A man and his son get into a car accident. The man dies, and the son is badly injured. The son is taken to the hospital for emergency surgery. The distinguished surgeon takes one look at the boy and says, “I can’t operate on him! He’s my son!”

How the riddle asks, can that be?

Even today, 40% to 75% of people can’t figure out the answer—the surgeon is the boy’s mother.

The riddle is a reminder that, like it or not, humans are unconscious inegalitarians. Just as the human eye has a blind spot in its field of vision, our unconscious minds can contain hidden biases that can guide our behavior. The study of unconscious bias was pioneered by Harvard psychologist Mahzarin R. Banaji. Banaji argued that all the efforts to promote “gender blindness”—the guiding HR philosophy of the ‘80s and ‘90s—were necessarily doomed because even if we fervently believe that we are treating people fairly, our unconscious has different perceptions.

As Banaji maps out in her 2013 book, Blindspot, these biases are survival mechanisms dating back millions of years. Humans are continually bombarded by information, an estimated 11 million bits of it per second. But we can process just forty bits a second, so our brains take shortcuts. Rather than consciously evaluate the appearance of a ravenous tiger—a process that could take several seconds—our ancestors’ unconscious biases flashed danger so they could flee faster.

These unconscious or implicit biases still kick in whenever we see someone not like ourselves. They also reinforce the basest stereotypes—we associate men with work and women with family. The riddle above trips us up because we don’t instantly think of women as surgeons.

Minute as they are, these unconscious biases have a compounding effect. Scientists at Rice University created a computer simulation of a company with an entry-level workforce evenly divided between men and women. Each worker was assigned a random performance score, but one that was programmed with a miniscule, 1% bias against women. The result? By the time you reach the top level of the company, it is 65% male.

Joanne Lipman has written a refreshing book on gender parity in the workplace, That’s What She Said: What Men Need to Now (And Women Need to Tell Them) About Working Together. She understands that any productive discussion about men and women in the workplace must start by drawing awareness to these biases.

The comment that the dancer Ginger Rogers did everything that her partner Fred Astaire did, but “backwards and in high heels” takes on new meaning when you consider the implicit biases against women harbored by males and females alike.

As Lipman observes:

These biases start young. Moms routinely overestimate the crawling ability of their sons, and underestimate that of their daughters. Parents of male two-year-olds ask Google, “Is my child a genius?” two and a half times more often than do parents of two-year-old girls.

Once children hit school age, teachers—even female teachers—subconsciously believe boys are better at math than girls. In one study, when a group of teachers graded math tests with no names on them, the girls outscored the boys. But when another group of teachers graded the same tests with names, the results were reversed: they gave higher grades to the boys than the girls. All of the teachers, by the way were female.

As these kids move into middle and high school, the biases are compounded. Not only students, but their parents, are often biased against teenage girls as leaders, a survey of almost twenty thousand teenagers found. By the time these children enter college, it’s almost impossible to break free of those constraints. Women are 33 percent more likely to earn a college degree than men. Yet professors are more likely to call on men than on women, and female students are more likely than male students to be interrupted, even though they speak less than their male classmates…

Both women and men prefer male bosses. In one study, respondents who agreed with the statement “I have no preference for a male boss or a female boss” nevertheless were willing to take a salary that was $3,400 lower to work for a man rather than a woman. They literally took a voluntary pay cut in order to avoid working for a woman…

Research has found that female economists are twice as likely to be denied tenure as their male counterparts—largely because when women and men collaborate on publications, evaluators give the men as much credit as if they’d written the article alone, while the women get no credit at all…

At every level, women are 15 percent less likely than men to be promoted, according to a joint survey of 118 companies and more than 300,000 men and women by McKinsey & Company and Leanin.org. In part that’s because men get promoted on the basis of their potential, but women are promoted only if they’ve proven themselves with past performance…Social scientists have calculated that a woman must be two and a half times more competent than a man to be viewed as his equal…

In one controversial study published in 2014, scientists concluded that more people die in hurricanes named for women than men—because we take “female” hurricanes less seriously.

In light of the depth and ubiquity of these biases, it is remarkable women have made as much progress as they have. They certainly help explain why the gender gap persists.

Awareness alone only achieves so much. Lipman, like Helena Morrissey, founder of the 30% Club, an organization that encourages companies to strive for 30% female representation on corporate boards, argue that the incentive to interrupt these biases has to be tied to the bottom line. Reframed, the argument becomes: When unconscious biases prevent women from assuming more leadership roles, everyone loses.

As we have discussed in these pages, hiring more women is not only the right thing to do in our increasingly transparent and diverse world, it is good for business (see WILTW 8 February 2018). There is a growing body of evidence showing that companies with women in senior positions perform better than those without. Although this is correlation, not causation, to an investor that distinction should not matter. If a business is tackling gender-related management issues, the chances are that it is dealing well with other risks and opportunities.

Companies willing to actively bust unconscious biases will enjoy a competitive advantage. Indeed, a growing number of firms are already experimenting with “blind resumes”, a practice inspired by the classical music world. As Harvard’s Iris Bohnet reminds us in What Works: Gender Equality by Design, in the 1970s, female musicians comprised just 5% of the top orchestras in the U.S. Today, women are 35% of the total players.

Progress wasn’t achieved through the gradual realization that women can (obviously) be great musicians, it was catalyzed by the introduction of “blind auditions”, first by the Boston Symphony Orchestra and soon followed by others. When players performed behind a curtain or a screen, the female musicians were 50% more likely to advance to the next round of auditions. The orchestra directors needed a “workaround” to overcome their gender bias.

In an increasingly complex world, strategic hiring is not about finding another one like those you already have, it is about deliberately seeking out the difference. In a 2016 speech, Andy Haldane, the Chief Economist at the Bank of England, challenged CEOs to consider whether their recruitment process really did encourage diversity:

There are two candidates for a position, A and B. They do a test based on attributes useful for the hiring organization. These tests might be state of the art, including all the diverse attributes one would wish for in an organizationcognitive, interpersonal and experiential skills.

In this test out of 10, candidate A scores 8 and candidate B scores 4. Which one should be hired? The answer is easy. The evidence points strongly to A as the candidate best meeting requirements for the job. They have quite literally, ticked the right boxes. But let’s add a twist. What if the answers A gets wrong are the ones B gets right? And what if the questions existing employees get wrong are also the ones B gets right? In other words, what if candidate B brings skills to the organization which otherwise do not exist?