Marvin Smith, project manager for Microsoft’s Entertainment & Devices division, co-wrote this article.
Diversity is a subject of great importance to recruiters. Most organizations have some stated diversity goals. Larger employers have director-level positions and even entire departments devoted to the goal of increasing diversity. Corporations are estimated to spend over $8 billion annually on diversity programs. This all has a purpose, presumably that more diversity is better.
The theoretical “business case for diversity” is that in a global and racially or ethnically diverse marketplace, a marketer that employs a racially or ethnically diverse workforce is better able to understand the demographics of the marketplace it serves and is thus better equipped to thrive in that marketplace than a company whose employee demographics do not match their market’s.
Sounds good. So recruiters must strive to create a diverse pool of candidates and deliver the same to their employers.
Now comes a study that establishes that efforts to improve diversity, instead of delivering the purported benefits, manage to accomplish quite the opposite. The study, by Robert Putnam of Harvard, establishes that people in ethnically diverse settings aren’t drawn together, much less work better. Across workgroups in the United States, as well as in Europe, diversity (in terms of ethnicity, age, and other factors) is generally associated with lower group cohesion, lower satisfaction, and higher turnover.
This is consistent with observed facts. There is precious little proof that all the resources and effort devoted to improving diversity does anything more than give some people a warm glow and diversity trainers a paycheck. Hard evidence that diversity produces a wonderland from which flow creativity, innovation, and all sorts of other good things, is not there.
An Expensive Distraction
The case for diversity has never been strong. First, unless an employer engages in active discrimination, its workforce should automatically be diverse, given that the population is diverse. So a diversity program is basically a stand-in for affirmative action.
Affirmative action, at least, is based on legislation with a clearly stated purpose of redressing discrimination, but diversity appears to be built on a foundation of nothing more than good intentions, lacking a clear end goal. Wikipedia mentions that “no objective research support has ever been found for the ‘diversity business case.'” The Putnam study expands on an earlier study by Thomas Kochan of MIT that found no positive effects of gender or racial diversity on business performance.
The trouble with programs driven by good intentions is that they tend to cause facts to be given short shrift. What is the rationale that leads to the conclusion that having a workforce that is racially or ethnically diverse automatically creates more innovation or productivity? Logic would suggest that should be a function of the skills the workforce possesses. We aren’t exactly short of assessments to measure skills, so why use race and gender as a proxy for these?
Advocates of diversity claim that a workforce represents more than just skills, but also includes the sum total of their experiences, which are more likely to be diverse when the members of the workforce are diverse. That may be true if there was any proof for this claim, and then it would only have relevance if the experiences are somehow able to impact the results that an organization expects of its workforce.
More important, this only matters if the organization is in a position to take advantage of the diverse experiences of its workforce. It might well be the case that a racially or ethnically diverse workforce is better than one that isn’t, but without proof, how can anyone know? In the absence of facts to support the benefits of the program, diversity is little more than an expensive distraction for recruiters.
Diversity metrics are not much use either. They are primarily process metrics that focus on items like the proportion of candidate pools that are diverse and how well an organization has met its diversity goals in terms of recruitment, promotion, and turnover.
Redefining Diversity to Deliver Value
Diversity programs will not be abandoned as a consequence of the Putnam study. It would be the rare head of HR who has the courage to do so. Questioning the benefits of programs like diversity (which is really more of a cause) is generally not well-received.
Like with most good causes, calling them into question is typically interpreted to mean that the challenger is opposed to them. In the case of diversity, one risks being labeled as ignorant and narrow-minded at best or bigoted and implicitly racist at worst.
A critic of the Kochan study said the conclusions could only have been reached if those doing the study were not diverse (which was not true) since the results “defied logic.” No evidence to refute the results was offered.
These studies are not poorly conducted efforts with weak results; the Kochan and Putnam studies build on earlier studies and are the results of years of effort and draw on vast amounts of data.
Challenging a position is not the same as opposing it. If diversity is an organizational goal, then organizations should do what it takes to make diversity programs effective (i.e., deliver value). The first step should be to recognize what value diversity can deliver.
It’s a great vehicle for building trust between people. Trust is a necessary precondition for people to work effectively together. That is a desired business outcome. For organizations that are global, because of where they do business, diversity can help foster a greater understanding of other cultures and that in turn can promote trust.
This is not a contradiction of what was written above. A diversity program must have building trust as its focus, not some meaningless goal of building a diverse workforce for its own sake. How can this be accomplished? The Putnam study suggests that for a diverse group to be more cohesive requires assimilation. The group must have fewer differences and more in common with each other. It must become less diverse in terms of its values, purpose, and goals. That’s a tough pill to swallow; the core of most diversity programs has been the promotion of differences and the necessity of accepting them.
Some organizations have managed to do so. The Army is one such example. Admittedly, the Army is different in many respects from other employers, but much of the success of diversity in the Army has to do with an emphasis on shared identities and clarity of purpose. Contrast that with the typical company where it’s estimated that fewer than 5% of employees even know the mission statement (quick: what is your employer’s mission statement?).
Having a diverse workforce can also make it easier to recruit talent that is increasingly drawn from a diverse, global pool. A diverse organization is more attractive to potential employees. This has the value of addressing a real economic problem (limited supply), not a social one. But for the full value to be realized, the program must include a dedicated and focused effort at assimilation.
The evidence is overwhelming that promoting diversity without emphasizing assimilation is counterproductive. The Putnam study includes another interesting example. During the Civil War, the Union Army noted that among its units the greater the diversity (in terms of age, hometown, occupation, etc.), the higher was the desertion rate, that is, turnover. Given that the risks of punishment for desertion were very low and the casualty rate was very high, the only powerful force preventing desertion was loyalty to one’s fellow soldiers.
This fits well with present-day research by Gallup that shows that organizations where the proportion of employees who have a close friend at work correlates negatively with the level of turnover. That is, the higher the proportion of employees who have a close friend at their place of work, the lower the turnover for the employer.
What has this to do with diversity? The majority of a person’s friends tend to be from within their own racial or ethnic group. The exception to this is the Army, where the average soldier has many closer interracial friendships than the average civilian of the same age and social class. Clearly, shared purpose and identity are a factor in making diversity work. That doesn’t happen on its own. It requires an organization to dedicate itself to the goal of getting a diverse workforce to share and accept its values and purpose.
This does not mean employees need to be clones of some corporate ideal, but that they share the employer’s goals and not elevate their differences above it. It also means that employers also need to expand the definition of diversity to include people who have diverse viewpoints and experiences, regardless of race and ethnicity.
Sticking to a definition of diversity that is limited to demographics only betrays a lack of understanding of what being diverse means. The result of diversity should be different decisions that could change outcomes of business decisions. Diversity can result in better business decisions.
What Works
Diversity programs can deliver some value, but the programs need to adopt a different perspective and goals and explicitly recognize that without a focus on assimilation, they cannot succeed. Despite all that spending on diversity programs, the number of racial harassment cases filed with the EEOC has increased by 500% in the last decade. Given the weight of the evidence, employers would do best to rethink diversity and expanding the scope of the programs beyond just improving the recruitment numbers.
For recruiters, it means changing diversity recruiting to include some screening of applicants to assess whether they either share the identity and purpose of the workgroup they will join or demonstrate a willingness to do so.
The prescription for making diversity work has been known for a long time. E Pluribus Unum can apply to companies as well. The author wasn’t referring to employers but certainly had diversity on his mind.