In past months, we’ve looked at how our data gathering and analyzing processes skew our cars, medical systems, and many more things to work primarily for men.
Following all of this talk about data’s gender gap, what can we do to begin righting its wrongs? There is no one-stop easy answer, rather a laundry list of ways to begin overhauling systems and diversifying our workplaces. It will involve intention and effort at every stage of every organization and require rethinking many priorities.
Some might question if it’s even worth it, and to that I would ask: Would you rather take time to review and change your operations or disenfranchise and endanger more than 50% of the population? Though there are a myriad of ways to begin the work to narrow data’s gender gap, today we’ll primarily look at two: altering hiring practices and intentionally gathering and considering data from a wide variety of sources.
While there have certainly been large strides made in recent years to encourage women to join STEM fields, the number present in the industry is still not representative of the general population. Women make up 47% of the United States’ labor force, but as of 2017 they only held 24% of positions in the technology and information science fields. According to Adam Rowe’s 2018 tech.co article “What Women’s Restrooms Can Tell Us About Tech Conference Sexism”, “An empty women’s restroom functions as the canary in the coal mine of the tech industry.” The article was largely inspired by commercial product manager Lin Classon’s tweets about having restrooms at conferences to herself:
This problem can easily become cyclical. When women do not see themselves represented in speaker lineups, they may become more reluctant to join the industry. On the other hand, when there are fewer women present in the workforce, there are fewer to pick from when organizing events. The most sensible first step here, in my opinion, would be to intentionally overrepresent the number of women in the industry at conferences. Doing so would serve to further normalize women working and pioneering projects in the data and information fields.
In order to properly change the industry’s approach to research and data collection, we must first ensure that the workforce is representative of all genders. However, fixing the gender gap does not merely involve leaving opportunities open for people of varying backgrounds– those in power must actively recruit them. A friend recently told me that a position had opened up on his team at work. He and his coworkers were all white men and he hoped that his superiors would bring in someone who was not white or not a man or both. However, they ended up hiring another person who shared everyone’s identities. When my friend pressed his supervisor about why he decided to hire another white man, his supervisor replied, “We left the position open for ages and no one with a different background who was qualified applied.” This statement contains a number of issues that contribute to gender gaps everywhere, and especially skew the one within the data world.
It is not enough to simply leave positions open to those of different genders (and races, sexual orientations, abilities, etc.), we must intentionally seek out those with different backgrounds to fill them. If the majority of those working on a team are men, a woman may feel unwelcome in that space. She might question what kind of workplace culture led to an all-male team, and if her contributions might be second-guessed by others due to her gender. When only one or a handful of women are present in a workplace, they may feel tokenized. By deliberately recruiting a representative population of women, an organization is showing a base level of commitment to welcoming and including people with different viewpoints and genders. According to LinkedIn’s 2018 Gender Insights Report, women apply to 20% fewer postings than men while on a job hunt.
It is not certain whether this is simply due to women being more selective and particular than men in their job hunt, or if they are less likely to apply to a listing they do not precisely fit the requirements for than men. Either way, recruiters can make an effort to seek out women with backgrounds that sound intriguing for the positions they are hiring, and ask those they know to refer non-male candidates they believe would be up for the job.
In order to begin closing data’s gender gap, we must take a look at the information we already have regarding our structures, vehicles, and systems, and see where the missing pieces are. To revisit a quote I have used previously, Londa Schiebinger of the Gendered Innovations project at Stanford University offered in a 2015 interview with NPR: “My suggestion here would be that all engineers, architects, urban planners, automobile designers go back and look at their standards,” Then ask: “What is the basic standard that things are engineered for? Who is the assumed ideal subject or user?” To engineer things in an equitable way like she suggests, we must ensure that we have gathered data on as diverse a population as will be using the product or system in question. If all citizens in a city will be contenders for using public transportation, we must make sure that the transit system is built to benefit every type of person in that city, as equitably as possible.
We must intentionally gather data on women (and people of color, disabled people, etc.) and utilize it in the same way that we would use data gathered from men. As of the 2010 US census, 49.2% of the US’s population is male. To favor data pulled from less than half of all citizens is careless. When we consider white men, who make up 29.4% of the country’s population, as the default within the default, the way we structure our world becomes downright negligent.
While I do not think this column will cause a global upheaval, it is my hope that those reading it will take a moment to think through their workflow and the organization of their own workplaces, and consider if they are truly as inclusive as they could be. Systemic change begins on an individual level, and every organization touched is a success.
Is there a topic you would like to see me cover in future columns on data’s gender gap? Send me a message on LinkedIn or at firstname.lastname@example.org.