Yes, Virginia — there really is still a pay gap.
How awareness, transparency, and workplace culture can help close it.
I was recently drawn to another Medium article, by Jackie Luo on the subject of pay equity and how transparency, even in every day conversation, can be a positive step toward evening the scales. I was dismayed (though not all that surprised) to see many comments (mostly from those of the male persuasion) questioning whether the gender pay gap is really still an issue. One asked “Is it really true that women are paid less than men for the same role? I feel like today many companies are very eager to hire women, especially in tech” and another suggested “this pay gap is a complete and total myth in tech.. id even go so far as to say Tech has over compensated and now men are being paid less.”
This is an important subject I’ve researched at length, particularly with regard to the gender wage gap in technology professions, when writing my master’s thesis on the implications of gendering of AI. Spoiler alert — neither of those two comments are founded by data and the short answer is yes, we really do still have a significant issue in pay equity between genders (and race/ethnicity). The longer answer is of course far more complicated than a simple “same pay for the same job” equation (which, by some measures is an issue making progress, albeit slowly).
The comments on the Medium article should have come as no surprise to me. In a recent survey conducted by Hired.com (2019), 72% of male tech workers believe there is a gender pay gap, whereas 90% of women tech workers do. In a different consumer survey, conducted by LivePerson, nearly half of all respondents erroneously believed that there is an equal mix of men and women working in tech and AI. While many professional workplaces (and professionals) today may view themselves as meritocratic, the average worker is too often blind to deeply rooted systemic biases in hiring, disbursement of assignments, compensation, and performance evaluations that are not apparent to individuals within the system, according to a paper by the American Bar Association, 2018. They mean to say that ignorance is not bliss: if the average person (both men and women) does not “see” there is a problem, or underestimates the gravity of the situation, we will never solve what is an actual problem.
72% of male tech workers believe there is a gender pay gap, whereas 90% of women tech workers do. -Hired.com
Some Data on Work and Wage Inequality
According to the U.S. Bureau of Labor Statistics (2018) there is a near equal balance of women and men within management, professional, and related occupations, though less than 40% of management positions are held by women. Women make up only 25.5% of computer and mathematical occupations, and within this category — a mere 21% of computer programmers and 19% of software developers are women, although the numbers are higher for systems analysts (39%) and web developers (32%). The World Economic Forumestimates that approximately 22% of the professionals working in the AI field today are women, yet reportedly make up only 15% of the AI researchers at Facebook, and only one-tenth of Google’s AI research staff. The Hired.com survey also reports that in 41% of cases, the candidate pool for tech jobs included only men, and that 60% of the time men are offered a higher salary than women for the same (technology) job.
On a macro level, over the past several decades there has been great progress in closing gender gaps in education, politics, and economics. However, the World Economic Forum reports that this progress has in fact slowed in the past few years and estimates that it will take over 200 years to attain gender parity across the categories of economic, political, education, and health globally. 200 years! That means not I, nor my daughter, nor my great, great, great, great (yes that’s 4 great’s) granddaughters will live and work in parity with their male counterparts.
The digital age has spurred optimism in some, yet there are concerns that the changes coming in the 4th industrial revolution will disproportionately affect women more than men, given that women tend to hold more of the jobs that are lower skilled and more likely to be automated. Currently, women use computers and digital communications in their work more than men, yet increases in wages related to digitization of the workforce disproportionately benefit men as much as 41% over women.
A Complex Problem
In fairness, the Hired.com survey found that the wage gap within the same job role is relatively small and closing (3% in 2019, down from 4% in 2018 for women as compared to men). Yet perhaps the question of “same money for the same job” is not the right question. The reality is that the overall wage gap is influenced far more by the different jobs men and women hold, and today men hold the predominance of higher-wage earning jobs in tech. This is why tracking gender disparities in computer science and especially the field of AI is particularly important. As one of the most in demand jobs around the world, continued lack of progress relative to gender equity could result in further disparities of the already significant gaps in economic equality.
The economic gender gap is largely driven by the jobs that women hold, and some might argue: well isn’t this a choice by women, when they take time off or work less hours during motherhood? Not entirely. A review of more than 98 empirical research studies indicates that there are systemic issues in the workforce that result in women’s disproportionate exclusion from promotion, roles of authority, and higher wage-earning positions. Stereotypical gender role associations can impact hiring decisions, particularly where one is predisposed to more traditional viewpoints on the types of skills and work that are congruent with each gender. In addition, there is known gender segregation within jobs today, which can further impede gender equality, whereby women are generally concentrated in roles of lower status in specific occupations (such as administrative and clerical) and are underrepresented in others (such as supervisory roles and engineering). This phenomena suggests there are persistent job and/or role stereotypes that are at least in part responsible for economic inequality between genders.
So the solution is to get more women into STEM education and the job market — right? Well, that’s a start. Certainly, there is a relationship between the population of women in STEM education and the hiring pool for technology jobs. However, the AI Now Institute suggests that focusing on the “pipeline problem” alone may serve to justify existing diversity problems, particularly by male leaders in the field. They, and others, call for more attention to how women are treated through all aspects of the workforce, including hiring, career progression, and departure. And when more than 65% of women tech workers surveyed by Hired.com say they’ve been discriminated against, compared to 11% of male tech workers, it’s not near enough a solution to just “get more women into tech.”
Social & Workplace Culture Matters
Emily Chang, in her book Brotopia, exposes a sexist and misogynistic culture of the Silicon Valley tech community as a primary reason for the low number of women coming in to tech, as well as the high number of those leaving. At Google, 20,000 employees walked out to protest the company’s handling of sexual harassment and misconduct. A recent class action law suit and press reports surrounding Microsoft’s female employees who expressed concerns over harassment and discrimination at the company, demonstrates this is not a regionally confined problem. As a woman working in the technical field for just over 20 years now, I have been fortunate to not have felt outright discrimination. Yet, like I wholeheartedly believe is the case for most female professionals (at least in tech), I have witnessed and also been subject to the subtle sexism, lewd comments, and otherwise hostile workplace issues mentioned in Chang’s book and the articles referenced.
Another key factor of the lack of women in technology-related careers today may lie in the lack of female role models. Notwithstanding the present gender imbalance in the field, there has been an undeniable omission of women role models in mainstream historical accounts of science and technology, despite many contributions to the field. In some cases, these “omissions” are intentional suppression. Take the very recent case of Katie Bouman, an assistant computer science professor at Caltech and Harvard postdoctoral fellow, who developed an algorithm that was used to capture the first image of a black hole. Within hours of Bouman’s social media posts to celebrate the image, where she specifically highlighted the team who was responsible, some attempted to discredit her contributions, suggesting that she herself had very little to do with the work and that the praise was instead owed to her male (white) colleagues. A 2013 study by Young, Rudman, Buettner, and Mclean found that a female professor of Science, Technology, Engineering, and Mathematics (STEM) serving as a positive role model positively affected the attitudes, identification, and aspirations of female STEM college students without affecting that of male students and even inverted stereotypes of STEM as a masculine venture. Thus, representation by more women in technology, both in media and the real world, could be a key factor to breaking the current stereotype of technology as solely a male endeavor.
Equality Takes Everyone
Solutions to complex problems are rarely a “silver bullet.” One positive comment I noted in the string from Luo’s article stated: “Fair compensation is, and always has been, a group effort. … Identifying blatant disparities between incomes and dealing with the problem is a huge step forward in ensuring that bias loses its role in wage determination. The only factors that should play a role is experience, ability, and compatibility.” Indeed, the idea of “transparency” and continuing to raise awareness that we really do still have a problem to solve, are great first steps. Deep educational changes (including pedagogy), role modeling and breaking stereotypes, encouraging better social norms and behaviors, creating better work environments, and ensuring fair laws (in all countries) will each contribute to progress, or equally prevent it. Economic equality goes beyond gender and as the above comment suggests, it will take all of us — particularly those who are presently in the privileged classes — to solve it, hopefully much faster than the predicted 200 years.
All citations are hot linked within the article. This article contains exerts from my original Master’s thesis Societal Implications of Gendering AI, available at ResearchGate.