Virality: What Makes Narratives Go Viral, and Does it Matter

Kai Gehring and Matteo Grigoletto

No 12064, CESifo Working Paper Series from CESifo

Abstract: Understanding behavioral aspects of collective decision-making remains a core challenge in economics. Political narratives can be seen as a key communication technology that shapes and affects human decisions beyond pure information transmission. The effectiveness of narratives can be driven as much by their virality as by their specific persuasion power. To analyze political narratives empirically, we introduce the political narrative framework and a pipeline for its measurement using large language models (LLMs). The core idea is that the essence of a narrative can be captured by its characters, which take on one of three archetypal roles: hero, villain, and victim. To study what makes narratives go viral we focus on the topic of climate change policy and analyze data from the social media platform Twitter over the 2010–2021 period, using retweets as a natural measure of virality. We find that political narratives are consistently more viral than neutral messages, irrespective of time or author characteristics and other text features. Different role depictions differ in terms of emotional language, but political narratives capture more than merely valence or emotions. Hero roles and human characters increase virality, but the biggest virality boost stems from using villain roles and from combining other roles with villain characters. We then examine the persuasiveness of political narratives using a set of online experiments. The results show that narrative exposure influences beliefs and revealed preferences about a character, but a single exposure is not sufficient to move support for specific policies. Political narratives lead to consistently higher memory of the narrative characters, while memory of objective facts is not improved.

Keywords: narrative economics; climate change policy; social media; virality; political economy; media economics; text-as-data; large language models (search for similar items in EconPapers)
JEL-codes: C80 D72 H10 L82 P16 Q54 Z1 (search for similar items in EconPapers)
Date: 2025
References: Add references at CitEc
Citations:

Downloads: (external link)
https://www.ifo.de/DocDL/cesifo1_wp12064.pdf (application/pdf)

Related works:
This item may be available elsewhere in EconPapers: Search for items with the same title.

Export reference: BibTeX RIS (EndNote, ProCite, RefMan) HTML/Text

Persistent link: https://EconPapers.repec.org/RePEc:ces:ceswps:_12064

Access Statistics for this paper

More papers in CESifo Working Paper Series from CESifo Contact information at EDIRC.
Bibliographic data for series maintained by Klaus Wohlrabe ().