Last year's uproar over Microsoft CEO Satya Nadella’s statement that women shouldn’t ask for raises, but instead, “trust in karma” for their rewards, has highlighted a tension that exists in corporate and investment communities: we believe that we operate in a meritocracy, but the evidence increasingly suggests that we do not. Instead, our systems for evaluating people and investments are skewed in unintended ways by biases that work against women and minorities.

For example, female-led startups find it harder to obtain funding than those led by men; women are often not hired or promoted at the rate of men; and the 2009 passage of the Lilly Ledbetter Fair Pay Act in the U.S. is only necessary because — even in the 21st century — women are not paid equally for the same work.

I became interested in the question of meritocracy because of my own research over the past three years, examining the emergence of a new field at the nexus of gender and investing. The leaders of this movement are working to understand how women-led ventures can get equal access to capital and how capital allocation can encourage more equal outcomes for women in organizations.

An idea that has come up over and over in the course of my research is that the collective belief in meritocracy is part of the problem. Sharon Vosmek, CEO of Astia — a non-profit organization focused on helping women participate fully in high-tech entrepreneurship — reflected this idea in a presentation at Stanford last year, describing the high-tech world as, “so captured by the myth of meritocracy… that it has really held back dialogue” on gender. The bottom line: the belief that we operate in a meritocracy takes us off the hook from examining the potential for bias in our evaluations of women-led businesses and women leaders.

Some evidence on bias

The biases that devalue women are not normally intended at all. Indeed, many of the skewed outcomes for women are produced through supposedly neutral or objective processes. The problem is, the way we see the world is shaped by our frames of reference, and, as Stanford Sociologist Cecilia Ridgeway has pointed out, gender is a primary frame that shapes how we value peoples’ contributions, without us even knowing it. That is why such stereotypes are called implicit biases.

Let’s start by thinking about how people are recruited, promoted and rewarded in organizations. In their studies of recruiting processes, Psychologist Rhea Steinpreis and colleagues found that when evaluating identical applications, both men and women prefer to hire a candidate named Brian over one named Karen by a ratio of two to one. These same results have been found for names that are seen as more African American orHispanic, and for resumes that signal homosexual orientation relative to those that do not, as Rotman Professor of Strategic Management András Tilcsik and others have shown. [Editor’s Note: read our interview with András on page 108 of this issue.]

Furthermore, even when the underlying quality of candidates is the same, applications vary. Studies of recommendation letters by anthropologists Frances Trix and Carolyn Psenka show that those written on behalf of female candidates are shorter, more likely to focus on personal characteristics, less likely to emphasize specific accomplishments, and more likely to subtly raise doubts (e.g., “it’s amazing how much work she is able to accomplish, given her family obligations!”) In this case, an equally qualified female candidate might actually appear to be less qualified because her recommendation letter is weaker. Either way, the results are the same: the man is hired more frequently than the woman.

Once on the job, the evidence suggests that there are differences in promotions and salary, as well. One of the essential underpinnings of the class action lawsuit against Walmart is the difference in salaries that increases as seniority increases, such that in 2001, female regional vice presidents earned on average 33 per cent less than men of the same rank. Elsewhere, research analyzing how people use performance evaluations to determine compensation shows that women and minorities with equal scores to those of white men receive lower salaries and smaller wage increases. The result: women have to be more qualified in order to get the same rewards.

At a recent conference on gender and finance, one of the participants told me, “I feel like my formal education process was a huge disservice to me as a professional, because it was really the only meritocratic environment I will ever be in. When I entered the workforce, I was so used to being rewarded for doing hard work and asking questions that it took me a couple of years to realize that it [meritocracy] wasn’t going to happen.”

One might contend that it is the context that leads to these biased results, and that in a more objective setting — say, the realm of investing — these effects should disappear. Unfortunately, they do not. Take the venture capital (VC) world of Silicon Valley, which has often been heralded as the ultimate meritocracy. According to one VC partner, “if you don’t get funded, it’s because your idea is stupid.” However, people are increasingly questioning why such a meritocracy leads to a situation where only six-to-eight per cent of venture capital investment goes to women-led businesses.

Is this a supply problem? Indeed, there are fewer women in Engineering programs, and therefore fewer high-tech startups led by women. On the other hand, we can’t ignore the demand side of the problem. Even the women-led startups that do get launched face higher barriers to funding than those started by men. A recent study in the Proceedings of the National Academies of Science by Alison Woods Brooks and colleagues examined nearly 100 startups in live-pitch competitions, and found that male-led companies were 60 per cent more likely to be funded. Perhaps, the argument goes, women start lower-quality companies, and the difference in funding represents a difference in quality. So, in web-based experiments, these scholars presented the exact same PowerPoint pitch presentation, narrated by either a female or a male voice. The result: investors were more than twice as likely to recommend funding the business pitched by the male voice, even though the scripts were exactly the same.

 

The paradox of meritocracy

Could it be that the very idea of a meritocratic culture exacerbates unequal outcomes? A series of experiments by Emilio Castilla and Stephen Benard found that organizational contexts that promote meritocracy are most likely to produce non-meritocratic outcomes — such as when men receive higher bonuses than equally-performing women. This occurs, they argue, because the idea of meritocracy gives evaluators moral credentials that convince them that they are unbiased, precluding them from being on the lookout for bias. Thus, the paradox of meritocracy is that a belief in it can lead to even more inequality, rather than less.

As one Silicon Valley participant in my research said, “those who are the biggest offenders of what a meritocracy should be are those who are screaming the loudest that there is a meritocracy.” Echoing this, another of my interviewees said about the VC world, “meritocracies are noble and worthy goals, but they’re absolute myths. The only thing that meritocracy serves in Silicon Valley is as great validation if you’ve made it; it justifies your success. You are just that much smarter than everyone else.”

If we succeed, we like to believe that it is due to our own merits and, thus, it is in our interest to believe that the system that enabled us to succeed is meritocratic. Indeed, as Jacki Zehner, a former senior executive at Goldman Sachs, reported about its internal research on diversity, “the feeling that the firm was a meritocracy was much more likely to be held by those in the majority [white, male, heterosexual] group. Non-majority members were more likely to say that there were ‘hidden rules’ for success, and that it was harder to get the right opportunities.”

Rutger’s Professor Nancy DiTomaso’s rich exploration of race and privilege in the U.S. could be extended to offer some insight about these gender dynamics. According to DiTomaso, “racial inequality is most often assumed to be the result of racism or discrimination, and providing equal opportunity within a context of individual effort and achievement has been offered as the primary solution.”

According to this logic, meritocracy should fix problems of racism, because most people are consciously committed to colour- or gender-blindness and to equal opportunity. But this very framing of the issue, she argues, “contributes to the inability of most whites to see the nature of their own participation in the creation and reproduction of racial inequality. Whites assume that other people are racists, but not them. They assume that equal opportunity embodies fairness, but they live lives of advantage — of unequal opportunity.” Thus, she concludes, we can have racial inequality in our society without racism per se.

Though the forces at work in sexism and racism are not the same, one could extend DiTomaso’s argument to suggest that explicit sexism need not be present to explain unequal outcomes. Instead, we only need to understand how dominant groups attempt to reinforce privilege — by promoting myths of meritocracy based on equal opportunity. If we lay the blame on overt racists or sexists, we divert attention from understanding the implicit biases built into our own frames of reference and in decision-making procedures that lead dominant groups to reinforce their privilege.

 

Why progress is so difficult

Given that gender is a primary frame that instantly (and sub-consciously) shapes our interactions with others, one solution to gender bias is to take gender out of the equation. When some symphony orchestras switched to auditioning candidates behind a screen — so that judges could hear the musician playing but not see any physical characteristics — the percentage of new female hires increased by between 25 and 46 per cent.

This approach may be difficult to replicate in business, where decisions depend on interaction with candidates or leaders of firms seeking investment. On the other hand, research by Roberto Fernandez, Isabel Fernandez-Mateo and others suggests that much of the bias is introduced in the early screening stages of decisions — for instance, those decisions about whether a job or investment candidate should even be in the consideration set. As a result, if ‘gender markers’ can be eliminated from the initial screen, we may see progress.

Another fix could entail teaching women to pitch better and present themselves better in interviews. This is a central recommendation of Sheryl Sandberg’s best-seller, Lean In, which underpins the discourse around the supposed confidence gap for women. Because women are not normally socialized to negotiate on their own behalf, there may be benefits to this approach. On the other hand, research suggests that when women lean in and act more confidently, they are likely to be pushed back. This is consistent with the well-established research that women can be seen as competent or nice, but not both, and that  women are penalized for succeeding in ‘male’ tasks.

One can almost understand why Microsoft’s Nadella told women not to ask for raises: perhaps he instinctively knows that such requests will not be viewed favourably coming from a woman. Trusting in karma does not seem like a good alternative, but the furor around his comments highlights the dilemma: damned if you do, damned if you don’t.

 

Towards a true meritocracy

Despite the challenges described herein, there is plenty we can do to get closer to a true meritocracy. Based on my research to date, following are eight suggestions.

1) Where possible, remove gender-identifying data. 

As indicated, it is impossible to avoid direct interaction when making hiring or venture-investment decisions. However, in the initial screening to create ‘short lists’, it may be possible to remove names and other gender-identifying information when evaluating candidates.

2) Include quantitative information (i.e. GPA or financial performance indicators) wherever possible.

When there is ambiguity with respect to credentials, members of the dominant group will benefit. Studies have shown that if grade point averages are added to otherwise identical resumes, the bias in selecting the male candidate is diminished.

3) Require adequate representation on short lists. 

Be proactive in soliciting diverse applicants — for jobs, business plan competitions, pitch nights, etc. When the National Football League put its Rooney Rule in place in 2002 (penalizing teams for not interviewing minority candidates), it increased the number of African American coaches from six to 22 per cent in just four years. Note that teams were not required to hire minorities — just to consider them, and this was enough to improve outcomes. Embracing this principle might mean actively inviting or seeking out candidates, because evidence suggests that women do not even apply for certain opportunities due to their anticipation of bias.

4) Train anyone in an evaluative position to understand implicit bias. 

If men and women alike accept that we are implicitly biased, we can do more to reflect on how that might be shaping our decisions. For example, if someone says that a particular female job candidate has sharp elbows, we can call that out as gendered and adjust our conclusions; or if someone says that the female CEO of a new venture ‘doesn’t have leadership qualities,’ we should think again.

5) Adjust for bias in recommendations and elsewhere. 

To compensate for implicit bias elsewhere in the system, Good Capital — an investor in social entrepreneurs — gives female-led startups extra points in the evaluation process. In their view, this simply gets these ventures back onto a level playing field. Venture capitalists might do well to follow suit.

6) Examine systems and procedures for unintended bias. 

Even workplace systems that appear to be gender neutral — such as a strict seniority promotion criterion — can reinforce the status quo of male privilege. Similarly, employee referrals are one of the most effective ways to recruit, because the candidates have already been screened for appropriateness by a trusted employee. However, the risk is that a company will continue to hire more people just like its existing employee base, rather than promoting diversity.

7) Create more transparency and monitoring. 

The good news is, when criteria are clear and evaluation processes systematic, bias can be reduced. Such approaches limit the discretion of decision makers to choose based on personal beliefs or preferences, which can decrease inequality. Of course, such transparency can be a tough pill to swallow: in an effort to make progress on gender equality, Google recently released data revealing just how far it has to go. While this brought the company lots of negative press, at the same time, it puts greater pressure on them to make progress.

8) Hold people accountable for results. 

Evidence suggests that training and awareness are not enough to make change. Inequality only declines when these programs are accompanied by incentive systems that demand accountability. As the Google example shows, transparency and accountability are tightly linked.

The bottom line is that gender is a highly salient primary frame, and each of us needs to be more aware of its effects and adjust our decision-making procedures accordingly.

You may also have noticed that the term ‘affirmative action’ does not appear in this article. Regrettably, this term has come to signify that underrepresented groups should receive preferential treatment. To be clear, this is not what I am advocating: my recommendations collectively serve to get all job candidates, employees and entrepreneurs onto a level playing field when it comes to hiring, promotions and investment decisions — taking into account the privilege from which many benefit and the implicit biases that we all possess. Rather than being related to affirmative action, they are more about ‘acting affirmatively’ to create actual equal opportunity.

If we want to make meritocracy real, we first need to recognize how the implicit biases in our frames of reference and in our procedures unfairly devalue women’s contributions and reinforce the privileges of dominant groups. It is my hope that leaders will embrace these first steps to moving from meritocratic myth to meritocratic reality.

---

This article first appeared in the Spring 2015 issue of Rotman Management magazine. Published in January, May and September, each issue features thought-provoking insights and problem-solving tools from leading global researchers and management practitioners. Subscribe today