Access isn’t everything. As “tech jobs” in fields like software engineering and data science have become among the most lucrative and exciting in the current economy, concern about who holds these jobs, and reaps their rewards, has grown, and not only among industry watchers and experts.
Technology companies feel increasing pressure from investors, the public, and their own employees to report on—and improve—the proportion of women and underrepresented minorities in their ranks, particularly in the high-status engineering and technical roles core to the business. Countless headlines promise to reveal “[insert company]’s diversity problem,” and countless others extol new fixes for the “STEM gender gap.” But all too often we approach gender and other disparities in technology as crises of opportunity, which is only part of the picture.
Much discussion about technology careers centers around diversifying the pool—that is, ensuring that more white women, and more Black and Latinx men and women, have access to the education, skills, and networks that make them viable candidates for jobs in technology. Coding camps and vocational training programs try to do this, as do traditional computer science departments that focus on recruiting more women and people of color.
With a more diverse range of people interviewed and hired into entry-level roles, we might anticipate that over time the upper levels of the hierarchy in technology functions and companies will start to shift. In a virtuous circle, public perception about who is suited for working in technology will evolve, making the field even more appealing and welcoming to traditionally underrepresented groups. Change the pipeline of prospective tech leaders and experts, and the leaders and experts will start to look different.
Yet it isn’t clear that such results emerge, at least on a broad scale, from this tactic. Not only do women remain underrepresented in technology, compared to their presence in the overall labor force, they are especially scarce in upper-level jobs and in the most lucrative subfields.
But this is a story we’ve heard before, and from which technology leaders can learn. More than twenty years ago, the consulting industry rectified a gender imbalance in its entry-level cohorts—yet today, less than 20% of managing directors and partners at major professional service firms are women. The pipeline has not been moving women and men along the career ladder at similar rates. When a focus on STEM education, training, and hiring comes at the expense of diagnosing and addressing the leaks that drain women, companies end up with an ever-shrinking pool of candidates for leadership roles. (Indeed, even organizations that focus on education and opportunity are increasingly aware that their efforts must be paired with culture change on the part of organizations hiring their graduates.)
To be clear, efforts to diversify the pool of qualified candidates and to eliminate biases from hiring are important. Practices like blind resume reviews, standardized evaluation rubrics, and outreach to underrepresented candidates can help get more women in the door. But they aren’t the solution to the “push factors” that prevent women from advancing, or spur them to leave their jobs or the industry altogether.
Researchers have identified how and when those factors manifest, and what measures have been shown to be effective in mitigating them. In addition to these concrete approaches, leaders have the opportunity to shape their teams and organizations in ways that mitigate the gender or other identity-based barriers that prevent employees from contributing and achieving at the level of their true ability.
One innovative approach lies in challenging a powerful assumption, what we might call the tech sector’s claim to meritocracy. When companies describe themselves as meritocratic, or declare that its people-management processes are merit-driven, it’s actually more likely that male employees are being paid more than women who receive the same performance ratings. The researchers who uncovered this phenomenon point out that meritocracy-focused missions and values statements may be inadvertently encouraging bias—by making managers feel that, simply by virtue of their membership in a meritocratic organization, they do not need to reflect on their thinking. In essence, managers may feel overly confident in the quality of their decision-making and fail to examine their judgements about people—no matter how many unconscious bias workshops they’ve attended.
Leaders certainly shouldn’t throw out meritocracy as a value, a vision, and a desired state, but they can shift attention and resources to the effort needed to achieve it, like improving transparency in compensation, formalizing evaluation processes, and creating “social accountability” among managers by sharing data on, for instance, promotion patterns by race and gender across different departments.
Understanding meritocracy as an aspirational state means recognizing that work remains to be done. Getting women into technology careers isn’t the end result companies should be striving for—it’s creating the conditions that prevent them from “leaking” out.
 Ammerman, Colleen and Boris Groysberg. Gender & Careers in the 21st Century. Harvard Business Review Press (forthcoming).