Estimating the Effects of Believing Conspiracy Theories

• In Event: Misinformation and Conspiracy Theories

Abstract

Misinformation has grown pervasive in American politics, and the rise of groups like QAnon reveal that conspiratorial thinking is neither a rare nor a benign phenomenon. Unfortunately, assessing the impact of conspiratorial beliefs is complicated by potential endogeneity. To clarify: confirmation bias would suggest that it is easier to accept a conspiracy theory that denigrates one's political opponents. The analyst's challenge is to separate the effects of the new belief (that a political figure was part of a conspiracy) from the effects of the original attitude that the conspiracy theory was reinforcing. While some scholars might encourage the use of randomized experiments to escape this conundrum, ethical guidelines weaken the potential treatments one could employ. We use an alternative approach, combining survey questionnaire design with instrumental variables methods. We use data from two original surveys that included questions that aimed to create valid instruments for beliefs in conspiracy theories about (a) illegal voting in the 2016 election, and (b) Russian involvement in U.S. nuclear policy and elections in 2016. We then examine how these beliefs impacted attitudes towards Trump's election integrity commission, the Mueller investigation, and presidential candidates. After controlling for endogeneity, we conclude that accepting a conspiracy theory has meaningful and independent effects on beliefs about policies and public figures.

Authors

• Scott J Basinger, University of Houston

• Yongkwang Kim, University of Houston