White men continue to disproportionately occupy top-level leadership positions in organizations. This fact frustrates economists who believe it’s suboptimal not to draw from the other 66% of the labor pool comprising white women and people of color. This “glass ceiling” persists even as corporations spend tens of millions of dollars to change it.

Past researchers have speculated about causes. Aside from the obvious “old boys network,” another cause has purportedly been the reluctance of women to support and hire other women (the queen bee syndrome) and for people of color to support and hire other people of color (reminiscent of crabs in a basket pulling down those who climb highest). These tendencies have been documented by research.

Now a new study suggests…

…another reason ethnic minority and women leaders may impede the advancement of other ethnic minorities and women is because they are penalized in the form of lower performance ratings when they engage in “diversity-valuing behavior,” or behavior that promotes demographic balance within organizations (e.g., behaviors such as hiring and promoting ethnic minorities and women).

Overall, white women and people of color do not receive poorer performance ratings, on average, than white men. But this might change, the researchers hypothesized, if white women managers and managers of color seek to hire other white women and people of color.

Since white women and men and women of color are members of low-status groups in US society, negative stereotypes exist that devalue their competency. Past research has shown when low-status group members are seen as similar to white males, they are viewed as belonging to the “in-group.” But when a low-status group member hires another low-status group member, this might change, since,

by engaging in diversity-valuing behavior, non-White and female leaders highlight their low-status demographics, which leads others to stereotype such low-status leaders as incompetent.

On the other hand,

Stereotyping White men who engage in diversity-valuing behavior is much more challenging than stereotyping non-Whites and women because doing so would threaten one’s beliefs about society and the status quo in which Whites and men maintain the position as highly valued members of society…. Likewise, there are very few (if any) negative stereotypes that can be applied to White men …. High-status group members are afforded idiosyncrasy credit, and, therefore, given freedom to deviate from the status quo….

The researchers conducted an initial study of 350 executives, including vice presidents, directors, CEOs and other C-level leaders attending a week-long executive development program. The executives came from 26 different industries and 20 different job functions. Ten percent were executives of color and 31% were women.

Each executive was rated by their boss on 1) their performance as a leader and 2) their leadership competence. Each executive was also rated by their peers on their diversity-valuing behavior.

The researchers controlled for pay grade, organization size, leader education level, leader organizational level, native vs non-native English speaking, and boss and peer familiarity with the leader.

Since the researchers were focused on how gender and race each operated in the workplace, the results were analyzed using gender (women vs. men) and race (non-white vs white) categories only. Consequently, no results were reported for combined groups such as white men or women of color.

Women were reported by their peers to engage in more diversity-valuing behavior than men. Likewise, executives of color engaged in more diversity-valuing behavior than whites.

As anticipated, women, who were rated high in diversity valuing behavior by their peers had lower performance ratings from their boss. The same result was found for race. Executives of color who had high peer ratings of diversity-valuing behavior were rated lower in their performance by their boss.

To the contrary, both men and white executives had slightly higher performance ratings when they were rated higher in diversity-valuing behavior.

The same pattern appeared for ratings of competence. Those women executives and executives of color who were rated high in diversity-valuing behavior by their peers were rated lower in competence by their boss. The researchers used a statistical technique to tease out the impact of competence ratings on performance ratings and found that bosses’ perceptions of competence accounted for much of the difference in how performance was rated.

This field study, then, supported the researchers’ theory that low-status group executives would be viewed as less competent, and thus rated as performing more poorly, when they engaged in diversity-valuing behavior. High status group executives, on the other hand, were not penalized for diversity-valuing behavior. Indeed, they came off slightly better.

The researchers then did a second study, this time in the laboratory using 307 adults employed in the United States.

Participants were given a scenario in which a hiring team of 4 people could not decide between two job candidates and a senior executive was brought in to make the decision. Participants were shown photos of the two candidates. Depending on the experimental condition, Candidate 1 varied by race and gender but always came from a low-status group, i.e. either a man or woman of color, or a white woman. Candidate 2 was always a white man.

Participants also saw a photo of the senior decision-maker. Depending on the experimental condition, this decision-maker was either a white man, or from a low-status group similar to Candidate 1.

In the scenario, the decision-maker either picked the white male, saying “Candidate 2 had the highest scores so we are going with Candidate 2.” Or, the decision-maker picked Candidate 1, saying the identical words, but adding that the candidate “increases the racial and gender balance of our leadership team.” Notably, in every case the candidate was said to have the “highest scores.” Participants were then asked to rate the competence and job performance of the executive decision-maker.

The results backed up the findings of the first study. Both women and executives of color were judged to be less competent and to have performed more poorly when they showed behavior that valued diversity (selecting the low-status group candidate). When they chose the white male candidate, their competence and performance ratings were high. Furthermore, the researchers again found that perceptions of competence accounted for ratings of performance.

White male decision makers were rated essentially the same, regardless of the candidate they chose.

With this second, laboratory-based, study the researchers contend that there was no “old boys network” in operation. Indeed, participants did not know one another, and so there was no network at all. The two studies taken together, one in the field and one in the laboratory, strongly back the researchers’ contention that low-status group executives are associated with negative stereotypes of incompetency when they display diversity-valuing behavior, and consequently their performance is viewed less favorably.

The researchers point to some disturbing implications:

One … is that engaging in a low level of diversity-valuing behavior may be a critical prerequisite for the upward social mobility of women and non-White leaders.
Indeed, they “found across both studies that some of the highest competence and performance ratings were awarded to non-White and women leaders who engaged in low levels of diversity- valuing behavior.

And another implication,

[T]he glass ceiling may actually become stronger, …with each ethnic minority or woman leader hired…[b]ecause we found that diversity-valuing behavior tends to be a somewhat acceptable behavior for White men, yet an illegitimate behavior for ethnic minorities and women…

And finally,

[O]ur findings suggest … that low-status demographic groups cannot “lift themselves up by their own bootstraps,” but, rather, must rely on members of high-status demographic groups …

They suggest some solutions. One, ironically, is to have “a White male spokesperson for the diversity office.” Another is to,

measure and reward the degree to which people hire and promote individuals who are demographically dissimilar from themselves. [R]ewarding such demographic unselfishness would naturally correct the demographic imbalances throughout organizations, as members of demographic majorities would tend to hire and promote members of demographic minorities.

To us, CSWAC, these findings point to a specific predicament for directors of diversity, who, almost invariably, are white women or people of color. Typically, the people who hire are part of an established white and male ingroup. Generally, the hiring organization aspires to become more diverse, especially in the upper ranks. The director of diversity will be valued and considered competent to the extent that they share the values and goals of the ingroup.

But when it comes time to produce significant change, by advocating for the hiring of a person of color for an upper management slot –something that fulfills the mandate of increasing diversity – the director of diversity is likely to be viewed as less competent, and their performance is likely to be devalued for doing what they were hired to do. If the hiring parties have a favored white candidate –perhaps a loyal employee one step down the chain – the director of diversity risks being viewed now as not sharing the goals and values of the ingroup.

To be warned is to be forearmed. Clearly women and people of color need to understand these dynamics. Many do. More importantly, senior management and board members need to ask themselves how much they value becoming multiracial. A serious change effort needs to incorporate a new understanding and affirmative valuation of everyone who can bring in diverse talent.


From Hekman, D. R., Johnson, S. K., Foo, M. Der, & Yang, W. (2017). Does diversity-valuing behavior result in diminished performance ratings for non-white and female leaders? Academy of Management Journal, 60(2), 771–797.