In the first part of this two-part article, we examined the “tech diversity” problem, highlighting the technology sector’s startlingly lack of diversity and its especially abysmal track record of recruiting and hiring Black and Latino workers. In this post, we examine some of the steps industry leaders have already taken to diversify a predominantly white and male workforce. Specifically, we explore how the industry is attempting to fix the diversity problem at three different but critical stages: the stage that happens years before candidates ever meet recruiters (or don’t get an opportunity to meet recruiters); the recruitment stage (e.g., on campus recruitment efforts); and the hiring stage.
High-tech Investments in Public Education
The tech industry is now investing in public education at unprecedented levels, often with an implied mandate to ensure that next-generation workers are indeed ready to work in an increasingly tech-driven world.
The most widely known initiative is the Bill and Melinda Gates Foundation’s College Ready Program. Fueled by the Gates’ own personal fortunes, the foundation—in addition to its overseas initiatives—has positioned itself as a major force in public education across the United States. At the center of the Gates’ College Ready initiative is a conviction that public education needs to be reformed in order to ensure that no child really is left behind and a conviction that a rigorous STEM curriculum is part of the solution. Among other initiatives, the program is investing in courseware and game-based forms of learning to ensure children can learn at their own pace and free up teachers’ time to work one-on-one with a higher number of struggling students. The Gates Foundation is also committed to the increased adoption of analytics in teaching and learning (e.g., providing schools with the tools needed to more effectively measure and monitor what is actually happening in the classroom and to more effectively measure short- and long-term student success).
While many public educators have criticized the Gates Foundation, in part because it has played a major role in the rapid roll out and adoption of the still controversial Common Core Standards curriculum across the US, a growing number of schools and school districts are turning to the foundation to support their efforts. A survey of the Gates Foundation’s 2014 awarded grants reveals that while the Foundation is committed to funding a range of projects, the distribution of funds favors projects that either seek provide advanced technology training to students and/or their teachers or to bring “big data” solutions to education problems.
Despite the Gates Foundation’s efforts to transform public education and invest in charter schools that share the foundation’s mandate, however, there is no clear indication that the foundation’s efforts have or hold the potential to directly increase diversity in the tech industry itself. Here, it is worth nothing that Microsoft, like nearly all the companies mentioned in the first part of this article, has historically had a poor track record on diversity (in 2015, only 1% of its employees identified as Black).
Other privately funded initiatives, while smaller in scale, promise to tackle the tech diversity issue more directly. In 2012, the New York City Department of Education opened the Academy for Software Engineering. The publicly operated school, kickstarted with support from Fred Wilson of Union Square Ventures (a major player in the city’s technology start-up scene), represents a bold attempt to change who has access to quality STEM education. Unlike other STEM high schools in the city, which are “screened” and “selective” and as a result, usually off limits to students with average to low GPAs and test scores, the Academy for Software Engineering accepts students with GPAs and test scores at and even below the citywide average. In others words, a student coming from a poor performing middle school who may have been ill prepared isn’t unfairly disadvantaged. At the Academy for Software Engineering, admission is based primarily on students’ desire to learn about software engineering. While the initiative is new and has yet to produce a graduating class, the school’s diversity statistics are impressive. While 9% of students identify as white and 11% identify as Asian, 46% identify as Hispanic and 29% as Black. By contrast, at the city’s most elite STEM high school, Bronx High School of Science, only 3% of students identify as Black and 6% identify as Hispanic. Of course, it is yet to be seen whether the Academy for Software Engineering will, over time, also be able to place its students in elite computer science and engineering programs across the nation.
Whether one is looking at the broad educational reforms being pushed by the Gates Foundation or the more modest experiments being executed by individual educators and tech investors at schools like the Academy for Software Engineering, there is no question that the tech industry has a role to play in public education and that the industry’s investment in elementary, middle school and high school education may ultimately prove to be the most effective way to solve the tech diversity problem.
Rethinking Traditional Approaches to Recruitment
Recruitment of new employees, especially in the technology industry, happens through several avenues. First, there are traditional on-campus job fairs. However, depending on the status of the college or university and its reputation in the tech industry, major players in the industry (e.g., Facebook, Yahoo, Google and Apple) may or may not ever come to campus.
Of course, even when students gain access to on-campus recruiters, race and ethnicity often continue to pose obstacles. Recruiters, as discussed in the first part of this article, are frequently looking for specific types of candidates—namely, candidates who match the “profile” of already successful workers in their workforce. In this respect, minority candidates must fight against recruiters’ embedded assumptions about what a successful job candidate looks like.
Of course, on this account, there are many pro-active steps that tech leaders can take, beginning with an assessment of where they are doing their on-campus recruiting and with what expectations or assumptions in mind. That said, there’s reason to believe that even when the tech industry changes its criteria, the tech diversity problem persists. To illustrate, consider Google’s hiring practices.
As Google has openly discussed, when they started to look at the numbers, they found no clear indication that a degree from one of the top-ranked computer science or computer engineering programs in the country was necessarily a guarantee that an employee would outperform a hire with a less prestigious degree. Google’s Senior Vice-President of People Operations has also famously declared that a high GPA is not necessarily an indicator that a worker will perform well on the job in the long term. “GPAs are worthless as a criteria for hiring, and test scores are worthless,” says Laszlo Bock, “We found they don’t predict anything.” Despite the company’s alleged dismissal of two of the most common hiring criteria—degree pedigree and GPA—and open effort to hire graduates from a wide range of universities, including state universities, there is no sign that Google’s hiring policies have impacted its tech diversity problem. 61% of Google’s employees are white and only 2% identify as Black. Unfortunately, this suggests that education and recruitment are not the only obstacles. In order to solve the tech diversity problem, then, we need to transcend hiring bias using new and innovative solutions, including those that are products of 21st-century tech expertise.
Can a Tech Solution Fix the Tech Diversity Problem?
Throughout 2015, there has been a growing focus on “big data” and its applications in both education and the workplace. Put simply, “big data” can refer to any large set of data that is analyzed to help inform decisions or practices. Recently, there has been growing speculation that “big data” may also offer a solution to the tech diversity problem.
Since humans are biased and training them to not be biased has often proven to be an upward battle, some companies are now handing recruitment over to algorithms. As one recent study published in the Harvard Business Review emphasizes, humans are good at identifying hiring needs but far less well equipped to evaluate the results. Theoretically, an algorithm should be bias-free and able to select candidates based on factors such as “potential to perform” or “capacity to learn independently” rather than traditional predictors, such as background, education and assumed “cultural fit.” And there is at least some indication that algorithmic hiring may provide a viable solution. For example, Infor Talent Science, a software company that offers organizations the tools needed to carry out hires based on data collected from potential employees, has analyzed data on over 50,000 hires and found that using their software increases African American and Hispanic hires by 26%.
Of course, as already noted, at Google, which uses its own software to carry out hires, there is no clear indication that hiring algorithms can address the tech diversity issue on their own. While there may be a number of reasons why algorithms might fail to diversify the tech workforce, it seems likely that the problem is at least partially rooted in the fact that algorithms are carrying forward some of the biases that have prevented minorities from moving forward in the tech industry all along. Can an algorithm be biased? Unfortunately, like humans, they can be. For example, if among other factors, the algorithm includes data collected from an oral exam score, some candidates (e.g., candidates working in a second or subsequent language) will be more likely to be deselected than native English speakers.
In the final part of this three-part post, we will examine why training and mentorship matter—arguably more than ever before—and how the tech industry can use learning management systems to address the tech diversity problem by supporting minority workers during the onboarding stage and throughout their careers.