Average US adult saw only about three fake news pages during the 2016 election

Evidence

Based on known social media estimates, they wagered that a typical share on Facebook generates about twenty page visits. Given the known 38 million shares of fake news stories in their database, that amounts to about 20 × 38 million, which equals 760 million page visits. If we divide 760 million by the adult US population in 2016 (about 248 million Americans) we arrive at about three visits to fake news websites per adult. That’s not much. — Sander van der Linden, Foolproof, 517

Categorization

• SEED: This back-of-envelope arithmetic — shares times page-views per share divided by population — is a clean, teachable example of how thin assumptions underlie even expert disinformation-exposure estimates. Could anchor a skeptical piece on quantitative claims about misinformation prevalence.
• QUESTION: The ‘20 page visits per share’ multiplier is doing enormous work here. Is it empirically grounded or an assumption by assumption? Worth chasing the original Allcott-Gentzkow paper for how the number was derived.
• XREF: connects to: Americans in 2016 each read roughly three fake news stories in the election build-up — Shoshana Zuboff, The Age of Surveillance Capitalism, 1611; Fake news exposure in 2016 probably could not swing election results — Sander van der Linden, Foolproof: Why Misinformation Infects…, 520; Multi-platform exposure could mean daily fake news contact for average voters — Sander van der Linden, Foolproof: Why Misinformation Infects…, 526 (+1 more)

Broader Connections

• Americans in 2016 each read roughly three fake news stories in the election build-up #corroborates (1.00) — Shoshana Zuboff, The Age of Surveillance Capitalism, 1611
• Fake news exposure in 2016 probably could not swing election results #qualifies (0.85) — Sander van der Linden, Foolproof, 520
• Multi-platform exposure could mean daily fake news contact for average voters #contradicts (0.80) — Sander van der Linden, Foolproof, 526
• Self-reported fake news exposure greatly exceeds model-based estimates #contradicts (0.80) — Sander van der Linden, Foolproof, 525