When trying to determine whether an organization is diverse, people often look to “the numbers”—that is, statistics about the relative representation of different groups. Implicit in this interpretation is that a threshold exists in terms of relative group representation at which an organization crosses over from not being diverse to being diverse. This research seeks to determine where members of different social groups draw this threshold—a concept we refer to as “the diversity line.”
The dominant groups (Men, White folks) had lower thresholds for diversity than non-dominant participants (Women, Black folks, Latinx folks) when presented with organizations of different make-ups that were randomly assigned. The authors tested different mediators, such as to see if people adjusted their threshold for diversity in response to contextual information (to justify a lower threshold of the outgroup in the relevant contextual baseline). The story for dominant groups was straightforward – they set thresholds of diversity that worked to serve their hierarchy enhancing goals.
While this study was extremely well thought out (and included a fun application of fitted Gaussian Curves), the study is, like many others, inherently flawed because it looks at men vs. women independently of race, which only included White, Black and Latino. A lot of research lacks an intersectional approach, as well as Asians (hello!) and non-binary folks. But the point of the research is to demonstrate how Social Dominance Theory plays out: members of dominant and nondominant groups generally have competing motivations when it comes to the existing hierarchy. dominant groups are typically motivated to preserve or enhance the hierarchical social order, whereas nondominant groups are more likely to be driven to attenuate the hierarchy, pushing toward a more egalitarian outcome.
If you are an organization that is setting targets for diversity – and I believe targets can be effective in a portfolio of diversity management strategies in many types of organizations – then you must include a variety of perspectives in that discussion. I would hate to see a group of homogenous leaders from a dominant group set a target, as this research indicates it would likely not meet the standards of non-dominant groups.