Organizational Diversity Should be More than Skin-Deep

Aaron Halliday, PhD

There are a multitude of reasons why organizations should be hiring more diverse workforces. Beyond any legal imperative, one category of reasons is as Gandhi put it “to be the change we wish to see in the world”. The motive, here, is that organizations strive to ensure that their organizations are not discriminating on the basis of protected classes like race, religion, and sex because things like racism and sexism are immoral, they harm society, our economy, and violate human dignity. This rationale has motivated some countries to enact legislation that attempts to level the playing field and promote fairer hiring and people operating practices. Simply put, it’s the right thing to do. However, as persuasive as the moral argument is it will never be enough to convince everyone. The omnipresent concerns over the bottom line and return on investment demand an evaluation of whether dedicated diversity programs are feasible, effective, and economically worthwhile. These concerns are all warranted and worth paying attention too as it was estimated that organizations spend a staggering $8 billion annually on diversity (Hansen, 2003). Which has motivated a great deal of research into the many financial benefits that being perceived as an organization that values diversity often bring.

Given that passions often run deep concerning issues pertaining to issues like race, sex, gender, age, and religion organizations and the people that comprise them are often wary to discuss these issues and when they do manage to be brought into the forefront the conversations are often embroiled in the emotions that are so deeply tied to these subjects. In this regard, discussions tend to predominantly focus on the moral and emotional arguments for (and against) diversity. As a scientist and organizational psychologist, it’s exactly this reason why I find it incredibly important that we take an objective look at the scientific evidence and try to understand diversity and what roles diversity plays in driving organizational performance and day-to-day people operations.

While the emotional and societal arguments promoting diversity rightfully focus on protecting individuals from human rights violations and promoting human dignity they tend to focus on superficial demographic characteristics (the extent to which a single workgroup or team is heterogeneous on characteristics such as age, gender, ethnicity, functional background, and organizational tenure; Lawrence, 1997; Tsui, Egan, & Xin, 1995) often termed surface-level diversity; whereas the strongest objective (business) case for leveraging diversity in teams to promote overall productivity tends to be found more in deep-level diversity, or differences within workgroups or teams with respect to attitudes, personality, values, cognitive processes and specialized intelligences (e.g., Barrick, Stewart, Neubert, & Mount, 1998; Jehn, Chadwick, & Thatcher, 1997; Harrison, Price, Gavin, & Florey, 2002). Not all diversity is identical and by taking the time to understand the differences you’re more likely to keep your people happy and productive while allowing your organization to make the most out of your existing and upcoming diversity initiatives.

The effects of surface-level diversity

The scientific study of diversity’s impact within organizations actually began with the study of surface level diversity. Given the historical context, this seems to be the natural default assumption of what it means to be of a minority group that would, in the aggregate, diversify a larger group: to be readily and easily be identifiable as different from the majority group in some visual or otherwise quickly discernable way. The research conducted in these early days produced something of a mixed bag of findings that led some to prematurely conclude that the business argument for promoting diversity was questionable. The 1980’s and 90’s produced quite a bit of research actually showing no effect or very complex effects of the diversity of age, sex, and ethnicity with regards to various outcomes.

Many scientific works showed that minority groups tended to serve communities that were historically underserved and actually perform better under these circumstances. For example, minority physicians were consistently found to be more eager to service minority groups including those located in rural locations and the patients regularly reported greater satisfaction when receiving care from these practitioners (for example, Moy & Bartman, 1995; Cantor, Miles, Baker, & Baker, 1996; Komaromy et al., 1996). Research similarly showed that learning environments could also be enhanced by having a more diverse student body as such variation tended to challenge stereotypes and cultural assumptions (For example see Witla et al., 2003).

Later research demonstrated surface-level diverse groups to perceive their information as being more diverse which tended to promote more discussion regarding the task at hand (relative to less diverse groups). However, this early research showed many negative outcomes of promoting merely superficial diversity, especially in team settings. Such surface-level diversity were regular and it was not uncommon to see reports indicating that surface-level diversity promoted such outcomes as social isolation (Kirchmeyer, 1995), reduced team cohesion (O’Reilly et al., 1989), poorer communication (Zenger & Lawrence, 1989; Larkey, 1996), higher turnover (Jackson, et al., 1991; O'Reilly et al., 1989; Tsui, Egan, & O'Reilly, 1992), reduced commitment (among majority members: Tsui et al., 1992), and even lower performance ratings (Kraiger & Ford, 1985). These findings were explained in terms of a great many of the scientific theories that were developed in an effort to understand the roots of racism and prejudice that were developed by psychologists of an even earlier age: from when America was still reeling at the tail end of the civil rights movement into the mid to late 80’s: social identity theory (Tajfel, 1978), the similarity-attraction paradigm (Byrne, 1971), and social- and self-categorization (Turner, 1982). These early days of research focusing on demographic diversity in teams seemed to paint an uncertain if not somewhat grim picture of human interaction in the workplace when it came to diversity. However, it was a necessary step for research to take in order to more clearly understand the processes at work when we think of the impact of diversity in a work setting.

The effects of deep-level diversity

As it usually does with science, when met with ambiguity and counter-intuitive findings researchers consider the evidence and experiment, and eventually stumble upon a clearer picture of what happens to be truly at work. When organizational psychologists and managerial scientists were originally considering the benefits that diversity would likely bring to an organizational setting they were using demographic factors like race, religion, sex, sexuality and other surface-level factors as stand-ins or indicators of deeper-level diversity. It may seem to be a natural assumption that individuals stemming from different cultures or backgrounds or social economic standings would tend to be quite different with regards to the attitudes they hold, their personality traits, what they tend to value and how they express their values, and what expertise they may have and the means by which they work through problems. However, as is typical for humanity, in making these assumptions humanity missed the mark and made one another out to be way more different on the basis of external features alone than we actually are. Rather than your race, religion, or sex, diversity in functional/occupational background and education were found to improve performance, creativity, and innovation (Bell, Villado, Lukasik, Belau, & Briggs, 2010).

Similarly, workgroups were found to perform better when the outcome was innovation and when the factors that diversified individuals within the team tended to revolve around task-relevant knowledge. When researchers specifically focused on how workgroups were composed they found that workgroups can either be composed in a way that multiple (deep- and surface-level) attributes can converge and align to form demographic faultlines (for example, a medical team of two female nurses and two male doctors) or, alternatively, they can be composed in such a way that cross-cut so that attributes tend to be mixed, random, and uncorrelated with one another (e.g., a male and a female nurse with a male and a female doctor). Although this is an overly simplified example, it moves to show that both workgroup compositions would have 50% diversity where it comes to sex, but a workgroup arranged around faultlines is likely to promote intergroup bias whereas cross-cut would not. A recent meta-analysis confirmed this finding, discovering that demographic faultline strength is negatively related to social integration and resulting performance (Thatcher & Patel, 2011).

Additional factors at play

Some of you reading this may be still hanging onto a lot of those concerning findings pertaining to surface level diversity potentially wreaking havoc in team environments. This may motivate some employers to expressly request their people to ignore any differences within their workgroup. Unfortunately, despite their good intentions doing this actually has the revers effect. By attempting to ignore group differences (rather than acknowledge and understand them) performance actually tends to drop. Simply put, by asking group members to ignore individual differences within a group you are more likely to be met with a worse performing team (Philips, Northcraft, & Neale, 2006).

So what do we do about the potential negative impact of surface level diversity while trying to be ethical people with integrity, acting as good stewards of our organization and society? Recent research has also shown that these effects may not be so profound in real world contexts as they may be in the lab. Just as preventing people from discussing and addressing their individual differences was found to decrease performance, allowing people the opportunity to get to know one another as human beings was found to yield the opposite effect. Research shows that the length of time individuals spend working together can have profound effects on a team particularly when it comes to diversity. Over time, people engage in meaningful interactions and develop a deeper understanding of the individuals they’re working with. The effect this has is quite convenient; it tends to weaken the (often negative) effects of surface-level diversity while simultaneously enhancing the (often positive) effects of deep-level diversity (Harrison, Price, & Bell, 1998). This indicates that it may be a great idea to create teams with particularly diverse people who have a longer history of working together rather than to build a highly diverse team of unfamiliar individuals. Regardless, encouraging your people to become familiar with one another beyond a superficial context is highly recommended as it promotes a number of positive organizational outcomes.

This latter work on diversity in the workplace all seems to be echoing the sentiments of many of the minorities fighting for equality during the civil rights movement. Many stating that they are complex people and their race, sex, or any single facet of who they are should not be the defining property that they are judged upon. After all, Martin Luther King Jr. famously stated in his I Have a Dream speech “I have a dream that my four little children will one day live in a nation where they will not be judged by the color of their skin, but by the content of their character.” This same sentiment is heard in modern research pertaining to intersectionality and similar works that acknowledge being a woman isn’t necessarily the same experience as being a gay, black, Muslim, woman. We all have our differences but given the complexity of people, our differences probably shouldn’t be focused on in isolation, but rather should be understood by taking in the composite of everything. This requires more work. It may require personality testing to understand your people at a granular level. It may require your hiring personnel to carefully vet people of unfamiliar educational and professional backgrounds who can perform the task with unconventional or unfamiliar methods.

It will certainly take more time to do correctly. However, it leads to a deeper understanding of an individual. In a work-related context this is to say it is infinitely better to intimately know your people so as to build a team with directed strategic intent rather than to hire an abundance of visual minorities because you have heard it will provide you with untold gains in reputation, performance, and profitability. The former is doing people operations right, the latter is merely a different form of exploitation.

The take-home message

Although there are many objective and emotional arguments for the continued promotion of surface-level diversity within organizations a strong case is similarly made for the promotion of deep-level diversity highlights factors pertaining to less readily apparent characteristics. This argues that all efforts to meaningfully promote both deep and superficial diversity should be wary of recruitment and selection efforts that deliberately weed out all but one ideal prototypical individual. Research may indicate that highly conscientious people are more likely to be high performing under most contexts of contemporary work. But by using psychometric instruments to assess a wide number of variables and to use this to constrain your candidate pool to a produce a number of (perhaps very diverse on the surface but) highly uniform, identical, cookie-cutter candidates that differ little from existing personnel you’re in many ways objectively failing to promote a more diverse organization and you’re less likely to reap many of the performance and fiscal benefits that comprehensive diversity programs should be leveraging.

The evidence never presents a crisp and uniform message, either. Surface- and deep-level diversity are not entirely detached from one another. The challenge that people and organizations face lies in keeping in mind that while surface-level diversity is quickly and readily apparent it’s a horrible stand-in for understanding deep-level diversity and fails to recognize the contributions of any individual. This makes it more work for organizations to understand what’s really feeding into diversity programs as an input but shows a clearer path for realizing the benefits of these programs. Although it’s statistically less likely, it’s entirely possible that a room of five old white guys is more deeply diverse than a room of five individuals each of a different race, religion, and gender. You may be hedging your bets by guessing that the more surface-level diversity present in a group that it will similarly be more likely for the group to also be deeply-diverse; but the truth of the matter is nobody really knows unless you’ve taken the time and effort to assess the characteristics of each individual. This presents serious challenges where organizations are attempting to rapidly hire people or staff an organization of single-serving contract workers that may never have the opportunity to get to know the individuals they’re employed to work with. Which makes a strong argument for slower, long-term hiring, investing in assessments of employee attitudes, values, beliefs, personality, and thought processes/intelligences. It may be easier to only use readily apparent data like race and sex, but it’s also ineffective. In order to understand someone you have to give it more than a cursory glance and that requires an investment in time and energy. Research indicates that rather than placing the entire focus on promoting surface-level diversity (something I think most of us agree is great to promote for a better and more equitable society) we should additionally take into consideration what each individual has to contribute to the gestalt of the whole (the team, department, organization, culture, etc.).

It is fantastic to hire more minorities and to bring diversity into our workplaces, but it is even better to recognize our people for the individuals they are and to have a deep and meaningful understanding of our newest addition to the team,

“Maria” – who she is, what she’s an expert in, and her understanding of the world. If you’re going to promote diversity the evidence shows that you shouldn’t cut corners or do it in a lazy way. You have to commit to it and go more than skin-deep.


Aaron Halliday is an Organizational Psychologist and Data Scientist.  He uses his education and experience to provide consulting grounded in a framework of evidence-based management to help organizations and individuals overcome challenges, prevent problems, and promote organizational health operations using a rigorous data-driven approach.