van Erkel, P., Hameleers, M., Farooq, A., Gattermann, K., Tulin, M., van den Hoogen, E., & de Vreese, C. (2026). The hostile misinformation effect: How ideological congruence drives the assessment of misinformation targets. Political Communication, 1–22. https://doi.org/10.1080/10584609.2026.2671760
Summary
This paper extends the classic hostile media effect into the misinformation domain, theorizing and testing a hostile misinformation effect: citizens systematically believe their political in-group is targeted by misinformation more than their out-group. Drawing on motivated reasoning and social identity theory, the authors use a three-wave panel survey conducted around the 2024 European Parliament elections in Germany, the Netherlands, and Poland (N=4,045) to demonstrate the effect robustly and across contexts. They further show that the bias is amplified by political interest, strength of party identification, ideological extremity, and right-wing orientation, and — contrary to attribution-bias expectations — is not stronger among electoral losers. The bias is associated with lower media trust and lower perceived electoral integrity, suggesting it has tangible downstream consequences for democratic legitimacy.
Key Contributions
- Conceptualizes and names the hostile misinformation effect, bridging hostile media effect research with misinformation perception scholarship.
- Provides cross-national, panel-based evidence during a real high-stakes EU election in three distinct media-political contexts.
- Maps individual-level moderators (interest, identity strength, ideological extremity, right-wing orientation) of susceptibility.
- Suggests that inflated public estimates of misinformation prevalence may partly reflect identity-driven perceptual bias rather than actual exposure.
- Offers preliminary evidence linking biased misinformation perceptions to erosion of media trust and electoral integrity beliefs.
Methods
A pre-registered three-wave online panel (Kantar) fielded before, during, and after the June 2024 EP elections in Germany, the Netherlands, and Poland, with quota sampling on age, gender, and region. Respondents rated on 7-point scales how much they felt their preferred and least-preferred parties (and left vs. right blocs) were targeted by misinformation. The data were stacked so each respondent contributed paired in-group/out-group ratings, analyzed via linear multilevel regressions with country fixed effects. Moderators included political interest, propensity-to-vote (proxy for party ID), folded ideological extremity, left–right position, and winner/loser status operationalized via vote-share change (with EP 2019 and subjective measures as robustness checks).
Findings
- 49.6% see their own party as targeted by misinformation vs. only 27.3% for the opposing party; a ~1-point gap on the 7-point scale (b=1.05, p<.001).
- Political interest strongly moderates the effect (b=0.28); the disengaged barely show it.
- Stronger party identification amplifies the bias (b=0.20); weak identifiers do not exhibit it.
- Ideological extremity strengthens the effect (b=0.25), and right-wing respondents show a larger gap than left-wing ones (b=0.13), with radical-right voters showing the largest interaction (b=1.58).
- Winner/loser status is null in the main specification; alternative measures hint that winners (especially radical-right winners) show a stronger effect, contradicting attribution-based predictions.
- Effect replicates with ideological (rather than party) in-/out-groups, across waves, and across all three countries (slightly weaker in the Netherlands).
- Hostile misinformation perceptions correlate with lower media trust (b=−.08) and lower perceived electoral integrity (b=−.05).
Connections
This work connects to research treating “misinformation” as a contested, identity-laden label rather than an objective category, resonating with Hameleers2026-mc on perceptions of mis/disinformation and with broader debates about how partisans construe hostile information environments — see Humprecht2025-ml on cross-national resilience and Kalsnes2025-zb on partisan asymmetries. The finding that right-wing and radical-right voters exhibit the strongest bias links to studies of hyperpartisan and right-wing information ecosystems such as Rothut2026-or and Frischlich2025-vn, while the downstream effects on media trust and electoral integrity speak to concerns raised in Gattermann2025-yx and Starbird2025-jj about how misinformation discourse itself shapes democratic legitimacy.
Podcast
A research-radio episode discusses this paper: 🎧 MP3 · Spotify · Apple Podcasts