Average Patent Application Pendency

In its annual Performance and Accountability Reports, the USPTO publishes a statistic identified as the "Pendency time of average patent application." The reports describe the figure as the "average time (in months) between filing and issuance or abandonment of utility, plant, and reissue applications. This average does not include design patents." I have charted the PTO's reported average pendency below in the first chart. The chart shows that for 2011, the PTO was successful in keeping average pendency under three years.

120711_1926_AveragePate1

Unfortunately, the PTO's numbers are somewhat misleading. The main technical problem with the USPTO numbers is that any request for continued examination (RCE) is counted as an abandonment of an application and a refilling of a new application. Thus, an application that remains pending for four years before issuance and that includes an RCE filing will be counted in the PTO calculations as two applications with an average two-year pendency. The PTO reporting also includes other application "reworkings" such as continuation applications and divisional applications that tend to be processed somewhat faster than original non-provisional applications.

The chart below adds two additional pendency calculations: One calculation ignores RCEs in the pendency calculation – looking instead at the time from non-provisional filing to issuance.  Thus, under my calculation, an application that remains pending for four years before issuance and that includes an RCE filing will be counted as one application with a four-year pendency.  The second calculation also ignores RCE filings and additionally limits the data set to only original non-provisional applications (excluding continuations, divisionals, and CIPs). As expected removing the RCE artifact from the data adds considerably to the average pendency. (Note here that my chart is lacking because it does not include statistics for pendency of abandoned applications.)

120711_1926_AveragePate2

The final chart shows a pendency histogram for patents issued in October and November of 2011 (again ignoring RCE filings). The median pendency is 3-4 years. 98% issued within nine years of filing.

120711_1926_AveragePate3

11 thoughts on “Average Patent Application Pendency

  1. Clients trying to time costs due to monthly/yearly budgets and patent “reviews” are the ones that cause 99% of the EOT’s and delays in my experience.

    Word.

  2. Fine…clients, firms, applicants, pro se inventors…whatever. Your experience is not my experience. Not really to the point I was trying to make.

  3. FYI, “firms” rarely do that because they eat the EOT fees and clients don’t like to see delays caused by their service providers.

    Clients trying to time costs due to monthly/yearly budgets and patent “reviews” are the ones that cause 99% of the EOT’s and delays in my experience.

  4. As far as accountability for the PTO…would not a “real” or more “accurate” pendency measurement exclude applicant’s response time. There seems (to me) to be a lot of firms that take the entire 6 month period to respond or a least longer than the 3 month SSP. At least such a number would give the PTO a “real” number with which to work with to try and improve on. It seems that the PTO is being held responsible, pendency wise, for applicants who take a longer time to respond.

  5. The PTO has for many years also misleadingly counted other mere refilings of the same application as “new” applications to artificially restart their “pendency” clock, including old-fashioned continuations.

    So, how many continuations are still being filed instead of RCEs? Would they significantly further increase true application pendency numbers?

  6. MND – just to make sure that you got this. My measure is “real” pendency. The PTO, on the other hand, calcluations pendency as filing-to-RCE and separately RCE-to-issuance.

  7. You might then indicate that you are merely measuring time to RCE filing then, and not “real” pendency.

  8. Thanks guys – I updated the post to correct the ambiguity. I do not remove cases with RCEs from the data set. Rather, I simply consider RCE filing irrelevant to the pendency of an application. Thus, using my calculation, an application that remains pending for four years before issuance and that includes an RCE filing will be counted as one application with a four-year pendency. In that same situation, the USPTO would identify two applications with an average pendency of two-years.

  9. Good question – as the RCE “actual” pendency will greatly negatively influence the results from three ways – first will be the actual “real” count, second will be the removal of the false “start-over” counts and third will be the lengthening/removal of short non-RCE first counts.

    Of course, all three of these are actually related.

  10. By ignoring, do you mean removing those cases from the data entirely, or using their proper pendency (i.e., 4 years with RCE is a single 4 year pendency)?

Comments are closed.