December 7, 2018
10:30am to 12:00pm
Haines 279

The Markets, Organizations, and Movements Working Group is proud to announce:

Gabriel Rossman (UCLA): The Weakness of Influentials

The “influentials” or “hub influence” hypothesis argues that social network hubs have disproportionate effects on the success of innovations. The performative version of the theory is “viral marketing,” in which firms, public health campaigns, or other organizations target hubs to promote the diffusion of products, health behaviors, or other innovations. Some extant simulations have confirmed hub influence because the models were constrained to allow learning only along the social graph. However, this paper relaxes this assumption and shows that hub influence is not robust to the introduction of even a weak constant hazard function (e.g., advertising). To a lesser extent, non-spatial endogenous growth (e.g., bestseller lists) also disrupts hub influence. Since such practices as advertising and bestseller lists are ubiquitous, this suggests that hub influence applies only under highly restrictive scope conditions and that viral marketing efforts will therefore usually not be especially effective. Even if the modal adoption is primarily the result of network influence, there are no distinctively important structural positions in the network so long as there is any non-spatial influence. The collapse in the special efficacy of high centrality nodes occurs as a phase transition when the constant hazard is greater than zero. These results suggest a return to the two-step flow model, in which non-spatial sources of information seed diffusion widely with numerous opinion leaders, none of whom are individually important, but who can drive diffusion in the aggregate.