The percent of technical workers in the U.S. who are Black or Latinx hovers in the single digits, and women have remained stuck at about 21% of technical roles for the past five years. Read More…

When advocates for underrepresented people in tech argue for leadership to be more proactive in recruiting minorities and women, however, they hear a near-constant refrain: “we won’t lower the bar.”

The problematic mindset behind this sentiment has already been thoroughly analyzed and debunked. This mindset is not just “problematic” — it is actively undermining the goal of increasing the diversity of the tech workforce.

Research in psychology suggests that the pervasive anxiety about lowering standards that is generated whenever tech companies consider hiring Black, Latinx, women, and other underrepresented people is itself causing poor performance among these groups. Worrying that these individuals will fare poorly is actually causing them to fare poorly. Disturbingly, the anticipation that performance standards are at risk when hiring folks who aren’t white or Asian men actually produces a self-fulfilling prophecy that prevents more diverse talent from being hired. But how?

Decades of research pioneered by Robert Rosenthal, Distinguished Professor of Psychology, have shown that our expectations of a person materially impact that person’s actual performance. Rosenthal’s early research in the ’60s showed that the link between expectations and actual performance holds true even when the “people” being tested are rodents. In his landmark study, a group of standard lab rats were arbitrarily assigned to one of two groups — “dull” rats or “bright” rats. The fascinating results showed that the rats labeled as “bright” actually learned mazes more quickly than the “dull” rats! Rosenthal concluded that his students, who were responsible for the testing of the rats, had unconsciously influenced their rats’ performance, depending on whether they believed their rats were “dull” or “bright.”

Troublingly, the same research has been conducted and confirmed in schools. Students randomly assigned to a list marking them as “academic bloomers” with the potential for high achievement were treated differently than those who were not on the list. Their teachers (who were unaware of the deception) treated these students with more attention and gave them more support and encouragement than the others. At the end of the experiment the “academic bloomers” had achieved greater increases in IQ scores than their “non-gifted” classmates.

In tech, it’s highly probable that our explicit and implicit expectations of an individual are affecting their actual performance. If we simply believe that white and Asian men tend on the whole to be smarter and better as technical workers, those expectations are likely having a negative effect on the performance of anyone who isn’t a white or Asian man.

Conversely, due to our human tendency to rise to meet expectations when someone believes in us, we may in fact see white and Asian men outperforming everyone else, other things being equal. This effect has actually been proven in psychological research measuring something called stereotype threat, which is a condition that causes people to perform well or poorly according to whether their minds have been activated to remember a positive or a negative stereotype about a group that they belong to. Research on stereotype threat shows that this effect causes girls to underperform in math, for example.

Resume study after resume study shows that we assess people differently based on traits like gender and race, even when their background is identical and even (especially!) when we believe that our judgment is unbiased.

We have to assume that people who are constantly hearing the refrain that the bar won’t be lowered in order to accommodate their participation in tech understand that expectations for their performance are already low. Thus, the common understanding of who is expected to meet or exceed standards, and who is expected not to meet them, is itself a significant part of the problem. Management’s collective expectation that it will be difficult to find underrepresented talent that can “meet the bar” is creating a self-fulfilling prophecy that undermines the performance of anyone who doesn’t already look like the majority of workers in tech. If we want to effectively diversify the tech industry, leaders will need to begin by shifting their expectations.