The next installment of Professors of Paranoia may come from patent law?
The June 23 Chronicle of Higher Education has a piece Professors of Paranoia? which might be subtitled Professors Gone Wild! If the Chronicle ever gets a hold of some of the recent legal research on patent reform, they might have to do a sequel.
The Federal Circuit Bar Journal (Kimberly Moore, editor), which brought us Quillen/Webster 1 and 2, has now brought out the third installment: Cecil D. Quillen, Jr. and Ogden H. Webster, Continuing Patent Applications and the U.S. Patent and Trademark Office--Updated, 15 Fed Cir B J 635-677.
If you naively thought that models for patent grant rate should be required, as a constraint of logic, to be limited to a grant rate of no more than 100%, the Federal Circuit Bar Journal is not for you. From page 661:
When the USPTO Grant Rate is determined on the assumption that all continuing applications are an attempt to patent the subject matter of their parents and that the parent applications were abandoned in favor of the continuing applications, the overall Grant Rate for 1981 - 2005 is 95%. This calculation establishes an upper bound for Grant Rate estimates. The assumption on which this calculation is based is not true in all instances, e.g., when a divisional or other continuing application is filed and its parent is not abandoned. The result is that, in some instances, this calculation produces a Grant Rate estimate that is above 100%. Such a result is not realistic and results form the fact that not all parent applications are abandoned in favor of continuing applications that claim benefit of the filing dates of such parent applications, e.g., divisional applications in which the parent applications are not abandoned.
As they did in the first two papers, Quillen and Webster continue to neglect situations in which more than one patent is issued in a patent family, especially wherein a patent does NOT issue on the parent. Quillen and Webster continue to neglect PCT by-pass route applications.
***
In QW3, I get taken to task for ignoring a 2004 paper Organization for Econ. Co-operation and Dev., Patents and Innovation: Trends and Policy Challenges 1 (2004).
available at www.oecd.org/dataoecd/48/12/24508541.pdf in my 2004 papers on patent grant rate in JPTOS and JIP, which concerned QW2.
In reality, the OECD paper cites ONLY TO QW1 (not to QW2, of concern to me in my 2004 papers) AND
--> the 2004 OECD paper cites to QW1 for a point about foreign prosecution yielding different results than US prosecution [!!!]
--> the 2004 OECD paper fails to cite to Clarke's 2003 paper in JPTOS, which demolished QW1
--> the 2004 OECD paper has nothing to do with the issues I raised in my 2004 papers [that QW'2 methodology was wrong both legally and numerically]
**
Although QW3 raises the bogus issue that Clarke's 2003 paper did not talk about patent grant rate, QW3 failed to cite my second paper in JIP, which discussed the difference between "allowance rate" and "issue rate."
from my second paper in JIP:
As noted in our previous article, there was some divergence in the raw data for applications filed and applications allowed between the papers of Quillen/Webster and Clarke. This divergence was insufficient to account for a divergence in grant rate between 85% (adjusted Quillen/Webster in QWII) and 74% (Clarke). There has been a suggestion that the divergence in grant rate arises from a difference in use of "allowed" applications (Quillen/Webster) and "issued" applications (Clarke).
(...)
The difference in basic grant rate percentage between using "allowed" applications and "issued" applications is of the order of 2%. The difference in using "allowed" applications and "issued" applications does not account for the difference in the numbers for corrected grant rate between Quillen/Webster (85%) and Clarke (74%).
***
Separately, my second paper in JIP pointed to an inconsistency in the way Quillen/Webster and Lemley/Moore were distinguishing "allowed" and "issued" applications:
Third, in working with data on issued patents to identify the number of patents based on continuing applications, Clarke merely did a more thorough job of what was attempted in the QWII. As stated in the footnote of Lemley and Moore: "The 85% number provided in the revised Quillen et al. study is based on actual data about the applications that ISSUE based on continuations," [emphasis added], QWII made their correction based on data on ISSUED applications, not on ALLOWED applications. If the use of data on ISSUED applications means an assumption that every allowed continuation resulted in a patent, then it is an assumption made by both Quillen/Webster and Clarke. Because it is assumed by both, this assumption is not a basis to select the results of Quillen/Webster over those of Clarke, as was done by Lemley and Moore.
[IPBiz post 1689]
The Federal Circuit Bar Journal (Kimberly Moore, editor), which brought us Quillen/Webster 1 and 2, has now brought out the third installment: Cecil D. Quillen, Jr. and Ogden H. Webster, Continuing Patent Applications and the U.S. Patent and Trademark Office--Updated, 15 Fed Cir B J 635-677.
If you naively thought that models for patent grant rate should be required, as a constraint of logic, to be limited to a grant rate of no more than 100%, the Federal Circuit Bar Journal is not for you. From page 661:
When the USPTO Grant Rate is determined on the assumption that all continuing applications are an attempt to patent the subject matter of their parents and that the parent applications were abandoned in favor of the continuing applications, the overall Grant Rate for 1981 - 2005 is 95%. This calculation establishes an upper bound for Grant Rate estimates. The assumption on which this calculation is based is not true in all instances, e.g., when a divisional or other continuing application is filed and its parent is not abandoned. The result is that, in some instances, this calculation produces a Grant Rate estimate that is above 100%. Such a result is not realistic and results form the fact that not all parent applications are abandoned in favor of continuing applications that claim benefit of the filing dates of such parent applications, e.g., divisional applications in which the parent applications are not abandoned.
As they did in the first two papers, Quillen and Webster continue to neglect situations in which more than one patent is issued in a patent family, especially wherein a patent does NOT issue on the parent. Quillen and Webster continue to neglect PCT by-pass route applications.
***
In QW3, I get taken to task for ignoring a 2004 paper Organization for Econ. Co-operation and Dev., Patents and Innovation: Trends and Policy Challenges 1 (2004).
available at www.oecd.org/dataoecd/48/12/24508541.pdf in my 2004 papers on patent grant rate in JPTOS and JIP, which concerned QW2.
In reality, the OECD paper cites ONLY TO QW1 (not to QW2, of concern to me in my 2004 papers) AND
--> the 2004 OECD paper cites to QW1 for a point about foreign prosecution yielding different results than US prosecution [!!!]
--> the 2004 OECD paper fails to cite to Clarke's 2003 paper in JPTOS, which demolished QW1
--> the 2004 OECD paper has nothing to do with the issues I raised in my 2004 papers [that QW'2 methodology was wrong both legally and numerically]
**
Although QW3 raises the bogus issue that Clarke's 2003 paper did not talk about patent grant rate, QW3 failed to cite my second paper in JIP, which discussed the difference between "allowance rate" and "issue rate."
from my second paper in JIP:
As noted in our previous article, there was some divergence in the raw data for applications filed and applications allowed between the papers of Quillen/Webster and Clarke. This divergence was insufficient to account for a divergence in grant rate between 85% (adjusted Quillen/Webster in QWII) and 74% (Clarke). There has been a suggestion that the divergence in grant rate arises from a difference in use of "allowed" applications (Quillen/Webster) and "issued" applications (Clarke).
(...)
The difference in basic grant rate percentage between using "allowed" applications and "issued" applications is of the order of 2%. The difference in using "allowed" applications and "issued" applications does not account for the difference in the numbers for corrected grant rate between Quillen/Webster (85%) and Clarke (74%).
***
Separately, my second paper in JIP pointed to an inconsistency in the way Quillen/Webster and Lemley/Moore were distinguishing "allowed" and "issued" applications:
Third, in working with data on issued patents to identify the number of patents based on continuing applications, Clarke merely did a more thorough job of what was attempted in the QWII. As stated in the footnote of Lemley and Moore: "The 85% number provided in the revised Quillen et al. study is based on actual data about the applications that ISSUE based on continuations," [emphasis added], QWII made their correction based on data on ISSUED applications, not on ALLOWED applications. If the use of data on ISSUED applications means an assumption that every allowed continuation resulted in a patent, then it is an assumption made by both Quillen/Webster and Clarke. Because it is assumed by both, this assumption is not a basis to select the results of Quillen/Webster over those of Clarke, as was done by Lemley and Moore.
[IPBiz post 1689]
posted by Lawrence B. Ebert at 4:42 PM
0 Comments:
Post a Comment
<< Home