Why Discrimination is Hard to See at the Individual Level

images-1Readers’ comments in response to the Ellen Pao vs. Kleiner Perkins Caufield & Byers verdict, such as those in the New York Times, made me realize how little the general public seems to understand about discrimination and why it is so difficult to see and prove at the individual level. I posted a comment and decided to flesh it out a bit more here.
First, discrimination can be rather subtle these days, especially in contexts where people generally know better than to make overtly sexist comments. This means that discrimination can take awhile to catch on to, as any particular incident might seem too small to warrant a complaint, or would make one seem overly sensitive or paranoid for pointing it out.
Second, by the time an individual catches on or has had enough — often when it culminates in being passed over for promotion or a decision to quit — the treatment is likely to have had the effect of making the individual seem unvaluable and unlikeable to others in the organization. Self confidence and performance, social relationships and reputation, and mental health suffer as a result of even mild forms of mistreatment over time, just as a drip of water on the forehead is irritating but, repeated over time, can drive victims insane. Once a case has been brought to trial, defendants may be able to successfully play the “personality” card and claim the plaintiff is an individual failure and/or villain.

Finally, every individual is different, so defendants can argue the plaintiff was unique in some way, and that those uniquenesses motivated her treatment, not her sex. She was going through a personal crisis; she wasn’t a “team player;” she was unpleasant and difficult; her performance was mediocre. Indeed, it can be difficult to tell if such claims are true or are themselves evidence of discrimination. Personal relationships break down when someone endures stress over a prolonged period of time, it’s hard to be a team player when one is left out of the loop, and men with similar personalities and mediocre performance get promoted where women do not.

Discrimination is much easier to see, and to prove, when we zoom out and look at patterns.2 A lot of research, including my own, has zoomed out and systematically compared men’s and women’s experiences in the workplace and controlled for confounding variables like education, experience, work hours, performance, ambition, family status, and personality. This research reveals double standards that consistently disadvantage women at work (and men at home, but that’s a post for another day).3 Some studies show no overall sex difference in, say, experiences of gender harassment, but once sex is combined with personality, parenthood, or caregiving, double standards jump out and one can see how women are penalized for the same things that are lauded in men.

Until the public becomes more educated and aware of this, these forms of subtle discrimination are likely to go unnoticed, and trial by jury is unlikely to result in findings for an individual plaintiff. Class action lawsuits may be required to raise awareness and bring about change.


1 Examples abound, from not inviting female colleagues to social events where information and gossip is exchanged, to calling a woman by her first name but a man by his last, to asking women how their families are doing but talking “shop” with men, to expecting a hug from a woman but a handshake from a man, to describing a woman as “abrasive” and a man as “leaderly” for forcefully expressing their opinions.

2 For example, men comprise 94% of venture capitalists, 95% of Fortune 1000 CEOs, 84% of those employed in computer programming, and 70% of those employed by tech’s big four (Apple, Google, Facebook and Twitter). Women continue to earn 20 cents less than men on the dollar (based on median usual weekly earnings of full-time workers), and are paid and promoted less than men in the same occupations.

3 For example:

  • If a male candidate has more education and a female candidate has more experience, people choose the man and say education is more important; switch their qualifications and people choose the man and say experience is more important. People are unaware of this bias and strongly believe their decisions were based on merit.
  • Assertive women are significantly more likely than their less assertive counterparts to be “put in place” with demeaning comments and sexual harassment at work. Assertive men are treated with more respect than their less assertive counterparts.
  • Women with children are assumed to be less competent than women without children and than men, and are offered significantly less pay and seen as significantly less promotable, even with identical hours, career ambitions, and levels of performance. Men tend to experience positive career outcomes as a result of becoming a parent (unless they’re active caregivers).

Gender Equality: The Best of Times or A Stalled Revolution?

          bildeI presented the statistics below in the opening of my talk today at Harvard University’s Kennedy School of Government. There are mixed messages related to progress toward gender equality: On the one hand, women have overtaken men in education, are increasingly the primary breadwinners within households, and men are closing the gender gap in childcare; on the other hand, progress on the gender wage gap has slowed, occupations remain segregated by sex, and men continue to dominate positions of leadership.The private sphere appears to be more progressive than the public sphere at this point. Household decisions about education and the sharing of breadwinning and childcare responsibilities are moving toward equality while employment norms for pay, segregation, and promotion push against it. As men and women increasingly interact as equals in the home, they may come to expect more equality at work; current work norms, structures, and leaders, however, are often stalwarts of the status quo.

In the remainder (and bulk) of my talk I presented my research on how gender identities are actively constructed and regulated at work through mistreatment, from gender and sexual harassment to general disrespect. To the extent that progress toward gender equality continues and family and work roles become degendered, mistreatment based on sex should decline; to the extent that sex-based mistreatment is eradicated, gender segregation, inequality, and stereotypes should also recede.



In 2009-10, women earned:
* 57.4% of bachelor’s degrees
* 62.6% of master’s degrees
* 53.3% of doctorates

In addition, 84% of wives are as (61%) or more (23%) educated than their husbands.
40% of households with children under 18 have a female as the sole or primary breadwinner.


The gender gap in childcare has been steadily closing. Men now do 79% of the primary childcare that women do, and 70% of the secondary childcare that women do, for an overall ratio men’s to women’s childcare hours of 72%.



Progress on eliminating the gender wage gap has slowed. In the U.S., women continue to earn only 81.1% of what men earn (based on median usual weekly earnings of full-time wage and salary workers):

Women earn less than men in the same occupations (occupations with >5% of the U.S. male and/or female labor force, median usual weekly earnings of full-time wage and salary workers):


The labor force remains segregated by sex. Below is the distribution of approximately 80% of the male and female labor forces in occupations that represent at least 5% of those labor forces:


Women constitute nearly half of the U.S. workforce, but their proportion decreases up the organizational ladder. Men constitute an overwhelming 95.4 percent of Fortune 500 CEOs.

Bias as Incompetence

imagesIn an effort to avoid finger-pointing and defensiveness, those studying prejudice — the tendency to pre-judge others, usually based on stereotypes — moved away from the term prejudice, which came to be associated with dislike, hostility, and intentional harm, to the term bias: prejudice, but not necessarily with the dislike, hostility, or intention.

Bias is still an ugly word, and researchers went further to use “implicit bias” (or “hidden bias,” “unconscious bias,” “subtle bias,” “modern ___ism,” and “second-generation bias”) to convey the unwitting nature of many of our prejudices. Rather than bad we are misled, by a world that conditions us to more quickly associate certain characteristics with certain groups (see for example, the IAT) because groups differ in social status and roles.*

The concept of implicit bias has helped us make great strides in accepting and facing our prejudices. It has helped move the discussion away from judgment and blame toward awareness and understanding.

Yet “bias” still invokes motive and the need to interpret it. “Explicit bias” is intentional, driven by conscious endorsement of prejudicial beliefs, whereas “implicit bias” is unintentional, driven by accidental stereotyping.

Ultimately the motive behind biased decision-making and behavior is difficult to discern and irrelevant to the harm done. Focusing on the lack of nefarious motives for bias (or the presence thereof) might help the medicine go down (or not), but directs our attention to the virtue of the decision-maker rather than the competence of the decision.

With now abundant evidence on bias and its workings,** it’s time to reframe bias in terms of competence. Those who make use of the knowledge available to become aware of their social biases and to minimize the impact they have on their decisions and behavior are more competent to make decisions and to lead and manage others than those who are not so educated. Just as we wouldn’t entrust those with financial ignorance to manage our money, so we should not entrust those with social ignorance to evaluate and manage our employees and clients.

Gender and diversity competence — the ability to select and evaluate people and ideas on their potential and merits rather than on what they look like or where they come from, and the capacity to build and sustain diverse and inclusive organizations — is essential to effective leadership. Regardless of the motives professed.

* Groups that occupy high status roles are are stereotyped as “competent” (e.g., professionals) whereas low status groups are stereotyped as “incompetent” (e.g., the poor). Groups that are harmless or give care are stereotyped as “warm” (e.g., mothers) whereas groups that compete with us for scarce resources are stereotyped as “cold” (e.g., the rich). These stereotypes stem from and reinforce group segregation and inequality (see Fiske, Cuddy, Glick & Xu, 2002).

**We know, for example, that even with identical qualifications, women and minorities are significantly less likely than men and Whites to be evaluated as competent and to be hired, paid well, or offered career mentoring. We know that letters of recommendation are written differently for men and women and that people unwittingly use post-hoc definitions of merit to justify selecting a male over a female candidate. We also know that teams and boards with more women on them outperform teams and boards of mostly or all men, and that increased gender equality enhances the economic growth of nations. There are hundreds more studies documenting gender and racial biases in selection and promotion and the benefits of diversity and inclusion in groups and organizations.

Who’s in charge: Why aren’t there more women in leadership roles?

The UBC Alumni Association hosted a panel discussion last night of why there aren’t more women in leadership roles. Gloria Macarenko moderated the discussion with myself, Maninder Dhaliwal, Anne Giardini, John Montalbano & Martha Piper as panelists. We were given one minute to address the following question:

For decades women have been fighting for workplace equality, and in many sectors, progress has been made. But in boardrooms and executive offices across the country, there remains a noticeable absence of women in senior leadership positions. What lies behind this gender imbalance? Is it due to deeply-ingrained biases by those making the appointments? Or are some women choosing to forgo leadership opportunities for career paths that offer greater flexibility?

I wrote up my opening comments (below). It’d be nice to have a transcript of the other panelists’ comments — it was a great discussion! You can listen to the podcast here.

My answer to the question of why gender imbalance in leadership continues is both deeply-ingrained biases that we all share, including those who make appointment decisions, and women choosing to forgo leadership opportunities – these biases and choices feed off of each other. It is one thing to write laws and quite another thing to change our economic structures and cultural norms. So that is really the challenge we face.

We’re born into a world run by men and quickly learn to associate leadership with men and masculinity and to value masculinity above femininity. Stereotypes of what a leader looks and acts like, what men and women are good at and what interests them, become values – what a leader should look and act like, what men and women should be good at, and what should interest them. As we social psychologists put it: the descriptive becomes prescriptive. These descriptive stereotypes – these associations we all have – drive bias in training, selection & promotion, and prescriptive stereotypes – the values component to it – drive and shape workplace culture and the treatment that employees experience.

And it all starts in childhood on the playground. You have the confident and assertive girl called a b-word, like bossy or bitch, and you have the sensitive or caregiving boy called a g-word, like girly or gay. We have boys who challenge authority and break the rules being admired by their peers – they’re the future entrepreneurs, convicts, or both. We have boys who follow the rules and authority being good sons – they’re the next promotion, or president. On the other hand, the “good girl” is placed on a pedestal — protected, but disempowered — and the “bad girl” is vilified and free game for sexual harassment.

This socialization directs our interests and aspirations and we carry them with us into the workplace – sometimes more subtly than on the playground. Women often face a trade-off between love and power, and most choose love – they go where they are rewarded and appreciated, into female-dominated occupations, assistant positions, or if they can afford it, the home.

Men face a different set of challenges. For men, love and power often coincide. But most men aren’t going to get to the top. What happens to those who don’t? They may find other ways to feel like a man, some healthy, some not. Today, men’s involvement is needed more than ever in the home as wives have become breadwinners, but organizational structures and norms don’t support fathers as they do mothers. At this point, men now report more work-life conflict than women.

In sum, our cultural biases and institutions continue to encourage women to choose to leave or scale back at work and men to choose to leave or scale back at home. This dynamic is mutually reinforcing. We really need to talk about men also if we’re going to solve the problem for women.

Organizational Structures that Hurt Women’s Relationships

Today at the Harvard Business School’s Gender & Work Symposium there were several talks addressing how aspects of organizational structure and culture affect women’s relationships with one another and their tendency to embrace or distance themselves from their identity as women.

Few Women at the Top

Robin Ely presented her dissertation research on White professional women in large law firms in which she asked: Does it matter how many women are in senior ranks for the relationships among women lower down in the organization? She expected there to be more supportive peer and hierarchical relationships among women in more sex integrated firms and that the quality of within- group relationships would depend on where that group is positioned in the organizational hierarchy. Her results showed that:

  • When the proportion of women in senior management was relatively high, women felt indifferent or good about being a woman, believed senior women were good role models, and reported supportive and positive relationships with other women.
  • When the proportion of women in senior management was relatively low, women felt that being a woman was a liability, believed senior women were poor role models (acted too much like men or too sexually), and reported unsupportive and dysfunctionally competitive relationships with other women.

Michelle Duguid
and Denise Lewin Loyd presented research on the value threats that members of low status token groups experience to prevent them from advocating on behalf of other members of their group:
  1. Favoritism threat: The fear of being seen as illegitimately biased in favor of demographically similar others.
    • An experiment showed that women in a token situation (the only woman among four men) evaluated an equally qualified female candidate lower than a male candidate, whereas women in a majority situation (one of four women in a group with one man) evaluated the candidates similarly.
  2. Competitive threat: The fear of being compared to a more qualified demographically similar other.
    • Another experiment showed that women in a token situation reported more competitive threat than women in a majority situation and were less likely to select a qualified female candidate.
  3. Collective threat: The fear of a less qualified demographically similar other devaluing the demographic group.
    • A third experiment showed that women in a token situation reported more collective threat than women in a majority situation and were less likely to select a less qualified female candidate.
Gender Bias in Organizational Cultures
Belle Derks presented research on the “Queen Bee” phenomenon, or why senior women might work against the progress of female subordinates. A danger of this phenomenon is that criticism of women by women is especially damaging because it seems credible and women as a group can be blamed for their own lack of advancement.
Derks presented three studies demonstrating that “Queen Bee” behavior is:
  • A response to gender bias in the organizational culture.
  • Enacted by other minority groups. Devalued ethnic minorities are more likely to distance themselves from their ingroup and act like majority group members at work.
  • Rewarded — by women and by men. When given a choice of candidates — a man (masculine self-presentation, anti-affirmative action), a “Queen Bee” woman (masculine self-presentation, anti-affirmative action), and a “feminist” woman (pro affirmative action), the Queen Bee was the most popular candidate.

Expectations of Female Solidarity
Leah Sheppard presented research on observers’ interpretations of criticism and conflict between women at work. She began by noting that women are held to a higher standard than men in terms of looking out for and supporting same-sex others at work, or that there exists a prescription for solidarity between women that vilifies the women who violate this prescription:
  • Senior women are at risk of being viewed as “Queen Bees” if they criticize female subordinates’ work performance. In one study Sheppard showed that negative feedback from a senior partner in a law firm toward a junior member of the firm was assumed to stem from jealousy or threat when it occurred between women, but was assumed to stem from legitimate work concerns when directed toward male subordinates (whether from a senior woman or man), and was assumed to stem from sexism when directed toward a female subordinate from a man.
  • Women coworkers who experience conflict are assumed to not be able to repair their relationship. In another study Sheppard showed that conflict between coworkers was assumed to be significantly less repairable if it occurred between women than if it occurred between men or between a woman and a man.
  • Senior women, but not senior men, are viewed as unsupportive and cold if they respond curtly to advice-seeking from a junior woman. Sheppard had participants imagine a scenario in which a person sought support or guidance from a would-be mentor who responded curtly. This curt response could be attributed to a variety of causes, such as being busy, rude, disrespectful, or hostile. When the person seeking advice and the would-be mentor were both women, participants reported they would expect to feel significantly more disappointed and let down, and rated the would-be mentor as significantly more cold, than when the person seeking advice and the would-be mentor were both men or were a mixed dyad.