What is it about?
When news channels post videos on YouTube, they choose words and images that trigger emotional reactions in viewers. This study looks at one emotion in particular: moral outrage, the feeling of anger or contempt toward someone perceived as doing something wrong. We collected about 400,000 YouTube videos from major Korean and American news outlets throughout 2024 and developed a classifier that reads both the thumbnail image and the video title together to detect what kind of moral emotion the content expresses. We then measured whether that emotional framing affected how many people watched, liked, or commented on each video. The results were consistent across both countries: content expressing moral outrage toward others reliably increased all three forms of engagement, and the effect grew stronger as the behavior required more effort, from passive viewing to active commenting.
Featured Image
Photo by Adem AY on Unsplash
Why is it important?
Most prior research on moral emotions and online engagement relied solely on text. YouTube is a platform where the thumbnail image and the title work together to grab attention, so analyzing only one of them misses a big part of how emotional framing actually operates. This study is the first to build a multimodal moral emotion classifier for both Korean and English, making it possible to study the phenomenon beyond Western contexts. The finding that moral outrage drives not just views but committed behaviors like commenting has direct implications for how we understand polarization online: platforms and news outlets that optimize for engagement are effectively rewarding a rhetorical strategy that divides audiences into in-groups and out-groups.
Perspectives
What I find most striking about this work is how consistent the pattern turned out to be across two culturally distinct media environments. Korea and the United States differ considerably in political culture, media landscape, and audience behavior, yet the amplifying effect of other-condemning rhetoric showed up in both datasets with remarkable stability. That consistency suggests we are looking at something close to a platform-level dynamic rather than a culture-specific phenomenon. It also makes the concern about misuse more urgent. The classifier we release is designed to support future research on misinformation and troll behavior, but the broader implication is that the attention economy structurally rewards division, and that reward operates regardless of cultural context.
Jaehong Kim
Korea Advanced Institute of Science and Technology
Read the Original
This page is a summary of: Moral Outrage Shapes Commitments Beyond Attention: Multimodal Moral Emotions on YouTube in Korea and the US, April 2026, ACM (Association for Computing Machinery),
DOI: 10.1145/3774904.3792728.
You can read the full text:
Contributors
The following have contributed to this page
