Understanding Misinformation Sharing: A Formal Model & Experiment

• In Event: Poster Session: Innovations in Public Administration Research & Experimental Research
  In Poster Session: Poster Session: Experimental Research

Abstract

Misinformation spread on social media has generated broad interest. At the inception of the internet, scholars, policymakers, and citizens expected the platform to be an amplifying medium of democratic norms (Diamond and Plattner 2012). While social media has helped amplify marginalized voices (Tucker et al. 2017), it has increasingly harmed democratic values and institutions through misinformation (Tucker et al. 2018, 20). The literature agrees that misinformation benefits someone, but a central dilemma is the lack of clarity on whom it helps and how. Despite continued advancements in the literature theoretically and empirically, problems and inconsistencies remain, including a lack of shared understanding of what causes people to spread misinformation, who spreads it, and how to stop it. This article focuses on ways to alter the sender’s incentives to spread misinformation.

In prior misinformation scholarship, the empirical and theoretical models have a limited understanding of the motivations of senders in an economic sense. For example, investigations on the impact of reputation costs are mixed. Some scholars report a mitigation effect of willingness to spread misinformation through reputation costs (Acemoglu, Ozdaglar, and Siderius 2021; Altay, Hacquin, and Mercier 2020; Gao, Ling, and Liu 2018), while others find no significant effects (Sunstein 2017, 73). We present a formal model that assumes people share misinformation if it benefits them. The formal model illustrates the individual-level rational decision to share misinformation. We then introduce a research design for an incentivized lab experiment that mimics the decision to share misinformation under different conditions to explore the motivations and countervailing interventions thinking about candidates and ordinary people. A lab game experimental design is an appropriate way to evaluate motivations behind misinformation sharing because it gets around the ethical problems of exposing people to misinformation that may impact real-world political viewpoints while illuminating the mechanisms behind the motivations. Additionally, the lab setting allows us to control specific incentives and factors that are present in a typical online environment. By manipulating monetary incentives, the experiment ensures that all participants have common incentives to select their preferred candidate or to be elected. We will randomly vary the incentives they get from getting their preferred candidate or themselves elected, including verification and reputation costs. Understanding the cost-benefit calculations behind misinformation sharing could be vital to policymakers working to mitigate the problem.

Authors

• Christina Walker, Purdue University

• Tara Grillos, Purdue University