What determines the pecking order in a flock of birds? Or the page rank of a website in a Google search? Or which academics land the most prestigious jobs? Hierarchies are everywhere, and new research continues to advance our understanding of how and why some end up high and others low. As it turns out, if we look at the general patterns that emerge when hierarchies form, we can offer a single answer to each of these questions.
In a commentary this week in the Proceedings of the National Academy of the Sciences, SFI colleagues Simon DeDeo and Elizabeth Hobson* discuss advances in the science of social hierarchy. They put recent work by Daniel Larremore* and colleagues in the context of nearly a century of research into this most basic of social phenomena. They argue that if we boil down the patterns that researchers observe in biological, informational, and social hierarchies, we see that hierarchies tend to form in universal ways.
For some time, researchers studied rank and hierarchy through the lens of the horse race. In a horse race, jockeys accumulate points independently that determine their place in the order of things. Yet, as DeDeo and Hobson argue, in network systems, rank is a much more relative thing.
When we adopt methods that capture the more holistic ways that hierarchies form, we start to see more clearly the universal patterns that obtain in any system of rank.
Among the several universal qualities of hierarchies that DeDeo and Hobson elucidate, two stand out. First, as they form, hierarchies tend to create gaps between intrinsic value and relative value. Second, these gaps are not incorrigible. As DeDeo and Hobson conclude, hierarchies tend to form in ways that admit of re-engineering.
Read the commentary, "From equality to hierarchy," in the Proceedings of the National Academy of the Sciences (May 25, 2021).
*Simon DeDeo is a member of SFI’s external faculty and a former SFI postdoctoral fellow, now based at Carnegie Mellon University. Elizabeth Hobson is a former SFI postdoctoral fellow now based at the University of Cincinnati. Daniel Larremore is a former SFI postdoctoral fellow now based at the University of Colorado, Boulder.