Time to file comments on DOCX yet again

(Update:  nine comments got filed.)

This has been a busy week for Federal Register notices by the USPTO that relate to the ill-fated DOCX patent application filing initiative.  We already knew about two Federal Register notices about this:

• • 88 FR 36956 dated June 6, 2023, blinking a fourth time on the start date for the $400 non-DOCX penalty, and
  • 88 FR 37036 dated June 6, 2023 blinking a second time on the ending date for the ability to file an “auxiliary PDF” file by which the applicant may try to protect against USPTO errors in rendering DOCX files.

Now comes a third Federal Register notice:

• • 88 FR 37039 dated June 6, 2023, requesting comments on USPTO’s estimates of the burden that the DOCX application filing initative will impose upon applicants.

In this blog article I briefly discuss this third FR notice.  In a later blog article I will discuss this third FR notice in greater detail. 

This third FR notice is remarkable in many ways.

Estimating the percentage of filers who will switch from PDF to DOCX.   For two decades now, applicants and practitioners have been able to e-file US patent applications in PDF format.  The USPTO has earned the trust of applicants and practitioners over these two decades with the predictability and reliability of the PDF file format.

The USPTO launched its DOCX initiative in beta test in August 2016, and extended its availability to all filers in April 2020.  The USPTO has by now presented well over one hundred and fifty webinars, trying to convince applicants and practitioners to file in DOCX rather than PDF.  The USPTO has by now done over fifty email blasts, probably amounting to well over half a million distinct email messages, promoting the DOCX initative.  In its presentations, the USPTO tells the attendee (falsely) that filing in DOCX is somehow safer than filing in PDF.

Countering the USPTO’s webinars are webinars (recordings available here) presented by professional associations and practitioners, explaining the malpractice risks associated with use of the DOCX e-filing system.

The natural question one might ask is, how successful has the USPTO been in inducing filers to switch from PDF to DOCX?  In this blog posting I reported polling results suggesting that the success rate might be in the range of 20% to 40%.

The number of patent applications filed at the USPTO in recent years has averaged around half a million per year.  In this FR notice, the USPTO estimates that the number of applications subject to this DOCX requirement would be 411817.  (That number is smaller than the half a million number, in part because some kinds of applications such as design and plant and PCT are not subject to the DOCX requirement.)

By way of background, USPTO’s success rate with many of its past initiatives has been close to 100%.   USPTO’s effort to convince patent application filers to e-file in PDF rather than filing on paper, for example, reached 99% within one year.  This DOCX initiative, however, even after seven years of outreach and advocacy, has fallen far short of 100%.  How far short?

The FR notice says that the USPTO estimates that 247751 DOCX applications  would get filed per year.  This amounts to an estimated success rate for the USPTO of about 60% of filers.  The remaining 164066 applications (the remaining 40%) would be filed by applicants and practitioners who choose to pay the $400 non-DOCX penalty rather than subject their patent applications to the risks of using DOCX.

The USPTO does not explain where it got the 60% figure (which is far short of a 99% success rate for PDF).  I imagine the USPTO took into account the very limited success of its outreach efforts, along with the ever-increasing number of webinars being presented by patent firms and professional associations that warn of the risks of the DOCX program.

What costs the DOCX initiative would impose upon the patent community.  One set of costs gets imposed on those who do what the USPTO asks them to do, namely who choose to file in DOCX.  Those filers would, among other things, need to spend time comparing the USPTO-modified DOCX file with the applicant’s originally uploaded DOCX file, to try to detect validation and conversion and rendering mistakes by the USPTO.  The USPTO estimates that this would add 30 minutes to each e-filing session.  Multiplying this by an estimated value of the time of the filer ($435 per hour), this works out to an estimated burden of about $53 million per year.

A second set of costs gets imposed on those who choose to pay money to be able to continue to file in PDF, thus avoiding the risks of filing in DOCX.  The USPTO estimates this at a government-fee cost of about $49 million per year.  It gets this number by estimating what fraction of non-DOCX filers would file as large entities, small entities, or micro entities (62%, 29%, and 9% respectively) and multiplying those application counts by the amount of the penalty ($400, $160, or $80 respectively).

One thing that I find interesting about this is that even if the USPTO were to shift upwards or downwards its guess as to what fraction of applicants and practitioners would or would not choose to switch from the safe PDF to the unsafe DOCX, the overall cost to the patent community does not change very much.  It ends up being in the neighborhood of $100 million per year regardless of the percentage who choose to switch from the safe PDF to the unsafe DOCX.

We can then compare this with how much money the USPTO says it will save by implementing the DOCX requirements.  The only specific dollar amount the USPTO has provided for its supposed savings is $3.15 per patent application in an avoided cost for performing OCR (optical character recognition) on each patent application (84 FR 37398, July 31, 2019).  Of course this money gets saved only with respect to those patent applications that get filed in DOCX, which the USPTO estimates will be only 60% of the eligible applications (247751 per year).  Multiplying these two numbers together yields an annual savings to the USPTO of $780 thousand.

One is prompted to pause momentarily to compare the magnitudes of the various numbers.

The DOCX initiative is estimated to impose a cost on the patent community of about $103 million annually, or just over a tenth of a billion dollars per year.

The DOCX initiative is estimated to save the USPTO about $780 thousand annually.

These two numbers are strikingly out of balance.  The disproportionality is around 130 to 1 (cost imposed upon USPTO customers by DOCX compared with benefit to the USPTO gained from DOCX).

USPTO’s past lack of candor to the Office of Management and Budget.  The main point of this FR notice is to take a step in the direction of disclosing to the Office of Management and Budget the paperwork cost of a particular data collection, namely the paperwork cost of the DOCX initiative.  It is thus of great interest to compare the paperwork cost estimate in this FR notice with the USPTO’s estimate that it provided to the Office of Management and Budget in its past filings regarding the DOCX initiative.

One of the past filings is 84 FR 37398 (July 31, 2019), where the USPTO said:

The Office did not identify any monetized costs and benefits of the proposed rule, but found that the proposed rule has significant qualitative benefits with no identified costs.

Another past filing is 85 FR 46932 (August 3, 2020) where the USPTO said:

The Office did not identify any monetized costs and benefits of this Final Rule but found that this Final Rule has significant qualitative benefits with no identified costs.

It is difficult to reconcile USPTO’s representations to OMB in 2019 and 2020 that the DOCX initiative would impose “no identified costs” with USPTO’s representation to OMB in 2023 that the DOCX initiative would impose costs of a tenth of a billion dollars.

Time for comments.  This FR Notice sets a closing date of August 7, 2023 for filing of comments.  It will be important for patent applicants and practitioners to file comments in response to this Notice.