‘More than a feeling’ – Facebook reactions and the sharing of political posts during Scandinavian elections

Summary

This paper offers a comparative analysis of how Scandinavian political parties and leaders deployed Facebook during the Norwegian (2021), Swedish (2022), and Danish (2022) national elections, focusing on the affective affordances of “Love” and “Angry” Reactions and their relationship to sharing. Drawing on CrowdTangle data and negative binomial regression, the authors argue that emotional reactions function as more than expressive gestures: they shape virality asymmetrically. Anger drives mobilization, love dampens it, and populist parties — particularly the Sweden Democrats and the Norwegian Progress Party — disproportionately benefit from this dynamic, even in the relatively consensual multiparty systems of Scandinavia.

Key Contributions

• One of the first explicitly cross-country comparisons of Facebook Reactions in Scandinavian election campaigns.
• Empirical demonstration that discrete reaction types map onto different sharing dynamics, refining accounts of emotional virality.
• Evidence that populist communication style benefits from platform affective affordances even in low-polarization, multiparty contexts.
• A replicable methodological template pairing scatterplot-based outlier identification with negative binomial regression of share counts.
• Proposed coding categories (gratitude/celebration vs. criticism/confrontation/call-to-action) for future content analysis of emotional political posts.

Methods

Quantitative comparative analysis of Facebook posts by parliamentary parties and their leaders, collected via CrowdTangle for the six weeks preceding (plus one day after) each election, yielding 1,609 Danish, 1,676 Norwegian, and 2,629 Swedish posts. Parties were classified as populist or non-populist using Inglehart & Norris (2016), the 2017 Chapel Hill Expert Survey, and the PopuList. Three per-country negative binomial regressions predicted shares from Love and Angry reactions, populist status, and actor type (party vs. leader), supplemented by scatterplots and qualitative inspection of high-engagement outliers.

Findings

• Swedish actors posted roughly twice as much as their Norwegian or Danish counterparts; the Sweden Democrats averaged 7.3 posts/day.
• Gratitude and anniversary posts from mainstream parties (Social Democrats, Labour, Red Party) attracted the highest Love reactions.
• Populist parties (Sweden Democrats, Hard Line, Progress Party) drew the most Angry reactions, typically on immigration, energy, and adversarial framings of opponents.
• Denmark produced the largest single Love spike (a Mette Frederiksen thank-you post, ~14K hearts); Norway produced the most-shared post (Vedum on electricity policy, ~7,500 shares); Sweden displayed the most balanced Love/Angry profile.
• Angry reactions positively predicted shares in all three countries (strongest in Norway); Love reactions negatively predicted shares in Norway (~7%) and Denmark (~12%), non-significant in Sweden.
• Party-leader posts were shared less often than party posts (3–6% less); populist status carried an unexpected negative coefficient in Norway once reactions were controlled for.

Connections

This study sits alongside other work tracing how platform affordances asymmetrically amplify negative or out-group affect, notably Rathje2021-negativity-style findings echoed in Mosleh2024-op and the broader anger-and-virality literature engaged by Rossini2026-jn and Gaisbauer2025-by. Its focus on populist advantage in low-polarization multiparty systems complements comparative European work such as Knupfer2025-vt and Esau2025-tf, while its Scandinavian framing speaks directly to Kristensen2025-ni. The use of platform engagement metrics to infer mobilization also connects to algorithmic-exposure debates seeded by Bakshy2015-rn.

Podcast

A research-radio episode discusses this paper: Listen