Consumption of fake news is a consequence, not a cause of their readers’ voting preferences

Kahan, Dan M., Misinformation and Identity-Protective Cognition (October 2, 2017). SSRN, https://ssrn.com/abstract=3046603

Abstract: This paper synthesizes existing work on misinformation relating to policy-relevant facts. It argues that misinformation has the greatest power to mislead when it interacts with identity-protective cognition.

Keywords: Identity Protective Reading, Misinform

---
Reasoning axiomatically from general dynamics of belief formation, commentators have tended to identify public consumption of “fake news” as an important part of Donald Trump’s victory [...] Allcott and Gentzkow (2017) report that individuals were much more likely to represent that they had read, agreed with, and shared articles that were favorable toward their preferred political candidate or unfavorable toward that candidate’s opponent. This is evidence, then, that consumption of the stories (or story: the average number of fake news articles read by a voter, A&G calculated, was 1) were a consequence, not a cause of their readers’ voting preferences.

Check also: Polarized Mass or Polarized Few? Assessing the Parallel Rise of Survey Nonresponse and Measures of Polarization. Amnon Cavari and Guy Freedman. The Journal of Politics, https://www.bipartisanalliance.com/2018/03/polarized-mass-or-polarized-few.html