Humprecht, E., Esser, F., & Aelst, P. V. (2020). Resilience to Online Disinformation: A Framework for Cross-National Comparative Research. The International Journal of Press/Politics, 1940161219900126. https://doi.org/10.1177/1940161219900126
Summary
This paper proposes a theoretical framework for studying societal resilience to online disinformation across countries, arguing that structural conditions of a nation’s political, media, and economic environment — not just individual-level psychology — determine how vulnerable a society is to disinformation. The authors define resilience as a collective, structural characteristic reflecting both lower exposure to disinformation and citizens being less inclined to believe or spread it. They identify seven macro-level dimensions that weaken resilience, translate them into measurable indicators drawn from existing comparative datasets, and empirically test them across eighteen Western democracies. A cluster analysis produces three distinct groups, with the United States emerging as a uniquely vulnerable outlier whose disinformation dynamics should not be generalized to other democracies.
Key Contributions
- Introduces and operationalizes the concept of country-level resilience to disinformation, reframing disinformation as a matter of structural defense rather than only external threat.
- Provides a systematic macro-level theoretical framework mapping seven structural dimensions across political, media, and economic environments.
- Translates abstract structural conditions into concrete, quantifiable indicators using existing comparative datasets, enabling country ranking and typology building.
- Delivers an empirical classification of eighteen Western democracies into resilience clusters to guide future case selection.
- Cautions against extrapolating U.S.-centric findings, highlighting the exceptional and non-generalizable nature of the U.S. case.
Methods
The authors develop their framework from the literature and operationalize seven dimensions — societal polarization, populist communication, trust in news, public service media strength, audience fragmentation, advertising market size, and social media use — plus an outcome dimension (exposure to disinformation). Indicators were drawn from secondary sources including the Digital News Report (2018, 2019), V-Dem, the Timbro Authoritarian Populism Index, the Global Populism Database, a public service broadcasting strength index, and World Bank data. Measures were merged into seven z-standardized indices with internal consistency checks (Cronbach’s alpha ~.71 to .96), some inverted so higher values denote higher resilience. Analysis combined principal components factor analysis (varimax rotation), a two-stage cluster analysis (Ward’s algorithm, three-cluster solution), and OLS regression predicting reported exposure. The eighteen-country sample follows Hallin and Mancini’s media systems typology.
Findings
- Substantial cross-national variation: Northern/Western European countries (Finland, Denmark, Netherlands) score high on most resilience indices, while Spain, Italy, Greece, and the U.S. score low.
- Factor analysis produced two factors explaining 69% of variance — factor 1 (social media use, media trust, polarization, populism, PSB) at 49.7% and factor 2 (market size, fragmentation, PSB) at 19.1%; PSB loaded on both.
- Three clusters emerged: a high-resilience media-supportive/consensual group (Western Europe plus Canada), a low-resilience polarized group (Greece, Italy, Portugal, Spain), and the United States alone.
- The OLS model explained 83% of variance in reported exposure; media trust, social media use, and market size were significant predictors, though polarization and shared media ran opposite to expectations.
- An alternative V-Dem expert-coded outcome measure explained ~40% of variance, with populism and social media use significant and polarization and shared media pointing in the predicted directions.
Connections
This paper is a foundational statement in cross-national comparative disinformation research and connects directly to work applying or extending its resilience framework, including Humprecht2025-ml by the same lead author. Its structural, media-systems approach aligns with Heft2021-ky on cross-national digital news infrastructures, Benkler2018-lw on structural media resilience, and Tsfati2020-uo on trust and the spread of misinformation, while its comparative resistance to U.S.-centric generalization resonates with broader debates in Vosoughi2018-at and Lazer2018-mm on the science of disinformation.