterfeit statistic bemoaned by Lemley in the IPL&B article.
***
At least, Lemley and Moore rejected the 95% number of the first Quillen/Webster article. From their footnote 22:
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, Polk Wagner's November 2005 article, written more than a year and half after Lemley/Moore used the 95% number!
Within Gideon Parchomovsky, R. Polk Wagner, 154 U. Pa. L. Rev. 1 (Nov. 2005), one finds, surprisingly, that footnote 76 is to the first Quillen/Webster paper:
See, e.g., 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, 3 (2001) (indicating that once continuing applications are included, the patent approval rate is 95%). Quillen and Webster conclude that the PTO might ultimately approve as many as 97% of all patent applications. Id. at 13.
[Note: the number in the first Quillen/Webster paper is as high as 97%!]
The predicate to footnote 76 was the text:
Furthermore, patent applications convey little information about the
potential commercial value of the invention. For example, patentees do
a notoriously poor job of referencing prior inventions in their patent
applications. n75 Without information about competing technologies and
blocking patents, third parties cannot possibly determine the value of the patented invention. Finally, the potential signaling value of the patent application is further weakened by the lax and "patent-friendly" review given by the PTO (which approves nearly all of the applications that it receives). n76 Moreover, after passing this diminished level of scrutiny, a significant percentage of issued patents are declared invalid when challenged in court; hence, third parties cannot rely too heavily on the validity of issued patents that have not been exposed to litigation. Indeed, even Long acknowledges that in many
circumstances patent signals may be ambiguous, in which case their
value becomes suspect. n77
The problem with Wagner's text is both that it cites to false information (i.e., the first Quillen/Webster paper when the second Quillen/Webster acknowledged that the assumptions of the first were in error) but that it neglects to cite to other material information (e.g., it talks about a "significant" portion of patents found invalid, even though Lumney's paper shows that invalidity determinations by the courts are down).
Thus, ironically, Wagner's paper illustrates Wagner's own text:
For example, [law review professors] do a notoriously poor job of referencing prior [law review articles] in their [own law review articles].
Merely fyi, footnote 75 is to Robert P. Merges, As Many as Six Impossible Patents Before Breakfast: Property Rights for Business Concepts and Patent System Reform, 14 Berkeley Tech. L.J. 577, 589-90 (1999) (discussing the poor quality of patent applications in terms of the number and nature of prior art references).
Merges actually cited one of my papers on the patent grant issue, noting that Clarke's number was indeed preferred over the 85% number of the second Quillen/Webster paper. Thus, Polk Wagner and the University of Pennsylvania Law Review are way behind the times.
***
The paper "Probabilistic Patents" by Mark Lemley and Carl Shapiro cites to both the second Quillen/Webster paper and the Parchomovsky/Wagner but fails to note the disparity (i.e., 95% vs. 85%). The Lemley/Shapiro paper mentions the 85% number:
the overwhelming majority of patent applications, perhaps 85 percent, ultimately result in an issued patent – far more than in Europe and Japan (Quillen et al., 2003; NAS, 2004). [Note that the cite to the NAS report is merely an imbedded cite to QWII; the NAS did not independently generate the 85% number.]
posted by Lawrence B. Ebert at 9:03 PM
0 Comments:
Post a Comment
<< Home