In a groundbreaking investigation that challenges long-held assumptions about social media dynamics during political turmoil, new research spearheaded by the University of Cambridge’s Social Decision-Making Lab reveals a profound shift in what drives online engagement in moments of crisis. Contrary to the widely accepted notion that outrage and division fuel digital virality, the study finds that during significant political upheavals, positive emotions rooted in ingroup solidarity become the dominant force energizing social media interactions.
The 2024 U.S. presidential election provided a natural experiment to explore how digital behaviours evolve under stress. Within a critical two-week span in July 2024, the political landscape was rocked first by an assassination attempt targeting Donald Trump at a public rally on July 13, and shortly thereafter, by Joe Biden’s announcement suspending his re-election campaign on July 21. These seismic events served as a catalyst for an unprecedented shift in the tone and content of political discourse on social media platforms, primarily Facebook.
To understand this transformation, the research team amassed an extensive dataset comprising over 62,000 public Facebook posts from a wide array of U.S. political actors, media commentators, and institutional outlets aligned with both Republican and Democratic perspectives. Employing rigorous computational text analysis and sentiment categorization methods, the study dissected the emotional and thematic qualities embedded in these posts before, during, and following the crisis moments, seeking patterns in engagement metrics such as shares, comments, and likes.
Historically, negative sentiments—in particular anger, outrage, and hostile partisanship—have been recognized as potent drivers of social media virality, acting as “rocket fuel” that amplifies political content to broad audiences. This phenomenon is grounded in psychological mechanisms where threat and conflict stimulate heightened arousal and motivation to react online. Conventional wisdom thus anticipated an even more intense surge of antagonistic posts and consequent engagement during the 2024 crises.
However, the empirical evidence uncovered is strikingly counterintuitive. Following the attempt on Trump’s life, posts emanating from the Republican sphere that emphasized unity, shared identity, and ingroup cohesion experienced a remarkable 53% increase in engagement compared to preceding days. This uplift represented a significant 17 percentage point jump, signaling a pivot away from outgroup-focused antagonism toward inward-facing solidarity. Exemplars included messages from religious leaders expressing gratitude for Trump’s survival and media figures framing his resilience as a rallying symbol for America.
Simultaneously, content coded as hostile or adversarial towards Democrats saw a notable decline in interaction, with Republican posts attacking their political opponents waning in engagement by 23 percentage points. This suggests that the collective shock imposed by the assassination attempt dampened the usual appetite for divisive rhetoric, instead fostering a common psychological imperative to consolidate within-group bonds for protection and morale.
A parallel yet even more pronounced effect unfolded after Biden’s announcement on suspending his re-election bid. Democrat-aligned social media posts marked by themes of solidarity, respect, and shared values soared in popularity, generating 91% more engagement relative to previous levels—a dramatic spike of 71 percentage points. Revered political figures extolled Biden’s legacy, emphasizing continuity and collective purpose despite electoral uncertainty. Contrasting with Republican dynamics, criticism of the opposition party among Democrats continued to incrementally increase but at a less dominant rate, reflecting nuanced differences in crisis-driven social media behaviour across ideological lines.
Underlying this shift is a psychological framework in which ingroup love—the emotional preference and affection for one’s own community—surpasses outgroup hate as the principal determinant of attention and sharing during existential or high-stakes threats. These findings resonate with prior research from the same Cambridge lab, which analyzed social media responses during other acute geopolitical crises, such as Russia’s full-scale invasion of Ukraine in 2022. In that context, posts celebrating national unity and shared identity surged dramatically, while hostile commentary directed at the aggressor attracted comparatively less engagement.
Beyond empirical observation, the study advances a theoretical challenge to dominant narratives about digital polarization. While platforms like Facebook and Twitter are often criticized for exacerbating division through algorithmic amplification of anger and outrage, the new evidence illuminates the malleability of social media’s emotional ecosystem. At moments of profound uncertainty and threat, humans appear to gravitate toward social cohesion online, prioritizing collective resilience over factional conflict.
Moreover, the research highlights critical implications for the design of information environments and the potential to harness social media not merely as battlegrounds of partisan hostility but as venues for cultivating solidarity and mutual support under duress. Understanding these dynamics could inform platform moderation policies, media strategies, and interventions aimed at mitigating polarization while reinforcing social fabric during crises.
Key to these insights was the collaboration of interdisciplinary experts in psychology, data science, and political communication, who integrated sentiment analysis with discourse evaluation, engagement metrics, and contextual political knowledge. The lead author, Malia Marks, PhD candidate at Cambridge’s Department of Psychology, emphasizes the importance of viewing social media behaviour through both emotional and structural lenses to capture the complex interplay between online expression and offline realities.
Co-author and senior researcher Dr. Jon Roozenbeek reflects on the broader resonance of the findings: “Our study underscores that ingroup love, rather than outgroup hate, can dominate social media behavior in times of crisis—even in deeply polarized societies like the United States. This challenges prevailing assumptions and opens avenues for more hopeful perspectives on digital political engagement.”
In sum, this research redefines our understanding of digital political communication by demonstrating that collective solidarity emerges as the crucial viral driver during moments when political leadership is threatened, transforming social media from arenas of discord into bastions of communal support. The implication is profound: in the swirling chaos of modern politics, shared identity and unity still possess unparalleled power to shape public discourse.
Subject of Research: Social media engagement dynamics and emotional content shifts during political crises in the 2024 U.S. presidential election.
Article Title: Ingroup solidarity drives social media engagement after political crises
News Publication Date: 29-Aug-2025
Web References:
- https://www.pnas.org/doi/abs/10.1073/pnas.2512765122
- https://www.nature.com/articles/s41467-024-52179-8
References:
Marks, M., Kyrychenko, Y., Roozenbeek, J., et al. (2025). Ingroup solidarity drives social media engagement after political crises. Proceedings of the National Academy of Sciences. DOI: 10.1073/pnas.2512765122
Keywords: Social media engagement, political crises, ingroup solidarity, digital behavior, online virality, political polarization, 2024 U.S. election, emotional contagion, computational social science, political communication
