Written Testimony of Kweilin Ellingrud Partner, McKinsey & Company and co-author of Power of Parity work on the gender gap

Meeting of 5-18-16 - Promoting Diverse and Inclusive Workplaces in the Tech Sector

Good afternoon and thank you for the invitation to join you in today's discussion. I am Kweilin Ellingrud, a Partner with McKinsey & Company and co-author of our Power of Parity work on gender equality, which I led in partnership with our macroeconomic think tank, the McKinsey Global Institute. McKinsey is a global management consulting firm, and over the last decade we have invested in growing our expertise on gender equality, recognizing that advancing the status of women is a major driver of economic growth. Most recently, we formed a 5-year partnership with Lean In, partnered with Facebook on TechPrep, and published an article on "What it will take to Make the Tech Industry More Diverse" in the Harvard Business Review. My remarks today build on the insights from that work.

The EEOC has been critical in moving American workplaces towards greater openness, inclusivity, and freedom from discrimination. Given the EEOC's current examination of the technology industry, I will focus my remarks on three areas: 1) The significant talent shortage in technology and the current state by gender and race in that field, 2) What we know about the drivers of the gap by gender and race, and 3) Potential solutions to address that gap, both for the EEOC and for others who can help in this challenge.

CURRENT STATE OF THE INNOVATION CHALLENGE: THE TECHNOLOGY TALENT SHORTAGE, AND WOMEN AND MINORITIES IN THE TECHNOLOGY SECTOR

Technology is pervasive in all of our lives, and the rate of change is rapid. We are becoming increasingly dependent on technology, and it is disrupting some industries (as Uber has in transportation, and Amazon in retail) as well as changing the dynamics of others (such as Insurance). In fact, there is not a single industry that is not affected by technology. The technology workforce that we are talking about today is critical not just for the technology companies that we typically think of (Google, Facebook, Apple, IBM, etc.), but it is also critical for every other company that needs talent with technological and computing skills in order to be competitive.

Technology is critical at the company level, and it is also critical for the overall national economy. When used well, technology can improve standards of living, bring rapid development, and help reduce inequalities in prosperity. Given the power of technology both globally and here in the U.S., it is critical that women and minorities play a significant role in technology. Diversity in technology, and the inclusive workplaces that are required to get there, is important for three main reasons: 1) We will not fill the talent shortages that we face unless we include women and minorities more actively, 2) We will not harness the power of technology to tackle a broader set of societal challenges unless the technology workforce better reflects society, and 3) Equality is an important principle for all of us, and one that the United States and the EEOC is committed to. It is not only the right thing to do, it is the smart thing to do for the American people, our companies, and our nation as a whole.

So what is the Technology talent shortage?

The fastest growing job categories are computer job categories in the U.S., and the job growth rate of engineering and computer science jobs is twice the national averageⁱ.

While this is an exciting story, there is an impending talent crisis. By 2020, there will be 1.4 million computing jobs in the U.S., but only 400,000 suitably-skilled workers to fill them, meaning 1 million jobs will go unfilled or filled from foreign workers if current trends continueii. The labor gap makes it more difficult for U.S. businesses to grow and drive the economy. At the same time as we confront this talent gap, we face the current challenge that women and minorities are not well represented in technology. This is part of the problem today, and could potentially be part of the solution to addressing the talent shortage tomorrow.

How are women and minorities doing today in Technology?

The gender gap in technology is getting bigger, not smaller. In fact, women are earning fewer computer science bachelor's degrees today than 30 years ago. Women earned 37% of computer science degrees in 1984, and earn only 18% of them today.ⁱⁱⁱ Similarly in engineering, women receive about 20 percent, 24 percent, and 23 percent of bachelor's, master's, and doctor's degrees, respectively.^iv Similarly, Blacks and Hispanics make up 30% of the U.S. population, but earn just 20% of computer science bachelor's degrees.

Given the gender and racial gap in terms of who is earning degrees, it is no surprise that there is a lack of diversity throughout the technology talent pipeline. Based on a 2015 Women in the Workplace survey of large U.S. companies sponsored by Lean In and McKinsey, the Technology sector has 37% of entry-level roles held by women. Female representation drops down to 20 - 25% in the mid-management levels, and drops all the way to 15% in the C-suite.

Throughout the talent pipeline-from the 37% women at entry-level to the 15% representation in the C-suite, we see many of the same issues within technology as we see in other industries. There are three main issues:

1. Issue #1 is the promotion rate: Women have a lower likelihood of being promoted at every single level. So what starts off as an uneven talent funnel is leaky and gets even more disproportionate over time. Women are not leaving at higher rates, and in fact women in the C-suite leave at roughly half the rate as men in the C-suite. Instead, women are stalling within the talent funnel because they are less likely to advance.
2. Issue #2 is the mix of staff vs. line roles: Women tend to take on more staff roles vs. line roles over time than men. A staff role is a support function role-HR, Finance, Legal, etc. And a line role is a business or P&L role. While both are important, CEO and other senior leadership roles tend to be pulled from line or P&L roles over time. And what we see is a significant gender skew in women holding more staff roles over time than men. While two thirds of men have line roles both at the manager level and at the C-suite level, 57% of women at the manager level have a line role, but that falls to 43% of women in a line role in the C-suite.
3. Issue #3 is women have smaller and more junior sponsorship networks: Sponsors are those who create opportunities for others-and what we see is that sponsorship networks tend to look very different for women vs. men. Men tend to have sponsorship networks that are predominantly men. Women tend to have sponsorship networks that are either majority women or a mix of men and women. Women also tend to have smaller sponsorship networks. Given smaller sponsorship networks and less female representation across all levels, we end up with women who have both smaller and more junior sponsorship networks vs. me. That means women are less likely to have the champions they need proactively creating professional opportunities for them.^v

While these three issues are patterns that we see across all sectors, they are particularly acute in technology given the low diversity levels. In addition to not being well represented, women on average feel that the technology workplace is not an inclusive one. 38 percent of women in technology feel that their gender will make it difficult for them to advance in the future. This figure is the highest across all sectors surveyed. A study by NCWIT showed that women in the science, engineering, and technology (SET) fields feel significant pressure to put in extensive face-time, much more than women in any other sector (71% in SET, compared to 19% in all other sectors).

So in summary, we have low diversity and inclusion in the technology sector, and it is a more stark situation than many other industries.

DRIVERS OF THE GENDER AND RACIAL GAP IN TECHNOLOGY

What are the drivers of the gap?

There is a gap at two levels-both within the talent pipeline in terms of conscious and unconscious bias and even before the talent pipeline begins in terms of the education to employment pipeline. We need better awareness, access, self-confidence, and perceptions of the technology industry by women and minorities.

Conscious and unconscious bias in the workplace:

We see technology companies starting to publish their gender diversity results publicly in an effort to improve their performance. We also see some technology companies starting to implement blind resume screening in a push to get more diverse candidates. As has been well documented, resumes that are otherwise identical but have just the name changed to be either a traditionally female name or an ethnically diverse name drive lower perceptions of the candidate's skills and qualities.

Also in the category of unconscious bias, a number of companies have identified talent and other processes that are gender neutral on the face of it, but affect men very differently from women and minorities. One large tech company, for instance, has a policy where individuals nominate themselves for promotion, and promotion committees only review candidates when they feel they are ready for promotion. A seemingly logical approach to ensure the committee spends their valuable time only on candidates who are ready. What happened in practice, however, was that women tended to wait longer on average to apply for a promotion, and men on average applied much earlier. This resulted in a gender skew in the promotion committee that was never intended.

There are also companies that are using advanced analytics to understand and assess unconscious bias much more strongly throughout their people processes. They are searching for keywords in review memos and other sources for gender-skewed feedback on things like "abrasive style" and "lack of executive presence," for women vs. men.

Yet others are setting ambitious targets for women in leadership roles, and asking their leaders to help develop the sponsorship and retention programs to get there. Some cutting edge companies are increasing accountability and transparency against their diversity goals, and driving that accountability into each business unit and part of the company. Some companies beyond technology are putting as much as 15% of their senior leadership compensation at risk and tying it to hitting diversity targets.

It is often difficult to tell at the industry level the difference between conscious and unconscious bias. Both can be systematic and have very much the same results in terms of low levels of diversity in an industry. In the category of conscious bias, #Gamergate happened as recently as 2014, where sexism, harassment, and threats of rape and murder were made against a few women in the gaming industry. It is incidents like these that use online harassment to make a hostile work environment or even a hostile industry environment for some.

Creating a more balanced pipeline of technology talent:

To achieve a more diverse technology workplace and help close the talent gap, it will be critical to get more women and minorities in the technology pipeline. We will need more women and minorities to be aware of computer science and related topics, be willing to try it, have access to it (either in school or in an after-school program), consider majoring in it in college, pursue the degree, and ultimately enter a career in technology. Across STEM, the issue is particularly acute in Computer Science (CS), and I will share with you some of the more targeted statistics for women, Blacks, and Hispanics in CS. These insights are based on a survey that McKinsey did in partnership with Facebook to launch a program called TechPrep. We conducted interviews with over 2,400 students, parents, recent college grads, and employees working in computer science-related industries.

Awareness and willingness to try: Awareness is a big barrier for blacks and Hispanics, but willingness to try is the biggest barrier for women when it comes to computer science.

• Awareness is high for women, but much lower for black and Hispanics vs. Asians and Whites-this is a significant barrier given awareness is the first step
• Although 91% of girls are aware of computer science, only 44% say they are willing to try it
• This is quite different between the U.S. and Brazil, interestingly, which points to a set of a societal perceptions that are not specific to sex but rather culturally-specific. In the U.S., 73% of girls are familiar with engineering, and 33% say they are attracted to it. In Brazil, in contrast, only 53% of girls are aware of engineering as a profession, but 72% of those aware of it are interested in working in the field
• 53% of blacks and 54% of Hispanics are open to trying computer science. Participation in computer science among black and Hispanic high schoolers is increasing, but remains low

Access: Three quarters of U.S. public high schools do not offer computer science, and this leads to far less access for black and Hispanic students.

• Compared to white and Asian students, black and Hispanic/Latino learners have far less access to computer science classes or after-school programs (Hispanics have the lowest access and blacks the next lowest level of access). White and Asian learners, for instance, are 1.3 - 2 times more likely to go to a school where there is a robotics club offered
• Women, blacks and Hispanics are also less likely to know an adult who works in the technology industry, meaning they have less exposure to role models. Men are 1.5 times more likely to know someone working in the field vs. women

Confidence: Confidence is a significant issue for all women regardless of race-there is a ~30% confidence gap for women, but is not an issue for black and Hispanic men.

• Women of all races are not as confident as their male counterparts. Regardless of their ethnicity, women are more likely to be unsure of their own abilities in computer science than their male counterparts, with their confidence level at roughly 70% of that of men^vi
• The 30% tech confidence gap is much higher than in other fields-11% in business, 7% in humanities, and 5% in the social sciences. This is the primary reasons more women are not participating in the tech industry. 50% of boys in the survey felt they would be good at working with computers, but only 33% of girls said so
• Confidence is not an issue with Black and Hispanic/Latino students. In fact, they have a more positive perception of computer science compared to white and Asian students. Some 50% of blacks and 42% of Hispanics say they believe they would be good at working with computers, compared to 35% of whites and 35% of Asians
• Minority females are less confident than whites. Black and Hispanic girls have the lowest perception of their computer science abilities among all groups, at half the confidence level of white males, and two-thirds of the confidence of black males^vii

Perception: There is a significant perception gap between how women view computer science vs. men. Females are 2.5 times more likely than males to say that people who work in technology are "boring" or "not like me." Women also drop out of pursuing computer science at rates that are 1.2 - 1.7 higher than male counterparts at every stage of the journey. There is 50-65% dropoff for women at each stage of the journey between those who are willing to try a computer-science-related class, actually trying it, continuing to learn the subject, intending to pursue a degree in it, actually pursuing a degree in it, and finally pursuing a degree in technology.

SOLUTIONS TO ADDRESS THE GENDER AND RACIAL GAP IN TECHNOLOGY

So what are the solutions we should consider given the talent pipeline and education to employment pipeline challenges in terms of awareness, access, and confidence? As the EEOC looks ahead to the next fifty years, there are two main issues we should try to solve: Helping companies create more robust talent pipelines by identifying and eliminating unconscious and conscious bias in the technology industry, and supporting greater access to computer science in K-12 to create a more robust education to employment pipeline.

1) Helping companies create a more robust talent pipeline for women and minorities with greater representation at all levels of management. This includes identifying and eliminating both unconscious and conscious bias in the workplace, as is the EEOC's objective.

This could include more advanced analytics to understand systematic discrimination and patterns of practice both at the company and industry level.

2) Supporting greater access in K-12 to ensure that all children regardless of gender or race are exposed to computer science and have a chance to try it given awareness and access are key.

The key to diverse and inclusive workplaces is to tackle the issue farther upstream, and get more children exposed to computer science, excited about pursuing it, and ultimately entering the workplace. Today, computer science is not taught in three quarters of all public U.S. schools. Virtually no federal funding goes to support computer science offerings in K-12 schools. At the same time, today almost 500,000 unfilled U.S. jobs require some level of computer-science understanding, and there are only 50,000 computer science graduates a year^viii. Our economic future depends not just on economic and job growth, but also on ensuring we are educating our students such that they have the right skills for the jobs being created. Broader access and uptake in computer science is a critical factor in making that match happen between the demand and supply of talent in the U.S.

In an article in the Washington Post just a month ago, dozens of top business leaders (Apple, Facebook, Target, Walmart, AT&T, etc.), as well as a bipartisan coalition of 27 governors, called on Congress to help provide computer science education in all K-12 schools. The group argued that K-12 computer science courses are "essential to give children the skills they need to be successful in the modern economy and help fuel job and economic growth." Many would argue that teaching Computer Science helps critical thinking, logic, and creativity skills-all skills that will be important in our future workforce.^ix

Without equal access, it will be decades if not generations until we can close the gap for women and minorities in technology. As we discussed earlier, access is one of the biggest challenges we face, and the workplace of today and tomorrow requires technology and computer science skills in a very different way than the workplace of the last 50 years.

CONCLUSION

In conclusion, there is a significant gap to address for women and minorities in technology today at two different levels: in the workplace and in the education to employment funnel. The U.S. faces a significant talent gap in technology of 1 million jobs by 2020, and women and minorities could be a significant part of the solution. Addressing the barriers at recruiting and hiring is important but is not enough-there is a talent shortage in the pools of qualified candidates for technical jobs. We must address conscious and unconscious bias to create a more inclusive workplace as well as look upstream to broaden K-12 exposure to computer science and related areas.

Thank you for opportunity to share some thoughts with you today. I look forward to helping you and others create a more diverse and inclusive pipeline and workplace in technology.

Footnotes

ⁱ "What It Will Take to Make the Tech Industry More Diverse" by Susan Colby, Helen Ma, Kelsey Robinson, and Lareina Yee on March 15, 2016 in the Harvard Business Review.

ⁱⁱ "Technical fields of study" is defined as STEM degrees, but excluding life, earth, and social sciences. Growth over last decade of 1.5% CAGR. Conservative assumption that 75% of graduates enter computer and engineering occupations.

ⁱⁱⁱ "What It Will Take to Make the Tech Industry More Diverse" by Susan Colby, Helen Ma, Kelsey Robinson, and Lareina Yee on March 15, 2016 in the Harvard Business Review.

^iv Digest of Education Statistics, National Center for Education Statistics, March 14, 2016, nces.ed.gov.

^v "Women in the Workplace" survey of 118 companies and 30,000 employees, Lean In and McKinsey, Oct 2015.

^vi "What It Will Take to Make the Tech Industry More Diverse" by Susan Colby, Helen Ma, Kelsey Robinson, and Lareina Yee on March 15, 2016 in the Harvard Business Review.

^vii "What It Will Take to Make the Tech Industry More Diverse" by Susan Colby, Helen Ma, Kelsey Robinson, and Lareina Yee on March 15, 2016 in the Harvard Business Review.

^viii "Top business leaders, 27 governors, urge Congress to boost computer science education," Emma Brown, April 26, 2016, Washington Post.

ix "Top business leaders, 27 governors, urge Congress to boost computer science education," Emma Brown, April 26, 2016, Washington Post.