Abstract
John Stuart Mill advocated for increased interactions between individuals of dissenting opinions for the reason that it would improve society. Whether Mill and similar arguments that advocate for opinion diversity are valid depends on background assumptions about the psychology and sociality of individuals. The field of opinion dynamics is a burgeoning testing ground for how different combinations of sociological and psychological facts contribute to phenomena that affect opinion diversity, such as polarization. This paper applies some recent results from the opinion dynamics literature to assess the impacts of the Millian suggestion. The goal is to understand how the scope of the validity of Mill-style arguments depends on plausible assumptions that can be formalized using agent-based models, a common modeling approach in opinion dynamics. The most salient insight is that homophily (increased interactions between like-minded individuals) does not sufficiently explain decreased opinion diversity. Hence, decreasing homophily by increasing interactions between individuals of dissenting opinions is not the simple solution that a Millian-style argument may advocate.