In a previous post, I raised the question of whether body-cams should be adopted by all police departments. I suggested that, although a typical evening newscast would lead you to believe that police abuse of authority is a national crisis, what is often overlooked, is the fact that the majority of Americans express a great deal of respect for the police. Gallup polling has consistently shown rates at 60% or above since 1992. That is nearly a quarter of a century in which a sizable majority respects the legitimacy of authority exerted by police. Moreover, the US fairs well with other wealthy countries which consistently show that prosperity and effective policing go hand in hand. Yet, these same studies indicate that lack of confidence in the police is concentrated among disadvantaged groups, and not just in the United States. What this boils down to is the simple fact that concern for the legitimacy of police authority is a localized crisis, not a widespread pandemic. In other words, the diffusion of negative information or beliefs about the police is selective.
In this post, I want to revisit this idea in light of a new paper in the American Sociological Review by Henrich R. Greve, Ji-Yub (Jay) Kim, and Daphne Teh entitled “Ripples of Fear: The Diffusion of a Bank Panic”. Greve and his colleagues argue that negative information (e.g. protests against Wal-Mart) is selective in that it only spreads to some organizations and not others. Not all Wal-Mart stores experience protests, why? They emphasize that the academic literature examining selective diffusion has mainly focused on organizations as key players (i.e. active agents) in this process. In contrast to the adoption of innovations that may benefit an organization, negative or adverse reactions (negative information) spreads in a different way because the organization is not trying to spread it. Wal-Mart does not have an incentive to encourage protests. Diffusion of information is a selective process that occurs, but the organizations are passive agents. In contrast to positive information (or something that can spread that helps the organization), there is an external agent that is responsible for the diffusion (what they call external stakeholders). The question then is, if the organization itself not responsible for spreading negative information, how do external agents create selective diffusion paths? Why some Wal-Marts and not others? They propose a model whereby information about an organization is first judged by an external agent to be negative. Then, stigma spreads to other organizations through categorical judgment. That is, the external agent determines whether a separate organization is sufficiently similar to the organization for which information has been deemed negative to place it within the same category. Greve and colleagues argue that this determination occurs due to organizational similarity, community similarity, and network proximity. As this mechanism operates, selective diffusion paths are created.
So what does this have to do with perceptions of the police? A key part of their argument is that external agents will judge an organization as negative if the community in which that organization exists is similar to the community in which an already stigmatized organization is itself embedded. They develop this position by extending well-known issues concerning attribution bias and in-group/out-group preferences. As they argue: “These preferences influence judgements, because individuals with out-group bias tend to ascribe the cause of a negative event in another community to its members if they have very different…characteristics. Stakeholders often assume that a negative event occurring in a dissimilar community will be less likely to happen in their own community” (p. 401). This provides a plausible mechanism why the concern for the legitimacy of police authority is a localized crisis, not a widespread pandemic. The marked racial and socioeconomic differences across communities may create a situation in which external agents do not categorize negative information about a police department (e.g. accusations of excessive force) in another community to be applicable to their own local police department. Here, individuals may attribute blame to the individuals within those communities. In contrast, similar communities may perceive such events to be more likely to occur in their community. Since the cause is less likely to be attributed to internal characteristics (due to similarity), blame is more likely to be attributed to the organization. This is the process of selective diffusion. As mentioned, such a mechanism could generate an empirical trace that is consistent with the Gallup polling mentioned above.