Women are an inextricable part of the fabric of our country and the pattern for our future. Today, as we battle the coronavirus with our lives and quality of life at stake, women are fighting for us on the front lines – they make up nearly 80 percent of the workforce in health care. Women are also working hard outside of the medical field to keep their families afloat. Half (49 percent) of employed women in the United States serve as their family’s main breadwinner. Many more are helping their families make ends meet by earning supplemental income. Eighty-three percent of sellers on Etsy self-identify as women, for example. While broadband is enabling many women to continue working while quarantined, others have their hands tied.
The 2019 National Compensation Survey from the Bureau of Labor Statistics reports that only 7 percent of civilian workers in the United States, or roughly 9.8 million of the nation’s approximately 140 million civilian workers, have access to a “flexible workplace” benefit, or telework. And those workers who are able to telework are largely managers, other white-collar professionals and the highly paid.
Beyond telework, Americans on the wrong side of the digital divide – those without high-speed internet access – are at a major disadvantage in multiple critical areas of life. They also miss out on the opportunity to take advantage of distance learning, e-commerce, telemedicine, and access to life-saving information, especially during a crisis like COVID-19.
Globally, the proportion of men using the internet in 2017 was 12 percent higher than women, according to the International Telecommunication Union. There are also 200 million fewer women than men who own a mobile phone, per a March 2018 report from the Organisation for Economic Co-operation and Development.
Women are on the wrong side of the digital skills gap, too. According to a 2019 McKinsey report, “worldwide, 40 million to 160 million women — 7 to 24 percent of those currently employed — may need to transition across occupations” by 2030, due to automation. To successfully overcome this challenge, women “need to be skilled, mobile, and tech-savvy, but women face pervasive barriers on each, and will need targeted support to move forward in the world of work.” The report notes that it will be vital for women to develop “the access to and knowledge of technology necessary to work with automated systems, including participating in its creation.”
Unfortunately, only 26 percent of computing jobs are held by women (this number has been on a steady decline for years). Twelve percent of engineers at Silicon Valley startups are women. And more than 20 percent of women developers over the age of 35 are still in junior positions.
Augusta Ada King, Countess of Lovelace, was an English mathematician who many believe published the first algorithm. If there were more Ada Lovelaces building algorithms today, maybe these digital decision engines would be more regardful of women.
We may not know everything about algorithms, but we should recognize that inherent and even unconscious bias that are present during the development process have a very real impact on users and society as a whole.
I once met a chief technology officer in Europe, and I talked to him about discrimination by algorithm. He immediately countered with the objectivity of the math in a data set, and I explained that data sets are gathered by people. I confess, I still have joy when the light hit his eyes, and he said, “And if only men assemble the data sets, they would not realize that there is bias.”
Basically, if more women were hired by the big tech platforms, then those women would actively participate in building algorithms and piecing together the data sets. This is how you work to prevent gender bias. Lack of diversity is a recipe for skewed data sets.
We need more women, more people of color, and more diversity of thought to create a world that is accessible to everyone.
My favorite example of discrimination-by-algorithm is “Mrs. Apple” being given a lower credit limit than “Mr. Apple” when they both equally own the same assets. Web programmer and author David Heinemeier Hansson — who has over 390,000 Twitter followers — tweeted that he received 20 times the credit line for an Apple Card that his wife did, despite them filing joint tax returns, being married for a long period and living in a community-property state. The response of Goldman Sachs was that the algorithm was vetted for potential bias by a third party and doesn’t even use gender as an input.
But as rebutted by WIRED’s Will Knight:
This explanation is doubly misleading. For one thing, it is entirely possible for algorithms to discriminate on gender, even when they are programmed to be “blind” to that variable. For another, imposing willful blindness to something as critical as gender only makes it harder for a company to detect, prevent, and reverse bias on exactly that variable.
With our lives upended by COVID-19, it’s increasingly clear that government, industry and NGO leaders all need to play a role in supporting the transition of millions of women into more tech-centered jobs, and every woman in this country needs access to broadband. Getting high-speed internet in the hands of more Americans requires policies that encourage continued broadband investment. Armed with education, training, and an internet connection, women have the best possible shot to continue supporting their families, even when they can’t show up to work in-person.
Katherine Johnson was the living computer relied upon by John Glenn in 1962 to calculate the orbital equations that would control the trajectory of the Friendship 7 mission, from liftoff to splashdown. There are many more Ms. Johnsons waiting to be discovered — they just need an opportunity.
Kim Keenan is co-chair of the DC-based Internet Innovation Alliance.
Morning Consult welcomes op-ed submissions on policy, politics and business strategy in our coverage areas. Updated submission guidelines can be found here.
Get the latest global tech news and analysis delivered to your inbox every morning.