The basis for the perception of this anti-woman bias is a paper by K. Jensen, B. Kovács, and O. Sorenson from the Yale School of Management entitled "Gender differences in obtaining and maintaining patent rights," published in Nature Biotechnology 36: 307-310 (2018) (https://bit.ly/3x1PYCb).¹

These researchers evaluated patents, prosecution histories, maintenance fee payments and forward citations (within and outside the Patent Office) for 2.7 million patents granted from 2001-2014. For these patents, women made up only 10% of patent inventors and 15% of life sciences inventors. Women inventors' claims were more likely to be rejected, less likely to be appealed, had a smaller percentage of claims allowed, were subject to more extensive prosecution, and required more amendments.

Post-grant women inventors' patents garnered fewer post-grant citations and were less likely to be maintained. Perhaps most surprisingly, these disparities were found to be more prevalent with life sciences patents (which had a disproportionate percentage of female inventors).

In performing these assessments, gender was identified by first name conventions (Michael more likely to be male, e.g.) using the Social Security Administration database and two commercial databases for gender frequencies (wherein 94.1% of names associated 95% of the time with one gender). In their evaluations these researchers determined a "proportion woman" statistic for each patent depending on the number of women named inventors; also added were controls for number of claims, number of inventors, and size of the assignee/applicant.

In addition to the overall findings set forth above, the paper reports seeing some effect for technology group/class that differed from rather than raw data. For example, without controlling for technology group/class "women predominant" application were 21% less likely to get allowed claims; this figure dropped to 7% when technology classification was taken into account. (The paper hypothesizes without data that these results could be due to Simpson's paradox: women may be applying to technology classes having lower intrinsic allowance rates).

In more fine-grained statistics, women-led inventor teams were 2.5% less likely to go to appeal; the number of allowed independent claims in these patents dropped by 20% when granted; 4.3% were less likely to be maintained; there were 11% fewer citations of these patents; and examiner citations were reduced by 3.5% compared to non-women predominant patents. Paradoxically the higher numbers of women inventors in life science fields does not improve these statistics better; in fact, the disparities were greater (more disadvantageous) than the overall results.

For example, patent applications naming all-women inventors were found to be 11% less likely to have a patent granted than all-men inventors and had 28% fewer citations.

For "rare" names (i.e., where gender cannot be easily inferred) the only party knowing applicant gender is the applicant (absent instances for example where there is an interview with the examiner).

Among inventors with common names, women had an 8.2% lower probability of having their application accepted than did men, whereas among inventors with rare names, women had only a 2.8% lower probability of acceptance than did men. The authors attribute to examiner bias 2/3rds of the observed gender disparities. But "patents authored by women with rare forenames, and who were therefore not easily identified as women, were cited approximately 20% more often than the average patent by a male inventor with a rare forename," all else being equal.

The authors suggest that these disparities could be ameliorated, if not substantially diminished, if the patent-prosecution process were made more "blind" to the identity of participants. They suggest that "patents and patent applications could list only the initials of the forenames of the inventors on patent applications and could require communication between examiners and applicants to occur on a platform that would maintain the anonymity of the applicants." They mention that this approach has worked in other situations, like applicants for orchestral musicians playing behind screens.

Substantially if not directly resulting from this study, Congress in 2019 first introduced the IDEA Act, which failed to get out of committee (see below). The current bill, introduced in March of this year, is thought to have a better chance, in part due to changes in the provisions of the bill.

The current bill² (S.632 in the Senate and H.R.1723 in the House of Representatives) provides that Title 35, Chapter 11 of the U.S. Code be amended to require the PTO Director to provide for the collection of demographic data, voluntary for applicants. The categories of such data includes gender, race, military or veteran status, and "any other Director decides to include." Importantly, the bill does not include information on sexual orientation, ethnicity, age, national origin, educational level and disability, which an earlier version introduced in the last Congress had (but within the Director's purview to include). This information is to be kept confidential and separate from application file and the Director is compelled to establish procedures to keep it that way, especially from examiners. This information is also to be exempt from the Freedom of Information Act (FOIA) and the Federal Information Policy Law (see Chap. 35, Title 44) does not apply.

The Director is also required to publish by January 31 of each year a Report comprising:

(1) the total number of patent applications filed, "disaggregated" with respect to this demographic information and by technology class number, technology class title, country of residence of the inventor, and state of residence of the inventor in the United States;

(2) the total number of patents granted, "disaggregated" with respect to this demographic information and by technology class number, technology class title, country of residence of the inventor, and state of residence of the inventor in the United States; and

(3) a discussion of methodology and summaries of aggregate responses.

The data shall be made public, allowing for data to be cross-tabulated to review subgroups but personal identifying data shall be kept confidential and not disclosed. The bill also requires a biennial report evaluating the data collection process, how accessible the data is to the public and what needs to be improved, no later than 2 years after the date of enactment and every 2 years thereafter.

These bills were introduced March 9, 2021. The House bill was introduced by U.S. Rep. Nydia M. Velázquez, D-N.Y., and for U.S. Rep. Steve Stivers, R-Ohio, and in the Senate by U.S. Sens. Thom Tillis, R-N.C., and Mazie Hirono, D-Hawaii; Sens. Chris Coons, D-Del., and Patrick Leahy, D-Vt., are co-sponsors of the Senate bill. In the Senate, S.632 was voted out of Senate Judiciary Committee April 29, 2021, (on a vote of 14-7) and was made part of the U.S. Innovation and Competition Act (S.1260)³ on May 18.

In the House the bill was referred to the House Judiciary Committee Subcommittee on Courts, Intellectual Property, and the Internet and has not come up for a vote.

In press releases and on their websites these legislators rely on additional statistics pointing out the problem of gender disparity in patent procurement. These include statements that "only 22 percent of all U.S. patents list a woman as an inventor and that women only make up 13 percent of all inventors" (Sen. Tillis) and that "[s]tudies show that women and minorities apply for and obtain patents at significantly lower rates than their male, white counterparts. Only 22 percent of U.S. patents list at least one woman as an inventor and that women make up only 13 percent of all inventors. African American and Hispanic college graduates apply for patents at approximately half the rate of their white counterparts." Rep. Stivers notes in favor of the bill that "[t]here are more than 11.6 million businesses owned by women nationwide, and yet, only a fifth of patents list female inventors," and that "women-owned businesses generate over $1.7 trillion in sales." All sponsors agree that the issue requires Congressional intervention in the form (for now) of these bills being enacted.

One reason for the positive prospects for this bill's passage is because the PTO's Director has an express requirement to amass statistics on demographic indicia such as sexual orientation, ethnicity, age, national origin, educational level and disability. The categories were contained in the language of the earlier bill and were for that reason opposed in the Senate by inter alia Ted Cruz, R-Tex., and (purportedly) the Catholic Church.⁴

The bill is supported by BIO, IPO, Innovation Alliance, Invent Together, AWIS, AIPLA, Patent Progress and SIIA.

The statistical evidence seems to solidly support the conclusion that there are disparities in the number of women who are inventors, who apply for patents, and who are awarded patents as well as the "strength" of such patents (i.e., their scope and value in providing exclusivity to their invention sufficient to motivate investors to provide needed capital for commercializing their inventions). There is also evidence that there are fewer businesses relying on women inventor-predominant patents, as reflected in the lower rates of maintenance fee payments for such patents (it is rare that maintenance fees are paid for patents not protecting successful investments).

But it is reasonable to ask whether these statistics support the conclusion that the issue is in the Patent and Trademark Office, in particular that the discrimination is affected at the examiner level. Part of the deficiency is in assuming that the outcomes (fewer patents, more rejections asserted, more amendments needed to overcome them) is merely gender determinative, and that it is the result of the simple calculus that examiners are reluctant to grant patents to women. This equivalence ignores some other important demographic considerations; for instance, these studies do not control for the existence of male and female examiners, and do not show any differences between the examination practices, allowance rates, outcomes, and other characteristics between patent applications examined by male versus female examiners. The same can be said for the existence of male and female practitioners (lawyers and patent agents) and any disparities arising from gender differences for these actors. And then there is the issue of whether there is any difference for a male examiner when interacting with and responding to a male practitioner as opposed to a female, or the converse, and the dynamic of these interactions.

The bill avoids such questions by requiring that any demographic information included by the Director in the accumulated statistics be kept from examiners. Putting aside the flaws in such a regime with regard to whether there will be opportunities (ample or otherwise) for examiners to discern the gender identity of applicants, merely accumulating the statistics recited in the bill (or as expanded ) may provide information regarding whether discrimination continues to exist, but will not provide any insight into why it persists or why it has been diminished (other than the facile conclusion that if examiners don't know the gender identity of inventors the disparity will miraculously disappear).⁵

More worrisome is the possibility that the bill is just an example of "virtue signaling" intended to demonstrate an attempt by Congress to address this sort of discrimination, with the hope if not expectation that the information obtained by the Patent Office (and kept from the patent examiners) will identify sources of discrimination (such individual examiners or Patent Office culture?) to permit corrective action. It has salient characteristic of "doing no harm" because information is assumed to be beneficial (and more or better yet specific information directed to the issue will be even more beneficial). But, of course, the question that needs to be asked is not how or whether such information should be obtained; the question is how solutions will be provided for the information the reports (mandated by the bill) disclose.

Bipartisan legislation has been a unicorn in recent Congresses, but if bipartisan support holds, this bill provides the best, most direct route to passage. The bill has been touted by its sponsors as being pro-innovation, pro-American competitiveness and pro-competition, and thus has hallmarks of constituencies that under other circumstances would not be inclined to support legislation directed at addressing gender discrimination. Because it does not actually require the Director of the Patent and Trademark Office to amass statistics, any policy change has been conveniently postponed until such information has been acquired, and by then there will be a new Director who likely will be charged with the hard work of coming up with a plan to remedy gender-based disparities in her Agency. These features suggest that on the one hand the bill will be enacted into law, and on the other hand, for now it will make little difference.

1 See also, Jessica C Lai, "The Role of Patents as a Gendered Chameleon," Social and Legal Studies 30: 203-229 (2021)(https://bit.ly/2UF4l1S); Rembrand Koning et al., "Who do we invent for? Patents by women focus more on women's health, but few women get to invent," Science 372: 1345-1348 (2021)(DOI: 10.1126/science.aba6990); L. Santhanam, "Why are most inventors men?" PBS Newshour Sept. 26, 2016 (https://to.pbs.org/3BzDKUx).

2 https://bit.ly/36VTza7

3 The U.S. Innovation and Competition Act of 2021 brings together the Endless Frontier Act (introduced by Sen. Schumer, D-N.Y., and Sen. Todd Young, R-Ind., which passed the Senate Commerce committee in early May and includes more than 20 bipartisan amendments), the Menendez-Risch Strategic Competition Act of 2021, and the Brown-Toomey Meeting the China Challenge Act of 2021, as well as bipartisan legislation from the Senate Homeland Security and Government Affairs Committee, HELP Committee, Judiciary Committee, and Appropriations Committee.

4 See, A. Karpan, "Lawmakers Reintroduce Bill To Quantify Patent Diversity Gap," IP Law 360, March 10, 2021 (https://bit.ly/3rudJ4D). This article cites the percentage of patents having female inventors at 12.8% in 2020 (an "all-time high").

5 This is different from, for example, the symphony orchestra experience because inter alia there is (or should be) less subjectivity in examining a patent claim than in evaluating a musical performance.