The case against diversity targets (not a viable performance driver)

Google Memo and firing of engineer James Damore continue to provide us with numerous heated discussion topics. It seems that the other extreme is actively denying biology, sociology and existence of any notable gender differences. On the other hand, people who do not completely deny science, that is, those who express some understanding of these disciplines or those who at least respect free scientific discussion are in danger of getting linked to controversial subcultures and political extremes.

Scientists from aforementioned fields have spoken up for a reason. For example, Jordan Peterson, a professor of psychology at the University of Toronto, has taken an active role in rationalizing and making sense of what is going on (see Interview with Damore). However, desire for "diversity" and its implications often go beyond biology and psychology. Diversity targets are often backed by vague notions that heterogeneity is good for organizations. Approaches that have an essential and less covered say on this topic are economics and organization science. I have an attempt to extend this discussion to these fields while paying respect to the findings of psychology.

Is "diversity" beneficial?

People tend to have an abstract idea that "diversity" is beneficial for organizations. Of course, it is very reasonable to assume that unique individuals complement each other. Interaction among different individuals may result in more creative and even efficient problem solving. Entrepreneurial approaches to organizations and economics emphasize that successful entrepreneurs imagine opportunities that are mobilized by combining heterogeneous resources. Human capital has gained a crucial role in resource combinations of contemporary organizations. Careless theorization could suggest that the degree of differences ("diversity") among team members could constitute a driver for economic performance.

WRONG. The devil is in the details. Organizations are different by their products and services, their strategies and business models. They have different requirements for resources and management in order to produce different services and products. For decades, contingency research has challenged the idea of universal best practices. Significant part of management studies are about realizing that the world is not homogeneous. Instead, there exist contextual dependencies. When it comes to common sense, it is rather realism not to expect everything to be equally distributed. The burden of evidence lies on the one who suggests that some specific distribution, say, 50/50 or 30/70 between genders or any other fixed formula of "diversity" should be universally the optimal one.

There is no objective measure for "diversity"

A couple of essential philosophical notions must be made here. First, diversity is not something that could be reliably measured nor even captured by a uniform scale. There are endless amounts of attributes that differentiate people from each other. Many of them are of qualitative nature too. In addition to gender or age, these aspects could be political views, ethnicity or as well desire for soy products for example. It is impossible to define objective weighs for every diversity attribute in order to derive an ultimate variable.

It is exceptionally only in some specific occasions where these attributes can even be direct or strong drivers for performance. At least when it comes to customer surface, diversity can be beneficial for reaching customer base. Now we expect that the customer base is diverse and acknowledged by the organization. Some attributes also make individuals more or less fit for providing the best individual service, implying higher revenue. Consequently, no universal generalizations can be made. Benefits or costs of any specific diversity formula are conditional and context-specific. It is the entrepreneurs and managers who have specific knowledge and judgment about the ways to improve performance.

"Diversity" is not a viable performance driver

More importantly, commonly used diversity attributes such as gender or ethnicity make bad direct variables for explaining organizational performance. As said, entrepreneurs and managers imagine opportunities and plan their implementation. Subjectively they combine resources (including capabilities) that appear suitable for mobilizing the plans. It is not the aggregate degree of heterogeneity that defines performance, but the specific resources in relation to others and the context.

When it comes to humans, closer attributes that define desired capabilities or personalities often lean on psychology. For example, a famous Big Five framework uses variables called openness, conscientiousness, extraversion, agreeableness and neuroticism. Psychological research (e.g. 1, 2) has reported gender differences among these aspects. For example, women appear to score higher on agreeableness and neuroticism than men. That is very likely to explain why there may appear more men in roles where people are required to stand arguments, confrontation and stress.

Implications for organizations

It is the statistical and managerial implications that people tend to get wrong about differences among populations, as the Google Memo case has once again illustrated. Gender differences in personality traits imply that it is easier for a recruiter to find more suitable personalities from the pool of a specific gender. This does not mean that men or women with specific desired personality traits would be generally "better" in the specific task than the opposite sex with the same traits. Neither does it mean that the company should favor any gender or try to achieve any "diversity" target by default. When desired attributes match, gender becomes meaningless.

Organizations that care about performance are looking for direct and close drivers when combining resources. In other words, they want to ensure fit between human resources and purposes. Gender or any fixed diversity distribution is not generally a direct performance driver. Thus, performance maximizing organizations are not primarily interested in gender or other similar diversity attributes when they hire employees and combine teams. At least this would hold in the absence of diversity regulations and other external institutional pressures. In light of scientific understanding, biological differences in gender also drive differences in personalities. That is why the distant role of gender in performance does not, however, result in completely equal gender distributions organizations and society.

To sum up

As a synonym for resource heterogeneity, diversity is a prerequisite for success. In a similar logic, recipe of different ingredients is a way of succeeding in cooking. However, it is not diversity among ingredients that simply drives success. Rather, it is the role of each single ingredient, how they fit together and how they serve the specific purpose. Some ingredients may appear useful in many dishes, but that does not mean one single recipe would be universally the most suitable for every dish. As well, it would be nonsense to claim that there was a single universal formula of diversity that maximizes performance and prosperity. Before embracing diversity, one should carefully justify the function and relation to performance.