Why Old Habits Die Hard

So how are your diversity initiatives going these days? Perhaps Legal is reporting that you’re compliant with rules and regulations in all relevant jurisdictions, and HR can tote up members of various social groups who have been recruited and hired. Marketing can take you smoothly through campaigns that offend no one and target everyone. Maybe all employees have had diversity-appreciation workshops and sensitivity training, while higher-ups have worked on “cultural competence” skills. With all that happening, things must be going well...

“Not really,” is a common answer, even among leaders of organizations that are doing, or trying to do, everything right. Many, while justly proud of their progress in this area, live with the feeling that they should be doing more, or doing differently.

Diversity, our ability to live and work with people who differ from us, is as old as civilization. But “diversity,” the universally lauded effort to bring difference into once-homogeneous workplaces, is only in its late 20s. (Many date its birth in the corporate world, for example, to the Hudson Institute’s 1987 “Workforce 2000” report, which opened the eyes of many executives to demographic changes in their labor force.) And, like many human beings around that age, “diversity” today is a somewhat awkward and ungainly creature. It is full of hope, energy and promise, and proud of its achievements—but also a bit vague on specific accomplishments and actual returns on investments.

The case for diversity has never been stronger, nor more widely accepted. Diverse workforces are more innovative and, generally, more profitable. And the sheer number and variety of efforts to achieve it are also at an all-time high. No serious observer thinks that these efforts will, or should, slow down. Yet many are uncertain about what diversity means, how to reach the goal, and what measures indicate real success. Why? The answers lie in the difference between what seems obvious about diversity and what is actually true—and in the complexities of the human mind.

A central aspect of the uncertainty is the fact that when it comes to fostering diversity, we still don’t know what works. In an exhaustive review of work on reducing prejudice, published in 2009, the political scientist Donald P. Green of Yale and the psychologist Elizabeth Levy Paluck, now at Princeton, concluded that very few efforts had been proven effective. “Of the hundreds of studies we examine, a small fraction speak convincingly to the questions of whether, why, and under what conditions a given type of intervention works,” the researchers wrote. “We conclude that the causal effects of many widespread prejudice-reduction interventions, such as workplace diversity training and media campaigns, remain unknown.”

One reason it is hard to tell if a diversity effort is succeeding is that the definition of “diversity” is variable. Consider, for example, the sharp differences in the meaning of the term that emerged among 68 human resources executives from around the world during a 2009 series of working groups at the Center for Advanced Human Resource Studies at Cornell University. The conversation for New York executives focused on race and gender, while San Franciscans were more prone to discuss LGBT inclusion, and those from the United Kingdom focused on generational divides as well as gender. In China, meanwhile, the important groupings were expatriates versus homegrown talent, Beijing versus Shanghai, and urban versus rural. From elsewhere in the world comes more testimony that diversity means different things to different people, from India (where caste must be considered) to Brazil (where disability is keenly appreciated as a category for inclusion) to many other Latin American nations, where the need for more diversity regarding higher education (whether and where a candidate went to college) is often mentioned by recruiters and employees.

Diversity across time has been as variable as diversity across territories. Women were not included in many first-generation diversity efforts, which focused on racial and religious identities. And efforts to include gay people in diversity efforts, now standard in many parts of North America and Western Europe, were almost unheard of until recently. The same could be said about efforts to include people with disabilities. Sometimes it seems that just as your diversity materials catch up with the latest cultural change, a new one appears on the horizon.

New identities—the kind that were not on the radar a decade or two ago—arise from the same source as old, familiar categories like race, religion and gender. That source is individuals’ thoughts and feelings about themselves and others. Each of us has many different group affiliations. An employee may be “the older person” in the morning meeting, “the mom” at lunch, the “bean counter” on the afternoon teleconference and “the Asian American” in a train car filled with non-Asians, all in the course of one day.

As time passes, some identities count for less in concerns about diversity (for example, the distinction between Catholics and Protestants in the United States) and others count for more (like the distinction between straight people and gays). Some categories can exist for a long time before they are seen as social groups who were excluded in the past and should not be in the future. For example, the concept of “neurodiversity” so defines people with attention deficits, autistic traits, Tourette’s syndrome and other non-typical brain conditions.

All this explains why “diversity” is a moving target. It cannot be reduced to specific measures aimed at specific groups of people. What works in one place will fail in another; what is cutting edge in one time may be inadequate a decade later. This is why bean counting, pep talks and good intentions will not automatically lead them to the benefits of diversity. It simply isn’t enough to make sure that your organization has (or is trying to have) enough members of x or y or z social group. It isn’t effective simply to put the staff through the familiar rituals of sensitivity training or “cultural competence.” The reason: Diversity isn’t a static trait like pregnancy or a measurable organizational capacity like the Six Sigma process. It is a characteristic whose indicators change from place to place and decade to decade.

Of course, the challenges of diversity are not limited to the difficulty of pinning down what it means for your particular time and place. However you define the social groups that you wish to welcome into your organization, employees will resist including those whom they are not used to working with. A generation ago, this was seen as an unfortunate flaw that afflicted bad apples among us. Today, psychologists understand that resisting the unfamiliar isn’t a trait, like red hair, but a state of mind, like fear: We all spend some time in it. The reason is that we are all disposed to protect our “social capital.”

“Social capital”—networks, friendships, norms of behavior, levels of trust that make it easier for teams to function smoothly and individuals to be effective—is often described as a key to better performance. The reasons are obvious: Employees who don’t trust one another or know how to behave around each other aren’t as able to get things done.

As diversity increases, social capital decreases. Among the things that individuals of different backgrounds don’t have in common is an easy understanding of each others’ beliefs, customs, hot buttons and human ties. The pattern is well established by researchers, who have found it in neighborhoods (where neighbors are less trusting in diverse areas) and in organizations of various sizes.

Yet one lesson of diversity is that the comfort of social capital is not the best thing for an organization.

The sociologist Alejandro Portes of Princeton recently concluded that the familiarity, predictability and ease of understanding that characterize non-diverse organizations produce mediocrity at best—and disaster at worst. It turns out there is such a thing as too much comfort and too much trust in the judgment of people who are like you.

Portes’s comments stemmed from a study published last year in the Proceedings of the National Academy of Sciences, in which Sheen S. Levine, a sociologist at Columbia, and his co-authors created stock-trading markets in Southeast Asia and North America, and recruited business and finance professionals to trade on them for real profits (or losses). Some volunteers, picked at random, were placed in markets run by the majority ethnic group in their area (which they knew because they would see the identities of their fellow players). Others were assigned to markets with at least one member of an ethnic minority. The difference between ethnic sameness and diversity was striking: Stock pricing on the exchanges “fit true values 58 percent better in diverse markets,” the authors write. Moreover, the homogeneous markets were more vulnerable to bubbles and suffered worse crashes after inevitable corrections. In the hard numbers of balance sheets, the experiment had shown, as the researchers note, that ethnic diversity “may be beneficial not only for providing variety in perspectives and skills, but also because diversity facilitates friction that enhances deliberation and upends conformity.”

Gender diversity, like ethnic inclusion, causes the same sort of friction and yields the same sort of benefits, according to a recent study by economists Sara Fisher Ellison and Wallace P. Mullin. The partners analyzed eight years of employee survey results from a large (and unnamed) global company. The researchers were interested in data that showed how diversity affected the performance of each office.

Higher levels of gender diversity were also associated with less cooperation among employees. Social capital (Ellison and Mullin called it “social goods”) decreased with gender diversity.

Yet offices where workers perceived the firm to be more accepting of diversity were more cooperative than those whose employees are more skeptical. Employee morale and satisfaction were also higher in such offices.

The explanation? The authors believe “employees like the idea of a diverse workplace—and may therefore provide social goods more readily in a setting that they think is supportive of diversity—but are actually more comfortable in a homogeneous setting.”

Whatever the comfort levels, though, the study found that revenues for the firm’s typical office improved with gender diversity. That suggests, as the authors write, that the loss of social capital due to gender diversity was made up for by the boost it gave to behavior that helps the bottom line. In fact, their analysis suggests that an office in this firm that moves from single-sex to gender-diverse should have a revenue gain of more than 40 percent—though, as they note, this is an abstract number drawn from their data. Real-world performance would be affected by the other changes necessary to shift staffing so dramatically. Yet, as they understatedly put it, “the relationship uncovered in the sample is still of interest.”

The research suggests, then, that diversity exchanges some social ease for a more profitable and dynamic organization. That should be a no-brainer, especially given that morale in these stimulating places seems to be higher, despite—or perhaps because of—the friction. However, we do have brains, which interfere with our ability to appreciate a diverse environment. Diversity imposes its burdens right here, right now for employees, while its benefits will appear in the future. Emphasizing the business case for diversity may not help here—it can make it appear that the benefits are going to go only to far-off shareholders. Employees sometimes take an “if it ain’t broke, don’t fix it” attitude.

At its ugliest, this kind of resistance can be expressed in hateful words and acts. But, as someone almost invariably points out after every shocking incident, today’s environment has changed profoundly, in a simple and important way: Where leaders once condoned racism, sexism and other kinds of discrimination, in much of the world they now condemn it. For millions of people who wish to appear educated, modern and employable, it is important to be unprejudiced—and many are sincere in this goal.

Yet even as the conscious mind accepts the value of diversity, the unconscious mind tries to protect its social capital, quite outside our awareness. Much of the behavior that forestalls diversity is entirely unconscious. In our minds, dealing with those who are like us—who understand what we mean without needing much explanation, who share experiences and an outlook—feels comfortable and reassuring. It is not consciously hostile to those who aren’t at the table, but it has the same effect. The biases that keep us from including new kinds of people exert their effects outside of our field of awareness.

Consider, for example, the latest continuing education catalogue from the University of North Georgia, titled “Success: Why Follow When You Can Lead!” The publication features a photo of four people in business attire on a race track. A white man is breaking the ribbon at the finish line; another white man is close behind him. In third place is a woman, while in last place, just behind her, is an African-American man. No one in the school’s administration, it is safe to say, believes white men will or should win all of life’s competitions. Yet none found the photo problematic. Neither have other publications that used the stock photo over the years.

Similarly, very few drivers around the world would be comfortable with the thought that they’re less careful about the lives of some pedestrians than others. Yet research shows unconscious bias exists. In 2014, researchers at Portland State University and the University of Arizona repeatedly went to an intersection in Portland to see if race made any difference in the way drivers treated them as pedestrians (there were three white experimenters and three black, all dressed identically). Across 168 instances, nearly twice as many cars passed a black pedestrian waiting to cross the street before yielding as they did passing a white one. This meant the average wait time before crossing safely for the black men was more than 30 percent longer than for the white men.

This is the kind of bias that subtly works against diversity, as people lean—without being aware of it—toward the familiar, and away from the different. The effect is to reinforce the traditional relationships between groups of people in society, even among those who, in their conscious beliefs, say we need change.

An amusing example of this kind of unconscious inertia occurred recently at a discussion on, of all things, diversity, at the South by Southwest conference in 2015. The incident was viewed millions of times on social media. In speaking at length about the need for more women and non-whites in the tech sector, Eric Schmidt, Google’s executive chairman, several times interrupted and talked over his fellow panelist, Megan Smith, who is the federal government’s chief technology officer. What, someone asked in the Q-and-A afterward, did this say about the dynamics of a gender-diverse workplace? The answer was clear: Even among those determined to create a better future, old habits die hard.

Those aspects of diversity work that can be quantified—the number of women in upper management, or the number of employees who have gone through training—are necessary but not sufficient for success. The process also requires flexibility, openness to change, self-awareness and self-discipline. True diversity, which lets an organization welcome all talent and thus benefit from inclusion, is a state of mind.