Time to Turn the US Rejection Office back into the US Patent Office: Part 3

USPTO’s Office of Patent Quality Assurance (OPQA) results show that US patent examiners adhere tightly to performance incentives that value speed and volume over quality. See Parts 1` and 2 of this series. Here in Part 3, we drill into the quality element of the patent examiners performance and appraisal plan (PAP) to identify criteria that results in examiners more likely rejecting an allowable over allowing a rejectable claim.

Beginning with the 2010 reform to the patent examiners’ count system, USPTO management worked closely with the POPA union to incrementally and systematically downgrade the PAP’s emphasis on quality. Background, data and details are provided in Does the USPTO’s Roadmap to Improved Patent Quality Lead to Lake Wobegon?

First, USPTO management and POPA came to an agreement that the level of attention necessary to formulate proper rejections under 35 U.S.C. 102, 103 or 112 would be downgraded from Office of Personnel Management’s (OPM’s) Position Classification Standard advanced level – tasks capable of being independently performed only by experienced examiners – to OPM’s basic level – tasks assignable to trainee examiners which essentially require only cursory review and can be performed without any sound understanding of procedural and substantive patent law. 

To illustrate OPM standards, by way of analogy, a law firm may task a clerk to perform basic functions, a paralegal to perform advanced functions and an attorney to perform legal functions. Ensuring that the drawings are sequentially numbered is an example of a basic task that could be adequately performed by a clerk in a law firm or by a GS-7 entry level examiner. Drafting an obviousness rejection or responding to an obviousness requires a sound understanding of procedural and substantive patent law.  This task cannot be adequately performed by someone working at the clerk level. Yet the agency which put the word NO in inNOvation, reviews junior and primary examiners’ 103 rejections as basictasks that could be prepared without any sound understanding of procedural and substantive patent law.

Second, USPTO management and POPA came to an agreement that a “category 1” errors could not be called on 10 out of the 19 quality major activities (QMAs) – tasks that are required to be performed to properly reject or allow patent claims – unless the same error occurred several times in the same year after training and mentoring. Only a handful of office actions per examiner are reviewed each year and via POPA-management agreement, office actions should be randomly selected. The category  proviso in the 2010 PAP, available as Attachment 1 here, coupled with no ability to perform targeted reviews made it nearly impossible for managers to hold examiners accountable for errors in 10 of their 19 QMAs.

These changes to water down examination quality criteria were not prompted by any simplification of the patent law. Its noteworthy that these efforts to decrease patent examination quality preceded the 2012 creation of the PTAB and its Inter partes review (IPR) process ostensively intended to address poor quality patents. See Figure 1.

To keep up with production and pendency requirements, its understandable that examiners let quality slip on the “category 1” tasks and began to provide basic level cursory review when drafting rejections under 35 U.S.C. 102, 103 and 112.

Congress noticed. In 2014, hearings were held to address perceived declines in patent quality amidst allegations of timecard fraud. One of those testifying, former APJ William Smith, aptly noted in his 2014 Congressional testimony, “[t]here was a saying during my tenure at the PTO that a patent examiner never got fired for doing bad quality work, as long as they did a lot of it.”

The USPTO continued to give examiners mulligans for 10 of their 19 required tasks and allow GS-14 primary examiners to make statutory patent determinations at the basic performance level.

In 2019, then-IP Subcommittee Chairman Senator Tillis held hearings entitled Promoting the Useful Arts: How can Congress prevent issuance of poor quality patents? On page 9 of his testimony, witness Drew Hirshfeld, then-Commissioner for Patents, announced “beginning next fiscal year, there will be a new performance evaluation that will serve as a roadmap to improve patent quality.” 

The Quality Element of the 2021 PAP is shown in Figure 2. To their credit, one improvement was made. The Category 1 error label was removed for those 10 required tasks.

The USPTO devised a brand new non-error based assessment which it applied to seven of the 19 quality major activities (QMAs). 

Figure 3. QMAs evaluated under the non-error-based assessment standard

A non-error based assessment means that the examiner can fail to perform these tasks or improperly perform these tasks without facing any consequences.

Imagine an auto company producing a car that requires correct assembly of 19 components to run safely but by union-management agreement, only 12 of those 19 components are inspected to verify proper installation.

Imagine if 19 tasks are required to ensure the safe landing of an airplane but the Federal Aviation Administration only assessed the air traffic controllers’ performance for 12 of those steps.

Why would a patent examiner waste precious time checking the application for technical accuracy, treating information disclosure statement, evaluating case law or declarations, or promoting compact prosecution, if it’s not required? With the advent of non-error based assessments, patent examiners are no longer required to correctly evaluate the sufficiency of declarations or affidavits or make the record reasonably clear and complete. Boiler plate language like “applicant’s response has been carefully considered but found non-persuasive” is often seen as sufficient. OPQA survey results discussed in Part 2 show how the requirements of the PAP are at odds with the MPEP’s guidance to fully address and consider the technical and legal aspects of applicant’s response, including arguments, declaration and evidence. 

Impact on the Backlog

One would think that by loosening quality requirements, more patent applications would be examined faster. This is not the case. Despite an increase in the number of examiners, patent application production units (a measurement for work moved) has declined from 627,760 in FY2016 to 532,858 in FY2024.

First action pendency has risen from 16.2 months in 2016 to 19.9 months in 2024. For applications examined between Oct. 1, 2024, to June 25, 2025, first action pendency is at 23.3 months. This is far outside the agency ‘s obligation to issue first actions within 14 months from filing or provide patent term adjustment. 

Despite a decrease in filings, the backlog currently stands at 805,773, down from a January 2025 high of 837,928. While the new administration’s efforts have chipped away at some of the backlog, the overall trend lines are incompatible with America’s need for a functioning patent office to protect and promote our IP interests. 

Granted, other factors including declining IT tools and adoption of the flawed CPC classification system which often misroutes applications have also contributed to the increased backlog of unexamined applications. But churning out lots of sticky, low-quality rejections will not help America regain our competitive technological edge. 

One cannot help but wonder why, in parallel to implementing PTAB IPR proceedings, the USPTO would steadily decrease examiner’s attention on quality. Was it to create a stream of poorly examined patents to fuel future IPRs? Somerightfully argue that “[t]he current anti-innovation posture of the USPTO is so problematic that severe action appears necessary.”

The Government Accountability Office’s report titled “Patent Office Should Strengthen its Efforts to Address Persistent Examination and Quality Challenges” supports these concerns.

Part 4 of this series will offer a roadmap intended to turn the US Rejection Office back into a US Patent Office.