Friday September 2 2016
Despite an estimated investment by the US technology industry of $0.8bn - $1.2bn in diversity efforts over the past five years, women’s representation at tech firms has actually decreased. In the last 15 years representation of women has fallen a percentage point, to represent just 28 percent of the tech workforce today. Given women are 50.8 percent of the total population, this equates to a shortfall of 900,000 – 1Mn women for the tech industry. A new report by Intel and Dalberg Global Development Advisors, “Decoding Diversity: the Financial and Economic Returns to Diversity in Tech”, digs into the issue to help explain why. But there’s no question it’s a substantial missed opportunity: achieving representational gender diversity in leadership alone would be an estimated $320-390Bn in increased market value for tech firms.
First the “why”, as in: “why have diversity ratios and inclusivity been so slow to improve?” Two critical barriers have been less discussed before now: the effect of limited data, and the lack of evidence. Data on diversity rates remains minimal, such that we still don’t have an accurate picture of even basic statistics; for example, only 10% of tech companies representing 39% of employees report workforce diversity data via their EEO-1 forms today. And that data doesn’t include other useful metrics such as pay levels by gender or race, or exit rates and reasons. Without that data, there’s a poor sense of how much improvement is needed, and where the greatest gaps may lie. Just as critically, little evidence exists of diversity program effectiveness, and what evidence exists hasn’t been well shared or integrated.
Below, two illustrative examples of existing findings which need broader sharing and acceptance…
- Annual diversity training doesn’t work, based on a longitudinal study of over 700 U.S. companies looking at training for managers. According to the Intel and Dalberg study, “it may even decrease representation of African American women.”
- Sponsor employees, don’t just mentor them, because research by the Center for Talent Innovation (CTI) finds that it’s the proactive advocacy from sponsorship that makes a difference in promoting pay raises, high-visibility assignments, and promotions. But only 14% of women in tech report that they feel they have a sponsor.
…and a selection of areas where funding is critically needed to create the evidence base:
- What are the most effective strategies to slow minority turnover?
- How is diversity impacted by offices set up outside of Silicon Valley?
- In what ways can employee performance criteria most effectively control for bias?
- What is the ROI of recruiting at colleges with greater proportions of underrepresented minorities?
Improvements in diversity take time, but the solutions don’t have to be complex or costly. Harvey Mudd College in Southern California increased the number of women in their Computer Science program from 13 in 2010 to 117 in 2015, with initiatives such as:
- Redesigning the introductory CS course to focus on problem-solving real-life challenges, and renaming it from “Introduction to Programming in Java” to “Creative Approaches to Problem Solving in Science and Engineering Using Python”
- Hiring more female CS faculty and giving them a role in introductory courses
- Instituting simple classroom rules to keep experienced programmers from intimidating others in class
- These approaches have already paid off at other schools: for example, at the University of California in Berkeley, an introductory course was renamed from “Introduction to Symbolic Programming” to “Beauty and the Joy of Computing.” Female enrollment increased by 50% as a result, with that class having more women than men for the first time.
Companies could seek to adapt and replicate these and other successful diversity initiatives to capture the estimated $320-390Bn opportunity. But doing so takes a commitment to action (for example, the White House’s Tech Inclusion Pledge for U.S.-based companies), data on the state of diversity today, and evidence for where diversity investments can be most effective. Those would be a great place to start.
