Donovan, J. (2025). EXPRESS: A short history of misinformation-at-scale and efforts to mitigate it. Journal of Public Policy & Marketing. https://doi.org/10.1177/07439156251384249
Summary
Joan Donovan offers a critical historical account of how “misinformation-at-scale” became a defining public concern between 2016 and 2021, and how social media companies responded through evolving content moderation practices. Her central argument is that misinformation-at-scale — false information amplified by densely networked actors like politicians, journalists, and celebrities — is not a bug but a structural feature of engagement-driven, advertising-based platform business models. Donovan traces platforms’ arc from non-intervention, through behavioral moderation (notably “coordinated inauthentic behavior”) and aggressive COVID-era and 2020-election interventions, to a post-2021 retrenchment in trust-and-safety investment. She frames the underlying failure as platforms’ inability to deliver TALK (timely, accurate, local knowledge), and argues that meaningful mitigation requires a whole-of-society response rather than platform self-correction.
Key Contributions
- Introduces and operationalizes misinformation-at-scale as a sociotechnical event distinct from individual rumor, arising when densely networked actors amplify falsehoods into controversies with negative externalities.
- Introduces TALK (timely, accurate, local knowledge) as the public good that engagement-optimized platform design systematically fails to surface.
- Provides a comparative taxonomy of fourteen content moderation methods mapped to the actors (individuals, groups, platforms, governments) empowered to deploy them.
- Offers a historical periodization of platform moderation (2016–2021), from non-intervention to active moderation to retrenchment.
- Synthesizes platform governance, public health communication, and democratic theory to argue for whole-of-society mitigation.
Methods
Donovan combines comparative historical analysis of platform moderation policies (2016–2024) with ethnographic case studies of key misinformation events — Charlottesville, COVID-19, the 2020 U.S. election, January 6th, and the Hunter Biden laptop story. She develops conceptual and typological work (the fourteen-method taxonomy), draws on participant-observation with civil society groups such as the Disinformation Defense League and the Election Integrity Partnership, and synthesizes investigative journalism, congressional hearings, whistleblower disclosures (Frances Haugen), and prior scholarship.
Findings
- Nine of fourteen catalogued moderation methods target amplification rather than removal, reflecting platforms’ preference for limiting scale over policing content itself.
- Facebook’s CIB framework let it act on behavioral signals without adjudicating truth, repurposing existing spam and fraud tools — because addressing veracity conflicts with the ad-based model.
- Labeling Trump’s tweets reduced on-platform engagement but increased cross-platform engagement, a dynamic Donovan calls platform filtering.
- Platforms moderated medical misinformation more aggressively than political misinformation, using information displacement (e.g., COVID information centers) to surface authoritative sources.
- After 2021, Meta and X sharply cut trust-and-safety capacity (Meta’s 11,000 layoffs; Musk dismantling X’s civic integrity and ethics teams; Google’s 2023 cuts).
- Misinformation actors adapt rapidly through coded language, backup accounts, and migration to less-moderated platforms.
- Trump’s January 2021 deplatforming marked a watershed inversion, signaling platforms could challenge state authority.
Connections
This paper offers a macro-historical and structural framing that complements empirical and conceptual work on platform governance and coordinated manipulation, including studies of coordinated inauthentic behavior and coordinated link sharing such as Giglietto2022-b30e8b4e, Giglietto2017-4375de2f, and Giglietto2020-6278a4aa. Its emphasis on prebunking and civil-society mitigation strategies connects to inoculation and countermeasure research like van-der-Linden2026-jt and Lewandowsky2026-ob, while its critique of platform self-governance and trust-and-safety retrenchment speaks to work on data access and platform accountability such as Rieder2025-ju and Bruns2026-yv.