Cotropia Study of USPTO Allowance is Rate is Total Bunk – Methodology Can Easily Result in “Corrected Allowance Rates” far in Excess of 100%, the Lower the Actual Allowance Rate Goes

(My thanks to Domenico Ippolito for his research on this post)

In his article Getting Patents is Preposterously Easy under Obama, Timothy Lee asserted that the Patent and Trademark Office “needs a way to permanently reject patent applications” and called giving the Office that power a “common-sense reform.”[1] In support of this assertion, Mr. Lee refers to Apple’s “slide to unlock” patent, implying that the patent was allowed only because of exhaustion on the part of the PTO “after six years of back and forth.”  He goes on to say that 92% of patent applications eventually result in allowance, and that this high percentage shows that the PTO is over-“permissive” and contributing to “a proliferation of low-quality patents.”  None of these points survive a closer look.

Having decided a priori that the slide-to-unlock patent should not have issued, Mr. Lee then turns to the allowance rate of the PTO in general, accusing the Office of allowing 92% of all applications, which “may mean that the patent office approved applications that didn’t deserve a patent.”  As Mark Twain famously said, “there are three kinds of lies: lies, damned lies, and statistics.”  So let’s tread cautiously where statistics are involved, and consider carefully what they reflect.  The PTO conveniently provides a dashboard with statistics showing that 51% of applications are being allowed.[2] As an applicant, that statistic is important: it means there’s a 51% chance that you’ll get an allowance on a single round of examination, before substantial additional fees kick in.  However, since an RCE can be filed after a Final Rejection, the 51% figure does not accurately reflect the chance of an original application eventually issuing as a patent.  Fortunately, the PTO dashboard also provides that number: it’s currently 70%.

Why is the PTO’s adjusted allowance rate so much lower than the 92% reported by the article Mr. Lee refers to?  The answer is simple: because the 92% figure is not an allowance rate.  The University of Richmond (UR) report “corrects” the 51% allowance rate reported by the PTO by adjusting the number of application disposals.[3] Walking through a calculation, the “nominal” number of application disposals was 548,051.  That number was corrected by subtracting out the RCEs, leaving 390,143 disposals.  Taking 51% of 548,051 as the number of allowances, and 390,143 as the number of disposals, a partially corrected allowance rate of 74% was obtained.  While this number is substantially higher than the rate of about 68% shown by the PTO for 2012,[4] at least the UR report and the PTO agree that this is an allowance rate.

However, UR goes on to calculate an allowance rate “corrected … for all Refiled Continuing Applications (including CIPs).”  The number generated after this “correction” is of uncertain value, at best.  To generate the corrected value, the number of allowances is held constant, but the number of disposals is reduced by the number of “continuations,” “refiled continuing applications,” and other “continuing applications.”  In this manner, an “allowance rate” of 89% was calculated for 2012.  This rate shows that for every application that is finally disposed of, an average of 0.89 patents issues.  That sounds similar to an allowance rate, and the rate is close to 100%, so does that mean that virtually every application eventually issues as a patent?  No.  That statistic was already calculated as 68% or 74%, depending on the source.  In fact, this number doesn’t have an upper bound of 100%, so the fact that it is close to 100% doesn’t have any particular significance.  To demonstrate the point, assume a 50% nominal allowance rate out of 500,000 disposals.  Further assume that continuations are filed in 90% of the cases.  Then the “corrected” allowance rate would be 250,000 allowances divided by the 50,000 cases finally disposed of – yielding a “corrected” 500% grant rate.  The statistic generated in this manner – whether the result is 89%, 92% or 500% — has no particular significance either to the inventor considering filing a patent application or to the outside observer interested in determining if the PTO is overly-permissive, as Mr. Lee alleges.  Calling it an allowance rate is sloppy at best, and deliberately misleading at worst.

The truth of the matter is that the increase in RCEs, far from showing that applicants can bully the PTO into granting dubious patents, instead shows that patent prosecution is harder, is taking longer, is costing more, and is resulting in the issuance of patents with narrower claims.  Contrary to Mr. Lee’s assertions, the PTO’s successful efforts in reducing pendency have resulted primarily from an increase in Office personnel, not from reduced quality of examination.[5]


[3] See Patent Applications and the Performance of the U.S. Patent and Trademark Office, Cotropia et al., available at http://papers.ssrn.com/sol3/papers.cfm?abstract_id=2225781 (numbers used in the article are taken from Appendix C).

[4] http://www.uspto.gov/dashboards/patents/kpis/kpiAllowed.kpixml.

[5] See http://www.uspto.gov/dashboards/patents/kpis/kpiExaminers.kpixml (showing 6,642 examiners in November 2012 and 8,041 in March 2014); see also http://www.uspto.gov/patents/stats/examstaffing.jsp (showing 6,063 examiners in Q1 2009).

 

One Response to “Cotropia Study of USPTO Allowance is Rate is Total Bunk – Methodology Can Easily Result in “Corrected Allowance Rates” far in Excess of 100%, the Lower the Actual Allowance Rate Goes”

  1. JTS says:

    Nice post, thank you.

    Any PTO allowance-rate is a non sequitor. The PTO will be happy to collect issue fees and maintenance fees on narrow and worthless claims. There’s no way to measure how permissive or restrictive examination is, collectively, that wouldn’t be prohibitively expensive.