SOM study reveals gender inequity in patent process

By: LINDSAY DAUGHERTY

Three professors at the Yale School of Management recently published an analysis that revealed gender disparities in the patenting process. Kyle Jensen, Balázs Kovács and Olav Sorenson reviewed 2.7 million U.S. patent applications and found that women had overall less favorable outcomes than men. Their patents were less likely to be accepted, appealed if rejected or adapted, the study found.

According to the report, published in the journal Nature Biotechnology, women make up only 10 percent of U.S. patent inventors. When the United States Patent and Trademark Office released information on patent applications from 2000–2016 to the public domain about three years ago, the SOM professors saw an opportunity to explore the causes of this unnerving statistic, Kovács said.

The three men found that the gender disparity could not be attributed entirely to application rates. Across all fields, women inventors are 21 percent less likely than men to have their patent application accepted. This percentage decreased to 7 percent when controlled for industry, still a significant discrepancy.

“Our theory is that they don’t know who these [applicants] are, so they just look at the name,” Kovács said.

The applicant’s name is the only potential indicator of gender in a patent application, but the professors thought that gendered names might be enough to evoke biases. The researchers used information from the Social Security Administration and commercial databases to determine the probable genders of applicants based on their names. Then they collected data on the progress and ultimate results of the applications.

“It’s surprising just how ubiquitous the negative effects are,” Jensen told Yale Insights, a School of Management publication. “It’s not just across one metric. They are less likely to have their patents accepted. Fewer claims are granted. The claims are longer, less often maintained, less often cited.”

It is difficult to determine whether the results point to biases within the patent office or to a larger societal problem, Kovács said.

Take, for example, the disparity in independent claims — attributes of an invention that are listed by the inventor to be novel and therefore warrant a patent. The Patent and Trademark Office can return an application with requests for the modification of or removal of certain claims. The inventor then makes changes to her application and resubmits it. At the end of this process, female inventors will have 80 percent fewer independent claims than males.

Women also tended to write longer independent claims, which are more specific and limiting to their patent, according to the study. These disparities might suggest that women are less willing to negotiate the terms of their revisions. On the other hand, it could point to a bias in the Patent and Trademark Office if officials are holding applications from women to higher levels of scrutiny.

Regardless of the causes, the study revealed an issue in innovation previously not discussed, Kovács said.

“The goal is that the United States Patent and Trademark Office is aware that this is happening,” Kovács said.

The study has gotten some attention from journals, but the results have not yet been sent to the Patent and Trademark Office. Kovács is hoping the study might inspire the Office to consider its biases and conduct more research on the roots of the issue.

The SOM research team, however, is not quite satisfied with its work. For example, the team still wants to look into the effects of interviews in the process. Not all patent applications include interviews, but Kovács said that interviews seem to decrease the demonstrated bias. The team plans to investigate this claim with the same database used for the 2000–16 report.

Meanwhile, the gender gap doesn’t seem to be getting any better: During the 16 years studied, the data on patent applications remained fairly consistent.

Source >> https://yaledailynews.com/blog/2018/04/16/som-study-reveals-gender-inequity-in-patent-process/