The 4th Industrial Revolution (IR) for Women and Work — Some Observations
This post is by Soo Min Toh, Associate Professor of Organizational Behaviour and HR Management at the University of Toronto Mississauga.
With the headlines focused on Artificial Intelligence (AI) — and with the progressive intelligence of machine learning, and its abilities to perform the work of people more quickly and effectively — it is no wonder that there is great concern with what this would mean for the future of work. This is particularly so for all women workers, who are inevitably affected by significant technological advancements, automation, and computerization. As history suggests (e.g., see Ensmenger’s [2012] account of the rise and fall of women computers and programmers), the work of women is particularly precarious, and susceptible to changes across technological, social, and political climates. With the 4th IR, the impact on women is being felt in terms of employment opportunities, as well as barriers therefor. This paper seeks to outline what some of these opportunities and barriers are, and to a provide discussion of a few recommendations for practice and policy.
Opportunities for Women in Employment
Some have argued that “the future of AI will be female” (Searles, 2018). To remain relevant in the new world of work, people have to do the work that involves higher-order thinking — critical, creative, and imaginative thinking — and work that demands high levels of social and emotional skills. These social and emotional skills are often seen as female-oriented skills (Cortes, Jaimovich, & Siu, 2018). In fact, the occupations that are seeing explosive growth are the ones that involve high cognitive skills, such as work in STEM, and high interpersonal skills, such as with health care, social work, education, and psychology. This suggests greater returns to education for women. The 2016 World Economic Forum (WEF) report, “The Future of Jobs,” shows high growth areas in knowledge and creative work, including business and financial operations, management, and computer and mathematical work, as well as jobs for data analysts. Specialized sales representatives are to become critically important too. Technology has also enabled women to enter and participate in the economy because of the flexibility afforded by telecommuting, and new mobile applications that allow for self-managed work schedules in the gig economy.
In addition to greater job opportunities, there is also an ever-growing need and recognition that women and people of colour must participate in STEM jobs because of the growth in this area, and the need for a diversity of perspectives and interests. For instance, two women of colour recently discovered bias in existing facial analysis programs that are unable to correctly detect darker-skinned faces. This suggests a greater need for cultural and gender sensitivity, as well as a further commitment to equitable solutions. The business case for diversity is no longer the dominant argument for increasing diversity in organizations, but organizations are recognizing that they are losing out on half of the hiring pool if they fail to be inclusive. Many companies are now openly committed to hire more women into roles that have traditionally been held by men.
Yet, there are still many barriers and threats in the environment that are further sharpened by the 4th IR. These include: loss of jobs/work, barriers to entry and persistence, and an invisible bias through algorithms.
Barrier Faced by Women in Employment
Loss of jobs/work. The 2016 survey from the WEF reports that, although women are not more likely to lose their jobs to technology, they are less likely to work in an industry where their careers benefit from advancements. Women represent small numbers in the fast-growing STEM job families, but also low numbers within high-loss job families such as manufacturing and production, or construction and extraction, that will disproportionately affect men. However, female employment is also concentrated in low-growth or declining job families such as sales, business and financial operations, along with administrative work. In absolute terms, men will face nearly 4 million job losses and 1.4 million gains, approximately one job gained for every three jobs lost, whereas women will face 3 million job losses and only 0.55 million gains, more than five jobs lost for every job gained. In the 2015–2020 period, there is nearly one new STEM job per four jobs lost for men, but only one new STEM job per 20 jobs lost for women.
To make matters worse, men are more likely to know which jobs technology will replace — this according to a survey from the recruiting platform ZipRecruiter. This differential awareness gives men a head start in retraining for new emerging jobs and added advantage over women in obtaining jobs in growth areas.
Barriers to entry and persistence. Despite increasing numbers of women obtaining higher education, university enrollment of women in STEM fields remains low. Intentions and desire to enter into the fields have been suggested to be formed as early as middle school as girls develop attitudes and expectations about STEM education, culture, and work (Shapiro & Sax, 2011). The choice to concentrate in STEM is also related to a growing interest in mathematics and science, rather than one in enrollment or achievement. This suggests that current policies, focused on advanced-level course-taking and achievement as measures to increase the flow of students into STEM, may be misguided (Maltese, & Tai, 2011).
In the workplace, selection systems that favour men exclude the hiring of women in STEM (Ensmenger, 2012). This serves to grow what was already a strongly masculine culture in STEM. Even if jobs for women do increase and pay well, they are excluded by the workplace culture. One study (Athena Factor) found that women are more likely than men to leave (41% compare to 17%; Hewlett et al., 2008), suggesting that programs getting more women to enter the industry are futile if they ultimately choose to leave. Organizational factors, including work–life balance, organizational climate, and mentoring, also influence women’s career development and persistence in technologically-oriented jobs (Trauth, Quesenberry, & Huang, 2009).
Invisible bias through algorithms. Because humans build and train AI, it reflects our biases and assumptions, including those related to gender and race. These biases can become encoded in the algorithms because we are blind to them, or may result from missing areas in a dataset the engineer or researcher is not aware of. This can cause programs to arrive at poor decisions. For example, facial-analysis programs from major technological companies demonstrate both racial and gender biases — the darker the skin for women, the more serious the error rates were (up to an almost 50% error rate); LinkedIn had an issue where highly-paid jobs were not displayed as frequently in searches by women, because of the way its algorithms were written. The training of AI on deficient and already biased data further exacerbates the problem, particularly when the decisions derived from AI appear to be objective, better, more efficient: the superior solution to human fallibility and bias.
Some Solutions
According to the 2016 WEF report, “The conclusion is clear. If current industry gender gap trends persist[,] and labour market transformation towards new and emerging roles in computer, technology and engineering-related fields continues to outpace the rate at which 3 women are currently entering those types of jobs, women are at risk of losing out on tomorrow’s best job opportunities[,] while aggravating hiring processes for companies due to a restricted applicant pool[,] and reducing the diversity dividend within the company.” Yet, current solutions tend to rely more heavily on suggestions that women should do more (or less) of certain behaviors to improve their lot. This quite unfairly places the onus on women to adapt, “lean in,” or even change a system that is inherently biased against women and minorities. Instead, others believe that the focus should be on how we as a society change the systems, practices, and mindsets, such that equity and inclusion become normative in the workplace (e.g., Toh & Leonardelli, 2013).
For instance, proactive hiring and equitable compensation for women in otherwise male-dominated STEM jobs is absolutely key, as well as education and training policies to ensure that girls are not already disadvantaged in these fields at an early stage. AI4All (ai-4-all.org), for example, is a nonprofit working to create pipelines for underrepresented talent, through education and mentorship programs around the U.S. and Canada. These programs are meant to give high school students early exposure to a socially-minded style of AI. Ideas about leaders also need to be expanded, such that effective leaders are not just seen to possess masculine traits. Rather, there should be a realization that different types of leadership match the needs of the situation at hand. This intuition is the most critical component of effective leadership: not the gender or race of the leader.
Successful hiring, along with investments in STEM training for women, will be for naught if women end up exiting their professions for more welcoming and inclusive ones. Retention of women in STEM should thus also be a major priority. Workplace cultures that are dismissive, and even hostile, toward female scientists and engineers — that require hours and expectations unsupportive of working mothers, that instill a sense of isolation in the lone woman on the team, or that obscure channels of information, opportunities, and career advancement — turn women away, even after they have entered the fields. The Athena Factor report provides an example of successful programming at Alcoa — Women in Operations Virtual Extended Network (WOVEN). WOVEN seeks to provide opportunities for women employees in operations roles who are geographically dispersed: opportunities to meet and support one another virtually as they pursue their careers. WOVEN also develops a well-planned and well-monitored program for each participant, that offers exposure to different divisions and different types of management experience, with specific targets for participants to obtain managerial roles within six to eight years (Hewlett et al., 2008). The successful retention and career development of women in STEM would in turn help attract more women to these fields.
While STEM training is critical, we should not forget the “softer” skills. For people to stay relevant in the new world of work, we will need to lean in to the skills that make us most human, and do the things that machines cannot do well yet. Hence, in addition to STEM areas, critical thinking, creativity, imagination, and social skills will remain relevant and should be paid a premium. The growing call for students to be trained in the arts and humanities appears to respond to this need: education and training policies could be flexible enough to adapt to these evolving priorities. A growing need for workers in healthcare will also provide important opportunities for women affected by automation. Opportunities for training and retraining in this field should continue to be accessible to women. Currently, men affected by the 4th IR have not been as willing as women to transition into these so-called “pink collar” jobs. Women who are willing and able to adapt themselves into new areas will weather the impact of computerization better than men.
Eliminating bias should also be priority. Improving the training data used for AI is one way. Technologies are now available to test training data and algorithms for encoded biases. Organizations like Algorithms Justice League aim to raise awareness of the bias that is 4 programmed in written code. Organizations looking to adopt AI technologies in their products and decision making, particularly those related to hiring, should do their research. Before buying into a new technology, find out if the developers included diversity in their training data, and remain critical of potential biases against women or racial minorities in the results derived from such software (Huhman, 2017).
Finally, when organizations find themselves automating jobs, they should know that retraining is not the only option, and make women aware that starting from bottom in a whole new career is not the only way out. Instead of dismissing women whose jobs are obsolete, organizations might take a closer look at what new challenges these women can take on, and find opportunities for women to step up, rather than down, in their careers with the organization. Finding ways to tap into the creativity and commitment of these employees could reap significant rewards for all involved (Huhman, 2017).
Our role as researchers and thinkers is thus: to continue to test these ideas for impact, to discover new ways to foolproof our futures from the rapidly changing technological environment of the 4th IR, and to pay attention to the human stories behind these changes.
References
Cortes, G. M., Jaimovich, N., & Siu, H. E. (2018). The” End of Men” and Rise of Women in the High-Skilled Labor Market (No. w24274). National Bureau of Economic Research.
Ensmenger, N. L. (2012). The computer boys take over: Computers, programmers, and the politics of technical expertise. MIT Press.
Hewlett, S. A., Luce, C. B., Servon, L. J., Sherbin, L., Shiller, P., Sosnovich, E., & Sumberg, K. (2008). The Athena factor: Reversing the brain drain in science, engineering, and technology. Harvard Business Review Research Report, 10094, 1–100.
Huhman, H. (2017). New research shows AI is going to make the workplace much worse for women — Unless you step up. Url: https://www.inc.com/heather-r-huhman/new-researchshows-ai-is-going-to-make-workplace-much-worse-for-women-unless-you-step-up.html. Retrieved: April, 27, 2018.
Maltese, A. V., & Tai, R. H. (2011). Pipeline persistence: Examining the association of educational experiences with earned degrees in STEM among US students. Science Education, 95(5), 877–907.
Searles, R. (2018). The future of AI will be female. Url: https://qz.com/1175985/thefuture-of-ai-will-be-female/. Retrieved: April 27, 2018.
Shapiro, C. A., & Sax, L. J. (2011). Major selection and persistence for women in STEM. New Directions for Institutional Research, 2011(152), 5–18.
Toh, S. M., & Leonardelli, G. J. (2013). Cultural constraints on the emergence of women leaders: How global leaders can promote women in different cultures. Organizational Dynamics, 42(3), 191–197.
Trauth, E. M., Quesenberry, J. L., & Huang, H. (2009). Retaining women in the US IT workforce: theorizing the influence of organizational factors. European Journal of Information Systems, 18(5), 476–497.
World Economic Forum (2016). The future of jobs employment, skills and workforce strategy for the Fourth Industrial Revolution. Url: http://reports.weforum.org/future-of-jobs2016/ Retrieved: April 27, 2018.
