Optimal Signaling of Content Accuracy: Engagement vs. Misinformation
83 Pages Posted: 12 Oct 2017 Last revised: 10 Jan 2019
Date Written: October 11, 2017
Abstract
This paper studies information design in social networks. We consider a setting, where agents’ actions exhibit positive local network externalities. There is uncertainty about the underlying state of the world, which impacts agents’ payoffs. The platform can choose a signaling mechanism that sends informative signals to agents upon realization of this uncertainty, thereby influencing their actions. We investigate how the platform should design its signaling mechanism to achieve a desired outcome. Although this abstract setting has many applications, we discuss our results in the context of a specific one: misinformation in social networks. Agents in a social network engage with content that is possibly inaccurate. Their payoff is based on the direct satisfaction from engaging, the disutility from engaging with inaccurate content, and the positive externality that they derive from engaging with the same content as their peers in the underlying social network. The platform can commit to a signaling mechanism that sends agents informative signals based on the realization of the inaccuracy of the content, and influence agents’ engagement decisions.
The optimal (in terms of engagement/misinformation) signaling mechanism admits a simple threshold structure: the platform recommends that agents “Engage” with the content if its inaccuracy level is below a threshold and recommends “Do not engage” otherwise. For the mechanism that maximizes engagement, these thresholds depend on agents’ network positions, which we capture through a novel centrality measure. In the case where the platform seeks only to minimize misinformation (regardless of the induced engagement), common threshold mechanisms with identical thresholds across agents are optimal. This is in contrast to the engagement maximization setting, where when agents are heterogeneous in terms of their network positions, common threshold mechanisms induce substantially lower engagement than the optimal mechanisms. We also study the frontier of the engagement/misinformation levels that can be achieved via different mechanisms and characterize when common threshold mechanisms achieve optimal tradeoffs. Finally, we supplement our theoretical findings with numerical simulations on a Facebook subgraph.
Keywords: Social Networks, Misinformation, Fake News, Bayesian Persuasion, Online Platforms, Information Operations
Suggested Citation: Suggested Citation